Interplay between Notch signaling and cytokinesis in the Drosophila ...

HAL Id: tel-03139834https://tel.archives-ouvertes.fr/tel-03139834

Submitted on 12 Feb 2021

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives| 4.0International License

Interplay between Notch signaling and cytokinesis in theDrosophila sensory organ lineage

Mateusz Trylinski

To cite this version:Mateusz Trylinski. Interplay between Notch signaling and cytokinesis in the Drosophila sensory organlineage. Cellular Biology. Sorbonne Université, 2019. English. �NNT : 2019SORUS470�. �tel-03139834�

https://tel.archives-ouvertes.fr/tel-03139834

http://creativecommons.org/licenses/by-nc-nd/4.0/



https://hal.archives-ouvertes.fr

Sorbonne Université École doctorale Complexité du Vivant

Laboratoire Génétique du Développement de la Drosophile

Interplay between Notch signaling and cytokinesis in the

Drosophila sensory organ lineage

Par Mateusz Trylinski

Thèse de doctorat de Biologie

Dirigée par François Schweisguth

Présentée et soutenue publiquement le 25 janvier 2019

Devant un jury composé de :

Mme Agnès Audibert

Mr Buzz Baum

Mr Stéphane Vincent

Mme Sarah Bray

Mr Bassem Hassan

Mr François Schweisguth

Professeur Sorbonne Université

Professeur UCL

MCU ENS Lyon

Professeur Université de Cambridge

DR Sorbonne Université

DR1 CNRS

Présidente

Rapporteur

Rapporteur

Examinatrice

Examinateur

Directeur de thèse

Acknowledgements First and foremost, I thank François for mentoring me during all these years and teaching me the

virtues and the subtleties of analytical reasoning in experimental sciences. You taught me to formulate

clearly my hypotheses, to design rigorously experiments testing them and not just confirming the

model I had in mind, to analyze critically my results, to formulate again new hypotheses, and so forth.

To deconstruct each biological question into small pieces and then patiently, step by step, bring them

together into a larger picture.

I would like to warmly thank all the present and past lab members with whom I shared so

much time during these years. First of all: Gantas! These years would never have been so great

without you. Each morning, I knew that some good time was awaiting me while you were in the lab.

Not only I found in you a terrific lab mate, but also a close friend. Khalil, such a good friend as well!

Always willing to share a good discussion about politics and, not least, some pretty good music

(unforgettable Radio Moscow!). All these things helped me to go through this thesis. Lydie, for your

unconditional willingness to help with flies or anything and for your incredible capacity in seeing the

good in all things. Vanessa, for your equally unconditional mood of yours and your communicative

laugh! I will never forget your shouts when something was not working as you expected. Elodie,

you’ve always been really attentive and willing to give tips and advice, to this day. Fred, you taught

me so many tricks with fly genetics, I’m so grateful! My crosses would never have looked the same

without you. Franck, for all the tips in my early PhD days. Maria, for your unfailing cool attitude and

your kindness. Chloe, I’ll miss our discussions on blockbusters! Alexis, you were my first master

student, I hope you enjoyed as much as I did the few months you spent with fly ommatidia. Juan, your

quirky humor that always made me smile. Tin, your beers will be unforgettable and Jang-mi, our new

fly padawan always keen on discussing science.

Stéphane, you were my first fly master. You shared with me your passion with these little

flying objects and introduced me to the beauty of their genetics and their development. At each step of

my studies, you were there to give me advice or simply to discuss science. You helped in finding my

way in the vast domain that is Biology. I’m really grateful for that.

I want also to express my gratitude to my family, who tried to be close despite the distance,

and my friends. I’m particularly grateful to Paul. Our weekly lunches were these magical breaks far far

away from science and experiments.

Last, and certainly not the least, I deeply thank Alison for supporting me during this long

journey and not being jealous with my little flies.

Content

INTRODUCTION ............................................................................................................................ 1 PREFACE ........................................................................................................................................ 3 1. NOTCH SIGNALING: MECHANISM, REGULATION AND DYNAMICS............................ 5

1.1. MECHANISM OF NOTCH RECEPTORS ACTIVATION BY ITS LIGANDS: “TRANS-ACTIVATION” . 6 1.1.1. Interaction between Notch and its ligands ..................................................................... 7 1.1.2. Conformational change of the receptor induced by ligand binding ............................... 15 1.1.3. Gamma-secretase and S3 cleavage .............................................................................. 21 1.1.4. Transcriptional activity of NICD ................................................................................. 22

1.2. REGULATION OF NOTCH RECEPTORS BY THE ENDOSOMAL PATHWAY ............................... 25 1.2.1. Notch endocytosis and regulation of surface levels ...................................................... 26 1.2.2. Recycling of Notch receptors ....................................................................................... 28 1.2.3. Degradation of Notch receptors in lysosomes .............................................................. 30

1.3. LIGAND ENDOCYTOSIS AND TRAFFICKING ........................................................................ 31 1.3.1. Ligand endocytosis and generation of the pulling force ............................................... 31 1.3.2. Ligand endocytosis and ubiquitination......................................................................... 37 1.3.3. Ligand endocytosis and trafficking .............................................................................. 43

1.4. NOTCH SIGNALING DYNAMICS ......................................................................................... 45 1.4.1. Experimental strategies to study Notch signaling dynamics ......................................... 45 1.4.2. Time-scaled genome profiling of the Notch transcriptional response............................ 53 1.4.3. Dynamic encoding of NICD production kinetics .......................................................... 55

1.5. CONCLUDING REMARKS................................................................................................... 57 2. NOTCH SIGNALING DURING DROSOPHILA NEUROGENESIS ..................................... 58

2.1. LATERAL INHIBITION AND SELECTION OF NEURAL PROGENITORS ..................................... 58 2.1.1. Embryogenesis and neuroblast selection...................................................................... 59 2.1.2. Notum patterning and SOP selection ........................................................................... 60 2.1.3. Photoreceptor R8 selection in the eye disc ................................................................... 65

2.2. CENTRAL NERVOUS SYSTEM: NEURAL EXPANSION AND CELL TYPE DIVERSIFICATION ....... 69 2.2.1. Neuroblast self-renewal............................................................................................... 70 2.2.2. Cell type differentiation and diversification ................................................................. 70

2.3. PERIPHERAL NERVOUS SYSTEM: SPECIFICATION OF THE MICROCHAETE LINEAGE ............. 71 2.3.1. The microchaete lineage: an overview ......................................................................... 71 2.3.2. Molecular mechanisms of SOP asymmetric cell division .............................................. 74 2.3.3. Unequal segregation of fate determinants and directional Notch signaling .................. 76 2.3.4. Notch is activated in pIIa, but where?.......................................................................... 80 2.3.5. Notch is activated in pIIa, but when? ........................................................................... 88

3. THE ARP2/3 COMPLEX, A MOLECULAR PIVOT BRIDGING CYTOKINESIS AND NOTCH ACTIVATION?............................................................................................................... 91

3.1. DIVIDING IN AN EPITHELIUM ............................................................................................ 91 3.1.1. Overview of cytokinesis progression ............................................................................ 92 3.1.2. Epithelial cytokinesis as a multicellular process .......................................................... 95 3.1.3. Asymmetric furrow ingression ..................................................................................... 97

3.2. ARP2/3 ROLE DURING CYTOKINESIS ............................................................................... 100 3.2.1. The Arp2/3 complex, a major actin regulator............................................................. 100 3.2.2. Regulation of the Arp2/3 complex by NPFs and inhibitors ......................................... 103 3.2.3. Contact expansion regulated by the Arp2/3 complex: the “zipper” mechanism .......... 109

3.3. ARP2/3 ROLE IN NOTCH-DEPENDENT FATE DECISIONS .................................................... 110 3.3.1. Arp2/3 and WASp requirement in Notch signaling ..................................................... 110 3.3.2. Arp2/3, WASp and Delta recycling: a problematic model .......................................... 114 3.3.3. Alternative hypotheses ............................................................................................... 115

RESULTS ..................................................................................................................................... 120 4. PAPER 1: INTRA-LINEAGE FATE DECISIONS INVOLVE ACTIVATION OF NOTCH RECEPTORS BASAL TO THE MIDBODY IN DROSOPHILA SENSORY ORGAN PRECURSOR CELLS ................................................................................................................. 121 5. PAPER 2: THE ARP2/3 COMPLEX COUPLES CYTOKINESIS TO NOTCH ACTIVATION IN THE DROSOPHILA SENSORY ORGAN LINEAGE ................................ 142

DISCUSSION AND PERSPECTIVES ........................................................................................ 180 6. DISCUSSION ........................................................................................................................... 181

6.1. APICAL VS LATERAL NOTCH SIGNALING ....................................................................... 181 6.1.1. Notch receptors: is localization reflecting function? .................................................. 181 6.1.2. Why an apical pool of Notch receptors in SOPs? ....................................................... 182 6.1.3. The lateral contact: a simple matter of area?............................................................. 183

6.2. INTERPLAY BETWEEN CYTOKINESIS AND NOTCH SIGNALING.......................................... 184 6.2.1. Contact formation and exocytosis .............................................................................. 185 6.2.2. Dual mode of Arp2/3 activation ................................................................................. 185 6.2.3. Is lineage progression hiding a “competence window”? ............................................ 186

7. PERSPECTIVES...................................................................................................................... 188 7.1. A WIDELY-APPLICABLE TECHNIQUE TO IDENTIFY NOTCH ACTIVATION SITE IN VIVO....... 188

7.1.1. Testing the contact area/Notch activity correlation .................................................... 188 7.1.2. Notch signaling site and regulatory mechanisms ....................................................... 189

7.2. ARP2/3-MEDIATED DELTA ENDOCYTOSIS: A MOLECULAR BASIS FOR THE MECHANOTRANSDUCTION MODEL ....................................................................................................

7.2.1. Towards an array of Delta endocytosis mechanisms .................................................. 190 7.2.2. Probing the forces generated by WASp-Arp2/3-dependent Delta endocytosis ............. 191

REFERENCES............................................................................................................................. 193 LIST OF FIGURES ..................................................................................................................... 224 ABSTRACT AND RÉSUMÉ ....................................................................................................... 226

1

Part I

Introduction

2

3

Preface Young PhD students often hear that investigators in developmental biology are split into two

categories: those who know and those who do not know yet that they are working on the

Notch pathway. Albeit this anecdote appears presumptuous at first sight, years of research

showed that the Notch pathway is in fact reused over and over during any Metazoan

development and adult life. It appears that Notch signaling has become through evolution the

favorite cellular toolkit to mediate juxtacrine signal transmission in a wide range of cellular

and developmental contexts, including fate specification, fate maintenance, boundary

formation, patterning or morphogenesis. Recent years provided valuable insights into the

molecular mechanism of receptor activation, the role of context-specific regulators and the

function of signaling dynamics in eliciting appropriate outputs. Nonetheless, some

fundamental questions are still barely understood. Where are Notch receptors activated in

relationship with the biological context? How is Notch activity coordinated with the

progression of a developmental process? Addressing these questions would shed a new light

on our understanding of the cellular mechanisms that fine tune or dysregulate Notch signaling

activity in physiological or pathological situations, respectively.

The introduction of this thesis is divided into three chapters: the first chapter gives a

synthetic overview on our current state of knowledge about the core mechanism the Notch

pathway; the second describes the multiple uses of Notch signaling during Drosophila

neurogenesis with a specific emphasis on the thoracic microchaete lineage and the question of

Notch activation site; the last introduces the Arp2/3 complex as a potent molecular pivot

between Notch activation and cytokinesis in lineages.

4

5

Chapter 1

Notch signaling: mechanism, regulation and dynamics

Notch signaling is an evolutionary conserved juxtacrine pathway relying on a fairly simple

logic where one signaling input translates into downstream signaling output. In biological

terms, a single Notch receptor is activated by one ligand harbored at the cell surface of a

contacting cell and releases its intracellular domain in the cytosol, which translocates to the

nucleus and directs gene expression. Unlike most pathways, signal is not modulated

downstream the receptor and prior nuclear entry by intermediate kinases. Thereby, signal

transduction is conserved in a stoichiometric manner along the pathway and affects

transcription in the signal-receiving cell in a straightforward manner (Housden and Perrimon,

2014).

The Notch pathway is not only conserved among phyla but is used in an incredibly

wide range of processes during animal development and adult life. Depending on the

biological context, Notch can elicit diverse, and sometimes opposite, cellular responses,

including proliferation/cell death, differentiation/self-renewal, tumor formation/tumor

suppression, and so forth (Bray, 2016). Since the characterization of the core mechanism in

the 1990’s, extensive work conducted on all available model organisms revealed that Notch

signaling is subjected to an equally diverse array of regulatory mechanisms. Therefore, Notch

signaling can be viewed as a simplified module adapted by cells to transduce context specific

signals.

This first introductory chapter synthesizes the current state of knowledge on the core

mechanism of Notch activation and emphasizes how each step of this process can be

modulated to generate specificity. Second, it summarizes how the regulation of receptor and

ligand trafficking affects signaling. In particular, the role of ubiquitin-dependent ligand

endocytosis is outlined. Lastly, it focuses on the signaling dynamics of the pathway, both in

terms of transcriptional response and Notch activation kinetics.

Chapter 1. Notch signaling: mechanism, regulation and dynamics

6

1.1. Mechanism of Notch receptors activation by its ligands:

“trans-activation”

The core mechanism of the Notch pathway consists in a four-step process: following ligand-

receptor interaction at a cell-cell contact (1), the receptor undergoes a conformational change

(2) that initiates a sequence of proteolytic cleavages, leading to the release of Notch

Intracellular Domain (NICD) in the cytosol (3). NICD then translocates into the nucleus and

acts as a transcriptional coactivator (4) (Figure 1).

Figure 1. Overview of the sequence leading to Notch activation. See text for

details. Adapted from Bray, 2016.

This section describes the current molecular models for each of these steps and briefly

presents the unanswered questions. The non-canonical Notch pathway and ligand-independent

mechanisms of Notch activation will not be detailed here and have been reviewed in

Andersen et al., 2012 and in Palmer and Deng, 2015.

1.1. Mechanism of Notch transactivation

7

1.1.1. Interaction between Notch and its ligands

The receptor-ligand interface

Notch and its ligands of the Delta/Serrate/Lag family (DSL) are type I transmembrane

proteins that interact through their extracellular domains (ECDs). Notch receptor ECDs are

made of a succession of 29 to 36 EGF repeats followed by a Negatively Regulated Region

(NRR) (Kopan and Ilagan, 2009) (Figure 2A). On the other hand, DSL ligands are

characterized by three main structural motifs: a module at the N terminus of Notch ligands

(MNNL) consisting in C2 domain with lipid binding properties, a DSL domain analogous to

EGF repeats and additional EGF repeats (6 to 16) (Kopan and Ilagan, 2009) (Figure 2B).

Figure 2. Structure and organization of Notch receptors (A) and DSL ligands (B).

Adapted from Gordon et al., 2008.

Early molecular characterization of Notch domains showed that receptor-ligand

engagement (Fehon et al., 1990; Heitzler and Simpson, 1991) is mediated by the interaction

between Notch EGF11-12 (Rebay et al., 1991) and the DSL domain (Shimizu et al., 1999).

Recent technological advances in membrane protein crystallography shed a new light both on

structures of isolated receptor and ligand ECDs and on the structure of the receptor-ligand

complex (reviewed in Kovall et al., 2017). Analysis of crystalized receptor and ligand ECDs

challenged the common view where receptor and ligand structures resemble linear and rigid

threads that extend from the cell surface (Cordle et al., 2008; Hambleton et al., 2004). In the

contrary, they rather alternate rigid segments with flexible pivots and bent regions, suggesting


8

that ECDs might support different conformation and consequently present different binding

affinities depending on their conformation (Kershaw et al., 2015; Weisshuhn et al., 2016).

Resolution of a Notch1-Dll4 complex confirmed this idea and revealed that receptor

and ligand ECDs interact in an anti-parallel fashion (Luca et al., 2015), a conformation that

supports both trans-activation and cis-inhibition (i.e. when ligand and receptor interact on the

same cell surface, see “Cis interactions”) (Figure 3). In addition, in contrast with former

models, they unveiled that the MNNL directly interacts with EGF12. Authors speculated that

this second interaction site might play a key role in modulating binding affinity in response to

post-translational modifications. Whereas the residues involved in the MNNL-EGF12

interface of the Notch1-Dll4 complex are poorly conserved among other receptors and

ligands, the EGF11-DSL interface shows high sequence homology. Finally, they provided a

structural basis for the role of sugar-based post-translational modifications of the receptor in

modifying receptor-binding affinity (see “Receptor glycosylation” in this subsection).

Figure 3. Structure of the Notch1-Dll4 complex in trans and cis conformations.

Adapted from Luca et al., 2015.


9

Of note, the crystal of the Notch1-Dll4 complex was resolved with mutated ligands

displaying an enhanced binding affinity for the receptor, interactions of endogenous receptor-

ligand complexes being too weak in solution to allow crystallization. Intriguingly, although

these affinity-enhancing mutations were found in the DSL and MNNL regions, they did not

affect residues that directly contacted EGF11-12. Hence, these substitutions might stabilize

the DSL and MNNL modules in a conformation that would strengthen binding interactions.

This might suggest that wild-type ligands are more flexible and would undergo dynamic

conformational changes at the cell surface with differential binding affinities. However,

whether such changes occur in living cells and how that might affect receptor activation and

signaling dynamics remain to be tested.

Ligand specificity

Identification of multiple Notch ligands (Delta and Serrate in Drosophila, Dll1-4 and

Jagged1-2 in mammals, Figure 2B) raised the question whether they can activate Notch in a

similar fashion to trigger the same biological outputs. Early work in Drosophila showed that

Delta and Serrate are interchangeable depending on the context. On one hand, Delta and

Serrate are either replaceable by one another or redundant during early embryonic

neurogenesis and in the peripheral nervous system, respectively (note that compensation

experiments of Delta by Serrate in the embryo were based on Serrate overexpression. It is still

unknown whether Serrate elicits the same response as Delta in Drosophila when expressed at

Delta endogenous levels) (Gu et al., 1995; Zeng et al., 1998). Similar results were obtained in

the chick inner ear and during primary neurogenesis in Xenopus laevis (Adam et al., 1998;

Kiyota et al., 2001). On the other hand, activity of Fringe, a transferase that adds glycans on

Notch ECD (see “Receptor glycosylation” in this subsection), discriminates between Serrate

and Delta in their capacity to activate Notch during dorsal-ventral margin specification in

Drosophila wing and eye discs, suggesting that both ligands are not equivalent depending on

the cellular context (de Celis and Bray, 1997; Fleming et al., 1997; Panin et al., 1997).

Since ECD sequences of Delta and Serrate differ by the number of EGF repeats (9 and

17, respectively) and the presence of a cystein-rich domain in Serrate, one could expect that

these two proteins would interact differently with Notch, which might in turn translate into

different capacities to activate Notch (see below). However, such differential behavior has

also been reported when comparing two close paralogs, Dll1 and Dll4 (Preuße et al., 2015).

Homozygous Dll1Dll4ki mouse embryos, where Dll4 is knocked into the locus of Dll1, failed to

give rise to adult mice and exhibited contrasting phenotypes in Dll1-dependent processes. For


10

example, while the retinal neuroepithelium was normally formed, somitogenesis was strongly

impaired. This indicates that, despite similar sequences, Dll4 can fully replace Dll1 only in

specific developmental and cellular processes.

What could be the basis of these ligand-specific processes? One part of the answer lies

in the structure of the receptor-ligand complex. Inducing a point mutation in EGF8 of

Drosophila Notch, termed “jigsaw”, was sufficient to impair Notch-Serrate signaling and

decrease Serrate binding affinity for Notch, while neither Delta signaling activity nor its

binding affinity to Notch were affected (Yamamoto et al., 2012). Comparison between the

structures of Dll4-Notch1 and Jagged1-Notch1 complexes revealed that the interacting

surface between Jagged1 and Notch1 was smaller, consistent with Dll4 high affinity for

Notch1 in vitro, and confirmed the role of EGF8 in stabilizing Notch1-Jagged1 interactions

(Luca et al., 2015, 2017). These differences were then correlated with the differential

capacities of the ligands to drive the conformational change of Notch1, a key step leading to

receptor activation (detailed in subsection 1.1.2.). In brief, while Dll4 could activate Notch1

when forces as low as 4 pN were applied, Jagged-1 mediated activation required a tensile

force of 12 pN. These results led the authors to conclude that cells could discriminate between

different ligands through this differential mechanical requirement. Another implication of this

observation was that Notch1-Jagged1 complexes experienced two different states depending

on the force regime: unproductive/low forces and productive/high forces, raising the

hypothesis that unproductive Notch1-Jagged1 complexes might behave as sinks until higher

forces are exerted and, moreover, explain how ligands can “trans-inhibit” receptors in specific

contexts (Benedito et al., 2009).

However, how do these mechanical requirements translate into distinct NICD

productions dynamics and specific physiological outputs in the signal receiving cell remains

unclear and will be discussed in subsection 1.4.3.

Receptor glycosylation

Glycosylation on receptor ECDs plays a critical role in tuning the affinity between the

receptors and its ligands. In parallel, glycosylation, in particular O-fucosylation and O-

glucosylation, can also modulate Notch signaling by regulating receptor trafficking from the

ER to the surface and protein stability. As this will not be discussed in this introduction,

additional details can be found in Harvey and Haltiwanger, 2018.

The landmark discovery that Fringe, a modulator of Notch activity that discriminates

between Delta and Serrate signaling (Panin et al., 1997), is a glycosyltransferase was crucial


11

to appreciate the role of glycosylation in regulating Notch signaling (Brückner et al., 2000;

Moloney et al., 2000a). Since then, several forms of glycosylation have been characterized, as

well as their respective transferases, including O-fucosylation (Ofut1/POFUT1), O-

glucosylation (Rumi/POGLUT1), and O-GlcNacylation (EOGT/EOGT1) (reviewed in

Harvey and Haltiwanger, 2018). O-fucose and O-glucose residues can be further elongated by

GlcNac-transferases (Fringe/Fringe family) and xylosyl-transferases (Shams/GXYLT1-2),

respectively (Figure 4). These modifications are added on serine and threonine residues of

EGF repeats. Loss of function studies showed that O-fucosylation and O-glucosylation are

essential for Notch signaling (Acar et al., 2008; Okajima and Irvine, 2002; Shi and Stanley,

2003), while glycan elongation mediated by Fringe/Fringe family or Shams/GXYLT1-2

rather act as modulators of Notch activity (Harvey and Haltiwanger, 2018; Lee et al., 2013).

Finally, biochemical analyses of transferase activities helped in constituting a consensus

sequence associating each type of modification to an EGF repeat (Figure 4). This unraveled

an incredible number of putative sites, ECD of Drosophila Notch being predicted to carry 22

O-fucose, 18 O-glucose and 18-GlcNac sites (Moloney et al., 2000b; Rampal et al., 2007).

However, despite such complexity, recent progress has been made in understanding how these

modifications, alone or in combination with other glycosylated residues, influence receptor-

ligand interactions.

Figure 4. Types of glycan modifications found in Notch EGF repeats. Adapted

from Harvey and Haltiwanger, 2018.

First, structural studies provided functional evidence for glycosylation in receptor-

ligand binding, confirmed former mutational analyses and identified unknown glycan

modifications in the ligand-interacting surface (Luca et al., 2015). Among the key findings,


12

they revealed that O-fucose residue from Notch1 EGF12 directly interacted with the MNNL

domains of Dll4 and Jagged-1, thereby providing a structural basis for O-fucose requirement

in Notch signaling. Complementarily, modeling O-fucose elongation at this site by Fringe

transferases showed how glycan elongation can modulate receptor affinity for its ligand.

Thus, this O-fucose residue not only plays a critical role in the receptor-ligand interaction, but

also in its modulation. In addition, O-fucose residue from EGF8 was shown to directly

interact with Jagged-1 MNNL (of note, the residue from EGF8 affected in the jigsaw

mutation is distinct from the O-fucose site), highlighting the importance of EGF8 in the

Notch1-Jagged1 interaction. Nonetheless, despite the structural rationale they provided for

glycosylation function in receptor-ligand binding, these studies focused only on EGF8-12,

letting unsolved the role of modifications lying on the remaining EGF repeats, as well as the

combinatorial effects of these modifications on receptor-ligand interactions.

Second, taking advantage of recent advances in semi-quantitative mass spectrometry

methods, the Haltiwanger group built a comprehensive map of Notch glycan modifications in

Drosophila S2 and mammalian cells in presence or absence of Fringe transferases (Harvey et

al., 2016; Kakuda and Haltiwanger, 2017) (Figure 5). In S2 cells, they first showed that all

receptors do not present the same glycan modifications, implying that different pools of

receptors coexist in cells based on their glycosylation status (Harvey et al., 2016) (Figure 5A-

C). Similar results were obtained with endogenous Notch receptors extracted from embryos.

When characterizing the nature of the glycan modifications, they observed that while O-

fucose sites were modified to high stoichiometries, Fringe-mediated elongation occurred only

in a subset of these sites. Alike, most O-glucose sites were occupied while xylose elongation

was detected only in a subset of them. This suggests that glycan elongation is sequence and

presumably context specific. They later tested in mammalian cells how glycan modifications

mediated by the Fringe family transferases, Manic Fringe (MFNG), Lunatic Fringe (LFNG)

and Radical Fringe (RFNG), influence Notch1 activation by Dll1 or Jagged-1 (Kakuda and

Haltiwanger, 2017). They characterized three classes of glycan modifications: Dll1 and

Jagged-1 activity enhancing, Dll1 activity enhancing and Jagged-1 activity inhibiting (of note,

RFNG could not modify O-fucose sites leading to Jagged-1 inhibition) (Figure 5D).

Therefore, depending on the context, Fringe activity in mammals might result either in

activation or inhibition of Notch1 activity. Intriguingly, some functional sites lied outside the

EGF8-12 binding domain, raising the question of how these modifications affect the

interacting interface.


13

Figure 5. A comprehensive and functional map of glycan modifications on Notch

EGF repeats. (A-C) Glycosylation status of Notch receptors in Drosophila S2

cells. (D) Role of Fringe activity in modulating receptor-ligand affinity. Adapted

from Harvey and Haltiwanger, 2018.


14

All taken together, these structural and biochemical data suggest that generating

heterogeneity in the receptor glycosylation status is a potent mechanism in fine tuning Notch

activity levels in response to ligand exposure. However, many questions remain unanswered.

For example, how do the O-glucose residues, which do not interact directly with the ligand,

enhance receptor susceptibility to proteolysis (Acar et al., 2008; Luca et al., 2015)? Why only

a subset of sites is subjected to Fringe elongation? Are different glycosylation patterns of

receptors correlating with distinct subcellular localizations? Is glycosylation a reversible

process? It seems that, so far, only the tip of the iceberg has been reached in our

understanding of receptor glycosylation and many exciting discoveries are awaiting

researchers.

Cis interactions

Receptors and ligands of the Notch pathway not only interact in trans, but also in cis (Figure

6). Since this section is dedicated to trans-activation, the role of cis interactions will be here

only briefly described. More details can be found in del Álamo et al., 2011.

Figure 6. Mechanisms of receptor-ligand cis-inhibition. Adapted from del Álamo

et al., 2011.

Genetic studies in Drosophila first hinted that cis-interactions between Notch and its

ligands might have an inhibitory effect. In the wing pouch, Notch receptors could be activated


15

only at the border of domains with high levels of ligands but were inhibited within these

domains (de Celis and Bray, 1997; Micchelli et al., 1997). In addition, ectopic expression of

Delta or Serrate in domains where Notch is normally activated was sufficient to prevent

expression of Notch downstream targets. This phenomenon was subsequently referred to as

“cis-inhibition”.

Despite this initial observation, few biological processes were identified where cis-

inhibition plays a critical role in regulating Notch signaling. Moreover, the mechanism by

which ligands inhibit receptors stays largely unknown. The Elowitz group has provided

experimental and computational evidence that cis-ligands titrate receptors at the cell surface,

hence excluding models based on catalytic processes (Sprinzak et al., 2010). However, since

their experimental strategy was based on a cell culture assay where trans/cis ligands and

receptors were located at the same contact surface, it remains a matter of debate whether their

model can be applied to heterogeneous cell-cell interfaces where receptors could theoretically

interact with cis- or trans-ligands depending on their location in the cell.

Finally, it was recently reported on BioRxiv that Notch receptors could be cis-

activated in isolated cultured cells that express both Notch receptors and its ligands (Elowitz

et al., 2018). This observation was confirmed with several receptor-ligand pairs in different

cell types with either overexpressed or endogenous Notch components. To demonstrate the

biological relevance of cis-activation, authors inhibited Notch cis-activation in isolated neural

stem cells (NSCs) and observed decrease in the survival rate. Given that Notch activity is

required in vivo for NSC maintenance, they concluded that cis-activation might occur in vivo

and regulate physiological processes. Nonetheless, these unexpected observations left many

questions unanswered: how cis-interactions can lead either to inhibition or activation? How

do cis-ligands drive the receptor conformational change? How does the cell integrate trans-

and cis- signals? What are the biological processes subjected to autocrine Notch signaling?

Future work aimed at solving these questions is needed to ascertain that cis-activation is not

an artifact from cell culture and indeed forms another layer of regulation that fine-tunes Notch

activity.

1.1.2. Conformational change of the receptor induced by ligand binding

Structure of the Negatively Regulated Region

The Negatively Regulated Region (NRR) is a highly conserved domain among Notch

homologs. It consists in the succession of three LIN12-Notch repeats (LNRs, defined as A, B,


16

C) and a homo-dimerization domain (HD) carrying the cleavage site S2 that initiates the

proteolytic cascade (Figure 7). In most Notch receptors, the HD is initially cleaved by a furin-

like protease at an S1 site prior to protein addressing towards the cell surface (Logeat et al.,

1998). This leads to the formation of a heterodimeric dimer linked by non-covalent

interactions between the N- and C-termini of the two halves. Although this cleavage does not

seem required for receptor ability to be activated by its ligands (Kidd and Lieber, 2002), it

appears to regulate surface levels of mammalian Notch1 and Drosophila Notch (Gordon et al.,

2009a; Lake et al., 2009). How such regulation is mediated remains elusive.

Figure 7. Structure and force-dependent unfolding of the NRR. Adapted from

Lovendahl et al., 2018.

The NRR plays a pivotal role in receptor activation by preventing access of the S2 site

to proteases in absence of ligands. Pioneering biochemical studies on the Notch receptor first

showed that disruption of the NRR by deleting the LNR repeats resulted in ligand-

independent constitutive activation of these mutant Notch receptors (Kopan et al., 1996). In

addition, point mutations in the NRR sequence found in human T-cell acute lymphoblastic

leukemia or in developing invertebrates lead as well to enhanced Notch activity (Berry et al.,

1997; Weng et al., 2004).


17

In the same time, several studies proposed that ligand endocytosis might exert a strain

on the receptor which changes the conformation of the NRR and allows receptor proteolysis

(detailed in section 1.3.) (Parks et al., 2000). Alternative models were proposed to explain

how such strain might affect the NRR and uncover the S2 site, including (1) the strain relieves

the NRR from interactions with other proteins (oligomerization model) and (2) the strain

changes the structure of the NRR (conformational change model) (Parks et al., 2000).

Finally, resolution of the NRR structure of human Notch2 revealed how the receptor is

maintained in an auto-inhibited state in absence of ligands: by wrapping around the HD in a

mushroom-like shape, the LIN repeats bury the S2 cleavage site into a small hydrophobic

pocket, hence preventing ectopic proteolysis (Gordon et al., 2007). Consistent with the

“conformational change” model, this structure implies that upon ligand binding, a force is

exerted on the receptor to unfold the NRR. At last, authors proposed that unfolding the NRR

structure through LNR displacement would require substantial energy and would be more

consistent with a mechanical opening than an allosteric change of the whole ECD. These

findings were further extended to Notch1 and Notch3 (Gordon et al., 2009b; Xu et al., 2015)

(Figure 7).

Although the mechanotransduction model fitted with the structural data, back in 2007,

it remained to be experimentally tested whether applying forces on the receptor was indeed

sufficient to trigger its activation (see below) and, more importantly, to determine what might

be in vivo the molecular mechanism generating the pulling force (detailed in section 1.3).

Unfolding the NRR: the pulling force at stake

In recent years, much effort has been made to establish the range of pulling forces required to

trigger the conformational change of the NRR, and to relate this range with physiological

conditions of Notch activation. To this purpose, refined strategies based on force spectroscopy

or molecular tension sensors have been designed. A detailed analysis of each of these

methods can be found in Lovendahl et al., 2018 and a schematic overview in Figure 8. Here

will be summarized the main findings of these studies.


18

Figure 8. Overview of experimental strategies probing the force range required for

Notch activation. See text for details. Adapted from Lovendahl et al., 2018.

All taken together, these reports first demonstrated that a pulling force is indeed

required to activate the receptor. When bound to soluble ligands, receptors remained inactive,

while applying forces on the receptor led to its activation (Gordon et al., 2015). They also

determined the ligand-receptor rupture force (~19 pN) (Shergill et al., 2012) and the

activation force range of human Notch1 receptor by Dll1, Dll4 or Jagged1, comprised

between 4 and 12 pN (Chowdhury et al., 2016; Gordon et al., 2015; Luca et al., 2017; Seo et

al., 2016; Wang and Ha, 2013). This range is compatible with the work required to drive


19

protein conformational changes (4.1 pN/nm) (Lovendahl et al., 2018) and importantly, with

forces generated by endocytosis (2-10 pN) (Langridge and Struhl, 2017; Meloty-Kapella et

al., 2012). Second, they provided evidence that force-dependent NRR unfolding is

independent of ligand ECDs or receptor EGF repeats. Directly applying forces on the NRR

was sufficient to drive S2 site exposure to proteases and NICD release (Gordon et al., 2015).

Complementarily, a Notch1 receptor carrying an N-terminal SNAP-tag fusion could be

activated by a benzylguanine-conjugated magnetic plasmonic nanoparticle (MPNs) when a

magnetic field was implemented (Seo et al., 2016). This argued that the NRR conformational

change upon pulling does not require allosteric structural changes induced by receptor-ligand

interactions. Third, it appeared that the activation force range was independent of ligand-

oligomerization or local receptor concentration (Seo et al., 2016).

Although this activating force range fits with forces generated by endocytosis, it

remains to be established in vivo what cellular mechanism indeed generate the force. In

addition, as Dll4 and Jagged-1 require different pulling forces to activate Notch1 (resp. 4 pN

and 12 pN), it would be worth knowing upon what force regime all receptor-ligand pairs

become productive, and how such differences might affect Notch dynamics in relationship

with specific cellular contexts. Eventually, even though Seo et al. proposed that receptor

clustering does not affect the pulling force, one cannot exclude that above given thresholds of

clustered receptors or oligomerized ligands, the activation force might be altered.

Lastly, even though accumulating evidence support the here presented

mechanotransduction model where the pulling force is generated by the signal-sending cell,

understanding how plated or membrane-tethered ligand can activate Notch receptors might

challenge this view and uncover other sources generating the pulling force (Narui and Salaita,

2013; Sprinzak et al., 2010). Last but not least, the discordant fact that receptors can be

activated by soluble ligands in worms might reveal additional mechanisms by which receptors

can undergo conformational changes to unbury the S2 cleavage site (Chen and Greenwald,

2004).

S2 cleavage mediated by ADAM proteases

Following receptor conformational change, the S2 site becomes accessible and is

subsequently processed by A Disintegrin Metalloprotease (ADAM) endopeptidases, a large

family of membrane-bound zinc-dependent metalloproteases that hydrolyze extracellular

domains of multiple class-I transmembrane proteins (Brou et al., 2000; Mumm et al., 2000).

This processing leads to receptor shedding and formation of a transient truncated form of the


20

receptor termed Notch Extracellular Truncation (NEXT), later recognized and processed by

the gamma-secretase complex (Figure 7).

In flies and mammals, the S2 cleavage is mainly achieved by one ADAM protease,

Kuzbanian(Kuz)/ADAM10 (Lieber et al., 2002; van Tetering et al., 2009). Its catalytic

activity is highly regulated in cells to prevent ectopic cleavages of Notch receptors or of its

other targets, including N-Cadherin or the Alzheimer’s precursor protein (APP) (Lammich et

al., 1999; Reiss et al., 2005). First, Kuz/ADAM10 proteases are maintained in a latent state by

a pro-domain that is further removed in the secretory pathway (Groot and Vooijs, 2012).

Second, surface levels are also tightly regulated by the C8 family of tetraspanins, which

mediate their trafficking from the Golgi apparatus towards the cell surface (Dornier et al.,

2012; Haining et al., 2012). Finally, a recent structural study uncovered an additional layer of

regulation (Seegar et al., 2017). By analyzing crystals of ADAM10 ectodomain, Seegar et al.

showed that mature ADAM10s lie at the cell surface in an autoinhibited state where the

disintegrin and the cysteine rich domains mask the catalytic site. Adding a modulatory

antibody that recognizes the cysteine-rich domain was sufficient to unmask the catalytic site

and relieve autoinhibition. This self-regulatory mechanism was proposed to restrict substrate

engagement and avoid ectopic shedding. However, how these allosteric changes are

coordinated with the exposure of the receptor S2 site in vivo remains speculative.

Cellular location of the S2 cleavage site has recently become a matter of debate. While

it was generally accepted that the Kuz/ADAM10 acts at the cell surface, a recent report

challenged this view and proposed a bilateral endocytosis model where heterodimer

dissociation of the receptor is caused by simultaneous ligand and receptor endocytosis,

thereafter followed by ADAM10-mediated shedding in an endocytic compartment

(Chastagner et al., 2017). Nonetheless, how ADAM10 can recognize just one half of the

receptor heterodimer and how endocytosis of ligand and receptor can be coupled is yet to be

determined.

Of note, ADAM17/TACE, another member of the ADAM family that shares high

sequence homology with Kuz/ADAM10, has been also shown to cleave Drosophila Notch

and mouse Notch1 in cell culture assays (Brou et al., 2000; Lieber et al., 2002). Moreover,

dysregulation of its activity in vivo leads in mammals to ectopic Notch activation in a ligand-

independent manner and results in pathological disorders (Bozkulak and Weinmaster, 2009;

Groot and Vooijs, 2012). However, its requirement in physiological Notch activation during

animal development and adulthood is likely redundant with Kuz/ADAM10 activity, as flies


21

and mice deficient for ADAM17/TACE do not exhibit Notch-related phenotypes (Groot and

Vooijs, 2012; Lieber et al., 2002).

1.1.3. Gamma-secretase and S3 cleavage

The core complex

NEXT molecules are then recognized by the gamma-secretase complex and undergo a

regulated intramembrane proteolysis (RIP) at site S3 in the transmembrane domain, thereby

releasing NICD from the plasma membrane into the cytosol (De Strooper et al., 1999;

Schroeter et al., 1998; Struhl and Greenwald, 1999). The gamma-secretase complex is

composed by presenilin, the catalytically active subunit, and three regulatory and essential

components, Nicastrin, Aph-1 and Pen-2.

Complex or complexes?

In mammals, presenilin has two homologues, PSEN1 and PSEN2, and Aph-1 two isoforms,

Aph-1a and Aph1-b, the former existing upon two spliced variants. In theory, six different

gamma-secretase complexes can co-exist in cells and might present different catalytic

properties and substrate affinities. A recent biochemical characterization of these complexes

in cell lines provided support to this hypothesis by showing that each combination cleaves

NEXT molecules with different efficiencies (Yonemura et al., 2016). However, to date, little

is known on the relative abundance of these alternative complexes in vivo or on their potential

role in modulating Notch activity in a context specific manner.

Cellular location of the S3 cleavage site

The location of the S3 cleavage seems also crucial in regulating gamma-secretase activity.

Depending on the cellular context, the S3 cleavage has been shown to occur at the cell surface

or, following NEXT internalization, in acidic endosomes (Siebel and Lendahl, 2017).

Importantly, the cellular environment might determine the position of the S3 cleavage,

leading to the production of NICD molecules with different N-terminal residues and

consequently with distinct half-lives according to the N-end rule (Tagami et al., 2008).

However, relative contribution of short- versus long-lived NICD variants in signaling is not

currently understood.


22

1.1.4. Transcriptional activity of NICD

Nuclear import

After its release from the membrane, NICD translocates into the nucleus. Intriguingly, while

much effort has been made to describe and understand NICD role in transcription, the

mechanisms mediating NICD transport into the nucleus are largely unknown. Several

importins have been involved in NICD nuclear import in flies and mammalian cell lines

(Huenniger et al., 2010; Sachan et al., 2013), yet knowledge in this area remains scarce.

Interestingly, a recent study uncovered a regulatory role for Brat/TRIM3, a tumor suppressor,

in attenuating nuclear NICD transport mediated by importins alpha and beta (Mukherjee et

al., 2016). Moreover, as brat knockdown in fly neural stem cells leads to Notch-dependent

brain tumors and reduced expression of TRIM3 is found in many human glioblastoma, it

suggests that the nuclear transport machinery can act as a sink for NICD molecules to

modulate signaling levels under physiological conditions.

Formation of the NICD-CSL activating complex

Once in the nucleus, NICD cannot bind to DNA by itself and necessitates interacting with a

CSL (CBF-1, Suppressor of Hairless (Su(H)), Lag-1) DNA binding protein to locate at target

loci (Figure 9). Interactions with CSL proteins are mediated by the N-terminal RAM domain

and the Ankyrin repeats. Once formed, the complex recruits a co-activator, Mastermind

(Mam/MamL1-3), to initiate target gene expression (Bray and Gomez-Lamarca, 2018).

Two models were proposed to explain how the CSL molecules switch from an inactive

to an active form in presence of NICD. It was originally considered that, in absence of NICD,

CSL already resides on DNA and is bound to repressors. Following receptor activation, NICD

molecules compete with repressors and displace them to initiate transcription. However, since

NICD and repressors present similar affinities for CSLs, the molecular mechanism for such

replacement remained a mystery (Bray and Gomez-Lamarca, 2018).

A recent report provided support for an alternative model where preformed repressing

and activating complexes would compete for the same binding sites (Gomez-Lamarca et al.,

2018). Analysis of in vivo CSL dynamics at the E(spl)-C locus in Drosophila, one of the best

characterized CSL binding site, revealed that in absence of NICD, CSL molecules were only

transiently bound to DNA, while in presence of NICD, recruitment of CSL molecules at the

E(spl)-C locus was enhanced and their residency time was increased by the recruitment of

additional factors, including Mam and chromatin remodelers. Whether this competition


23

mechanism is conserved among the different loci subjected to NICD-CSL regulation remains

to be determined.

Figure 9. Overview of mechanisms regulating NICD transcriptional activity. See

text for details. Adapted from Bray, 2016.


24

Target selectivity

Another intriguing feature of the Notch pathway is the diversity of the cellular behaviors it

can trigger while using a single NICD/CSL/Mam complex to drive transcription. Notch

signaling can indeed activate different set of genes depending on the cellular context, this in

turn leading to context specific responses including differentiation, proliferation or cell death

(Bray, 2016). Moreover, only a restricted subset of optimal CSL binding sites lying in

euchromatic regions is occupied in various cell types in flies or in mammalian cell lines

(Skalska et al., 2015; Wang et al., 2014), hinting that specific mechanisms might be at stake

to recruit NICD at target genes. Taken together, this raises the following question: how is

regulated the selection of target sites by CSL-NICD complexes to elicit cell specific

transcriptional responses?

Exhaustive analyses are lacking but several reports have characterized context specific

mechanisms influencing CSL-NICD selectivity, which could be split among three categories:

the nature of the CSL binding sites, the combination in cis with other transcriptional factors

into cis-regulatory modules (CRMs) and the recruitment of additional factors that either

enhance or inhibit transcription. Each category will be briefly described here and further in-

depth analyses can be found in Bray and Gomez-Lamarca, 2018 and in Kovall et al., 2017.

CSL complexes bind to enhancers with conserved DNA consensus sequences across

the genome. A subset of them, so-called Su(H) paired sites (SPS), is organized in a head-to-

head orientation and are bound by CSL-NICD dimers (Bailey and Posakony, 1995; Hass et

al., 2015; Severson et al., 2017). As these dimers form by cooperative binding, thereby

increasing stability of the complexes, and given that most of the SPS map in highly

responsive enhancers, one might speculate that SPS promote more efficiently transcription

than monomeric CSL binding sites. Yet, the molecular basis for such differential behavior is

not currently understood. On the other hand, few studies reported that degenerated CSL

binding sites could also provide context specific selectivity (Swanson et al., 2010, 2011). In

the developing Drosophila retina, mutating back low-affinity CSL binding sites upstream the

Pax2 gene, normally expressed in the cone cells, towards the consensus sequence led to

ectopic Pax2 expression in additional cell types experiencing Notch activation (i.e.,

photoreceptors and primary pigment cells). Hence, sequence of the CSL binding sites can

influence target gene expression in many flavors.

In addition to CSL binding sites, numerous enhancers also contain other

transcriptional factor binding sites to integrate several upstream signals in CRMs. For

example, architecture of enhancers composed by a combination of E-boxes bound by


25

proneural factors and CSL sites is key to drive target gene expression in a selective manner

during fly neurogenesis (Castro et al., 2005; Cave et al., 2005, 2011). Genome-wide and case

studies also identified another class of transcription factors, the Runx family, whose binding

sites are often found in CRMs containing CSL binding sites (Hass et al., 2015; Terriente-Felix

et al., 2013). However, whether these transcription factors act by creating a permissive

chromatin state for CSL-NICD complexes to drive transcription or by recruiting in synergy

with CSL-NICD complexes the transcriptional machinery is currently not known.

Finally, selectivity in gene expression can also be achieved by local recruitment of

chromatin modifiers. p300, an histone acetylase, has been early shown to be recruited by

Mam and to mediate transcription of target genes (Oswald et al., 2001). More recently,

identification of nuclear partners of the CSL-Notch1ICD complex in human T-all cells

uncovered a large number of chromatin modifiers, including activators and repressors (Yatim

et al., 2012). Nonetheless, why some CSL-NICD complexes would recruit a given chromatin

modifier at specific loci is shrouded in mystery. Of note, while these modifiers can play a key

role in initiating transcription, they also might be involved in modulating the transcriptional

status of a gene, i.e. ON or OFF, once NICD has been degraded.

NICD stability

The core mechanism of Notch signaling ends with the degradation of the NICD fragment.

NICD molecules are supposed to be short-lived due to a PEST sequence localized at the C-

terminus of the protein. Post-translational modifications of PEST residues can additionally

modulate this half-life in a context specific manner, thereby prolonging or abbreviating gene

transcription (Bray and Gomez-Lamarca, 2018; Carrieri and Dale, 2017). To date, our

understanding on the respective role of these modifications in modulating Notch activity in

vivo is scarce.

1.2. Regulation of Notch receptors by the endosomal pathway

Genetic studies in Drosophila and pulse-chase assays in mammalian cell lines revealed that

Notch receptors are constitutively endocytosed and subsequently trafficked among endocytic

compartments till degradation in lysosomes. Impairments in receptor sorting or degradation


26

lead to ectopic Notch activity, either ligand dependent or independent, and are found in

biological disorders. This highlights the importance for signal-receiving cells to tightly

control the levels of available receptors for activation at the cell surface or at the limiting

membrane of endosomes.

This section follows the endosomal path from endocytosis to degradation and briefly

describes at each step known regulators of receptor trafficking (synthetized in Figure 10). In-

depth details can be found in Schnute et al., 2018.

1.2.1. Notch endocytosis and regulation of surface levels

Starting the journey with E3 ligases

Endocytosis is a general mechanism regulating surface levels of transmembrane proteins. In

particular, ubiquitination of Notch intracellular domains has been proposed to promote

receptor endocytosis in a ligand-independent manner (Le Bras et al., 2011). In Drosophila and

mammalian cell lines, three E3 ligases initiating Notch endocytosis have been identified so

far: Deltex (Dx)/DTX1, Suppressor of Deltex (Su(Dx))/Itch and Nedd4 (Chastagner et al.,

2017; Qiu et al., 2000; Sakata et al., 2004; Sorensen and Conner, 2010; Wilkin et al., 2004;

Yamada et al., 2011). They all have been shown to ubiquitinate NICD in in vitro assays,

although physical binding to NICD has only been established for Dx/DTX1 and Nedd4,

suggesting that Su(dx)/Itch requires an unknown intermediate to interact with NICD.

In vivo studies, on the other hand, point out that these E3 ligases are fine-tuners of

Notch activity rather than essential components of the pathway. First, dx and Su(x) flies

undergo full development and are viable (Fostier et al., 1998; Hori et al., 2004). Moreover, in

those mutant flies, Notch-related phenotypes are mild and mostly observed in adult wings. In

line with these observations, overexpressing mutated forms of Nedd4 in the fly wing produces

only mild phenotypes (Sakata et al., 2004). Second, most of in vivo data relies on the

observation of mislocalized Notch receptors in fly epidermises where either wild-type E3

ligases or their respective dominant-negative forms were overexpressed. Knowing that (1) Dx

and Su(dx) might also regulate the switch between receptor degradation and recycling

downstream in the endosomal path and (2) Nedd4 has many other targets than Notch, it

complicates the establishment of a correlation between the defects in receptor localization and

a specific cellular function of the E3 ligases. Third, an even more puzzling observation is that

the ubiquitin-transferase activity of Dx and Su(dx) does not seem required to promote

receptor endocytosis in cell lines (Shimizu et al., 2014). Instead, Dx and Su(Dx) would act as


27

adaptors recruiting the endocytic machinery in a clathrin dependent or independent way,

respectively (Schnute et al., 2018). Nonetheless, no data is currently available to support this

hypothesis.

Therefore, although regulating receptor surface levels seems in principle crucial to

avoid any excess in receptor activation, there is presently a paucity of knowledge about the

initiation of receptor endocytosis.

Figure 10. Notch trafficking from the cell surface to the lysosomes. The inset

illustrates sequential activities of the ESCRT complex during Notch

internalization in MEs. See text for details. Adapted from Schnute et al., 2018.


28

Notch endocytosis

The molecular mechanism by which Notch receptors are endocytosed from the cell surface

and incorporated in early endosomes (EEs) is poorly understood. Genetic studies in

Drosophila wing epidermis and S2 cell lines suggest that Notch receptors might take different

routes, depending on the presence or absence of a clathrin coat during endosome formation

(Shimizu et al., 2014; Windler and Bilder, 2010). Windler and Bilder showed in addition that

most clathrin-dependent endocytic events were dependent on the adaptor protein AP-2.

Intriguingly, AP-2 was not required for Notch signaling, indicating that AP-2 might

discriminate between receptors intended for signaling and those for internalization. A more

tantalizing model would be that AP-2 recognizes specifically the poly-ubiquitinated receptors,

thereby promoting endocytosis of “old” (presumably damaged and unfit for signaling)

receptors and indirectly preserving newly-synthetized ones at the cell surface. Whether such

quality control exists in cells for Notch receptors remains to be tested. Likewise, the

functional relevance of different endocytic pathways (clathrin-dependent or independent) for

Notch signaling is not established.

Intriguingly, how the endocytic machinery recognizes NICD of receptors intended for

internalization is still a mystery (Schnute et al., 2018). One current limitation of fly studies

investigating Notch trafficking resides in their experimental approach. First, mutating or

downregulating core components of endocytic pathways might generate unspecific effects

that impact Notch localization in an indirect way. Second, time scales of the endocytic events

are often missed. Defects in Notch signaling and/or mislocalization of Notch receptors are

assessed days following gene knockdowns or clone inductions, while trafficking events occur

within seconds. Whether the reported phenotypes are indeed related to the induced genetic

perturbation, or rather reflect pleiotropic effects affecting cells over a longer time period is

challenging to untangle. Generation of a new set of optogenetic tools allowing temporal and

spatial control of protein behaviors might provide a promising framework to study Notch

endocytosis and trafficking.

1.2.2. Recycling of Notch receptors

Numb, a negative regulator of recycling

Once at the limiting membrane of early endosomes, Notch receptors can either by recycled

towards the cell surface or be addressed towards the lysosomes for degradation. Recycling

can theoretically occur following two main routes routes: (1) cargos are directly addressed


29

from EEs towards the cell surface in a Rab4-dependent way or (2) cargos are first trafficked

into recycling endosomes (REs) prior Rab11-dependent surface sorting (Schnute et al., 2018).

Numb, a multi-domain protein distributed at the cell surface and in early endosomes,

has been initially shown in Drosophila to be an antagonist of Notch signaling (Spana and Doe,

1996). Its function is in particular well documented in the context of Notch-dependent

lineages undergoing asymmetric cell divisions (ACDs) (detailed in 2.3.3.). A series of reports

showed that Numb is asymmetrically segregated in the signal-sending daughter cell where it

turns off Notch signaling by inhibiting Notch recycling towards the cell surface and by

promoting its addressing towards the lysosomes (Couturier et al., 2012, 2013, 2014; Doe and

Spana, 1995; Rhyu et al., 1994). However, although several hypotheses based on the

characterization of Numb physical partners were proposed, the exact mechanism by which

Numb inhibits Notch recycling is currently unknown.

These findings on the antagonizing role of Numb in Notch signaling were further

confirmed during mouse neurogenesis and in mammalian cell lines (McGill et al., 2009;

Zhong et al., 1996, 2000), suggesting that its molecular function is well conserved among

phyla.

Recycling and fine-tuning Notch activation

Several recent case studies pointed out the role of recycling regulators in fine-tuning Notch

activity levels. In Drosophila, recycling through Rab4-positive compartments of Notch

receptors mediated by the cargo sorting regulator Rme-8 contributes in reaching final Notch

activity levels to properly specify wing veins (Gomez-Lamarca et al., 2015) (Figure 10). In

mouse, loss of COMMD9, a component of the recycling CCC complex, led also to decreased

Notch2 activity, increased intracellular Notch2 levels and caused phenotypes reminiscent of

Notch2-deficient mice (Li et al., 2015). Intriguingly, some proteins of the Bardet-Biedl

syndrome (BBS) complex, which normally associates with Rab11 endosomes to deliver

proteins to cilia, were also found to regulate Notch surface in mammalian cell lines via

regulated recycling (Leitch et al., 2014).

In brief, these studies not only highlight the role of recycling in modulating Notch

activity, but also the diversity of pathways by which this can be achieved. Of note, in the

aforementioned examples, it is not clear whether defects in Notch activity are due to

decreased Notch surface levels or to the lack of post-translational modifications occurring

during recycling and modifying receptor properties in binding its ligands.


30

1.2.3. Degradation of Notch receptors in lysosomes

ESCRT-mediated intraluminal entry in maturing endosomes

Receptors distributed at surface of EEs undergoing maturation, named maturing endosomes

(MEs), can no longer be recycled and are consequently intended for degradation in lysosomes

(of note, why some receptors will be recycled while other trafficked towards MEs is not

determined). MEs are defined by two characteristics: the acidification of their lumen and the

internalization of intraluminal vesicles (ILVs). This latter process is crucial to prevent any

ectopic Notch activity, as NICDs fragments can no longer be released in the cytosol (Schnute

et al., 2018).

Formation of ILVs containing Notch receptors is dependent on the endosomal sorting

complex required for transport (ESCRT) machinery (Figure 10). Four complexes (ESCRT-0,

I, II and III) compose the machinery and act sequentially to promote ILV internalization.

ESCRT-0 is first recruited to the ILV formation site, and then recruits ESCRT-I and ESCRT-

II to form a super complex hypothesized to generate inward membrane bending. ILV

abscission preceding its release into the lumen is mediated by ESCRT-III. Mutating

components of ESCRT-I, II and III leads to the accumulation of receptors at the ME limiting

membrane and strong ligand-independent Notch activity (Vaccari et al., 2009). Curiously,

impairing ESCRT-0 complex does not affect Notch activity, while receptors do accumulate at

the surface of MEs (Tognon et al., 2014). The molecular basis for this observation is currently

not known and questions whether other factors can recruit ESCRT-I, II and III to Notch

receptors.

Recognition of target cargos for ILV formation by ESCRT0, I and II is favored when

cargo intracellular domains are poly-ubiquitinated. In flies, Dx, Kurtz (Krz), a nonvisual beta

arrestin, and Su(dx) were also proposed to regulate the ubiquitination status of Notch

receptors at the ME limiting membrane (Hori et al., 2011; Shimizu et al., 2014). Whereas Dx

would mono-ubiquitinate Notch, hence preserving receptors for signaling, a Dx-Krz complex

binding to NICD would mediate its poly-ubiquitination, thereby promoting ESCRT complex

recruitment (Hori et al., 2011). In a temperature-sensitive manner, Su(dx) appears also to

promote receptor poly-ubiquitination (Shimizu et al., 2014). Noteworthy, these models are

built from experiments where housekeeping molecular functions are disturbed for several

days prior analysis, hence raising the question whether the observed defects in Notch

localization or signaling are specific or caused by indirect pleiotropic effects.

1.3. Ligand endocytosis and trafficking

31

Ending the journey in the lysosome lumen

Notch receptors are finally delivered into the lumen of lysosomes where they are degraded by

hydrolases in a highly acidic environment (Figure 10). This step consists in the fusion of ILV-

containing MEs to pre-existing lysosomes in a Rab7-dependent manner (Yousefian et al.,

2013).


For the past two decades, ligand endocytosis has emerged as the prevailing molecular

mechanism driving the conformational change in the receptor ECD and subsequent exposure

of the S2 site to ADAM10 proteases. Studies revealed that ubiquitination of the ligand

intracellular domain acts as a trigger initiating ligand endocytosis and is required for receptor

activation in many contexts. Importantly, recent work provided evidence that ligand

endocytosis generates forces consistent with NRR unfolding. Finally, alternative models

suggest that ubiquitin-dependent ligand endocytosis is not only required for receptor

activation, but also for ligand trafficking through recycling endosomes where post-

translational processing should render them active.

This section aims at summarizing current knowledge on ligand endocytosis and

discussing the significance of the aforementioned models in Notch signalling, as well as the

experimental data supporting them.

1.3.1. Ligand endocytosis and generation of the pulling force

Early evidence for ligand endocytosis requirement in Notch signaling

Published in 1997, the pioneering study linking ligand endocytosis and Notch signaling relied

on the analysis of flies lacking dynamin (shibire in Drosophila) activity (Seugnet et al., 1997).

Dynamin acts at the end of the endocytic process by pinching off endosomes from the

membrane. Thereby, impairing dynamin activity is sufficient to block endocytosis in cells.

Authors based their analysis on the bristle pattern of the Drosophila dorsal thorax and

observed that conditional inactivation of shibire, using a thermo-sensitive allele shits, during

bristle precursor selection resulted in bristle tufts instead of a regularly spaced lattice. This


32

phenotype was reminiscent of impaired Notch activity during lateral inhibition (see section

2.1.). Epistasis experiments further revealed that shibire was required upstream of the

proteolytic cascade leading to NICD release. Finally, authors analyzed borders between shits

clones and neighboring wild-type tissues to test whether shibire activity was required in the

signal sending or receiving cell. They noticed that 80% of the bordering bristles were located

in the mutant clone, indicating that loss of shibire activity made the mutant cells less sensitive

to the Notch inhibitory signal sent by the neighboring wild-type cells. Hence, this shows that

shibire is mostly required in the signal-receiving cell in this context. Even more puzzling is

the observation of bristles tufts in shits adult flies. As the inactivation period at restrictive

temperature also covered the sensory organ development, if shibire activity would be essential

for Notch signaling, one would expect to recover bald thoraces reminiscent of impaired Notch

activity in fate specification. Therefore, analysis of shi phenotypes in the Drosophila dorsal

thorax does not draw a strong link between ligand endocytosis and Notch activation, even

though the contrary is often found in the literature. More generally, as dynamin activity is also

required for vesicle budding from the Golgi apparatus, an alternative interpretation would be

that diminished levels in Notch activity are instead due to the accumulation of signaling

molecules (receptor, ligand, proteases involved in the proteolytic cascade) in the trans-Golgi

network. Finally, analysis of dynamin loss-of-function experiments is also problematic from

the theoretical viewpoint. As dynamin acts only once endosomes are internalized but still

membrane-tethered, how would pinching off the endosome from the membrane generate the

pulling force? One would rather expect that force transmission occurs during endosome

invagination, and not during endosome release from the cell surface.

Few years later, S2 and S3 cleavages were molecularly characterized and the

proteolytic sequence model leading to Notch activation was demonstrated, but the trigger

initiating the cascade remained hypothetical. In this context, a fly study from the Muskavitch

lab provided in vivo evidence for the role of ligand endocytosis in triggering the proteolytic

cascade (Parks et al., 2000) and confirmed former observations made with S2 cells (Klueg

and Muskavitch, 1999). In binary systems where ligands and receptors are expressed in

distinct contacting cells, Parks et al. observed first, that NICD and NECD localized in the

signal-sending and signal-receiving cells, respectively; and second, the colocalization of

Notch and Delta ECDs in endocytic compartments of the signal-sending cells. This indicated

that Notch receptors dissociated upon activation into two fragments and that NECD was taken

up together with DlECD in the signal-sending cell, referred by the authors as NECD trans-

endocytosis. In flies expressing a combination of Delta alleles whose gene products


33

accumulate at the cell surface in a temperature sensitive manner, trans-endocytosis was no

longer detected following a heat pulse at restrictive temperature. Finally, authors showed that

ectopic expression of a mutated form of Delta, DeltaC301Y, presumably defective for

endocytosis did not cause ectopic Notch activation in the developing wing, in contrast with

wild-type Delta. To conclude, although highly correlative, these data provided the first

evidence supporting the role of ligand endocytosis in promoting receptor cleavage.

Since then, much effort has been made to determine the molecular machinery

regulating ligand endocytosis (see 1.3.2.) and to establish the force range generated by ligand

endocytosis in vitro and in vivo (see below) (Figure 11). Nonetheless, direct evidence for the

conformational change induced by ligand endocytosis are still lacking. Demonstrating this

process would require the generation of new tools capable of sensing in vivo NRR unfolding.

Given that NRR structures have been resolved, rational design of FRET sensors might

provide a promising experimental strategy.

Figure 11. Ligand endocytosis as a molecular mechanism generating pulling

forces during receptor activation. Adapted from Weinmaster and Fischer, 2011.


34

Ligand endocytosis as a molecular mechanism generating the pulling force

As the mechanotransduction model gathered supporting evidence, it became needed to

demonstrate that ligand endocytosis does generate pulling forces, and importantly within the

range of those found in artificial experimental setups. To date, only two studies, one in vitro

and one in vivo in Drosophila (Langridge and Struhl, 2017; Meloty-Kapella et al., 2012),

demonstrated that ligand endocytosis can generate forces of approximately 2 pN. Their main

findings as well as their limitations will be discussed here.

In a complex system where Notch1 fragments fixed on beads were stabilized by

optical tweezers, cells expressing Dll1 at their surface were presented in close vicinity to the

beads to allow receptor-ligand binding (Meloty-Kapella et al., 2012). Forces generated by

Dll1 endocytosis were inferred from the bead displacement during a 60 second time window.

Although the average force was of 2.8 pN, hence comparable with the 2-10 pN range found in

artificial pulling systems, 30% of the cells exerted pulling forces below 2 pN and 20% failed

to exert any, while the 50% remaining cells scaled between 3 and 10 pN. Whether this

variability in force generation reflects the technical limitations of the tweezer-based system,

with half of the iterations being unsuccessful, or whether this reveals different flavors of Dll1

endocytosis that would depend on the number of formed receptor-ligand pairs, as this

parameter was not controlled in this study, remains to be determined. Later on, authors used

this system to identify the molecular components required for Dll1 endocytosis in the signal-

sending cells (discussed in 1.3.2.). Nonetheless, the relevance of these results regarding

physiological Notch activation had to be yet demonstrated.

Recently, an elegant study from the Struhl lab provided in vivo evidence that Epsin-

dependent endocytosis of Delta generates pulling forces of ~2pN in the Drosophila wing

imaginal disc (Langridge and Struhl, 2017). Authors designed a genetic framework allowing

ectopic expression of wild-type or chimeric receptor-ligand pairs in complementary clones

throughout the epithelium. Ability of each pair to mediate productive Notch signaling was

determined by the expression of Cut, a characterized Notch downstream target in this tissue,

at the clone border. The main part of their work consisted in discriminating between several

models for Epsin role in Notch signaling, and these findings will be discussed below. Lastly,

they used this assay to establish the pulling capacity of Delta endocytosis by replacing Notch

NRR by different versions of the von Willbrand factor (vWF) A2 domain in the chimeric pair

where Delta ECD (resp. Notch ECD) were substituted by the alpha and beta subunits of the

follicle stimulating hormone (FSH) (FSH receptor, resp.). Of note, the chimeric FSH-

Delta/FSHR-Notch pair was shown earlier in this study to induce Cut expression at the clone


35

border. The A2 domain is a force sensor requiring force application of ~8pN to uncover a

cleavage site subjected to ADAM-dependent proteolysis. Several A2 variants have been

characterized with either higher (unknown force requirement) or lower (~2pN) force

thresholds, thereby providing a range of force sensors. Cut expression at the clone boarder

was only observed with A2 variants displaying higher sensitivity to proteolysis, thus

indicating that Epsin-mediated endocytosis generates pulling forces of ~2pN.

Nonetheless, even though the title of the study is stating the opposite, authors did not

provide direct in vivo evidence for force dependent activation of Notch. First, substitution of

the NRR by the A2 variants in the wild-type Delta/Notch pair is not reported, interrogating

whether the observations made with the FSH-Delta/FSHR-Notch chimeric system are specific

to this pair. One could hypothesize that force transmission along receptor and ligand ECDs

depends on the molecular nature of the ECDs, as illustrated by the differential force

requirement of Dll4 and Jagged1 in activating Notch1. Hence, comparison with wild-type

Delta/Notch is needed to confirm the 2pN force range. Second, all the results presented in this

study rely on overexpressed ligand/receptor pairs. How this ectopic situation relates to

physiological Notch activation remains to be determined. One way to overcome this

limitation would be to generate chimeric Notch receptors at the endogenous locus using the

CRISPR technology and test their capacity in mediating Notch signaling in a similar clonal

strategy. Additionally, such approach would allow estimating the different force requirements

for Notch activation depending on the biological contexts.

Evidence for alternative activation mechanisms

The growing body of evidence supporting ligand endocytosis as the main molecular

mechanism in the mechanotransduction model has occulted alternative sources for the pulling

force suggested by recent studies.

Besides the specific case of Notch activation by soluble ligands in worms (Chen and

Greenwald, 2004), several studies on mammalian or fly Notch receptors provided convincing

evidence that a pulling force needs to be exerted on the NRR to initiate the proteolytic

cascade. Then, how to explain that plated or membrane-bound ligands ECDs can efficiently

activate receptors harbored at the surface of contacting cells (Narui and Salaita, 2013;

Sprinzak et al., 2010)? One possibility would be that receptor endocytosis can provide the

source of pulling forces, as suggested by the bi-lateral endocytosis model (Chastagner et al.,

2017). Alternatively, a tantalizing idea would be that shear stress forces at the cell surface

might drive NRR unfolding once receptors are bound to fixed ligands. Despite their


36

differences regarding the molecular mechanism, these two hypotheses imply that signal-

receiving cells can also generate pulling forces. So far, this possibility has not been directly

tested in vivo.

Shear stress forces have also been correlated with Notch activation in the context of

vascular homeostasis. Several studies concomitantly showed with their respective

microfluidic-based strategies that Notch1 receptors expressed at the cell surface of endothelial

cells are activated upon laminar flow induction (Fang et al., 2017; Mack et al., 2017;

Polacheck et al., 2017). Activation scaled with increased flow speed (Mack et al., 2017) and

was dependent on Dll4 and gamma-secretase cleavage (Polacheck et al., 2017). Polachek et

al. proposed the following mechanism where shear stress generated by the laminar flow was

inducing ligand endocytosis in the signal-sending cell, but as the only functional test was

based on dynamin inhibition, it remains debated whether shear stress can directly transmit

forces to the receptor-ligand pair.

Also confusing is the report in flies and fish of Notch receptors being activated in a

ligand-dependent manner at the surface of endosomes (Coumailleau et al., 2009; Kressmann

et al., 2015; Montagne and Gonzalez-Gaitan, 2014). In several studies, the Gonzalez-Gaitan

lab provided genetic and molecular evidence in the context of Notch-dependent asymmetric

cell division (ACD) that a subtype of Rab5 endosomes, marked by the endosomal protein

Sara (Smad Anchor for Receptor Activation), was asymmetrically segregated in the signal-

receiving daughter cell. These endosomes were carrying Notch receptors and ligands at their

surface, and dissociation of NECD and NICD fragments at their limiting membrane was

shown to be dependent on gamma-secretase and on the activity of an E3 ligase promoting

ligand endocytosis (Coumailleau et al., 2009). This led to the conclusion that directionally

trafficking Sara endosomes were part of the molecular machinery that biases Notch signaling

during ACD. It is still unclear whether S2 cleavage does occur at the surface of these

endosomes, and why receptors are only cleaved in the signal-receiving cell, as Sara

endosomes carrying receptors and ligands are detected in the mother cell prior to division and

in the signal-sending cell after ACD; nevertheless, Sara endosomes might support an

alternative mechanism for receptor activation independent of ligand endocytosis. Cis-

activation could provide a mechanistic rationale in this context, but additional work would be

needed to explain how force would be transmitted in a cis configuration and why this

activation mode would take place only in specific contexts.

Finally, contradictory results came from the use of synthetic receptor/ligand pairs

regarding the requirement of ligand endocytosis in Notch signaling. On one hand, Gordon et


37

al. and Langridge and Struhl showed that chimeric ligands lacking their intracellular tail were

unable to activate their associated receptors in cell culture assays and Drosophila imaginal

discs, respectively (Gordon et al., 2015; Langridge and Struhl, 2017). Additionally, Langridge

and Struhl reported that mutating lysines to arginines in ligand native or chimeric intracellular

tails to prevent ubiquitin-dependent ligand endocytosis inhibited receptor activation. On the

other hand, Morsut et al., in their approach to design synthetic signaling circuits using Notch

transmembrane domains flanked by ectopic ECDs and ICDs as signal transducers, observed

that surface bound ligand did not require endocytosis nor endocytic motifs in their chimeric

intracellular tails to trigger activation of their cognate receptors (Morsut et al., 2016).

Hence, ligand endocytosis might be just one among many other molecular

mechanisms generating the pulling force. Future work will be needed to understand why

ligand endocytosis predominates in most Notch dependent processes, and what advantages it

provides compared to other mechanisms. Finally, reinvestigating ligand endocytosis

requirement by other means than dynamin inhibition would be valuable to ascertain this

predominance compared to alternative mechanisms.

1.3.2. Ligand endocytosis and ubiquitination

Neuralized and Mindbomb, the E3 ligases promoting ligand endocytosis

Mono-ubiquitination of transmembrane protein ICDs is a common cellular mechanism to

initiate endocytosis. Ubiquitination requires sequential activities of three ligases, termed E1,

E2 and E3. While E1 and E2 ligases are used for a broad spectrum of transmembrane

proteins, E3 ligases are often specific to a subset of substrates.

In the context of ligand endocytosis in Notch signaling, the first E3 ligase being

characterized was Neuralized (Neur) (Figure 12A). Early evidence from the 80’s indicated

that fly embryos lacking Neur displayed neurogenic phenotypes similar to embryos deficient

for Notch (Lehmann et al., 1983; Weinmaster and Fischer, 2011). Sequence analysis first

suggested that Neur functioned as a transcription factor through its RING (really interesting

gene) domain, but further biochemical analyses of other proteins containing RING domains

ruled out this hypothesis by demonstrating that these domains encoded an E3 ligase activity.

In 2001, four studies provided genetic and biochemical evidence in Drosophila and Xenopus

that Neur/X-Neur also functions as an E3 ligase, is required in specific developmental

processes regulated by Notch signaling and is mediating Delta/X-Delta ubiquitination and

subsequent internalization through its RING domain (Deblandre et al., 2001; Lai et al., 2001;


38

Pavlopoulos et al., 2001; Yeh et al., 2001). Further investigations revealed that Neur was also

targeting the second Drosophila ligand, Serrate (Glittenberg et al., 2006).

Nonetheless, it remained unclear at that time whether Neur acted autonomously by

clearing Delta from cell surfaces to relieve Notch receptors from cis-inhibition (Deblandre et

al., 2001; Lai et al., 2001), or non-autonomously in signal-sending cells by promoting ligand

endocytosis in line with the trans-endocytosis model (Pavlopoulos et al., 2001). Subsequent

work in flies provided support to the second hypothesis with Neur acting in the signal-sending

cell (Le Borgne and Schweisguth, 2003; Weinmaster and Fischer, 2011). Concomitantly, two

studies inferred from their experimental data that the balance between ubiquitinated and non-

ubiquitinated ligands in cells expressing Neur might be key to generate a dual state where

cells can still signal to their neighbors while being cis-inhibited (Glittenberg et al., 2006;

Miller et al., 2009). This fine-tuning of Neur activity could be a mechanism particularly

relevant during lateral inhibition to prevent neighboring cells to signal back to the emerging

proneural cell and to lock its fate.

Finally, one peculiarity of Neur is its differential requirement in model organisms,

despite its conservation from worms to humans. While essential in flies, mice lacking the two

Neur homologues, Neurl1 and Neurl2, do not exhibit Notch-related phenotypes during

development, even though biochemical evidence support interaction between Neurl1 and

Neurl2 and Notch ligands (Weinmaster and Fischer, 2011). Whether other E3 ligases act

redundantly, such as Mib1 (see below), to compensate for losses of Neurl1 and Neurl2 is not

known. On the other hand, X-Neur exhibits an expression pattern restricted to neural fields in

early frog embryos, likewise Neur in flies, and can promote Drosophila Delta internalization

when ectopically expressed in fly wing discs, indicating that Neur activity is conserved

between flies and frogs (Deblandre et al., 2001). Unfortunately, the impact of Neur depletion

on Xenopus early development has not been reported. Finally, the role of Neur homologues in

regulating Notch signaling during worm, chicken or zebrafish developments has not been

documented. One possible explanation for this apparent discrepancy might be found in the

process of sub-functionalization. Besides its implication in Notch signaling, Neur has also

been shown to regulate epithelial cell polarity and germ stem cell maintenance in Drosophila

(Liu and Boulianne, 2017; Perez-Mockus et al., 2017a). Therefore, Neur homologues might

have specialized in one Notch-unrelated function in some species along evolution while other

E3 ligases would take over in promoting ligand endocytosis.


39

Figure 12. Neur and Mib1 functional domains. Ring (RF, RNG) domains mediate

ubiquitin addition on ligand intracellular tails. Adapted from Perez-Mockus et al.,

2017b and D’Amato et al., 2016.

The second E3 ligase, Mindbomb (Mib) (Figure 12B), was first discovered in a

genetic mutagenesis screen conducted in zebrafish. mib embryos displayed neurogenic

phenotypes, but the relationship with Notch signaling was not established at that time (Schier

et al., 1996). In 2003, Itoh et al. analyzed mib locus by positional cloning and revealed that

the gene product contains three RING domains, acts as an E3 ligase and promotes Delta

endocytosis through ICD ubiquitination (Itoh et al., 2003). Following that pioneering

molecular and functional characterization, two homologues of Mib, Mib1 and Mib2, were

found in flies, mice and humans and shown to ubiquitinate ligands of the Serrate family (Koo

et al., 2005a, 2005b; Lai et al., 2005a; Le Borgne et al., 2005; Pitsouli and Delidakis, 2005).

In both vertebrates and flies, mutational analyses pointed out that Mib1 is the main E3

ligase involved in Notch signaling, as mib2 animals either did not display pleiotropic Notch-


40

related phenotype and presented defects in a restricted number of cellular processes (Koo et

al., 2007; Mikami et al., 2015; Nguyen et al., 2007). Specifically, in mice, Neur homologues

and Mib2 cannot compensate for loss of Mib1 in processes regulated by Notch signaling

during development, hence suggesting that Mib1 is the main E3 ligase required for ligand

endocytosis in this model organism (Koo et al., 2007). In contrast, mib1 fly embryos did not

exhibit neurogenic phenotypes and animals developed till late pupal stages (Lai et al., 2005a;

Le Borgne et al., 2005). Developmental defects were observed in Notch-dependent

developmental events that were independent of Neur, including wing margin formation, leg

segmentation and eye specification. Drosophila Mib1 and Neur, unlike in mammals, have

been proposed to be interchangeable in certain contexts, i.e. during wing margin formation

(Lai et al., 2005a; Le Borgne et al., 2005; Pitsouli and Delidakis, 2005; Wang and Struhl,

2005) or fate specification in the sensory organ lineage (Lai et al., 2005a). Nonetheless, while

the functional conservation is demonstrated, several observations and technical limitations

balance this statement. First, all these experiments were based on the overexpression of Mib1

transgenes in neur mutant backgrounds using the UAS/Gal4 system. Whether such approach

reproduces physiological Notch activity levels is questionable. Second, despite the

overexpression drawback, clustered sensory organs were still observed in rescue assays

conducted in the notum, reflecting defective signaling transmission between emerging SOPs

during lateral inhibition (Lai et al., 2005a). Hence, Mib1 and Neur do not seem

interchangeable in this process. Third, Mib1 expression driven by neurPGal4 (Gal4 insertion

at the neur locus generating a null allele), a driver presumably more representative of

endogenous Neur expression levels, could not rescue the cuticular neurogenic phenotype of

neur embryos. This was not due to insufficient expression levels, as UAS-Neur constructs

could compensate the loss of Neur with the same driver (Le Borgne et al., 2005). This

observation provides a second example related to neurogenesis where Mib1 does not

recapitulate Neur function. Fourth, fate restoration has not been quantified in the sensory

organ lineage, letting uncertain the penetrance of Mib1 rescue in this context (Lai et al.,

2005a). All taken together, this series of remarks question whether Mib1 can indeed

compensate Neur activity when expressed at Neur endogenous levels and to what extent these

two E3 ligases are interchangeable in flies.

More generally, the requirement of Neur or Mib1 in specific contexts during fly

development might rather underlie that Neur and Mib1 mediate ligand endocytosis in a

different manner with different capacities in activating Notch receptors. In support of this

hypothesis, Neur and Mib families do not share any structural similarities (Weinmaster and


41

Fischer, 2011) nor recognize the same epitopes on ligand ICDs (Daskalaki et al., 2011). This

might result in the recruitment of distinct endocytic adaptors depending on the ICD

ubiquitination pattern, which would in turn modulate either the ligand endocytosis rate or the

generated pulling force. Alternatively, a recent structural study uncovered a bipartite

mechanism by which Mib1 binds to Jagged1 ICD (Guo et al., 2016), raising the possibility

that Neur and Mib1 interact with Delta or Serrate ligand families with different affinities.

Similarly, this would translate into differential ubiquitination kinetics and distinct endocytosis

rates. Lastly, specificity might also arise from interacting partners, as Mib1 has been shown to

directly bind with components of the endocytic machinery (Okano et al., 2016).

Finally, whether E3 ligases constitutively ubiquitinate ligand ICDs or in response to a

stimulus is largely unknown. To address this question, Hansson et al. compared the cellular

distributions of Jagged1 and a mutated form of Jagged1 that cannot bind to receptors, named

Jagged1(Ndr), in presence of absence of neighboring cells expressing Notch1 (Hansson et al.,

2010). Intriguingly, when signal-sending cells were isolated, both Jagged1(WT) and

Jagged1(Ndr) were detected at the cell surface while when the same cells were put in contact

with Notch1-expressing cells, only Jagged1(WT) was efficiently endocytosed in a Mib1-

dependent manner and signal was transduced to the signal-receiving cell. This led the authors

to conclude that ligand-receptor interaction induced ligand endocytosis. How binding to the

receptor stimulates Mib1 activity on the ligand intracellular tail and whether such signal

transduction occurs in vivo remains to be established.

From a wider perspective, such transduction mechanism leading to ligand

ubiquitination might be a potent mechanism to segregate bound from unbound ligands. This

would explain why among the total pool of endocytosed Delta molecules in the Drosophila

wing disc, Epsin-mediated endocytosis leading to Notch activation (see section below) affects

only a small fraction of them in the fly imaginal wing disc (Wang and Struhl, 2004). Such

model would also imply that several types of ligand endocytosis coexist in cells with different

abilities to generate pulling forces.

The endocytic machinery

Following ubiquitination of cargo ICDs, additional adaptor proteins bind to these ubiquitin

motives and thereafter promote cargo endocytosis in a clathrin dependent or independent way.

To date, the unique adaptor protein shown to be required in signal-sending cells to mediate

Notch signaling is Epsin (Chen et al., 2009; Overstreet et al., 2003, 2004; Tian et al., 2004;

Wang and Struhl, 2004). In flies, loss of Lqf (Liquid Facets, Drosophila Epsin) impaired


42

photoreceptor selection during eye patterning, lateral inhibition in the notum, wing margin

formation, wing vein specification and cardioblast selection during embryogenesis (Overstreet

et al., 2004; Tian et al., 2004; Wang and Struhl, 2004). Of note, in lqf clones, bristle tufts

could be observed, indicating that fate specification in the sensory organ lineage was not

disturbed (Wang and Struhl, 2004). This implies that other adaptors might replace Epsin in

specific contexts. Similar requirements for Epsin in signal-sending cells have also been

reported in worms and mice (Chen et al., 2009; Tian et al., 2004). Finally, the two studies

described in “Ligand endocytosis as a mechanism generating the pulling force” showed that

the force generated by ligand endocytosis was dependent on Epsin activity (Langridge and

Struhl, 2017; Meloty-Kapella et al., 2012).

Generally, there is a paucity of knowledge regarding Epsin downstream effectors in

the context of Notch signaling. Whether ligand-receptor ECD complexes are internalized

within clathrin coated or uncoated vesicles is not known. Actin polymerization is required,

but nucleating enzymes are not characterized (Meloty-Kapella et al., 2012). The role of BAR

proteins in membrane bending during endosome invagination has not been tested yet. One

explanation to this situation is that impairing proteins with household activities might

generate pleiotropic phenotypes, thereby complicating the dissociation of direct from indirect

effects.

Is ubiquitination always required?

A recent study in Drosophila challenged the here presented paradigm where Epsin-mediated

endocytosis of ligands depends on the ubiquitination of their ICDs by Neur or Mib1 (Berndt

et al., 2017). Authors provided genetic evidence that Delta can activate Notch in two other

modes independent of ubiquitination but either dependent on or independent of Neur. The

first mode was inferred from the expression in a ligand-dependent manner of a Notch activity

reporter in the notal part of the wing disc in absence of Mib1 and Neur. The second derived

from the analysis of Dl Ser clones encompassing sensory organ precursors and expressing Dl

constructs with mutated lysines or lacking Neur interacting domains. Lateral inhibition

defects observed in absence of ligands could only be rescued by the former construct. In

addition, the ubiquitin binding domain of Epsin was also shown to be dispensable in this

process. Therefore, Neur appeared to act in this context as an endocytic adaptor independently

of its E3 ligase activity.

Invoking evolutionary considerations, authors proposed that the first alternative mode

(ubiquitination and Neur independent) corresponds to the ancestral mechanism of Notch


43

activation (Gazave et al., 2009). Nonetheless, as Mib1 is present in the most ancient analyzed

phylum, this hypothesis remains debated. The second alternative mode (ubiquitination

independent and Neur dependent) might provide the rationale to the partial rescue of Neur by

Mib1 in Drosophila, as Neur would recruit specific endocytic effectors. Further investigations

are needed to estimate the relevance of these activation modes in other developmental

contexts or species and to decipher the molecular mechanisms underlying them.

1.3.3. Ligand endocytosis and trafficking

The recycling model

An alternative model, based on ligand recycling, was proposed for Epsin- and Neur/Mib1-

dependent (referred as Epsin-dependent hereafter) ligand endocytosis (Wang and Struhl,

2004) (Figure 13). Despite the fact that this model is nowadays ruled out by many studies and

by internal inconstancies, it is still presented in some articles as a preeminent mechanism for

Notch activation (Daeden and Gonzalez-Gaitan, 2018). Therefore, this model will be briefly

described here (a more in-depth analysis can be found in Weinmaster and Fischer, 2011) and

selected contradictory evidence presented in the next subsection.

Figure 13. Delta recycling model. See text for details. Adapted from Wang and

Struhl, 2004.


44

The main assumption of this model states that newly-synthetized ligands must enter

into a recycling loop prior to being able to activate receptors (Wang and Struhl, 2004).

Concretely, newly-addressed Delta molecules at the cell surface will first be endocytosed in

an Epsin-dependent manner and targeted towards a Rab11 recycling endosome (Emery et al.,

2005; Jafar-Nejad et al., 2005; Rajan et al., 2009; Wang and Struhl, 2004). Within this

compartment, Delta will be processed to become “active” and subsequently readdressed to the

cell surface to mediate Notch signaling. A milder version of this model eludes the processing

step, but instead assumes that recycling is needed to target Delta to the dedicated activation

site of Notch receptors at the cell-cell contact (Benhra et al., 2010).

Contradictory evidence

The central inconstancy of this model relies on the nature of the “active” form of the ligand.

So far, besides the original paper (Wang and Struhl, 2004), no study has provided any

experimental evidence for such Delta processing, including the subsequent reports supporting

this model (Emery et al., 2005; Jafar-Nejad et al., 2005; Rajan et al., 2009). Other criticisms

will be formulated in a dedicated subsection in the section 2.3, as these reports used the SOP

asymmetric cell division as a model.

The strongest set of in vivo evidence against the recycling-processing model came

from the Langridge and Struhl paper (Langridge and Struhl, 2017). In their clonal approach

presented above, they first showed that either wild-type or synthetic receptor/ligand pairs

depended on Epsin to mediate signaling. As it is unlikely that all the chimeric ligands would

require any additional processing within a recycling compartment to bind their counterparts, it

suggests that Epsin-mediated is only required at the activation step. In line with this

conclusion, authors reported that wild-type ligands and receptors can still be bound in absence

of Epsin, henceforth negating the requirement for additional ligand processing.

Corroborating data were also published by Shergill et al. (Shergill et al., 2012). In their

tweezer-based setup used to measure Dll1-Notch1 interaction strength, they did not observe

any difference between control conditions and Dll1 expressing cells with impaired Rab11 or

Epsin activities. Instead, recycling assays revealed that Rab11, but not Epsin, regulated Dll1

surface levels. Therefore, this suggests that impaired recycling might impact Notch signaling

by decreasing ligand surface levels rather than by modulation ligand-receptor binding affinity

through an uncharacterized process.

1.4. Notch signaling dynamics

45


Signaling pathways determine cell identities in both space and time. To date, the Notch

pathway has mostly been investigated regarding its spatial requirement during animal

development or adulthood, its core mechanism has been dissected in great detail but very little

is known on how the transcriptional output is deployed in time in response to Notch

activation. This problem can be subdivided in two parts and approached accordingly: either

by studying the dynamics of the transcriptional response or by analyzing the integration of the

Notch signalling activity into the gene expression profile.

Recent studies provided valuable insights into these two sides of Notch signaling

dynamics. Their findings will be briefly outlined in this section and several perspectives will

be discussed. A subsection will also be dedicated to the available or potential experimental

methods to study Notch signaling dynamics. Cross-talk with other pathways will not be

discussed here and has been recently reviewed in Ho et al., 2018.

1.4.1. Experimental strategies to study Notch signaling dynamics

Reporting expression of downstream targets

Gene specific reporters

Gene specific reporters are either fusion proteins or gene promoters driving fluorescent

protein expression. In most model animals, these reporters were based on genes from the HES

family and used to understand where Notch signaling is active as well as to map context-

specific gene responses (Imayoshi et al., 2013; Zacharioudaki and Bray, 2014). For example,

in Drosophila, Hes genes from the E(spl) (Enhancer of Split) complex are expressed in a

tissue specific manner (Zacharioudaki and Bray, 2014). Similarly, in mice, the requirement of

their homologues depends on context (Fischer and Gessler, 2007). Moreover, when combined

with live-imaging and mathematical modeling, these gene specific reporters proved to be

extremely powerful in understanding in vivo Notch signaling dynamics at the tissue scale

(Aulehla et al., 2008; Corson et al., 2017).

However, one obvious limitation of this method is the restricted number of genes that

genes that can be studied. Indeed, if one aims at understanding how cell identities are

acquired in time in response to Notch signaling, gene specific reporters will offer only a


46

partial view and will leave the following questions unanswered: are cells acquiring fates

shortly following Notch signaling, or does it require time-dependent downstream relays?

What are the feedback loops that either lock fates or maintain unstable states? Are they

coherent or incoherent? Are they context specific? One can always increase the number of

gene-specific reporters, assume that the studied set of genes recapitulates the whole process

and eventually infer regulatory transcriptional loops based on reporter expression patterns, but

the risk always remains that another key gene is missing.

Genome-wide methods

Dynamic genome-wide approaches might provide a powerful strategy to understand the

temporal expression pattern of Notch targets and their regulatory interactions. Recent

improvements in RNAseq protocols facilitate their application not only in cell culture assays

but also in tissues (Kolodziejczyk et al., 2015). To probe the dynamics of Notch target gene

response, if the onset of Notch activation is known or can be induced, one could perform

RNA extractions at fixed time points to assess the temporal pattern of gene expression. This

approach was recently applied to Drosophila muscle progenitors and their findings will be

discussed hereafter (Housden et al., 2013) (Figure 14). Conversely, if cells are not

coordinated, one can separate them in groups based on the expression of a Notch fluorescent

reporter through FACS sorting prior extraction.

Figure 14. Temporal profiling of Notch target gene expression (A) Principle. (B-

C) NICD decay following initial stimulus. Adapted from Housden et al., 2013.


47

Although appealing in theory, application of time-scaled RNAseq on tissues might

reveal to be difficult when cell numbers are low and/or tissue dissections are time consuming.

In addition, RNAseq protocols imply cell isolation, therefore tissue dissociation by proteases.

Whether this process affects cell transcriptomes and generates artifacts is currently poorly

understood. An alternative and not less powerful technique is the targeted DamID (TaDa)

genome-profiling technique (Marshall et al., 2016; Southall et al., 2013) (Figure 15). DamID

relies on the fusion of the bacterial enzyme Dam (DNA adenine methyl-transferase) to a DNA

binding protein of interest. For example, a Dam-Pol II fusion would be useful to imprint all

transcribed genes in the genome. Targeted expression in a specific cell type is achieved with a

conditional expression system, such as UAS/Gal4 in Drosophila. Following DNA extraction,

methylated fragments are enriched and subsequent sequencing will therefore identify DNA

sequences from the cell types expressing the Dam-fusion. Importantly, contrary to RNAseq

protocols, this technique does not require cell isolation, fixation or nucleus purification

(Marshall et al., 2016).

Figure 15. Principle and application of the TaDa technique. Adapted from

Marshall et al., 2016.


48

In the context of Notch signaling, DamID has been used further engineered to

characterize SPS binding sites at a genomic scale (Hass et al., 2015). The enzyme was split

into two inactive halves fused to a pair of proteins of interest (Notch1ICD with RBP,

MAML1 or p300). A similar approach can be used to assess where Notch transactivating

complexes localize in the genome at several time points following the onset of signaling, but

this will not provide information on gene expression. A more exciting strategy would be to

combine the rationale of Notch activation mechanism with split DamID, where Pol II would

be fused to one half of Dam while a synthetic Notch receptor whose ICD is replaced by a

complementary portion of the Dam protein to control Dam activity in a Notch-dependent

manner. Similarly, the whole Dam-Pol II fusion could replace NICD, but this would not allow

control of its expression levels. Since Dam-Pol II is toxic at high expression levels, such

constructs might be unfit in high Notch activity contexts. Furthermore, cell specificity will be

lost if the Notch-Dam-Pol II synthetic receptor is expressed at Notch endogenous levels. An

alternative approach to the split system would consist in the introduction of an orthogonal

activating system. On the TaDa side, the enzymatic activity of a Dam-Pol II fusion would be

inhibited by an additional peptide hindering Dam catalytic site. On the Notch side, a chimeric

Notch receptor would be generated in what NICD is replaced by the enzyme recognizing the

linker between the inhibiting peptide and Dam. Then, following signaling onset, the enzyme

would be released in the cytosol, activating the inhibited Dam-Pol II fusion. This would

prevent titration effects that might occur in the split system due to unwanted reconstitution of

the Dam-Pol II fusion at the ICD of unproductive Notch synthetic receptors.

To conclude, by using any of these methods, temporal evolution of the methylation

profile would inform on the gene expression pattern in response to Notch activation. By doing

so, TaDa would not only be cell type specific, but also Notch-dependent. In fact, this strategy

could be used in a broader range of cellular contexts using synthetic ligand/receptor pair to

drive Dam activity in a contact-dependent manner.

Assessing Notch activity dynamics in vivo

Gal4 chimeric receptors

Notch receptors where the intracellular domain is replaced by a Gal4 transcription factor have

been intensively used in cell culture assays to describe the kinetics of Notch activation and to

understand their influence on target gene expression (LeBon et al., 2014; Nandagopal et al.,

2018; Sprinzak et al., 2010) (Figure 16). NotchGal4-expressing cells are mixed with cells


49

expressing ligands under the control of an inducible promoter to control the timing of

activation. Alternatively, signal-receiving cells can be plated on fixed ligands and signaling

starts when the gamma-secretase inhibitor DAPT is removed from the medium. Once

receptors are activated, the cleaved Gal4 moiety translocates to the nucleus and drives

expression of a UAS-driven fluorescent reporter. Integration over is used as readout of the

NICD production dynamics. Lastly, by being orthogonal to Notch endogenous targets, this

method circumvents promoter-specific regulations and downstream feedback. Of note, this

strategy does not take into account the translocation time of endogenous NICD, which can be

modulated in certain contexts (Mukherjee et al., 2016).

Figure 16. Notch-Gal4 chimeric receptors as a tool to follow Notch signaling

dynamics. Adapted from Nandagopal et al., 2018.

The impact of different NICD production dynamics on target gene expression can also

be directly tested using this system. Artificial activities are generated by applying controlled

pulses of DAPT in the medium of cells that express truncated Gal4 receptors lacking their

extracellular domains (NotchΔECD-Gal4) (Nandagopal et al., 2018). By doing so, Gal4

moieties can be released from the membrane in a time-controlled manner (see below for

experimental applications).

Interestingly, while chimeric Notch-Gal4 constructs have been generated for more

than two decades in Drosophila (Lecourtois and Schweisguth, 1998; Struhl and Adachi,

1998), they only have been used for qualitative assessment of the effectiveness of S2 or S3

cleavages. As Notch expression levels are low in tissues and the detection in vivo of the

endogenous protein is difficult (see below), reintroduction of these tools in experimental


50

strategies might be valuable in quantifying NICD production dynamics in a living organism

over longer time periods and/or with a higher temporal resolution.

Synthetic reporters

Another versatile approach by which to analyze the kinetics of Notch endogenous signaling in

vivo relies on Notch-responsive synthetic fluorescent reporters. Upstream regions of these

reporters consist in multiple Su(H) binding sites combined with additional enhancers and a

ubiquitous promoter, so that expression of a fluorescent marker is induced upon NICD entry

in the nucleus (Zacharioudaki and Bray, 2014). This strategy has been efficiently applied in

flies and mice and, by contrast to Notch-Gal4 receptors, integrates NICD degradation rate, as

the reporter activity depends on endogenous NICD. Therefore, this method provides a good

readout of NICD activity in the nucleus, even if it cannot be used to discriminate between a

reduction in signaling and an increase of NICD degradation rate.

These reporters have been extensively used in Drosophila to determine the different

cell types undergoing Notch activation. The most commonly utilized, Notch Responsive

Element (NRE), is made of two SPS sites flanked by binding sites of a broadly expressed

transcription, Grainy Head (Grh), and a minimal hsp70 promoter (Furriols and Bray, 2001)

(Figure 17). Several variants with different fluorescent proteins have been generated (eGFP,

mCherry or a destabilized Venus-PEST) (Zacharioudaki and Bray, 2014). So far, due to weak

fluorescence, the NRE reporter activity has been mostly detected using antibodies targeting

the reporter protein. Recently, a study provided evidence that a destabilized super-folder GFP

(sfGFP) NRE variant can be used to quantify in vivo Notch activation in epidermal cells of

the fly notum over several hours (Hunter et al., 2016).

Figure 17. Expression patterns of NRE-GFP and Cut, a Notch target gene. Note

the broader pattern of the synthetic reporter. Adapted from Zacharioudaki and

Bray, 2014.


51

Nonetheless, the weak fluorescence could be overcome by simple means. For

example, increasing the number of NRE reporters per animal can enhance the signal-to-noise

ratio. However, this might also affect endogenous signaling. Indeed, reporters can act as sinks

for NICD molecules and compete with endogenous target genes. Another way to increase the

signal-to-noise ratio would be to fuse several fluorescent proteins separated by T2A

sequences downstream the promoter, instead of just one. This strategy has proved to be

successful in refining live imaging of E-Cadherin and Myosin II in the fly notum (Pinheiro et

al., 2017).

Lastly, the choice of the fluorescent reporter also requires much attention. As

fluorescent proteins display different maturation times, detection of the reporter expression

kinetics might be biased by this factor in fast developmental processes (Couturier et al., 2014;

Khmelinskii et al., 2012). Therefore, fast maturating fluorescent proteins should be favored to

avoid long delays between RNA synthesis and fluorescent detection.

Direct measurement

Recent work from our lab provided strong evidence that Notch activation rate can be directly

measured in flies expressing GFP-tagged receptors (Couturier et al., 2012). The strategy

consisted of inserting a GFP in the Notch intracellular domain to monitor both full-length

Notch and cleaved NICD fragments. The construction, named NiGFP, was introduced in flies

within a bacterial artificial chromosome (BAC) and was shown to rescue a Notch null

mutation. In NiGFP transgenic flies, following SOP division, GFP fluorescence could be

detected in vivo in the nucleus of the signal-receiving daughter cell, but not the signal-sending

one (Figure 18). Additionally, the onset of signaling and NICD accumulation rates in the

nucleus of the signal-receiving cell could be determined. Nevertheless, GFP fluorescence was

low in vivo and might limit more refined analyses of Notch dynamics. Tagging Notch

intracellular tails with several GFP proteins could address this shortcoming, but protein

functionality might be altered. More generally, further applications of this approach will

depend the generation of fast-maturating and bright fluorescent tags.

Eventually, it is worth noting that the main difference between this approach and the

aforementioned relies on the direct integration of NICD degradation rate in the quantification

of Notch signaling dynamics. Whereas Gal4 chimeric receptors solely reflect the NICD

production rate at the cell surface, and synthetic reporters measure NICD nuclear activity,

quantification of GFP fluorescence in the nucleus directly integrates both NICD production

and degradation rates. To estimate these processes, two complementary approaches based on


52

light manipulation could be used. First, measuring nuclear GFP fluorescence after FRAP

application on the nucleus (or FLIM outside the nucleus) would allow the estimation of NICD

entry/exit rates in/from the nucleus. Second, in flies expressing Notch receptors tagged with a

green-to-red photoconvertible fluorescent protein, the nuclear NICD pool could be

discriminated from surface or cytosolic Notch molecules, thereby allowing dual imaging of

NICD production (green fluorescence accumulating in the nucleus) and NICD degradation

(decay of the nuclear red fluorescence). Nonetheless, although simple in design, these setups

might be challenging to implement in vivo due to weak fluorescent signals and may need to

be combined with combination either chimeric Gal4 or synthetic reporters to quantify Notch

signaling dynamics.

Figure 18. Dynamics of Notch signaling using a NiGFP transgene. (a-b) Detection

of NiGFP in a fixed pIIa nucleus. (c) Detection of NiGFP in a pIIa nucleus in

vivo. (d-e) Signaling dynamics based on GFP fluorescence quantification in pIIa

and pIIb nuclei. Adapted from Couturier et al., 2012.


53

1.4.2. Time-scaled genome profiling of the Notch transcriptional response

To date, only one study investigated the temporal pattern of the transcriptional output in

response to Notch activation (Housden et al., 2013). Genome-wide approaches have been

applied to identify Notch target genes, but these strategies could only provide a snapshot of a

single transcriptional state, but not the logic of the underlying regulatory network (Chen et al.,

2018; Krejci et al., 2009; Palomero et al., 2006; Weng et al., 2006).

Figure 19. Types of gene expression profiles in response to an NICD stimulus.

See text for details. Adapted from Housden et al., 2013.


54

To obtain the expression dynamics of Notch target genes, Housden et al. performed

expression microarrays on Drosophila DmD8 cells at fixed time points following an initial 5-

minute pulse of NICD expression (of note, NICD was not detected after 30 minutes,

indicating that NICD was active in the nucleus between 5 and 20-25 minutes). They identified

154 differentially expressed genes that respond to the NICD pulse and analyzed a subset of

them that showed the strongest up or down regulations. These 57 selected genes were

distributed among 3 types of expression profile: early up-regulated (5- to 15-minute

response), late up-regulated (>30-minute response) and down-regulated genes (Figure 19).

The first group was mainly composed by genes from the E(spl) complex, while

developmental genes constituted the second group and negative transcriptional regulators the

third one. At first glance, one might think that E(spl) genes act as activating relays driving the

expression of downstream targets that change the cell behavior and/or fate and whose

expression might not necessarily depend on NICD. However, as E(spl) genes encode

transcriptional repressors, the authors hypothesized that they alleviate repression of late up-

regulated genes by repressing another repressor. Indeed, expression of some late up-regulated

genes depended on the E(spl)-dependent repression of Hairy, a transcriptional repressor

belonging to the group of down-regulated genes. As this feed-forward repression affected

only a subset of genes, this implies that additional layers of regulation are at play.

Nonetheless, this supports some interesting model whereby Notch-responsive genes are

maintained in a repressive state to buffer ectopic expression when Notch activity levels are

low or transient. Taking into consideration the fact that E(spl) genes display a context-specific

expression pattern, one might think that E(spl) not only function as relays, but also as points

of a railway track that orient information flow.

The authors also tested the influence of two other parameters, signal duration and

signaling dynamics, on the dynamics of the transcriptional response. A comparison between a

5-minute and a 30-minute pulse of NICD did not reveal significant changes in expression

profiles of most genes, suggesting that, in this context, extending the presence of NICD in the

nucleus does not affect the expression pattern of target genes. Said differently, the

transcriptional network downstream of Notch activation works autonomously once triggered.

Similarly, pulsed or continuous Notch activation did not differentially affect the dynamics of

gene expression, although mRNA levels were higher at some loci. Nonetheless, the authors

did not verify whether additional genes were expressed when changing these parameters (see

next section).


55

Beyond the experimental considerations specific to this study, the key information

here, in light of this section, is that expression of Notch targets requires time, transcriptional

relays and feed-forward interactions. Therefore, once a cell receives Notch signaling, the

acquisition of a new fate with a new set of cellular behaviors is not instantaneous and might

present context specific temporal requirements. This is somewhat obvious, i.e. cells do not

switch fates within seconds or few minutes, but this has never been demonstrated at the

genomic scale in the context of Notch signaling. Finally, as animal development is a time

constrained phenomenon, it would be interesting to determine how the response at the

transcriptional level to Notch signaling is coordinated with the progression of developmental

processes and to reveal the molecular mechanisms enabling fast or slow responses depending

on the temporal constraint of the process.

1.4.3. Dynamic encoding of NICD production kinetics

Why Notch ligands cannot always compensate for each other depending on the cellular

context and whether kinetics of NICD release from the cell surface influence the nature of the

transcriptional output seem at first sight as two separate questions. Yet, a recent study

connected these two questions and elegantly showed that Dll1 and Dll4 activate Notch1 with

different dynamics, which in turn translate into the expression of different Notch target genes

(Nandagopal et al., 2018) (Figure 20).

To reach that conclusion, Nandagopal et al. first observed in a cell culture assay that

Dll1 and Dll4 activate Notch1 with different kinetics, either pulsatile or sustained (NICD

production kinetics were probed using Notch-Gal4 chimeric receptors described above).

Moreover, Dll1 and Dll4 drive expression of different Notch direct target genes: Hes1

expression was induced by both ligands, while Hey1 and HeyL showed strong up-regulation

only with Dll4 (i.e., sustained activation regime). The authors then investigated the

relationship between the kinetics of NICD production and the nature of the transcriptional

output using a NotchΔECD-Gal4 expressing cell line. As stated above, activation dynamics

can be artificially generated with this construct by the timely removal of DAPT. First,

analysis of Hes1, Hey1 and HeyL expression profiles following DAPT removal revealed that

Hes1 responded early and transiently, between 0 and 2 hours after removal, while Hey1 and

HeyL expression was detected 3 hours after removal and was maintained afterwards. This

suggested the idea that pulsed activation of Notch can only trigger Hes1 expression, an early

up-regulated gene, but not Hey1/L, two late responding genes requiring a more sustained


56

activation regime. This hypothesis was corroborated when different durations of NICD pulses

were applied: Hes1 was expressed in a transient manner under any condition (5-, 15-, 30-

minute and continuous pulse) while Hey1 expression required a continuous pulse. Finally, the

authors provided evidence that this differential behavior of Dll1 and Dll4 was also pertinent in

vivo in the context of embryonic myogenesis, where both ligands direct opposite fates, and

was dependent on ligand intracellular tails, but not ECDs, raising the possibility that ligand

specificity arises from different modes of endocytosis.

Figure 20. Dynamic encoding of Notch activity elicits different biological outputs.

See text for details. Adapted from Nandagopal et al., 2018.

From the dynamic viewpoint, this study provides intriguing insights into the way

dynamics of NICD production are decoded by the genome. Identical cells can display

different gene expression patterns that specify distinct fates just simply because the dynamics

of NICD production differ. As the Notch pathway elicits a wide range of responses depending

on the cellular context, it would be interesting to estimate how much the dynamic encoding of

Notch activity described in this study contributes to this heterogeneity.

1.5. Concluding Remarks

57

1.5. Concluding remarks

To conclude, the logic of the Notch pathway where upstream (the ligand) and downstream

signals (NICD) are relayed in a stoichiometric manner appears to be modulated in many

different ways to trigger the expected cellular response at the right time and at the right

location. In a sense, synthetic circuits based on chimeric receptor-ligand pairs only reproduce,

in a crude way, the strategies employed by cells throughout development and evolution to

adapt this signaling module to a given biological context. Moreover, our perception of the

Notch pathway is limited by the extensive work conducted in model organisms, whose

findings are often extended to the whole animal kingdom. Evo-Devo approaches might

therefore provide exciting insights on how the Notch signaling module has been adapted not

only between cell types, but also between species and phyla.

Although only a small part of this knowledge will be useful in the following chapters,

the main aim of this chapter was nonetheless to give to the reader an integrated view of our

current knowledge of the pathway and a general context related to Notch signaling in which

my thesis is integrated.

58

Chapter 2

Notch signaling during Drosophila neurogenesis

The involvement of Notch in Drosophila embryonic neurogenesis was first discovered in the

1930’s in pioneering work of Poulson on embryos carrying deficiencies on the X

chromosome (Poulson, 1937). Among the different deficiencies he analyzed, one spanned the

whole Notch locus. He observed that endodermal and mesodermal tissue layers were lacking

in Notch deficient embryos. At the same time, the ectoderm being proliferated producing cells

belonging to the nervous system at the expense of the epidermis. Forty years later,

combination of mutational analysis with electronic microscopy confirmed this original

observation and revealed that the “neurogenic” phenotype is caused by an imbalance in fate

acquisition between neuroblasts and epidermal cells within proneural fields (Hartenstein and

Wodarz, 2013; Lehmann et al., 1981, 1983). Notch and other “neurogenic” genes (Delta,

Mastermind, E(spl)-C) were later proposed to single out neuroblasts in the embryonic

neurectoderm through a process termed “lateral inhibition” (Cabrera, 1990).

Since then, regulation of lateral inhibition by the Notch pathway appeared to occur in

multiple neurogenic processes throughout fly development. In parallel, molecular

characterization of neural lineages revealed that Notch signaling is also involved in fate

specification in the central and peripheral nervous systems. This chapter aims to give to the

reader an overview of these different processes. In a first section, it synthesizes the different

neural contexts where Notch signaling regulates lateral inhibition. Second, it briefly

summarizes the current knowledge on the role of Notch in larval neurogenesis. A last and

consequent section will be dedicated to the sensory organ lineage, the model I worked on

during my PhD.

2.1. Lateral inhibition and selection of neural progenitors

Since the original discovery of Notch role in selecting neuroblasts among a proneural groups,

several additional selection processes of neural progenitor selection were found to be

dependent on Notch. General principles of lateral inhibition will be discussed in the first


59

subsection dedicated to embryonic neurogenesis, while the two following subsections will

serve as variations of this common theme.

2.1.1. Embryogenesis and neuroblast selection

Figure 21. Selection of neural progenitors during embryogenesis: from

prepatterning genes to neuroblast. (a) Prepatterning genes expression pattern. (b-

c) Proneural clusters. (d) Neuroblast delamination after selection. (e) Notch-

mediated lateral inhibition. (f) From the proneural cluster to the NB lineage.

Adapted from Hartenstein and Wodarz, 2013.

Chapter 2. Notch signaling during Drosophila neurogenesis

60

Formation of a proneural field

Lateral inhibition occurs within a group of equivalent cells that can either adopt a neural or an

epidermal fate (Figure 21). During embryogenesis, these groups of cells, or “proneural

clusters”, are found in the ventral neurectoderm and their location in different embryonic

segments is dependent on prepatterning genes (Hartenstein and Wodarz, 2013; Skeath and

Carroll, 1992; Skeath et al., 1992) (Figure 21a). These prepatterning genes regulate the

expression of a set of proneural genes, acheate (ac), scute (sc) and lethal of scute (l’sc), that

belong to the Achaete-Scute complex (AS-C). These three proneural genes encode bHLH

transcription factors that specify the equivalency group (Figure 21b-c). Intriguingly, these

proneural genes are not always expressed altogether among all the proneural clusters (Skeath

et al., 1992). Whether this is responsible for the different types of neurons produced later by

the selected neuroblasts is not known.

Neuroblast selection

Once proneural clusters are formed, proneural genes enhance expression of Delta and bi-

directional Notch signaling starts between cells (Figure 21e). In this context, Notch receptors

exert an inhibitory effect on proneural factors and bias fate specification towards the

epidermal identity. NICD drives the expression of E(spl)-C genes whose gene products bind

to the promoter region of AS-C genes and inhibit their transcription. As signaling endures,

cells from the proneural clusters enter a bi-modal state where they express simultaneously the

proneural genes and their inhibitor. Eventually, one cell among the cluster “wins” the

competition by expressing more Delta and/or receiving less Notch signal from the neighbors.

As the selected neuroblast is the only remaining cell expressing proneural genes, it

upregulates Delta expression and sends a permanent inhibitory signal to lock in respective

fates (Campos-Ortega, 1994; Hartenstein and Wodarz, 2013). Finally, lateral inhibition ends

in this context with neuroblast delamination and its first division (Figure 21d,f).

Lateral inhibition has also been extensively studied in the context of sensory organ

precursor (SOP) selection and recent advances will be discussed in the next subsection.

2.1.2. Notum patterning and SOP selection

SOP selection, a classical lateral inhibition model

The understanding of Notch involvement in lateral inhibition arose with concomitant

discoveries in neuroblast or SOP selection (Hartenstein and Wodarz, 2013). SOPs produce a


61

wide variety of sensory organs through stereotyped lineages during fly development (Lai and

Orgogozo, 2004) (Figure 22). Like neuroblasts, SOPs emerge from proneural groups

characterized by proneural genes, either belonging to the AS-C complex or not (Jarman et al.,

1995; Simpson, 1990). Selection of SOPs then occurs via a competition mechanism mediated

by Notch signaling (Heitzler and Simpson, 1991; Simpson, 1990).

Figure 22. Diversity of sensory organs and lineages in Drosophila. Adapted from

Lai and Orgogozo, 2004.


62

Lateral inhibition has been mostly studied in the context of the microchaete lineage in

the fly notum, or dorsal thorax. SOPs giving rise to these structures are first selected among

proneural clusters expressing achaete and scute (Simpson, 1990). A recent report provided

experimental and computational insights into the selection process of microchaete SOPs

(Corson et al., 2017). Authors demonstrated that proneural cells experience a succession of

intermediary states where they both emit and receive the signal (Figure 23). This cell-intrinsic

bistability in turn leads to a gradual refinement of proneural clusters, organized as anterior-

posterior stripes in the notum, into single SOPs selected at the center of the stripe. These

results contrast with former models proposed for lateral inhibition where precursors are

selected in a sharp manner once a threshold of signal-sending activity has been crossed (Barad

et al., 2010; Lubensky et al., 2011). Finally, the model implied a positive feedback loop

between proneural activity and signal-sending activity (i.e. Delta expression) in the proneural

group and cis-inhibition of Notch as mechanisms permitting bistability and amplifying subtle

heterogeneities in cell states, although their biological significance has not experimentally

been tested in this study. More information on modeling lateral inhibition can be found in

Binshtok and Sprinzak, 2018.

Figure 23. Neural precursor selection occurs through gradual refinement of the

cluster. (A-B) Model recapitulating precursor selection. (C-D) In vivo cluster

refinement. Sc (Scute) is the proneural gene. Adapted from Corson et al., 2017.


63

Dynamic patterning of SOP selection in the notum

Proneural clusters where SOPs give rise to macrochaete or microchaete are determined in the

notum either by prepatterning factors (Gómez-Skarmeta et al., 2003; Simpson, 2007) or by

Notch-mediated self-organization (Corson et al., 2017), respectively.

In the former case, the situation is much like neuroblast selection during embryonic

neurogenesis. In the latter case, however, extrinsic cues do not provide the informational

position. As mentioned above, proneural clusters organize within narrow stripes along the

body axis. At the end of the patterning process, each hemi-notum displays five rows of

regularly spaced SOPs. Strikingly, at the beginning of the process, no stripe can be observed.

Careful examination of proneural activity and Notch signaling components revealed how the

five rows emerge (Corson et al., 2017) (Figure 24). First, a prepattern of Delta expression

with Notch activity at its sides sets the position for Ac and Sc expression in the presumptive

rows 1, 3 and 5. As SOPs are selected within these rows, Notch activity starts receding

between rows 1-3 and 3-5. Gaps generated by this decrease provide the positional information

for Ac and Sc expression in presumptive rows 2 and 4. Thus, dynamic Notch signaling

provides in this context a negative template for the sequential emergence of SOP rows in a

self-organized manner.

Figure 24. Patterning of the bristle rows in the notum combines prepatterning and

self-organization. Adapted from Corson et al., 2017.


64

To conclude, emergence of the bristle pattern in the fly notum provides an interesting

example where Notch signaling is used sequentially to provide the positional information for

proneural activity in the prospective rows and to single out SOPs once proneural stripes are

defined. Nonetheless, it still remains unclear why Ac and Sc start being expressed in between

rows of Notch activity.

Long-range signaling: what mechanism? what role?

In the notum, proneural domains exceed several cell diameters. This questions how cells that

are not in the immediate vicinity of the selected SOP are inhibited. A simple model would

assume that proneural cells at the boarder of the proneural domain would be inhibited first,

and so forth as one of the cells at the center of the domain will be selected. Such gradual

refinement is coherent with current models of lateral inhibition (Corson et al., 2017;

Lubensky et al., 2011).

Figure 25. Combination between short-range (A) and long-range (B) signaling is a

potent mechanism explaining lateral inhibition in large proneural clusters.

Adapted from Perez-Mockus and Schweisguth, 2017.


65

Alternatively, lateral inhibition might combine short-range and long-range Notch

signaling (Figure 25). This hypothesis has emerged fifteen years ago when thin actin-based

protrusions, or filopodia, containing Delta molecules have been observed on the basal side of

SOPs (de Joussineau et al., 2003; Renaud and Simpson, 2001). However, whether these

structures could contact distant cells and activate Notch was speculative at that time. A series

of reports from the Kornberg lab (Huang and Kornberg, 2015; Roy et al., 2011, 2014)

provided evidence that filopodia can indeed mediate Hedgehog or Notch signaling in other

developmental contexts. Concomitantly, a combination of in vitro experimentation with

modeling revealed that signaling in filopodia is presumably dependent on ligand diffusion

coefficient to compensate for small contact surfaces (Khait et al., 2016). Therefore, high

ligand turnover and/or diffusion length in filopodia might yield high levels of Notch activity.

All these reports considered, it became clear that filopodia might be a potent source of

ligands for Notch receptors. Recent studies provided support to a model where filopodia are

involved after SOP selection in refining the bristle pattern during notum morphogenesis

(Cohen et al., 2010) and in maintaining fates of distant epidermal cells prior to SOP division

(Hunter et al., 2016). Whether filopodia are required for the selection process of SOPs within

the proneural cluster is yet not known.

Another mechanism regulating Notch long-range signaling is based on the secreted

glycoprotein Scabrous (Sca). Sca was first discovered in the context of R8 selection in the eye

disc (see subsection below, Baker et al., 1990). In the notum, loss of Sca leads to

supernumerary SOPs with reduced spacing compared to wild-type cells (Renaud and

Simpson, 2001). This suggested that Sca promotes long range signaling, either in a Notch-

dependent or independent way. Discovery that Sca complexes with Notch favored the first

model (Powell et al., 2001). To date, the molecular function of Sca in Notch signaling is still

debated, as it has been proposed to act as a positive (Giagtzoglou et al., 2013; Li et al., 2003)

or negative (Lee et al., 2000; Powell et al., 2001) regulator.

2.1.3. Photoreceptor R8 selection in the eye disc

Morphogenetic furrow progression

The Drosophila eye compound is a sensory organ composed by 800 subunits, called

ommatidia, that are organized in a crystal lattice (Ready et al., 1976). Each ommatidium is a

complex multicellular structure made of eight photoreceptor neurons, four cone cells that

secrete the apical lens and additional pigment cells (Kumar, 2012). The crystal pattern of the


66

adult retina originates from the selection in a regularly-spaced pattern of R8 photoreceptors

during imaginal eye disc development. These are the first cells selected among each

ommatidium, the remaining cells being recruited and specified afterwards.

Selection of R8 photoreceptors is tightly linked to the progression of a morphogenetic

furrow that crosses orthogonally the whole epithelium from the posterior to the anterior

margin (Ready et al., 1976; Wolff and Ready, 1991). R8 cells emerge right behind the furrow

as it progresses anteriorly (Figure 26).

Figure 26. Morphogenetic furrow progression (red) and formation of ommatidial

rows (green) in the eye disc. Adapted from Spratford and Kumar, 2014.

Besides the classical features defining a furrow, i.e. apical constriction and tissue

invagination, this morphogenetic furrow is also characterized by the expression of a bHLH

proneural gene, atonal (ato), within an anterior stripe (Jarman et al., 1994). In absence of ato,

the morphogenetic furrow keeps progressing but R8 specification is lost (Jarman et al., 1995).

This led to the conclusion that Ato is the proneural gene in this context. Analysis of Ato

expression revealed that R8 selection occurred in a two-step restrictive process where first

Ato expression pattern is refined from the anterior stripe into intermediary groups (1) prior to

R8 selection (2) (Jarman et al., 1994, 1995).

Following this discovery and knowing that Notch signaling was required for

photoreceptor specification (Cagan and Ready, 1989), the analogy with neuroblast or SOP

selection through lateral inhibition became clear.


67

Notch signaling, patterning and selection of the R8 photoreceptor

The transition from the intermediary proneural group to the single R8 photoreceptor through

ato down-regulation looks much like classical Notch-mediated lateral inhibition (Figure 27).

In addition, patterning of photoreceptors into a crystalline lattice highly resembles a template-

based mechanism where the former row of specified R8 cells provides the positional

information for the next row by emitting a circular inhibitory field (Lubensky et al., 2011).

Then, two functions can be hypothesized for Notch signaling in this context: it is required for

R8 selection (“local lateral inhibition”) (1) and it constitutes the inhibitory signal that

positions the next row (“long-range inhibition”) (2).

Figure 27. R8 selection and patterning in the eye disc. See text for details.

Adapted from Frankfort and Mardon, 2002.


68

Using a Notch thermosensitive allele to probe in a timely manner the requirement of

Notch activity during the selection process, Baker and others demonstrated that Notch

signaling was indeed required for the selection of the R8 cell among the intermediary group

(Baker and Yu, 1997; Baker et al., 1996) (stage 2/3 in Figure 27). Mechanistically, Notch

activity appeared to inhibit Ato autoregulation in intermediary groups, presumably through

the expression of E(spl) genes. This led to the extinction of Ato expression in the intermediary

group except in the future R8 cell. To date, whether R8 selection occurs in a similar fashion

as SOPs in the notum (Corson et al., 2017) or by a preselection mechanism (Lubensky et al.,

2011) remains unclear. Further mutational analyses revealed that Delta is the key ligand in

this context (Baker and Yu, 1997). Intriguingly, although redundant with Delta, Ser

expression was detected in the morphogenetic furrow and in selected R8 cells. What non-

essential function Ser could mediate in this process remains a mystery.

The role of Notch signaling in exerting the long-range inhibition that serves as a

template for the next row of R8 photoreceptors is less clear. While first ruled out (Baker et al.,

1996), such role was assessed in later developmental studies (Baker et al., 1996; Sun et al.,

1998) or inferred in computational analyses describing the crystalline pattern formation

(Lubensky et al., 2011). In support of this “long-range inhibition”, signals emitted by Delta

expressing cells posterior to the morphogenetic furrow were shown to extend over several cell

diameters (Baker and Yu, 1997). The cellular basis (filopodia, cell packing) of this long range

signaling has yet to be determined. Similarly to the notum, Scabrous has been proposed as the

molecular effector of the inhibitory field (Baker et al., 1990; Lubensky et al., 2011), but its

molecular function has still not been defined.

More careful analysis of Notch phenotypes revealed an unexpected third function for

Notch signaling prior to R8 selection. While it was established that Ato expression in the

stripe anterior to the morphogenetic furrow was dependent on Hedgehog (Hh) and

Decapentaplegic (Dpp) signaling (stage 1, Figure 27), it appeared that Notch was required to

stimulate Ato expression levels at higher levels (Baker and Yu, 1997). Importantly, this

enhancement seemed to be required for the transition from the stripe to the intermediary

group and correlated with the shift of ato regulation by upstream regulators (Hh, Dpp)

towards autoregulation. This Notch-dependent upregulation was hypothesized to enhance

expression of Ato target genes at sufficient levels to trigger subsequent processes.

Nonetheless, whether such gating mechanism is taking place remains to be tested. More

confounding at that time was the fact that Notch signaling could either stimulate or inhibit

Ato expression depending on the step during the selection process. One explanation resides in


69

the signal transduction mechanism: while Ato inhibition during R8 selection required E(spl)-

C expression in a Su(H)-dependent manner, enhancement of Ato expression in the stripe by

Notch apparently did not require Su(H) (Ligoxygakis et al., 1998). Further reexamination of

the Su(H) phenotype supported the idea that Notch activation is required to alleviate Su(H)-

mediated repression (Bray and Furriols, 2001; Li and Baker, 2001). A second explanation lies

in the mode of Ato regulation. Isolation and characterization of ato enhancers uncovered that

Ato expression within the stripe or in intermediary groups is mediated by two distinct

regulatory regions (Sun et al., 1998) (Figure 27). One appealing, but, hypothesis is that the 3’

enhancer responsible for Ato expression within the stripe integrates Hh and Dpp positioning

signals and a Notch-mediated Su(H) “derepression” to elevate Ato expression levels in order

to trigger the switch in Ato regulation towards the second 5’ enhancer specific to the

intermediary groups that depends on Ato auto-regulation and is sensitive to E(spl) inhibition.

As a result, intermediary groups form only at equal distance from R8 cells of the former row,

while Ato expression becomes sensitive to the long-range Notch-mediated inhibitory signal

(Figure 27).

To conclude, just like in the notum, Notch signaling is reused several times during the

R8 selection process to trigger the transition from the stripe to the intermediary groups (1), to

select R8 by inhibiting immediate neighbors (2) and to presumably set the position of the next

intermediary groups emerging from the stripe (3). Hence, Notch-mediated patterning and

lateral inhibition seem to be coordinated. How this emerged during evolution still remains

poorly understood (see discussion in Gazave et al., 2017).

2.2. Central nervous system: neural pool expansion and cell type

diversification

Following neural precursor selection during embryonic neurogenesis, Notch signaling is

reused several times to maintain and specify cell fates along neural lineages when quiescent

embryonic neuroblasts of the central nervous system (CNS) reactivate at larval stages.

Depending on the step when Notch is required, fate specifications can either lead to stem cell

renewal or to diversification of differentiated cell types. These two cases will be described in

the following two subsections.


70

2.2.1. Neuroblast self-renewal

In the larval brain, neural stem cells are subdivided in two categories: type I and type II

neuroblasts (NBs). They differ by the nature of the lineages they initiate. Type I NBs divide

asymmetrically to self-renew and produce differentiated glial mother cell (GMC). The GMC

then divides to produce either neurons or glial cells (Brand et al., 1993; Jarman et al., 1993).

On the other hand, type II NBs also divide asymmetrically to self-renew but instead of a

GMC, produce an intermediate neural progenitor (INP) (Bowman et al., 2008). These INPs

undergo in turn a succession of self-renewing ACDs generating GMCs that terminally divide

to generate two neurons or glial cells. Although in fewer number than type I NBs, type II NBs

makes a significant contribution to larval neurogenesis due to this transit-amplification. In

both lineages, despite these differences, daughter NBs receive Notch signaling from the

newly-born GMC (Zacharioudaki et al., 2012).

The requirement of Notch signaling is different in the two subtypes. While absolutely

required to maintain identity of type II NBs (Bowman et al., 2008), Notch activity appears to

play only a minor role in type I NB self-renewal (Almeida and Bray, 2005; Homem et al.,

2015; Wang et al., 2006). This discrepancy might be due to the expression of Asense (Ase), a

bHLH transcription factor, in type I but not type II NBs (Bowman et al., 2008). One

hypothesis would be that Ase buffers Notch activity in type I NBs by antagonizing E(spl)

genes (Southall and Brand, 2009) and regulates other mechanisms that maintain type I NB

identity after division (Zacharioudaki et al., 2012); nonetheless the basis for the differential

integration of Notch signaling between NB lineages is not known. In addition to self-renewal,

Notch signaling regulates also growth of the daughter NB following division of the mother

type II NB (San-Juán and Baonza, 2011; Song and Lu, 2011).

2.2.2. Cell type differentiation and diversification

NB lineages are characterized by one (type I) or a succession (type II) of asymmetric cell

division(s) that is followed by the terminal division of a GMC giving rise to two neurons or

two glial cells. Often considered as a symmetric division (Homem et al., 2015), the terminal

division of the GMC can generate daughter cells with distinct fates. Analysis of neural

lineages revealed that Notch is also activated differentially between daughter cells of a GMC.

In the ventral CNS, directional Notch signaling between the terminal daughter neurons is

2.2. CNS: neural pool expansion and cell type diversification

71

crucial to generate the diversity of secondary neurons (Truman et al., 2010). Similar results

were obtained with medulla NBs, whose lineages are identical to type I ones ((Li et al., 2013).

A recent report provided evidence that such role for Notch in diversifying cell fates among a

common cellular type is also conserved in gliogenesis (Ren et al., 2018).

Hence, Notch signaling appears to regulate cell fate decisions in NB lineages from the

top, i.e. the stem cell in type II lineages, to the bottom, i.e. the neuronal or glial identities.

Such preponderance, although intriguing at first sight, might be simply due to the fact that

Notch signaling, being a juxtacrine pathway, is the perfect signaling toolkit to specify fates in

an intra-lineage fashion. Thus, NB lineages can undergo their developmental program

autonomously without being dependent on external cues to dictate fate decisions.

2.3. Peripheral nervous system: specification of the microchaete

lineage

As in the CNS, Notch signaling regulates cell fate decisions in various developmental

contexts in the peripheral nervous system (PNS), including ommatidium formation,

diversification of peripheral neurons or fate specification in sensory organ lineages. This

section will focus specifically on the role of Notch in one the sensory organ lineages: the

microchaete lineage of the notum. Phenomenology of the lineage will be presented, as well as

a brief synthesis on the ACD of sensory organ precursor cell (SOP). The two last subsections

will each emphasize an aspect of SOP division that is debated or poorly understood and that

will correspond each to a Chapter of the Results section.

2.3.1. The microchaete lineage: an overview

Phenomenology of the lineage

Following their selection by lateral inhibition during early metamorphosis, SOPs undergo a

stereotyped series of ACDs that finally give rise to a four-cell mechanosensory organ called

microchaete (Fichelson and Gho, 2003; Gho et al., 1999; Hartenstein and Posakony, 1989;

Roegiers et al., 2001a) (Figure 28). SOPs first divide asymmetrically along the body anterior-

posterior axis to generate a posterior pIIa cell and an anterior pIIb cell. Two hours later, the

pIIb undergoes its own ACD in a neuroblast-like manner along the apical-basal axis. It


72

produces an apical pIIIb cell that remains at the level of other epidermal cells and a basal glial

cell that undergoes apoptosis (Fichelson and Gho, 2003). The pIIa cell divides as the SOP

along the body axis approximately one hour after pIIb ACD and gives rise to the socket cell

and the shaft, i.e. the two external cells of the mechanosensory organ. pIIIb then divides

asymmetrically along the apical-basal axis two hours after being produced and generates the

neuron and the sheath cell, i.e. the internal cells of the mechanosensory organ. Hence, two

types of ACDs occur in the microchaete lineage: following the anterior-posterior axis (SOP,

pIIa) and along the apical-basal axis (pIIb and pIIIb). Of note, the microchaete lineage, named

hereafter the SO (sensory organ) lineage, is just one variant among many different sensory

organ lineages found in flies (Lai and Orgogozo, 2004) (Figure 22).

Figure 28. The thoracic microchaete lineage. (a) Dorsal thorax, or notum. (b)

External and (c) internal cells of the mechanosensory bristle. (d) Lineage

overview. Red (resp. blue) outlines Notch activation (resp. inhibition). Adapted

from Schweisguth, 2015.

Once the four cells composing the future organ are made, they undergo coordinated

morphogenetic processes to form a functional mechanosensory organ.

2.3. PNS: specification of the microchaete lineage

73

Early evidence for Notch requirement

In 1990, while the Notch field was exploding, Hartenstein and Posakony provided pioneering

evidence supporting the requirement of Notch activity in fate specification in the SO lineage

(Hartenstein and Posakony, 1990). They made use of the thermosensitive allele Nts of Notch

to estimate the temporal requirement of Notch activity in the SO lineage. While early

inactivation of Notch during pupal development impaired lateral inhibition, thereby leading to

an excess of bristles on the adult notum, bald cuticles were recovered when the temperature

shift was performed after SOP selection. Although this experiment clearly demonstrated the

requirement of Notch in specifying the SO fates, notions of intralineage signaling and ACD

were not considered. In fact, by analogy with the lateral inhibition model, it was hypothesized

that the neuron was inhibiting the three remaining cells.

Figure 29. Unequal segregation of Numb during SOP ACD in pIIb. Adapted from

Schweisguth, 2015.

This changed with the discovery of the unequal segregation of Numb in one daughter

cell following SOP and NB ACD (Rhyu et al., 1994) (Figure 29). Then, it became clear that

fates in the SO lineage are internally regulated by ACDs. Of note, Numb was already

identified as a fate determinant in SO lineages, but its role in biasing fates towards the neural-

like one was unclear (Uemura et al., 1989). A last brick was missing: the link with a signaling

pathway. The connection with Notch was made few years later when Numb was shown to

directly bind Notch and to antagonize its activity. Additionally, live-imaging of Partner of

Numb (Pon, Lu et al., 1998), a Numb adaptor protein in SOPs, revealed that pIIb, pIIIb, the

socket and the neuron were the cells inheriting Numb (Gho et al., 1999). This led to a new era

when the regulators that orchestrate directional Notch signaling in the SO lineage started

being characterized.


74

2.3.2. Molecular mechanisms of SOP asymmetric cell division

Polarization of the SOP in response to PCP

ACD is step-wise process where an initial symmetry breaking event triggers the subsequent

orientation of the spindle along the division plane and the unequal segregation of cellular

materials that include cell fate determinants, organelles or cytosolic content (Knoblich, 2008;

Schweisguth, 2015). SOPs lying in the notum, a monolayered epithelium, display an apical-

basal polarity with distinct domains. Their symmetry breaking mechanism relies on the

polarization of polarity proteins prior to ACD, a strategy found again and again in metazoans

(Cowan and Hyman, 2004; Knoblich, 2008) (Figure 30).

Figure 30. Polarization of polarity proteins prior to SOP division ensures unequal

segregation of Numb and Neur and orientation of the mitotic spindle. Adapted

from Schweisguth, 2015.

The polarity proteins Par3 (Bazooka (Baz) in flies), Par6 and the atypical protein

kinase C (aPKC) were shown to accumulate at the posterior pole of interphasic SOPs in

response to the Fz/Dsh (Frizzled/Dishevelled) PCP signaling (Bellaïche et al., 2001; Bellaıche

et al., 2001; Besson et al., 2015; Roegiers et al., 2001b) (Figure 30). Of note, PCP signaling is


75

only responsible for the anterior-posterior orientation of the ACD and impairing its

components does not prevent ACDs for occurring. How the asymmetric mode of division is

rescued at mitosis in PCP mutants is still a mystery. Recently, Meru, a RASSF9/10 homolog

that polarizes posteriorly prior to ACD, was proposed to bridge the Fz/Dsh pathway with Par3

recruitment at the posterior pole (Banerjee et al., 2017). Similar to PCP components, Meru is

dispensable for ACD.

In interphase, Par6-aPKC is maintained in an inactive state in a complex with Lgl

(Lethal (2) giant larvae) (Wirtz-Peitz et al., 2008). Upon mitosis entry, the mitotic kinase

AuroraA phosphorylates Par6 and mediates dissociation from Lgl. Released Par6-aPKC

dimers then associate with Par3 and form the active Par3-Par6-aPKC complex which excludes

Numb from the posterior cortex by the coordinated Par3-dependent recruitment of Numb

(Wirtz-Peitz et al., 2008) and its aPKC-mediated phosphorylation (Smith et al., 2007; Wirtz-

Peitz et al., 2008) (Figure 31). Thus, Numb anterior polarization is controlled both spatially

and temporally. Nonetheless, it remains unclear how Par6 and aPKC are recruited posteriorly.

Indeed, they are sequestered by Lgl, which is uniformly distributed at the cortex, and cannot

interact with polarized Par3 prior AuroraA activation.

Figure 31. Phosphorylation-based exclusion of Numb from the posterior pole

prior to SOP division. See text for details. Adapted from Schweisguth, 2015.

In parallel, an anterior complex was also shown to form prior ACD in response to

Pk/Stbm (Prickle/Strabismus) PCP signaling and to be required for Numb anterior recruitment

(Bellaïche et al., 2004; Bellaıche et al., 2001; Schaefer et al., 2000). This complex is made of

the polarity protein Dlg (Disc-large), Pins (Partner of Inscuteable) and Gai. In synergy with


76

Meru and Fz/Dsh signaling, Pins excludes Par3 from the anterior cortex and thereby enhances

symmetry breaking (Figure 30).

Intriguingly, although two mechanisms were found to direct Numb anterior

localization, i.e. the posterior phosphorylation-based exclusion and the anterior positive

recruitment, it is still unclear why Numb and Pon form a cortical crescent at mitosis entry and

do not accumulate at endosomal cell surfaces or at apical membranes. An entry point to this

matter might reside in the regulation of Pon distribution by the mitotic Polo kinase (Plk) in

NBs (Wang et al., 2007; Zhu et al., 2016). Plk-mediated phosphorylation is indeed required

for anterior localization of Pon. Therefore, Pon might be the key effector recruiting Numb

anteriorly. More recently, Pon and Numb were shown to form liquid condensates at the basal

cortex of NBs and liquid droplets in Hela cells through atypical interactions between Numb

PTB domain and Pon N-terminus (Shan et al., 2018). Authors speculated that physical

characteristics of this condensate, coupled with the membrane anchoring of the Numb-Pon

complex (Knoblich et al., 1997; Lu et al., 1999), explain why Numb and Pon distribution is

crescent-shaped.

Spindle orientation

The polarizing cues driving Numb anterior localization are reutilized during ACD to align the

spindle with the anterior-posterior division axis. While randomly distributed upon mitosis

entry, centrosomes are recruited to the anterior and posterior poles at mitotic entry (Bellaïche

et al., 2001; Schweisguth, 2015). One centrosome is recruited anteriorly by polarized Pins and

Gai in a Mud-dynein-dependent manner whereas Fz/Dsh attract the second centrosome via

Mud (Bowman et al., 2006; Johnston et al., 2013; Ségalen et al., 2010) (Figure 30).

In this way, unequal segregation of Numb is coupled to spindle orientation along an

anterior-posterior division axis.

2.3.3. Unequal segregation of fate determinants and directional Notch

signaling

Numb, the negative regulator

Numb is an evolutionary conserved protein involved in a plethora of cellular processes that

include ubiquitin-mediated protein degradation, vesicle trafficking, and cell adhesion (Gulino

et al., 2010). Drosophila Numb acts as an endosomal adaptor through its phospho-tyrosine


77

binding (PTB) domain (Guo et al., 1996; Hutterer and Knoblich, 2005) and as an endosomal

sorting protein through its two C-terminal Eps15 homology regions, termed DPF and NPF

(Gulino et al., 2010; Hutterer and Knoblich, 2005).

Figure 32. Time-lapse images illustrating the relocalization of Numb from the

anterior crescent (B-E’’) to subapical endosomes in pIIb (F-G’). Adapted from

Couturier et al., 2013.

During ACDs in the SO lineage, Numb is asymmetrically inherited by the signal-

sending cell (Guo et al., 1996). At cytokinesis, Numb relocalizes from the anterior crescent to

subapical sorting endosomes (Couturier et al., 2013) where it is thought to inhibit Notch

recycling (Cotton et al., 2013; Couturier et al., 2013) and promote its addressing towards

lysosomes (Couturier et al., 2014) (Figure 32). The current view on Notch inhibition by

Numb involves a third protein, Sanpodo (Spdo). Spdo is a four-pass transmembrane protein

expressed specifically in multiple lineages where fate acquisition is Notch-dependent (Dye et

al., 1998; O’Connor-Giles and Skeath, 2003; Park et al., 1998; Skeath and Doe, 1998). It

physically interacts with Notch (O’Connor-Giles and Skeath, 2003) and is required in the

signal-receiving cell for Notch activation through the recruitment the g-secretase complex

(Upadhyay et al., 2013). Spdo also physically binds to Numb and thereby negatively regulates

Notch in the signal-sending cell (O’Connor-Giles and Skeath, 2003). A combination of live-


78

imaging and mutational analyses provided support to a “recycling inhibition” model

(Schweisguth, 2015) whereby Numb prevents Notch recycling towards the cell surface via

Spdo (Cotton et al., 2013; Couturier et al., 2013) (Figure 33). The molecular details of this

model have yet to be determined.

Figure 33. Current model for Numb function in pIIb. See text for details. Adapted

from Couturier et al., 2013.

Neuralized, the positive regulator

Neur is an E3 ligase promoting endocytosis of Notch ligands (see section 1.3.2. for details). In

the SO lineage, it is both required to single out SOPs during lateral inhibition and to promote

Delta/Notch signaling during lineage progression (Yeh et al., 2000) (of note, Delta is the key

ligand in this context (Zeng et al., 1998)).

Figure 34. Unequal segregation of Neur in pIIb during SOP ACD. Adapted from

Le Borgne and Schweisguth, 2003.


79

Just like Numb, Neur is segregated unequally in pIIb during SOP ACD (Le Borgne

and Schweisguth, 2003) (Figure 34). This bias in Neur distribution is considered to increase

the signal-sending activity of pIIb, thereby reinforcing intralineage fate specification.

Regulation of Neur asymmetric anterior localization shares similarities with Numb and Pon as

Neur anterior crescent was shown to be dependent on PCP and the components of the

anteriorly polarized complex (Le Borgne and Schweisguth, 2003). However, a mechanism

based on posterior phosphorylation-dependent exclusion does not seem to apply to Neur.

Finally, the molecular machinery leading to Neur anterior accumulation is not known.

Is unequal segregation of Numb and Neur relevant, after all?

Unequal segregation of fate determinants is a powerful mechanism to bias fate decisions in an

intralineage fashion, yet in the context of the SOP ACD, the relevance of Numb and Neur

unequal inheritance has not been directly tested.

Regarding Numb, it remains unclear to what extent increased Numb concentration

and/or differentially regulated Numb activity contribute to the inhibition of Notch recycling

specifically in pIIb. The fact that Numb redistributes at cytokinesis in subapical sorting

endosomes in pIIb, but not in pIIa, raises the question whether an additional regulator is

present in pIIb to drive this cell-specific relocalization. This question could be directly

addressed by analyzing fate specification, Notch activity and Numb distribution in pIIa-pIIb

pairs expressing Numb mutants unable to bind to Par3 (Wirtz-Peitz et al., 2008) or aPKC

(Smith et al., 2007).

On the other hand, several lines of evidence suggest that Neur asymmetric inheritance

in pIIb is not crucial. First, loss of Numb leads to robust bidirectional Notch signaling along

the lineage. As the formation of Neur anterior crescent is independent of Numb (Le Borgne

and Schweisguth, 2003), this means that pIIa can efficiently signal to pIIb in absence of

Numb despite a reduced Neur concentration. Second, Neur knockout can be compensated by

uniformly expressing Mib1 in SOPs to restore fate specification (Lai et al., 2005b). Therefore,

uniform E3 ligase activity does not induce cell fate transformations. Taken together, these

data rather indicate that unequal segregation of Neur provides a non-essential mechanism

increasing the robustness of binary fate decisions. Challenging the system by decreasing

Notch or Delta expression levels might uncover such role.


80

2.3.4. Notch is activated in pIIa, but where?

Contact area and Notch signaling

As Notch signaling is a juxtacrine pathway, it was long assumed that signaling activity could

be affected by the contact area. Recently, two in vitro studies provided experimental evidence

supporting this idea (Khait et al., 2016; Shaya et al., 2017) (Figure 35). In the latter report, to

address this question, investigators designed an elegant micropattern-based device allowing

the precise control of the contact area between a signal-sending and a signal-receiving cell.

Using trans-endocytosis levels as a readout for Notch activity, they observed that Notch

activation levels scale with contact area. From a biological standpoint, these results emphasize

the possibility that the location of Notch activation site at the interface of contacting cells

might have a quantitative effect on signaling. Therefore, elucidating this location might not

only explain how cells elicit strong Notch activity but also shed a new light on known Notch

regulators regarding the organization of the activation site.

Figure 35. Correlation between contact area and Notch signaling activity. Adapted

from Shaya et al., 2017.

In the SO lineage, although investigators were mostly making use of this model to

perform genetic screens to characterize novel Notch regulators (Acar et al., 2008; Giagtzoglou

et al., 2013; Go and Artavanis-Tsakonas, 1998; Jafar-Nejad et al., 2005; Le Bras et al., 2012;

Mummery-Widmer et al., 2009; Rajan et al., 2009), several models were proposed over the

years for Notch activation site during SOP ACD. In fact, these models were mostly built to

interpret the phenotypes observed when newly-characterized regulators were mutated rather


81

than to directly address where Notch is activated. The supporting evidence will be presented

and discussed below as well as the implications resulting from the hypothesized signaling

subcellular site. Developing novel experimental strategies to directly tackle the location of

Notch activation site along the pIIa-pIIb interface comprises Chapter 4 from the “Results”

section.

Adherens junction, apical microvilli and Delta recycling

Detection of Notch and its ligands at the level of the adherens junction (AJ) adhesive belt and

the above-located apical region in epithelia led investigators to assume that Notch receptors

are activated at this location (Banda et al., 2015; Mizuhara et al., 2005; Sasaki et al., 2007).

This idea is further reinforced by the reported physical interaction between Notch and E-Cad

(Sasaki et al., 2007).

As mentioned earlier, SOPs lying in the notum display an apical-basal polarity.

Likewise, the pIIa-pIIb interface is polarized and characterized by an apical region marked by

Crumbs and Par6, subapical AJs and a basolateral domain (Benhra et al., 2011; Founounou et

al., 2013). It is noteworthy that furrowing is asymmetric during SOP ACDs, leading to an

apically positioned midbody at cytokinesis (subsection 3.1.3, Founounou et al., 2013).

Therefore, the pIIa-pIIb interface is characterized by a small apical domain located above the

midbody and a large basolateral area below the midbody.

On the receptor side, experimental data supporting Notch activation at the level of AJs

and/or the most-apical region mostly relies on the detection of endogenous Notch both in

fixed tissues (Benhra et al., 2011; Cotton et al., 2013) and in vivo (Couturier et al., 2012)

(Figure 36). Nonetheless, no functional assays were designed to test whether these receptors

contribute to signaling, thereby assuming that localization mirrors function.

Figure 36. Notch localizes at the AJ level at cytokinesis in fixed tissues (A) and in

vivo (B). Adapted from Benhra et al., 2010 and Couturier et al., 2012.


82

The array of evidence appears to be denser on the ligand side. Several studies

suggested that Delta trafficking, in light of the recycling model (subsection 1.3.3, Wang and

Struhl, 2004) towards the apical domain was reflecting the signaling site (Benhra et al., 2010;

Emery et al., 2005; Jafar-Nejad et al., 2005; Rajan et al., 2009). The seminal observation

reporting that Delta-containing Rab11 recycling endosomes were forming specifically in pIIb

at cytokinesis raised the possibility that Delta undergoes a recycling cycle prior to being

readdressed to the signaling site (Emery et al., 2005) (Figure 37A). However, careful analysis

of this paper rather suggests that this recycling event plays a minor role in tuning Notch

activity. First, authors indicated several times that in Rab11GFP expressing flies, formation of

large Rab11-positive compartments was detected only in a small subset of pIIa-pIIb pairs,

which were subsequently selected for the analysis of Delta trafficking. Thus, the importance

of this mechanism in SO lineages is uncertain. Second, whether inhibiting Delta recycling by

the expression a dominant-negative Rab11 protein affects cell fate acquisition is not reported.

Third, symmetrizing the formation of large Rab11 compartments in pIIa and pIIb by genetic

means only mildly perturb cell fate specification. Alternatively, even though these Rab11

endosomes still form in pIIb in absence of Neur, one might consider that they appeared

through evolution to handle the bulk of Neur-mediated Delta endocytosis in pIIb.

Despite these limitations, the Delta trafficking model is rapidly gaining acceptance. In

the same issue of Developmental Cell, the component of the exocyst Sec15 was reported to

mediate Notch signaling in the SO lineage (Jafar-Nejad et al., 2005). In sec15 cells,

trafficking of Notch, Delta and Spdo was defective and proteins accumulated at the basal side

of pIIa and pIIb. Nonetheless, the authors chose to only emphasize the role of Sec15 in Delta

trafficking towards the apical side of the pIIa-pIIb interface to match with the Rab11-

dependent recycling model (Figure 37B). Some data of this study instead argue that signaling

defects might be due to impaired exocyst-mediated budding from the Golgi apparatus

(Yeaman et al., 2001). Indeed, Spdo was shown to accumulate in LVA-positive

compartments. In such a scenario, signaling would be abolished due to sequestered Spdo

instead of mistrafficked Delta. Finally, as exocytosis is broadly impaired in sec15 cells, one

might hypothesize that pIIa-pIIb contact formation during furrow ingression is defective due

to insufficient membrane material deposition.


83

Figure 37. Evolution of the Delta recycling model in the SO lineage. See text for

details. Adapted from Emery et al., 2005, Jafar-Nejad et al., 2005 and Rajan et al.,

2009.


84

This model was further supported by a proposed connection between the Arp2/3

complex, WASp and Delta trafficking (Rajan et al., 2009). The authors found that Arp3, an

essential subunit of the Arp2/3 complex (section 3.2), is required for Notch activation in the

SO lineage. Together with WASp, an activator of Arp2/3 that also regulates Notch signaling

in this context (Ben-Yaacov et al., 2001), they appeared to regulate the formation of an actin

rich structure (ARS) at the pIIa-pIIb interface and microvilli covering the apical surfaces of

pIIa and pIIb (Figure 38). In parallel, they also mediated Delta recycling towards the apical

side of the ARS and/or microvilli (Figure 37C). Alternative hypotheses, including Delta

endocytosis or contact formation/expansion during cytokinesis, were not thoroughly tested in

this study. Collectively, these data extended the original recycling model where henceforth

Delta-containing vesicles are addressed towards the apical domain (AJs and/or microvilli) of

the pIIa-pIIb interface in a WASp-Arp2/3-Sec15-dependent manner. Whether the ARS is

necessary for signaling is not tested in this study. This model will be discussed more in-depth

in the last Chapter of this Introduction describing the relationship between the Arp2/3

complex and Notch signaling (subsection 3.3.2).

Figure 38. An actin rich structure (top panels) containing microvilli (bottom

panels) forms at the pIIa-pIIb interface. Adapted from Rajan et al., 2009.


85

The Delta recycling loop was at last closed when Neur-mediated endocytosis was

proposed to provide Rab11 endosomes in Delta molecules (Benhra et al., 2010). In this study,

Neur activity was found to trigger the endocytosis of Delta proteins that essentially localized

laterally. Later on, investigators made use of MDCK cells to test the capacity of Neur to drive

basal-to-apical transcytosis. It is noteworthy that MDCK cells are a model cell line derived

from canine kidney and display elevated endogenous transcytosis activity. Moreover, MDCK

cells used in this study were in interphase, while Delta trafficking occurs in pIIb in the

specific context of cytokinesis (Emery et al., 2005). Lastly, Neur or Delta homologs are not

expressed in this cell line. Therefore, one should be careful while transposing experimental

results between biological systems. Neur2 was found to trigger Dll1 basal-to-apical

transcytosis in MDCK cells and this finding was extrapolated to pIIb.

Taken together, these studies imply that Notch is activated either at the level of

subapical AJs or apical microvilli, thus in a reduced contact area. Importantly, they also

assume that Delta needs to be recycled prior to activating Notch. Recent work oppositely

contradicts such requirement (subsection 1.3.3).

The lateral contact

A much weaker, but not less important, body of evidence suggests on the contrary that Notch

receptors are activated by Delta on the lateral side of the pIIa-pIIb interface. Consistently,

Neur-mediated endocytosis of Delta occurs at this location (Benhra et al., 2010).

Whereas Notch is mainly detected apically in wild-type pIIa-pIIb pairs, receptors

appeared to accumulate laterally in mutant backgrounds impairing Notch trafficking

(Couturier et al., 2012). In numb cells, Notch lateral localization might reflect the

readdressing of receptors in pIIb towards the interface to drive ectopic signaling in pIIb

(Figure 39). In spdo cells, this ectopic lateral distribution might account for the accumulation

of unprocessed receptors, thereby correlating with abolished Notch signaling in spdo SO

lineages (Figure 39). However, alternative explanations can account for these phenotypes. In

both numb and spdo backgrounds, lateral detection of Notch might simply reflect an excess of

Notch molecules caused by impaired Numb-mediated degradation or Spdo-dependent

endocytosis of Notch prior to SOP ACD. More generally, this versatility in interpreting

protein distributions in multiple genetic backgrounds highlights the difficulty to draw strong

and unambiguous conclusions from this type of data. Thus, Notch activation site should be

determined without relying on genetic correlations.


86

Figure 39. Notch is detected laterally in vivo in numb and spdo backgrounds. See

text for interpretation. Adapted from Couturier et al., 2012.

Finally, as the lateral domain of the pIIa-pIIb interface is a large contacting surface,

this location might be particularly adapted to elicit high levels of Notch activity in pIIa, if

required.

Directionally trafficking Sara endosomes

An unexpected location for Notch activation emerged from the discovery and characterization

of directionally trafficking Sara (Smad anchor for receptor activation) endosomes during SOP

ACD (Coumailleau et al., 2009). Sara endosomes, which belong to the family of Rab5 early

endosomes, were found to incorporate Notch and Delta molecules prior to ACD, localize at

the central spindle during anaphase and be directed at late telophase in pIIa. Once inherited, it

is speculated that Notch receptors are activated at the endosomal surface in a Neur- and Delta-

dependent manner (Figure 40).


87

Figure 40. Directionally trafficking Sara endosomes carrying Notch and Delta

molecules are inherited by pIIa. Adapted from Coumailleau et al., 2009.

Although the molecular mechanism biasing Sara endosome inheritance in pIIa has

been resolved with great details during SOP ACD (Derivery et al., 2015; Loubéry et al.,

2017), some aspects of the model remain elusive and question its significance in fate

specification throughout the SO lineage. First, the mechanism by which Notch can be

activated at the endosomal surface by ligand endocytosis needs to be elucidated. Notch

cleavage was observed at the surface of enlarged endosomes expressing the dominant-

negative Rab5Q88L to help visualize NECD and NICD moieties (Coumailleau et al., 2009).

Whether Notch is efficiently processed at the surface of endogenous Sara endosomes is not

reported to date. Ligand-dependent cis-activation (Elowitz et al., 2018) might solve this

matter, yet it remains to be experimentally addressed. Second, in contrast with other

biological contexts (Kressmann et al., 2015; Montagne and Gonzalez-Gaitan, 2014), loss of

Sara does not affect fate decisions in the SO lineage (Coumailleau et al., 2009). Depleting

zygotically Sara enhances neur knock-down phenotype in the notum (Loubéry et al., 2017),

albeit the reason for this synergy in unclear. Indeed, impairing Sara endosomes could increase

the total pool of available Neur, since Notch signaling at the Sara endosomal surface is Neur

dependent, and thereby rescue the neur phenotype. Lastly, authors observed that the Sara-

positive Rab5Q88L compartment was randomly segregated in pIIa and pIIb and subsequently

made use of it to assess whether ectopic localization in pIIb drives a pIIb-to-pIIa fate change

(Coumailleau et al., 2009). This was indeed the case in 66% of pIIb cells, yet this phenotype

might be due to impaired Numb activity at the surface of Rab5 sorting endosomes (Couturier

et al., 2013). More importantly, when the Sara-Rab5Q88L enlarged endosome was found in

pIIb, the pIIa fate was normally specified. This indicates that Sara endosomes do not

contribute significantly to Notch signaling in pIIa.


88

Collectively, these remarks suggest that Sara endosomes are dispensable for fate

specification in the SO lineage but might play a significant role in other biological contexts

(Kressmann et al., 2015; Montagne and Gonzalez-Gaitan, 2014). Consequently, this implies

that the bulk of NICD in signal-receiving cells originate from the pIIa-pIIb interface.

2.3.5. Notch is activated in pIIa, but when?

Lineage progression and theoretical temporal requirement for Notch activation

As outlined above (subsection 1.4.2), Notch transcriptional response is orchestrated in time

and requires time-demanding relays mediated by E(spl) genes (Housden et al., 2013). Using

Drosophila muscle progenitors as an experimental model, authors showed that activated

Notch targets are divided into two categories: early- and late-responding genes. They also

uncovered that the expression of some late targets relied on a negative feedforward loop.

Considering that late-responding gene expression peaks approximately one hour after the

NICD stimulus, one can now appreciate that specifying fates is a time-demanding process and

might be highly time-constrained with regards to Drosophila developmental time.

The SO lineage clearly illustrates such constraint. As mentioned earlier (subsection

2.3.1), the division rate following first SOP division ranges from 2 to 3 hours. In particular,

after being produced, pIIb divides two hour later. As it is the single signaling source for pIIa,

this means that the pIIa fate must be specified within this time window. From the signaling

standpoint, this requirement questions whether Notch receptors need to be activated in pIIa

prior to a time limit to enable correct target gene expressions and fate acquisition.

Evidence for Notch activation at cytokinesis

Making use of the Nts allele, Remaud et al. determined the temporal requirement of Notch

activation in pIIa by switching flies from the restrictive to the permissive temperature at fixed

time points following SOP division (Remaud et al., 2008). First defects in fate specification

(approx. 10%) were observed when flies were shifted between 15 and 30 minutes after SOP

division. Adding an additional 15-minute delay to this shift (i.e., 30-45-minute total delay)

increased the proportion of abnormal lineages to approximately 35%. This percentage kept

increasing as Notch activation was set further from SOP division, till recovering only mutant

lineages with a 90-minute delay (Figure 41). This clearly demonstrates that Notch receptors

need to be activated at cytokinesis to ensure pIIa fate specification in a reproducible manner.

Authors later attempted to link this result with a competence window related to cell cycle


89

progression. Evidence supporting this hypothesis were based on the observation that pIIb cells

were more responsive to ectopic Notch signaling during S phase than G2 phase. However,

how an ectopic NICD input relates to endogenous signaling experienced by pIIa is not clear.

Second, as the S-G2 transition occurs ~75 minutes after SOP division, this hypothesis does

not account for fate defects observed with early shifts in the Nts experiment. A more

parsimonious interpretation would assume that delayed Notch activation prevents the

expression in pIIa of the whole set of genes required for fate specification.

Figure 41. Notch activation at cytokinesis is required to specify the pIIa fate. See

text for details. pI corresponds to SOP. Adapted from Remaud et al., 2008.

Live-imaging of GFP-tagged Notch firmly demonstrated that Notch receptors are

indeed activated at cytokinesis in pIIa (Couturier et al., 2012). The GFP being inserted on

Notch intracellular tail, investigators could monitor in vivo both the full-length receptor and

the cleaved NICD fragment following receptor activation (Figure 42). At cytokinesis, GFP

fluorescence was shown to accumulate in pIIa nucleus once furrowing was completed.

Figure 42. Notch is activated at cytokinesis. Adapted from Couturier et al., 2012.


90

Therefore, as soon as a contact surface is established between the SOP daughter cells,

signaling starts. When linked to the requirement of Notch activation at cytokinesis (Remaud

et al., 2008), this observation suggests that the interplay between SOP cytokinesis and Notch

activation is a key mechanism to ensure fate specification prior to the next round of ACDs.

How the two are related will be discussed in the next introductory chapter and experimentally

tested in the Chapter 5 of the Results section.

91

Chapter 3

The Arp2/3 complex, a molecular pivot bridging cytokinesis and Notch activation?

The requirement of Notch activation at SOP cytokinesis to ensure pIIa fate specification

questions the interplay between the mechanism driving cytokinesis and the core mechanism

mediating Notch processing. Given that Arp3 is required for Notch signaling in the SO

lineage and that the Arp2/3 complex regulates contact expansion during cytokinesis,

investigating the role of Arp3 in the context of SOP cytokinesis might reveal the nature of this

interplay.

This chapter is divided into three sections. The first one succinctly describes the

molecular mechanisms at stake during cytokinesis occurring in an epithelium. The second

section introduces the Arp2/3 complex, its molecular activity and its role in cytokinesis. The

last section discusses the nature of Arp2/3 requirement in the SO lineage.

3.1. Dividing in an epithelium

Cytokinesis can be defined as the last step of cell division where different cellular materials

(chromosomes, organelles, cytoplasmic content) are segregated into the two daughter cells. Its

molecular mechanisms have been intensively studied in isolated cell systems, including the

budding yeast or mammalian cell lines. However, dividing in an epithelium presents

additional challenges for a cell: not only it has to generate forces to split itself into two

daughter cells but also to preserve the barrier function of the epithelium by generating new

adhesive contacts and, not the least, to endure the pressure exerted by the neighboring cells

while dividing. On top of that, SOPs need to generate a contact surface that sustain Notch

signaling between pIIa and pIIb.

This section briefly describes the different steps occurring during cytokinesis and later

focuses on the specificities found in epidermal cell or SOP cytokineses of the fly notum.

These latter subsections are linked with SOP ACD to present how cytokinesis can influence

Chapter 3. Arp2/3, Notch activation and cytokinesis

92

Notch signaling in the pIIa-pIIb fate decision. A more detailed and integrated view of

cytokinesis can be found in Glotzer, 2017.

3.1.1. Overview of cytokinesis progression

Anaphase onset and cytokinetic ring positioning

Cytokinesis starts upon metaphase-to-anaphase transition when cyclin B is degraded and

Cdk1 inactivated after activation of the anaphase promoting complex/cyclosome (APC/C)

(Glotzer, 2017). This switch in kinase activity at anaphase onset enables the activation of

several cytokinetic regulators, including the centralspindlin complex (Figure 43). This

complex then relocalizes at the spindle midzone and promotes cortical recruitment and

activation of the RhoGEF (guanine exchanging factore) Ect2/Pbl (Pebble) at the division

plane (Lekomtsev et al., 2012). Ect2/Pbl in turn recruits and activates the GTPase RhoA to

trigger the assembly of the cytokinetic ring (Basant and Glotzer, 2018).

Of note, although the requirement of the cytokinetic ring is valid in many cell types,

alternative mechanisms through polarized cell migration enable cytokinesis independently of

the formation of an actomyosin ring (Dix et al., 2018; Neujahr et al., 1997).

Figure 43. Cytokinetic ring positioning at anaphase onset. Adapted from Basant

and Glotzer, 2018.


93

Cytokinetic ring assembly and contraction

Once activated, RhoA orchestrates the recruitment of the key components forming the

cytokinetic ring, including the MyoII kinase ROCK, formins mediating filamentous actin

polymerization and the mulitfunctional scaffold protein anilin (Piekny and Maddox, 2010;

Thieleke-Matos et al., 2017). RhoA activity leads to the formation of bipolar MyoII

minifilaments and the additional recruitment of tropomyosin, a-actinin and septins (Glotzer,

2017). Anchoring of the actomyosin ring to the membrane is mediated by RacGaps, FERM

(4.1 protein – ezrin – radixin – moesin)-domain containing proteins, anilins and F-BAR

proteins, depending on the context (Figure 44).

Figure 44. Overview of the molecular mechanisms at stake during anaphase

elongation. (a-b) Metaphase-anaphase transition and anaphase elongation. (c)

Anchoring of the cytokinetic ring to the plasma membrane. (d) Polar relaxation.

See text for details. Adapted from Ramkumar and Baum, 2016.

Contractility that drives ring closure is mediated by RhoA-activated Rho- and citron-

kinases whose activity regulate the MyoII-dependent sliding of actins filaments. This induces

membrane furrowing at the division plane. Progression of ring closure till the formation of the


94

midbody depends on the cycling between active and inactive forms of RhoA (Basant and

Glotzer, 2018).

Several cellular mechanisms that also shape cellular architecture accompany ring

constriction. Anaphase cell elongation is dependent on forces generated during mitotic spindle

elongation (Scholey et al., 2016), albeit the specific molecular mechanisms at stake in an

epithelium are not known. Polar relaxation through kinetochore-localized PP1-Sds22

signaling facilitates anaphase elongation and couples chromosome segregation to anaphase

progression (Rodrigues et al., 2015) (Figure 44). At the cytokinetic furrow, coordination of

constriction progression with vesicle exocytosis to deliver membrane material at newly-

forming membranes is key for cytokinesis completion (Frémont and Echard, 2018; Neto et al.,

2011; Schechtman, 1937). In the remainder cortex, Rac1-dependent Arp2/3 activity is thought

to counteract forces exerted by the contractile ring (Loria et al., 2012).

Midbody formation and abscission

As the cytokinetic ring closes, a narrow intracellular bridge filled with antiparallel

microtubule filaments is left between the two daughter cells. Following cytokinetic ring

maturation, a circular structure termed the midbody forms within the bridge and is actively

remodeled through exocytosis (Frémont and Echard, 2018). In parallel, microtubule filaments

are depolymerized. Finally, the ESCRT-III complex is recruited by the midbody to drive

membrane scission and resolve the daughter cells’ membranes (Frémont and Echard, 2018;

Glotzer, 2017) (Figure 45).

Figure 45. Midbody formation and abscission. Adapted from Glotzer, 2017.


95

3.1.2. Epithelial cytokinesis as a multicellular process

Non-autonomous MyoII activity in neighboring cells

Neighboring cells in the first place can be considered as a negative constraint from the view

point of the dividing cell. Firstly, during anaphase elongation, the dividing cell must at the

same time push on the neighbors located at the poles and resist to the pressure they exert in

return on the cell body. Second, neighboring cells can directly induce a tension via adherens

junctions (AJs) on the constricting cytokinetic ring (Founounou et al., 2013).

Figure 46. MyoII (red) activity in neighbors helps in juxtaposing the dividing cell

membranes during cytokinetic ring constriction. Adapted from Herszterg et al.,

2013.

Recent work in the fly notum challenged this view and provided evidence that MyoII

activity in immediate neighbors to the newly-forming AJ cohesive contact is necessary to set

proper contact geometry (Founounou et al., 2013; Herszterg et al., 2013) (Figure 46).

Constriction of the cytokinetic ring exerts a pulling force on the neighboring cells and drives

local recruitment and activation of MyoII in Rho-kinase dependent manner at the level of the

AJ belt. This localized activity facilitates the apposition of the invaginating membrane and

controls the length of the future AJ interface (Herszterg et al., 2013). These findings were

recently augmented by a study resolving the molecular mechanisms that trigger this local

MyoII activity (Pinheiro et al., 2017). Pulling forces exerted by the ring elongate the AJs


96

where the dividing membrane ingress, thereby diluting locally AJ concentration. This

promotes on the side of the neighboring cell self-organized actomyosin flows that ultimately

lead to the recruitment of MyoII at the basis of the invagination. Thereby, cytokinesis

progression is coordinated with AJ remodeling and formation.

This mechanism resolved in the notum is exclusively based on dividing epidermal

cells. Therefore, whether neighboring cells contribute to set the geometry of the AJ interface

between pIIa and pIIb is not known. Assuming that signaling occurs at this level (Benhra et

al., 2010, 2011; Rajan et al., 2009), one could hypothesize that such contribution would be

crucial to promote the formation of elongated AJs, thereby increasing the contact surface to

sustain high Notch signaling levels (Shaya et al., 2017). A similar hypothesis could be

formulated regarding the lateral side of the pIIa-pIIb interface, although this model is based

on the analysis of the apical side of epidermal cells where AJs are mostly detected and

organized within a belt. Whether MyoII activity in neighbors is also required to properly set

the geometry of the lateral contact, which differs by polarity protein composition and AJ

organization from the apical adhesive belt, between daughter epidermal cells or pIIa and pIIb

is not known.

Autonomous contact expansion

In vivo analysis of epidermal cell cytokinesis in the notum also uncovered a cell-autonomous

mechanism that promotes contact expansion during cytokinetic ring constriction (Herszterg et

al., 2013). Activities of the small GTPase Rac1 and the Arp2/3 complex in the dividing cell

are required for AJ elongation, and in some cases for the formation of the adhesive contact, by

promoting a wave of actin polymerization above and below the apically positioned midbody

(see next subsection). This actin bulge has been proposed to draw newly-formed membranes

closer to allow AJ formation and consequently to push away the intervening membranes of

the neighboring cells (Herszterg et al., 2013, 2014) (Figure 47). Therefore, Rac1 and Arp2/3

activities also regulate the contact geometry between daughter cells and could influence

Notch signaling levels in pIIa in a quantitative manner (Shaya et al., 2017). As Arp3 is

required for Notch activation in pIIa (Rajan et al., 2009), this hypothesis will be discussed

with greater details in the last section of this chapter.


97

Figure 47. Cell-autonomous Rac-dependent Arp2/3 activity ensures neighbor

membrane withdrawal and contact expansion. Adapted from Herszterg et al.,

2014.

3.1.3. Asymmetric furrow ingression

Apical midbody positioning

Epidermal cell and SOP cytokineses in the notum are characterized by an apical positioning

of the midbody due to asymmetric furrow ingression (Founounou et al., 2013; Herszterg et al.,

2013). The basal initiation of furrowing does not seem to respond to extrinsic cues in this

context and appears to rely only on the initial basal bias in MyoII distribution along the

cytokinetic ring (Founounou et al., 2013). After ring closure, the midbody is shifted apically

and localizes just below the AJ belt (Figure 48).

Figure 48. Asymmetric furrowing in the notum is caused by asymmetric MyoII

distribution. Adapted from Herszterg et al., 2013 and Herszterg et al., 2014.


98

Recent work on abscission during epidermal cell cytokinesis in the notum provided

evidence that neighboring cells stay attached to the midbody following telophase and thereby

generate a physical discontinuity in the contact surface separating the daughter cells (Daniel

et al., 2018; Wang et al., 2018) (Figure 49). Thus, asymmetric furrowing generates two

contact surfaces, above and below the midbody, that are physically separated. Moreover,

these two contact surfaces are composed by different polarity proteins, including Crumbs, E-

Cad above and basolateral markers below the midbody.

Figure 49. Two physically separated contact surfaces above and below the

midbody at cytokinesis. Adapted from Daniel et al., 2018.

Relevance in Notch signaling

Asymmetric furrowing observed during SOP cytokinesis can influence Notch signaling in

several manners. First, by generating two contact surfaces with different areas, it might dictate

where Notch activation site will be located. Indeed, as the lateral contact (i.e. below the

midbody) appears to be much larger than the apical one (i.e. above the midbody), one could

assume that signaling predominantly occurs where the contact area is the largest, that is to say

laterally (Shaya et al., 2017). Therefore, mechanisms setting the geometry of the lateral

contact might be particularly relevant regarding Notch activity levels in this context.

Second, the two contact surfaces present also different properties in terms of protein

composition and cortical cytoskeleton organization. The apical contact surface is

characterized by apical polarity proteins, such as Crumbs, and AJs coupled to an actomyosin


99

belt while the lateral one displays basolateral markers, including Dlg and NrxIV (NeurexinIV)

(Benhra et al., 2011; Founounou et al., 2013) (Figure 50). In theory, activation of Notch

receptors might depend on these different membrane contexts. Experimentally, E-Cad and

Crumbs have been shown to directly regulate Notch activity in Drosophila imaginal wing

discs (Nemetschke and Knust, 2016; Sasaki et al., 2007) Thus, specific and local functional

requirements might favor one contact surface against the other.

Figure 50. The pIIa-pIIb interface is polarized along the apical-basal axis.

Adapted from Le Bras et al., 2011.

Third, asymmetric furrowing might also fulfill the requirement of Notch activation

following telophase. As the lateral contact is the first to be generated, Notch receptors located

laterally could be activated there in the first place.

Beyond which hypothesis is valid among the three proposed here, this subsection also

emphasizes the importance of knowing where Notch is activated. It would help in orientating

further investigations to understand how local environments can influence Notch activity and

how basic cellular processes, such as cytokinesis, are coordinated with cell fate decisions that

are dependent on Notch signaling.


100

3.2. Arp2/3 role during cytokinesis

Work in flies and cultured cells revealed that the branched actin network initiated by the

activity of the Arp2/3 complex is recurrently used to expand the contact area between two

contacting cells, and more specifically during cytokinesis. As the subunit Arp3 is required for

Notch activity in the SO lineage, it becomes tempting to speculate that Arp2/3 links Notch

signaling to cytokinesis by generating large contact surfaces between daughter cells.

However, Arp2/3 regulates many other cellular processes that might also affect Notch

signaling.

This section aims at introducing the Arp2/3 complex, its molecular function and the

molecular mechanisms leading to its activation. Lastly, its role in contact expansion will be

outlined and connected with the small GTPase Rac1.

3.2.1. The Arp2/3 complex, a major actin regulator

Molecular function

The Arp2/3 complex is an actin nucleator conserved in almost all eukaryotic cells (Rotty et

al., 2013). It is composed of seven subunits: the two actin-related proteins Arp2 and Arp3 and

the five additional structural components ARPC1-5. The Arp2/3 complex nucleates actin

filaments in a unique fashion by branching daughter actin filaments to mother filaments with

a characteristic angle of 70° (Pollard, 2007) (Figure 51). By doing so, it generates a branched

actin network that regulates a plethora of cellular processes (see below).

Figure 51. Structure of the Apr2/3 complex under its active or inactive form. See

text for details. Adapted from Goley and Welch, 2006.


101

Extensive work on the structure of the Arp2/3 complex and the actin branch it forms

uncovered the sequence by which the daughter filament is initiated and branched to the

mother filament (Gournier et al., 2001; Pizarro-Cerdá et al., 2017; Pollard, 2007; Robinson et

al., 2001; Rouiller et al., 2008; Volkmann et al., 2001). Upon activation by a nucleating

promoting factor (NPF, see subsection below), the Arp2/3 complex undergoes a substantial

conformational change. This allows the interaction of the subunits ARPC2 and ARPC4 to the

mother filament together with the initiation of the daughter filament whose first subunits of

the pointed end are made of the subunits Arp2 and Arp3 (Figure 51). The daughter filament

further extends from the branch through barbed-end polymerization until abortion by capping

proteins (Michelot et al., 2013). The role of the other subunits during this process is not firmly

established. ARPC3 is proposed to bridge Arp3 to the mother filament, although its loss has

only minor consequences on the complex activity (Pizarro-Cerdá et al., 2017). The N-

terminus of ARPC5 is hypothesized to tether Arp2 to the rest of the complex (Pollard, 2007).

The exact role of ARPC1 remains unclear (Pizarro-Cerdá et al., 2017).

ATP binding to and hydrolysis by Arp2 and Arp3 appear also to play a key role in

Arp2/3 activation. Nucleotide binding to Arp2 and Arp3 drives a small conformational

change that correlates with increased affinity for NPFs (Espinoza-Sanchez et al., 2018;

Pollard, 2007). The role of hydrolysis is less clear. Although it is necessary for daughter

filament elongation, it is currently not known whether it is related to actin nucleation or

dissociation of branches (Pollard, 2007).

Finally, recent studies revised the widely accepted view of a single canonical seven-

subunit Arp2/3 complex and provided support for a diverse array of functionally distinct

complexes (Pizarro-Cerdá et al., 2017). First, Arp2/3 complexes can present “exotic”

compositions in focal adhesions of chicken smooth muscles with Arp2, Arp3 and ARPC1

associated to a-actinin and vinculin, or to ARPC3 and vinculin (Chorev et al., 2014). Second,

several subunits of the complex present different isoforms in many species, raising the

possibility of isoform-specific complexes (Pizarro-Cerdá et al., 2017). Not only such

complexes were found but were also shown to elicit different cellular functions (Kühbacher et

al., 2015) and to present different efficiencies in actin nucleation (Abella et al., 2016). Hence,

context specific composition of the Arp2/3 complex might present a potent mechanism

regulating Arp2/3 activity and the associated cellular output (Figure 52).


102

Figure 52. Diversity of Arp2/3 complexes found in cells. Adapted from Pizarro-

Cerdá et al., 2017.

Cellular function

As most cytoskeleton regulators, the Arp2/3 complex is implicated in a wide range of cellular

processes that can be found from the nucleus to the cell surface. A short and non-exhaustive

list is given here to apprehend how often the branched actin network is reutilized in cells

(Figure 53).

Figure 53. Overview of cellular processes regulated by the Arp2/3 complex. Its

nuclear function is not represented. Adapted from Rotty et al., 2013.


103

In nuclei, Arp2/3 plays a critical role in repairing DNA double-strand breaks (Caridi et

al., 2018; Schrank et al., 2018). It is required in hetero- and euchromatic regions for the

migration of double-strand breaks to repair sites located at the nuclear periphery for

homology-directed repair. In the cytosol, Arp2/3 regulates the endosomal pathway at several

steps, including endocytosis, segregation of surface subdomains, formation and fission of

vesicles budding from recycling endosomes and endosomal motility (Simonetti and Cullen,

2019). At the cell surface, it can either maintain cell shapes by stabilizing adherens junctions

or deform them by inducing the formation of lamellipodia, filopodia or podosomes, thereby

conferring motility to the cells (Rotty et al., 2013; Swaney and Li, 2016).

Therefore, the Arp2/3 complex can elicit functions spanning from the molecular to the

cellular scale. Such pleiotropy implicates context specific regulators that help coordinating all

these different processes within a single cell. These specific regulators are described in the

next subsection.

3.2.2. Regulation of the Arp2/3 complex by NPFs and inhibitors

Mechanisms of activation

The idea that the Arp2/3 complex requires additional activators to mediate actin nucleation

arose from the observation that purified Arp2/3 complexes present a weak nucleating activity

on their own (Welch et al., 1998). On the contrary, when the surface protein ActA of Lysteria

monocytogenes was added to the medium (a protein required for the Arp2/3-dependent

motility of L. monocytogenes in infected cells), the nucleating activity was strongly enhanced

(Welch et al., 1998). Since this seminal discovery, numerous endogenous activators termed

nucleating promoting factors (NPFs) were characterized and can be split into two categories

(Figure 54).

Figure 54. Summary of Arp2/3 interactors. Adapted from Rotty et al., 2013.


104

Type I NPFs are defined by a WCA domain (WASp-homology 2 domain/ Verpolin-

homology domain, Cofilin-homology domain/Central domain, Acidic domain). The CA motif

binds to Arp2/3 while the W motif to monomeric actin (Figure 55). This category is

constituted by WASp (Wiskott Aldrich Syndrome protein), N-WASp (neural WASp),

SCAR/WAVE (suppressor of cyclic AMP repressor/WASp-family verpolin-homologous

protein), WASH (WASp and SCAR homologue), WHAMM (WASp homologue associated

actin, membranes and microtubules) and JMY (junction-mediating and regulatory protein).

Type II NPFs lack the WCA domain but possess acidic regions at their N-terminus binding to

Arp2/3 and binding domains to F-actin (Figure 55). They include cortactin and the

hematopoietic cortactin-like HS1 (Rotty et al., 2013). They are weak activators, and act by

stabilizing branches. Nonetheless, they can enhance Arp2/3 activity in synergy with a type I

NPF (Helgeson and Nolen, 2013).

Figure 55. Domain structure of NPFs. See text for details. Adapted from

Campellone and Welch, 2010.

Activation of Arp2/3 by type I NPFs involves the interaction of the CA region with

various subunits of the complex, thereby driving the conformational change required for

nucleation (Campellone and Welch, 2010; Espinoza-Sanchez et al., 2018). The WH2 transfers

G-actin to the barbed ends of Arp2 and Arp3 to initiate daughter filament elongation.

Following the first polymerization round, type I NPFs presumably detach from the primed

Arp2/3 complex and drives additional elongation cycles (Campellone and Welch, 2010).

Functional specificity and activation of NPFs

Although all type I NFPs share a C-terminal WCA domain, they all differ regarding their N-

termini. This sequence specificity confers to each NPF a proper regulatory mode which

translates into the regulation of specific processes. These different modes will be succinctly

presented for each type NPF and are summarized in Figure 56. Noteworthy, NPFs also have

cellular functions unrelated to Arp2/3 activation. Details can be found in Tyler et al., 2016.


105

Figure 56. Activation modes of type I NPFs. See text for detailed description.

Adapted from Campellone and Welch, 2010.

WASp and N-WASp

WASp and N-WASp N-termini are defined by a GTPase binding domain (GBD) that

maintains the NPF in an auto-inhibited state by intramolecular interaction with the WCA

region and a proline-rich domain (PRD) adjacent to the WCA region implicated in

intermolecular interactions (Campellone and Welch, 2010). Of note, WASp is specific to the

hematopoietic lineage in mammals while N-WASp is found in most cell types, whereas there

is only one WASp homolog in flies. The rest of the paragraph will focus on N-WASp

activation, which presumably resembles the one of Drosophila WASp.

Alleviation of N-WASp auto-inhibition was mainly found to be dependent on the

membrane-tethered small GTPase Cdc42 (Kim et al., 2000). More specifically, Cdc42-GTP

appeared to bind with either active or inactive N-WASp with higher affinity than Cdc42-GDP

(Leung and Rosen, 2005). Additionally, binding of a GBD subdomain to membrane PI(4,5)P2

enhances N-WASp-mediated activation of Arp2/3 (Tomasevic et al., 2007). Alternatively, N-

WASp can be stimulated by the SH3-containg proteins Nck1 and Nck2 (non-catalytic kinases


106

1-2) (Tomasevic et al., 2007), the F-BAR protein Cip4/Toca-1 (Fricke et al., 2009; Ho et al.,

2004; Takano et al., 2008) and an interactor of the kinase Abl, Abi1 (Innocenti et al., 2005).

Finally, tyrosine kinases can also stabilize active N-WASp through phosphorylation of the

GBD (Torres and Rosen, 2006) (Figure 56).

A last layer of N-WASp regulation is found with the prolin-rich protein WIP (WASp

interacting protein). WIP binds to N-WASp as well as to G- and F-actin. Intriguingly, while it

stabilizes the auto-inhibited state of N-WASp in vitro, it enhances N-WASp activity in vivo to

promote actin polymerization (Campellone and Welch, 2010). Recent studies revealed that

WIP rather acts as a scaffold protein required for the recruitment of context specific effectors,

including type I myosin and Nck (Donnelly et al., 2013; Lewellyn et al., 2015; Sun et al.,

2017).

Integration of these various inputs by N-WASp drives numerous Arp2/3 mediated

processes occurring throughout the cell, including membrane ruffling, formation of

lamellipodia and filopodia, membrane invagination during endocytosis, endosomal motility

(Campellone and Welch, 2010), DNA-double strand repair (Caridi et al., 2018; Schrank et al.,

2018). In flies, WASp function was studied with less details but was notably found to regulate

myoblast fusion (Massarwa et al., 2007; Schäfer et al., 2007), Notch activation in neural

lineages (Ben-Yaacov et al., 2001), exocytosis in larval salivary glands (Tran et al., 2015) and

E-Cad endocytosis in the fly notum (Georgiou et al., 2008; Leibfried et al., 2008).

WAVE

In contrast to WASp proteins, WAVEs (3 homologs in mammals, SCAR in flies) do not

possess a GBD and are constitutively active when isolated in vitro. Their regulation arises

from additional proteins interacting with its Scar-homology domain (SHD), including Brk1,

Abi1, Nap1 (Nck-associated protein 1) and Sra1 (specifically Rac-associated-1) (Campellone

and Welch, 2010). When these WAVE regulatory complexes (WRCs) are reconstituted in

vitro, WAVEs no longer display any activity (Ismail et al., 2009). The recent resolution of the

WRC structure provided a structural basis for WAVE inhibition within the complex (Chen et

al., 2010). It revealed that interactions of the WCA domain and a “meander region” of WAVE

with Sra1 buries the actin- and Arp2/3-binding sites into a hydrophobic pocket.

Similarly to WASp proteins, WAVEs can also be activated by a small membrane-

tethered GTPase, Rac1 (Eden et al., 2002), acidic phospholipids (Oikawa et al., 2004) and

phosphorylation (Sossey-Alaoui et al., 2007), presumably altogether in a cooperative manner

(Lebensohn and Kirschner, 2009) (Figure 56). The structural studies mentioned in the former


107

paragraph also uncovered how Rac1-binding and phosphorylation release the WCA region of

WAVE from the hydrophobic pocket. Rac1 competes with the WAVE meander region to

interact with Sra1 and, once bound, is thought to induce a conformational change in the

meander region and alleviate by an allosteric switch the sequestration of the WCA domain.

Likewise, phosphorylation is hypothesized to disturb the interaction between Sra1 and the

meander region as conserved tyrosines are found in the associated contacting surface. More

recently, a second activating binding site of Rac1 on Sra1 was found opposite to the

interacting surface with the meander and appeared to be required for WRC activation (Chen et

al., 2017). Nonetheless, the mechanism by which the two binding sites cooperate to drive the

allosteric change is not elucidated. Author speculated that this dual requirement for Rac1

binding might underlie a regulatory mechanism sensing Rac1 concentration at the membrane.

WAVEs are predominantly involved in processes affecting cell shapes. WAVE-

dependent Arp2/3 activation stimulates formation of lamellipodia and protrusions at the

leading edge of migratory cells (Alekhina et al., 2017). In flies, WAVE is also critical to

mediate lamellipodium formation (Gautier et al., 2011), organize protrusions (Evans et al.,

2013; Georgiou and Baum, 2010), drive myoblast fusion (Richardson et al., 2007) and

regulate cell shape during morphogenesis (Del Signore et al., 2018; Zallen et al., 2002).

WASH

In many aspects, WASH resembles WAVE regarding its structure and its regulation. WASH

proteins neither possess a GBD and are maintained inhibited within a large complex

composed by FAM21, SWIP (strumpellin and WASH interacting protein), strumpelin and

CCDC53 (coiled-coil domain containing protein 53) (Derivery et al., 2009; Gomez and

Billadeau, 2009). Both WRC and WAVE complexes display similar morphological

characteristics, including size and shape, although protein composition differ and amino-acid

sequences are not related (Jia et al., 2010). It was proposed that WASH and WRC complexes

are ancient regulators of the Arp2/3 complex that diverged along evolution and specialized in

a subset of cellular processes (Rotty et al., 2013).

WASH complexes are activated by different mechanisms. While RhoA has been

shown to promote WASH activation in flies (Liu et al., 2009) (Figure 56), this was not the

case in mammals (Jia et al., 2010). Instead, activation of the WASH complex was found to be

dependent on WASH polyubiquitination mediated by the E3 ligase MAGE-L2-TRIM27 (Hao

et al., 2013).


108

WASH complexes are involved in a restricted and specific array of cellular processes.

They promote Arp2/3-mediated actin polymerization at the vicinity of endosomes to drive

endosomal motility, vesicular fission and maintenance of the maturing endosome shape

(Alekhina et al., 2017). Surprisingly, WASH is dispensable for fly development and adult life

(Nagel et al., 2017), raising the question of its redundancy with other NPFs in this model

organism.

WHAMM and JMY

WHAMM and JMY are type I NPFs specifically found in mammals. WHAMM, like WAVE,

is active when purified in vitro. To date, the proteins composing its anticipated inhibitory

complex have not been identified (Figure 56). WHAMM proteins carry a coiled-coil (CC)

domain that interacts with microtubules protofilaments (Shen et al., 2012). Intriguingly, this

binding hides its WCA region and prevents activation of Arp2/3 (Liu et al., 2017). Thereby,

WHAMM might play a pivotal role between the microtubule and the actin cytoskeleton in

specific cellular processes (Campellone et al., 2008). JMY sequence harbors in addition of its

WCA region two WH2 domains. This additional WH2 tandem can promote polymerization of

unbranched actin filaments (Coutts et al., 2009; Zuchero et al., 2009). Moreover, binding of

G-actin to the WH2 tandem masks a bipartite nuclear localization signal (NLS) which, when

uncovered, relocalizes JMY to the nucleus to regulate p53-dependent gene expression

(Zuchero et al., 2012). Mechanisms regulating JMY activity are to date not established.

WHAMM and JMY elicit unique functions among the NPFs. WHAMM is found in

the Golgi network and in the ER where it regulates structure maintenance and vesicular

transport (Campellone et al., 2008). Together with JMY, they both play a central role in

autophagosome biogenesis (Coutts and La Thangue, 2015; Kast et al., 2015).

Mechanisms of inhibition

In contrast with NPF-mediated activation, the Arp2/3 complex can also be directly inhibited

(Figure 54). Several negative inhibitors were identified over the past years and all present

specific mechanisms. Cofilin and its homolog GMF act by debranching the Arp2/3 complex

from the mother filament (Blanchoin et al., 2000; Gandhi et al., 2010). Coronins stabilize the

open and inactive state of the Arp2/3 complex (Cai et al., 2005, 2008; Sokolova et al., 2017).

Gadkin sequesters Arp2/3 complexes at endosomal surfaces (Maritzen et al., 2012). PICK1

inhibits Arp2/3-mediated actin polymerization via an unknown mechanism (Rocca et al.,


109

2008). Arpin, similar to coronins, stabilizes the inactive conformation of Arp2/3 (Dang et al.,

2013; Sokolova et al., 2017).

Recent structural analysis of the Arp2/3-inhibitor interaction showed that Arpin,

Cororin and GMF all bind to Arp2/3 at distinct sites, suggesting that the Arp2/3 activity might

be regulated in a context specific manner by combinations of inhibitory ligands (Sokolova et

al., 2017).

3.2.3. Contact expansion regulated by the Arp2/3 complex: the “zipper”

mechanism

The branched actin network polymerized by Arp2/3 complexes can be viewed as the slider of

a zipper that draws contacting membranes close enough to enable junction formation (i.e. the

zipper’s teeth) and contact stabilization (Cavey and Lecuit, 2009) (Figure 57).

Figure 57. Arp2/3-mediated contact expansion as a zipper-like mechanism.

Adapted from Herszterg et al., 2014.

In several studies, branched actin filaments were found in lamellipodium-like

structures at the edges of spreading cells on substrates or at locations where contact was

expanding between contacting cells (Ehrlich et al., 2002; Helwani et al., 2004; Kovacs et al.,

2002; Yamada and Nelson, 2007). Cell-cell contacts were further stabilized by formation of

homophilic E-Cad junctions. In addition, the small GTPase Rac1 was detected at the leading

edges of expanding contacts and regularly correlated with Arp2/3 activity (Ehrlich et al.,

2002; Engl et al., 2014; Yamada and Nelson, 2007). As contact expands, Arp2/3 and Rac1

activities are further inhibited and restricted to the edges by E-Cad-recruited a-catenin

(Benjamin et al., 2010; Hansen et al., 2013) and E-Cad-stimulated RhoGAPs (Kitt and

Nelson, 2011), respectively.


110

In flies, contact expansion is also regulated by Rac1 and/or Arp2/3 during epidermal

cell cytokinesis (Herszterg et al., 2013), retina morphogenesis (Del Signore et al., 2018) and

cellularization during embryogenesis (Stevenson et al., 2002; Zhang et al., 2018). Thus, the

Rac1-Arp2/3 pathway appears to promote contact expansion in many different contexts by

providing pushing forces at the expanding edges.

Although SCAR/WAVE requirement has been tested in only one of these studies (Del

Signore et al., 2018), one could assume based on SCAR/WAVE is the intermediate that

integrates Rac1 activity to promote Arp2/3-dependent actin polymerization.

3.3. Arp2/3 role in Notch-dependent fate decisions

Arp2/3 and WASp were found to be required for Notch signaling in a subset of cellular

processes during fly development, and in particular in the SO lineage. A model based on the

regulation of Delta trafficking by Arp2/3 and WASp has emerged, although several

observations contradict this view and indicate that their function in Notch signaling should be

reinvestigated.

This section presents the original experimental data linking Arp2/3 and WASp to

Notch signaling, discusses the currently accepted model and proposes alternative models.

Thereby, this section also serves as an analytical introduction to the Chapter 5 of the Results

section.

3.3.1. Arp2/3 and WASp requirement in Notch signaling

Arp3 and WASp in the SO lineage

WASp requirement in fate specification during SO lineage progression was initially found in

the seminal paper characterizing the WASp fly homolog (Ben-Yaacov et al., 2001). Authors

first reported that mutant flies where zygotic WASp function was lost developed until the

pharate stage but did not hatch from the pupal case. Careful observation of these WASp

pharates uncovered a deficit in microchaete and macrochaete numbers in many epidermises,

including the head, the abdomen, the notum and the eye (Figure 58). Further analysis of the

SO lineages revealed that lateral inhibition was not affected in WASp flies, whereas cell fate

changes reminiscent of impaired Notch signaling were found following SOP and pIIIb ACDs.

3.3. Arp2/3 role in Notch-dependent processes

111

Although the molecular function of WASp was not determined in this study, authors provided

evidence that WASp was not required for correct Numb and Pon segregation in pIIb.

Additionally, they showed that WASp is epistatic to Numb, reinforcing the idea that WASp

plays a central role in Notch activation in this context. Noteworthy, another epistasis

experiment highlighted an intriguing relationship between WASp and the Notch antagonist

Hairless (H). Removal of one copy of H in a WASp-deficient background was sufficient to

promote fate restoration in a large number of interommatidial SO lineages (Figure 59). One

interpretation might be that loss of WASp does not abolish Notch signaling in the SO lineage

but rather diminishes signaling activity to levels where active NICD-Su(H) complexes are too

few to outcompete H-Su(H) repressor complexes in the signaling receiving cell nucleus. A

following mutational analysis confirmed that WASp activity in the SO lineage required its

CA domain, thereby setting the bridge with the Arp2/3 complex (Tal et al., 2002). Notably,

this study provided the initial evidence that the Arp2/3 complex is required for cell fate

acquisition in the SO lineage by reporting bristle losses in arpc1 mutant heads.

Figure 58. WASp is required for correct bristle development in multiple

epidermises (A-B, head; C-D, abdomen; E-F, thorax; G-H, eye). Adapted from

Ben-Yaacov et al., 2001.


112

Figure 59. Decreasing Hairless activity is sufficient to rescue the WASp

phenotype. (D) Control eye. (E) WASp eye. (F) WASp Hairless+/- eye. Note the

rescue of the bristle phenotype. Adapted from Ben-Yaacov et al., 2001.

Arp3 was isolated in a genetic screen designed to identify novel regulators of the

Notch pathway (Rajan et al., 2009). This screen made use of the SO lineage of the fly notum

as a functional readout for impaired Notch signaling. Authors found three lethal alleles

causing bristle loss that were found in a restricted cytological region later associated to the

arp3 locus. Like WASp, bristle losses were correlated with cell fate changes after SOP and

pIIIb ACDs. Further functional characterization indicated that Arp3 is required upstream of

the Notch proteolytic cascade and does not affect Numb or Neur segregation. Nonetheless,

authors did not generate arp3 mitotic clones in the SO lineage to ascertain whether Arp3 was

required in signal-receiving or in signal-sending cells. To demonstrate this point, they used

alternative epithelial tissues that include the ovarian follicle cells and the imaginal wing disc.

Finally, they excluded that Arp2/3 and WASp were required for Delta endocytosis and

proposed a model where Arp2/3 and WASp regulate Delta recycling towards the apical side

of the ARS based on Delta trafficking assays and phalloidin immunostainings performed in

control and mutant backgrounds (see also 2.3.4. and Figure 37). Limitations of this model will

be discussed in the next subsection.

Arp2/3 and WASp only required in lineages?

An intriguing observation is that WASp and Arp2/3 appear to be required for Notch activation

only in lineages. The original WASp paper in flies did not report any Notch-dependent

cytokinesis-independent phenotype in pharates, including wing margin formation, vein

morphogenesis, segment formation in legs or lateral inhibition, but instead uncovered a role

for WASp in different lineages during embryogenesis (Ben-Yaacov et al., 2001). They first

found a similar requirement in fate specification in embryonic SO lineages as in the notum.


113

They also observed that a Notch-dependent fate decision in the RP2 neuroblast lineage was

impaired in embryos lacking maternal and zygotic WASp (Figure 60). Lastly, they reported

that loss of WASp, by contrast, mimicked a Notch gain-of-function phenotype in a

mesodermal lineage thought to produce from a single progenitor the future pericardial cell

(PC) and the DA1 muscle founder cell (Park et al., 1998). The authors noted a decrease in

DA1 founders and in DA1 daughter cells. Reexamination of mesodermal lineages revised this

relationship between the PC and the DA1 founder and showed that DA1 muscle founders are

arising from another progenitor, while the PCs in fact are generated by a PC founder which

shares a common progenitor with the DO2 founder muscle cell (Han and Bodmer, 2003).

Therefore, loss of WASp should be reinvestigated in light of this in mesodermal lineages

giving rise to PCs. Finally, considering that WASp also regulates myoblast fusion in embryos

(Massarwa et al., 2007; Schäfer et al., 2007), one might speculate that the reduced number of

DA1 nuclei in WASp embryos accounts for defective muscle formation. Consistently with this

hypothesis, analysis of WASp muscles generated by DA1 or DO2 founders revealed a

decreased number in nuclei per syncytium and the abortion of fusion after a single round of

founder cell fusion with the fusion-competent myoblast (Massarwa et al., 2007; Schäfer et al.,

2007). Taken together, although WASp role is unclear in mesodermal lineages, these data

suggest a conserved role for WASp in mediating Notch signaling in neural lineages. In

mammals, N-WASp might also play a central role in neurogenesis in both neuronal cells and

neural stem cells (Jain et al., 2014; Liebau et al., 2011), although the connection with fate

specification during lineage progression has not been tested.

Figure 60. The neuronal fate is favored in absence of WASp in the RP2 lineage.

Adapted from Ben-Yaacov et al., 2001.


114

Similar to WASp, components of the Arp2/3 complex seem to affect Notch signaling

only in the SO lineage. Loss of Arp2 does not impair Notch-dependent expression of the

Notch target genes Wingless and Senseless in the wing margin (Legent et al., 2012) (Figure

61). Although the overall wing shape is strongly disturbed, veins appear to form normally in

arpc1 wings (Gohl et al., 2010) (Figure 61). Role of the Arp2/3 complex in type II NBs is not

documented. During mammalian neurogenesis, as the Arp2/3 complex might be involved in

mediating Notch signaling in asymmetrically dividing radial glial cells (Chou and Wang,

2016; Dong et al., 2012).

Figure 61. Impairing Arp2/3 function does not impair Notch signaling during vein

morphogenesis (A-B) or wing margin formation (F-F’’’). sop2Q25st corresponds

to a arpc1 background in B. arp2 clones are marked by the GFP in F’’’. Adapted

from Legent et al., 2012 and Gohl et al., 2010.

As one key feature of lineages is cytokinesis and knowing that Notch receptors are

activated at cytokinesis during SOP ACD (Couturier et al., 2012), one might speculate that

Arp2/3 and WASp molecular functions are to link SOP cytokinesis with Notch activation in

the signal-receiving cell (Chapter 5).

3.3.2. Arp2/3, WASp and Delta recycling: a problematic model

The model currently accepted for Arp2/3 and WASp roles in activating Notch assumes a

mechanism independent of cytokinesis (Rajan et al., 2009). Authors proposed that Arp2/3 and

WASp drives the formation of an actin rich structure (ARS) at the pIIa-pIIb interface and

regulates the trafficking of Delta towards the apical side of this structure (see subsection

2.3.4. for a detailed description of the model).


115

The first limitation of this model appears when one considers the different modes of

ACD occurring in the SO lineage: half of the lineage, that gives rise to the external cells of

the organ, together with the SOP divide along the anterior-posterior axis while the second

half, that produces the internal cells, divide along the apical-basal axis like neuroblasts. As

Arp2/3 is also required for fate specification following pIIb and pIIIb “NB-like” ACDs, it is

unclear how apical microvilli could be involved in signaling as the pIIIb-glial cell and the

neuron-sheath interfaces does not have this type of apical-basal polarity. Consequently, the

idea that Delta must be addressed to the apical side of the ARS seems less plausible in this

context.

The second limitation is the discrepancy between the requirement for Notch activation

at cytokinesis (Remaud et al., 2008) and the kinetics proposed for Delta trafficking in pIIb.

Delta molecules were pulse-chased for one hour prior their detection at the apical side of the

ARS. Moreover, pulse-chases were started at random time points after SOP division. This

means that Delta apical addressing occurs between one hour after SOP division and pIIb

mitosis entry, a time window that does not fit with Notch signaling onset at cytokinesis.

Finally, it was unclear whether apical Delta molecules were still located in endosomes or

were distributed at the pIIb cell surface.

Finally, this model assumed original recycling model where Delta recycling is

necessary to prime the ligand for activation (Wang and Struhl, 2004). Not only has no such

“activating” modification been found to date, but several lines of evidence strongly contradict

the requirement of Delta recycling to activate Notch receptors (Langridge and Struhl, 2017;

Meloty-Kapella et al., 2012).

Together, these remarks rule out the proposed molecular functions for Arp2/3 and

WASp in Notch signaling, although they highlight a role for actin and Arp2/3 in regulating

Delta distribution to activate Notch receptors.

3.3.3. Alternative hypotheses

Contact formation and/or expansion

The most straightforward hypothesis is that the Arp2/3 complex together with WASp regulate

Notch activation at cytokinesis by generating the contact surface between pIIa and pIIb

(Figure 62). If Arp2/3 and WASp were required for contact formation, signaling would be

completely abolished between pIIa and pIIb. Conversely, if Arp2/3 and WASp were rather

regulating contact expansion in accordance with the described role of Arp3 in epidermal cell


116

cytokinesis (Herszterg et al., 2013), pIIa-pIIb interfaces would be formed but with small

contact surfaces in arp3 or WASp cells. Signaling levels would then be decreased, as Notch

activity depends on the contact surface area (Shaya et al., 2017), and presumably insufficient

to drive the expression of all target genes in pIIa. Thus, Arp2/3 and WASp would link

cytokinesis to Notch signaling through the regulation of the surface contact area during

furrowing.

Despite being parsimonious and elegant, this hypothesis is based on the assumption

that WASp activates Arp2/3 to drive contact expansion. Although not impossible, the

literature rather indicates the associated NPF in similar processes is SCAR/WAVE (Alekhina

et al., 2017). Moreover, SCAR activity appears to be required to form the ARS and expand

the pIIa-pIIb contact surface (King et al., 2010), albeit it is not required for fate specification

in SO lineages (Zallen et al., 2002). Therefore, it is important to assess WASp involvement in

the expansion of the pIIa-pIIb contact to test this hypothesis.

Figure 62. Alternative functions for the Arp2/3 complex at cytokinesis. See text

for details. Adapted from Herszterg et al., 2014, Kaksonen and Roux, 2018 and

Kapus and Janmey, 2013.


117

Delta endocytosis

WASp-dependent Arp2/3 branching activity was found to generate the pushing forces driving

membrane inward invagination during endocytosis (Kaksonen and Roux, 2018; Mund et al.,

2018; Picco et al., 2018) (Figure 62). In the fly notum, this function seems to be conserved as

Arp2/3 and WASp promote E-Cad endocytosis in epidermal cells (Georgiou et al., 2008;

Leibfried et al., 2008). Thus, in addition to contact expansion, Arp2/3 and WASp might also

be required to promote Delta endocytosis at the pIIa-pIIb interface.

Nonetheless, this hypothesis was ruled out by Rajan et al. based on an antibody uptake

assay performed in arp3 and arpc1 backgrounds (Rajan et al., 2009). They observed Delta

being internalized in arp3 and arpc1 pIIa-pIIb pairs while Delta was mainly detected at the

membrane when Dynamin activity was impaired. Careful analysis of this experiment reveals

out that authors only addressed whether Arp2/3 is essential for Delta endocytosis in pIIa and

pIIb. First, as endo-Delta vesicles were not quantified, it remains unclear whether loss of

Arp2/3 activity affects specifically a subset of Delta endocytic events. Second, the uptake

assay was performed at a random time point following SOP division in each condition. As

Notch is activated at cytokinesis, this means that that the requirement of Arp2/3 and WASp

roles in Delta endocytosis within this time window remains untested. Last, precise

interpretation of arp3 and WASp phenotypes in the notum already indicates that these two

actin regulators are not essential effectors of Neur-mediated Delta endocytosis. Indeed, lateral

inhibition, which depends on Neur and Delta, is unaffected in arp3 or WASp tissues. This

reinforces the idea that Arp2/3 and WASp are only required in Delta- and Neur-dependent

processes involving cytokinesis. Mechanistically, Arp2/3 recruitment and concentration at the

newly-forming contact during furrowing might be necessary for its activation by WASp and

the subsequent endocytosis of Delta (Chapter 5).

Anchoring Notch receptors to the cortex

To end this introduction, a final speculative hypothesis will be proposed. As the

mechanotransduction model relies on pulling forces exerted by the ligand on the NECD, it

implies that the membrane where the receptor is tethered must resist this pulling force.

Otherwise, the receptor would simply follow its bound ligand and the force would not be

transmitted to the NNR (Figure 62).

Therefore, Arp2/3 activity in pIIa might polymerize a cortical actin network, such as

the ARS, to which Notch receptors would be anchored to resist forces exerted by Delta

endocytosis. In the same line, Arp2/3 could mediate Notch endocytosis as well accordingly

3.4. Axes of the thesis

118

with a bilateral endocytosis model (Chastagner et al., 2017), yet simultaneous coordination of

ligand and receptor endocytoses is lacking a mechanistic basis.

Although tantalizing, this hypothesis is undermined by the current lack of knowledge

on Notch association with the actomyosin cytoskeleton at the cell surface.

3.4. Axes of the thesis

This thesis interrogates the interplay between Notch signaling and cytokinesis in the sensory

organ lineage around two questions, each being addressed by a Chapter written in the form of

a paper:

Question 1: Where are Notch receptors activated along the pIIa-pIIb interface at

cytokinesis? (Chapter 4)

Question 2: What is the molecular mechanism ensuring Notch activation at cytokinesis?

(Chapter 5)

119

120

Part II

Results

121

Chapter 4

Paper 1: Intra-lineage Fate Decisions Involve Activation of Notch receptors

basal to the midbody in Drosophila Sensory Organ Precursor Cells

Chapter 4. Results. Intra-lineage fate decisions involve activation of Notch receptors basal to the midbody in Drosophila SOPs

122

Article

Intra-lineage Fate Decisions Involve Activation ofNotch Receptors Basal to the Midbody in DrosophilaSensory Organ Precursor Cells

Highlightsd Photo-tracking fluorescent Notch receptors reveals where

nuclear Notch comes from

d A specific subset of receptors contributes to signaling in

asymmetric cell division

d Signaling is restricted to sister cells to ensure intra-lineage

fate decision

Authors

Mateusz Trylinski, Khalil Mazouni,

Francois Schweisguth

[email protected]

In BriefWhere Notch receptors are activated at

the cell surface is not known. To address

this, Trylinski et al. used photo-bleaching

and photo-conversion approaches in

Drosophila. Only a specific subset of

receptors, located basal to the midbody,

were shown to contribute to signaling and

binary fate decision in the context of

asymmetric cell division.

Trylinski et al., 2017, Current Biology 27, 1–9August 7, 2017 ª 2017 Elsevier Ltd.http://dx.doi.org/10.1016/j.cub.2017.06.030


123

Current Biology

Article

Intra-lineage Fate Decisions Involve Activationof Notch Receptors Basal to the Midbodyin Drosophila Sensory Organ Precursor CellsMateusz Trylinski,1,2,3 Khalil Mazouni,1,2 and Francois Schweisguth1,2,4,*1Institut Pasteur, Paris 75015, France2CNRS, UMR 3738, Paris 75015, France3Universit!e Pierre et Marie Curie, Cellule Pasteur UPMC, Rue du Dr Roux, Paris 75015, France4Lead Contact*Correspondence: [email protected]://dx.doi.org/10.1016/j.cub.2017.06.030

SUMMARY

Notch receptors regulate cell fate decisions duringembryogenesis and throughout adult life. In manycell lineages, binary fate decisions are mediated bydirectional Notch signaling between the two sistercells produced by cell division. How Notch signalingis restricted to sister cells after division to regulateintra-lineage decision is poorly understood. Moregenerally, where ligand-dependent activation ofNotch occurs at the cell surface is not known, asmethods to detect receptor activation in vivo arelacking. In Drosophila pupae, Notch signals duringcytokinesis to regulate the intra-lineage pIIa/pIIb de-cision in the sensory organ lineage. Here, we identifytwo pools of Notch along the pIIa-pIIb interface, api-cal and basal to the midbody. Analysis of the dy-namics of Notch, Delta, and Neuralized distributionin living pupae suggests that ligand endocytosisand receptor activation occur basal to the midbody.Using selective photo-bleaching of GFP-taggedNotch and photo-tracking of photo-convertibleNotch, we show that nuclear Notch is indeed pro-duced by receptors located basal to the midbody.Thus, only a specific subset of receptors, locatedbasal to the midbody, contributes to signaling inpIIa. This is the first in vivo characterization of thepool of Notch contributing to signaling. We proposea simple mechanism of cell fate decision basedon intra-lineage signaling: ligands and receptorslocalize during cytokinesis to the new cell-cell inter-face, thereby ensuring signaling between sister cells,hence intra-lineage fate decision.

INTRODUCTION

Cell-cell interactions mediated by the Notch receptor regulatecell fate in a wide range of developmental processes during an-imal development. One conserved function of Notch is to regu-late binary fate decisions between sister cells within various

cell lineages in insects [1–5] and mammals [6, 7]. Notch recep-tors are mechanosensitive transmembrane proteins with anextracellular ligand-binding domain and an intracellular domainacting as a membrane-tethered transcriptional coactivator [8].In the absence of ligands, receptors are in an auto-inhibitedstate. Ligand binding allows cellular forces to pull onto ligand-re-ceptor complexes and unmask a cleavage site in the extracel-lular domain of Notch [9, 10]. Extracellular cleavage producesa membrane-tethered form of Notch, which is then further pro-cessed to release the Notch intracellular domain (NICD) thatlocalizes to the nucleus and regulates gene expression.Whereas the mechanism of receptor activation is well estab-

lished, where ligand-dependent activation of Notch occurs atthe cell surface remains unknown. Because both receptorsand ligands accumulate at adherens junctions (AJs) in manyepithelia [11–16] and because Notch signaling appears to befunctionally linked with E-cadherin [16–20], it is often assumedthat signaling takes place at the level of AJs. Nevertheless, earlierstudies have suggested that receptor activation may occurat various locations, depending on developmental contexts,including apical cell surface [21], cell-cell junctions [16], basal-lateral membrane [22], and basal filopodia [23]. Because thesize and geometry of the contact site may regulate both thestrength and the range of signaling [23, 24], it is key to determinewhere at the cell surface Notch activation takes place.Sensory lineages in Drosophila are well suited to study Notch-

dependent intra-lineage fate decisions [25]. In the dorsal thorax,or notum, sensory organ precursor cells (SOPs) are specifiedwithin a single-layered epithelium in early pupae [26] and divideasymmetrically to produce sensory organs [27]. SOPs first divideasymmetrically along the fly body axis to generate an anteriorpIIb cell (precursor of the internal cells) and a posterior pIIa cell(precursor of the external cells) [28]. Fate asymmetry resultsfrom the unequal segregation of Numb and Neuralized (Neur)into pIIb [29–31] so that Notch is activated in pIIa and inhibitedin pIIb [32, 33]. Numb inhibits the recycling of Notch and pro-motes its late-endosome targeting in pIIb [32, 34–38], whereasNeur regulates the endocytosis of Delta (Dl) in pIIb [30, 31]. Usinglive imaging of GFP-tagged Notch, we previously showed thatNICD was produced in pIIa during cytokinesis [32]. However,where NICD came from was not determined.Here, we show that NICD is produced mostly from a subset of

receptors located basal to the midbody during cytokinesis. This

Current Biology 27, 1–9, August 7, 2017 ª 2017 Elsevier Ltd. 1

Please cite this article in press as: Trylinski et al., Intra-lineage Fate Decisions Involve Activation of Notch Receptors Basal to the Midbody in DrosophilaSensory Organ Precursor Cells, Current Biology (2017), http://dx.doi.org/10.1016/j.cub.2017.06.030


124

suggests a simplemechanism linking cytokinesis to intra-lineagefate decision via directional signaling by this pool of Notch.

RESULTS

Two Pools of Notch along the pIIa-pIIb InterfaceWe first re-examined the dynamics of Notch in living pupae usinga GFP-tagged Notch, NiGFP, which we generated by CRISPR-mediated homologous recombination (HR) (see STARMethods).Low Notch levels were detected in SOPs prior to division and atthe apical pIIa-pIIb interface during cytokinesis [32] (Figures 1A–1A’’ and S1A–S1A’’). We also confirmed that directional signalingoccurred during cytokinesis, as measured using nuclear NiGFPlevel (Figure 1B; NICD levels reached a plateau at t30 in pIIa). Nu-clear Notch levels appeared to correlate with the transcriptionalactivation of the E(spl)m8-HLH gene, a direct Notch target [39]:a similar kinetics was seen for a bacterial artificial chromosome(BAC)-encoded GFP-tagged E(spl)m8-HLH transgene (Fig-ure 1C; see STAR Methods). This indicated that Notch is rapidlyactivated after division to regulate the pIIa/pIIb decision.

In epithelia, cytokinesis can be viewed as a multicellular pro-cess, with both dividing cell and its immediate neighbors regu-lating the formation of the new contact region [40]. As reportedearlier [41], we observed that epidermal cells formed stablefinger-like protrusions separating the two SOP daughters atthe level of the midbody between t15 and t30 (Figures 1D and1E; arrow in Figure 1D’). Thus, the pIIa andpIIb cellswere in direct

contact both apical to the midbody (Figures 1D and 1E), wherenew AJ complexes form (Figures S1A–S1A’’), and basal to themidbody (Figures 1D’’ and 1E). Using the midbody as a positionlandmark (detected as a faint spot using iRFP670nls; FiguresS1B–S1B’’), we defined an apical and a lateral pool of Notch. Api-cal NiGFP transiently accumulated [13, 32], reaching a peak!15min after anaphase (Figure 1F), whereas low levels of NiGFPwere detected basal to the midbody (Figures 1A’, 1F’, and S1A’).

Delta and Neuralized Localize Basal to the Midbodyduring CytokinesisThe observation of these two pools of Notch raised the questionof their relative contribution to the production of NICD. Severalobservations suggested that the lateral pool of Notch mightcontribute to signaling. First, ectopic activation of Notch innumb mutant cells correlated with an accumulation of Notchbasal, but not apical, to the midbody during cytokinesis (Figures1F–1G’’) [32]. Thus, Numb appeared to inhibit the recycling ofNotch toward the lateral pIIa-pIIb interface [34–36], and thislateral Notch might contribute to ectopic signaling in pIIb (Fig-ure S2A). Second, blocking g-secretase activity in Presenilin(Psn) mutant cells led to increased NiGFP levels basal to themid-body (Figures 1H–1H’’), consistent with lateral Notch remainingunprocessed in Psn mutant cells. Third, in fixed nota, endoge-nous Dl was primarily detected along the lateral pIIa-pIIb inter-face, basal to the midbody, and not at the level of AJs (Figures2A–2A’’) [31, 42, 43]. Similarly, a GFP-tagged Dl generated by

A D G

G’

G” H”

H’

H

D’

D”

A’

A”

B

E

F

F’C

Figure 1. Two Pools of Notch along the pIIa-pIIb Interface(A–A’’) NiGFP (green) transiently accumulated at the apical pIIa-pIIb interface at t15 in living pupae. The pIIa/pIIb nuclei (outlined in A’) were marked with

iRFP670nls. A faint NiGFP signal was detected along the lateral interface at t15 (arrow, A’; the apical surface area of the pIIa and pIIb cells are indicated in A). In

these and other panels, the ‘‘apical’’ and ‘‘basal’’ labels refer to the x,y sections showing the apical and ‘‘lateral’’ populations of Notch receptors (see E).

(B andC) Quantification of nuclear NiGFP (B) andGFP-E(spl)m8-HLH (C) levels during cytokinesis (t0; metaphase-anaphase transition). Nuclear NiGFP (nR 21 for

each time point) and GFP-E(spl)m8-HLH (n R 18) reached a plateau at t30. Data are represented as mean ± SEM.

(D–D’’) Morphology of the pIIa-pIIb contact at t30 (BazCherry, magenta; DlgGFP, green; iRFP670nls, red). Two contact regions were seen apical and basal to the

midbody. Note the absence of pIIa-pIIb contact at the level of the midbody (arrow in D’).

(E) Schematic representation of the apical and lateral contact regions along the pIIa-pIIb interface.

(F and F’) Quantification of the apical (F) and lateral (F’) pools of NiGFP in wild-type (green; nR 21) and numbRNAi (red; nR 27) pupae. Significant levels of NiGFP

were transiently detected apical to the midbody in wild-type pupae. The silencing of numb led to NiGFP accumulation specifically at the lateral interface. Data are

represented as mean ± SEM.

(G–G’’) NiGFP (green; iRFP670nls; marking the pIIa/pIIb nuclei, red) accumulated in dots along the lateral interface (arrow) at t15 in numbRNAi pupae.

(H–H’’) The loss ofPsn activity inmutant clones led to the accumulation of NiGFP (green) along the lateral pIIa-pIIb interface in livingmosaic pupae (arrow; pIIa/pIIb

nuclei were identified using iRFP670nls, red; mutant cells were marked by the loss of a RFP marker, white; only isolated mutant SOPs were studied).

In these and all other panels, anterior is left (i.e., pIIb, left, and pIIa, right) and scale bars are 5 mm. See also Figure S1.

2 Current Biology 27, 1–9, August 7, 2017



125

CRISPR-HR [26] was clearly detected along the lateral pIIa-pIIbinterface in fixed nota (Figures 2B–2B’’). By contrast, DlGFP wasbarely detected along the pIIa-pIIb interface in living pupae (Fig-ures 2C–2C’’; quantified in Figures S2C and S2C’). This differ-ence between live and fixed samples suggested that mostDlGFP molecules are in a non-fluorescent state in SOPs dueto its rapid turnover relative to GFP maturation. Consistentwith this, inhibition of Dl endocytosis using a stabilized versionof Bearded that acts as a strong antagonist of Neur (BrdR)[44–47] increased the direct fluorescence of DlGFP (Figures2D–2D’’; BrdR efficiently blocked Notch signaling; Figure S2B).Importantly, DlGFP strongly accumulated basal to the midbody(and also, to a lesser extent, apical) at t15 upon inhibition ofNeur (Figures S2C and S2C’). Thus, Neur appeared to regulatethe endocytosis of Dl from the lateral membrane. Because theendocytosis of Dl by Neur is essential in pIIb for the activationof Notch in pIIa [31], receptor activation most likely occurredbasal to the midbody. Consistently, Neur localized transientlyalong the lateral pIIa-pIIb interface from t5 to t15, before accu-mulating apically (Figures 2E–2E’’, S2D, and S2D’; Neur dy-namics was monitored using a functional GFP-tagged versionof Neur). Together, these data suggested that Dl is endocytosedin a Neur-dependent manner from the lateral membrane of pIIbto activate the lateral pool of Notch in pIIa.

NICD Is Produced from Receptors that Turn OverRapidlyWe next addressed where Notch is activated, i.e., where NICDcomes from. We previously reported that GFP-tagged Notch

was present at the plasma membrane, early endosomes, andlate endosomes but that receptors accumulating in late endo-somes were not detected by direct fluorescence in living fliesdue to quenching at low pH [38]. However, because GFP re-sponds reversibly to pH changes [48], GFP-tagged NICD pro-duced from receptors located at late endosomes should befluorescent in the nucleus. To address whether late endosomalNotch contributed to NICD production, we took advantage of adual-tagged Notch marked intracellularly with both GFP andCherry, NiGFP4Cherry5 [38] (Figure 3A). Because it takes longerfor Cherry to become fluorescent relative to GFP, due to slowerprotein folding and/or chromophore maturation [38, 49], this dualtag acts as a pH-sensitive protein timer for Notch: ‘‘young’’ re-ceptors located at the plasma membrane and in early endo-somes are mostly GFP-positive and Cherry-negative whereas‘‘old’’ receptors accumulating in endo-lysosomes are GFPquenched and Cherry-positive (Figure 3B) [38]. These wouldtherefore produce GFP- and Cherry-positive NICD. To testwhether NICD originates in part from late endosomes, wemeasured the fluorescence of GFP- and Cherry-tagged NICD.We found that NICD was GFP-positive but Cherry-negative inpIIa (Figures 3C–3E; no signal was detected in pIIb). This indi-cated that receptors localizing within late endosomes did notparticipate to signaling in pIIa. However, it is conceivable thatGFP-positive and Cherry-negative receptors exist at the limitingmembrane and contribute to the generation of NICD. It alsoindicated that NICD half-life was shorter than Cherry maturationhalf-time [49] and that NICD was produced from receptors thatturn over too rapidly to become Cherry-positive.

A

A’

A” B” C” D” E”

B’ C’ D’ E’

B C D E

Figure 2. Ligand Endocytosis from the Lateral pIIa-pIIb Interface(A–A’’) Dl (anti-Dl intracellular domain, green) localized mostly below the midbody (marked with MyoII, not shown; CadGFP, magenta). High-magnification views

of the pIIa-pIIb interface (arrows in A and A’; pIIa and pIIb nuclei were marked with iRFP670nls, not shown) are shown as insets.

(B–C’’) DlGFP (green) was detected in fixed nota (B–B’’) and in living pupae at t15 (C–C’’). The pIIa and pIIb nuclei were marked with iRFP670nls (red in C–C’’; not

shown in B–B’’). A faint DlGFP signal was detected along the pIIa-pIIb interface in living pupae. By contrast, a clear anti-GFP signal was seen basal to AJs (Cad,

magenta in B–B’’) and midbody (MyoII; not shown) in fixed nota.

(D–D’’) DlGFP (green) specifically accumulated along the lateral interface (arrow in D’) in living neur > BrdR pupae at t15, suggesting that Dl is endocytosed in a

Neur-dependent manner from the lateral pIIa-pIIb interface.

(E–E’’) Distribution of NeurGFP (green) at t15 in living pupae. NeurGFP localized both apical and basal (arrow in E’) to the midbody (iRFP670nls, red).

See also Figure S2.

Current Biology 27, 1–9, August 7, 2017 3



126

Tracking the Origin of NICD Using Selective Photo-bleachingWe next sought to determine the relative contribution of apicaland lateral Notch to NICD production in pIIa using photo-bleach-ing. In principle, selective bleaching of one specific pool of Notchcould reveal its contribution to NICD production through a corre-sponding decrease in nuclear fluorescence in pIIa (Figure 4A).This approach had two pre-requisites. First, selective bleachinghad to be achieved along the z axis. This was possible at posi-tions where the pIIa-pIIb interface was tilted relative to the laserpath (Figures 4B–4B’’). Second, the exchange rate of Notch be-tween the apical and lateral contact regions should be slow rela-tive to the timescale of nuclear Notch accumulation. Otherwise,bleaching one pool would also result in a fluorescence decreaseof the other pool, thus preventing us from tracking back theorigin of bleached NICD. We found that bleaching apical NiGFPat t15, t20, and t25 efficiently bleached the apical pool of NiGFPwith no significant effect on lateral NiGFP fluorescence at t30(Figures 4C, 4C’, and S3A–S3B’; the effect of photo-bleachingwas measured at t30 because NICD levels reached a plateauat t30 in pIIa). Conversely, repeated bleaching of lateral NiGFPefficiently reduced fluorescence levels at this location with nosignificant effect on apical NiGFP fluorescence (Figures 4E,4E’, and S3C–S3D’). We could therefore use photo-bleachingto examine where NICD comes from. We first found that bleach-ing the lateral pool at t15, t20, and t25 decreased nuclear NiGFPfluorescence in pIIa at t30 (Figures 4E and 4F). Thus, receptorslocated in the bleached area were processed to generateNICD molecules translocating to the pIIa nucleus within 5–15 min. A single photo-bleaching at t15 was sufficient todecrease nuclear fluorescence at t30 (Figures S3G and S3H),indicating that receptors present along the lateral interface att15 produced a fraction of NICD found at t30 in the nucleus.By contrast, bleaching the apical pool of NiGFP had no signifi-cant effect of the level of nuclear GFP in pIIa (Figures 4D, S3E,and S3F). Thus, receptors apical to the midbody did not signifi-cantly contribute to signaling.

We repeated these photo-bleaching experiments in pupallegs, which can be imaged as cross-section views (Figure 4G),hence providing better spatial resolution along the apical-basalaxis. SOPs divide around the same time in pupal legs and notaand GFP-tagged NICD similarly accumulated in pIIa (FiguresS4A and S4A’). Also, a lateral pool of NiGFP was detected alongthe pIIa-pIIb interface (Figure 4H). Photo-bleaching this lateralpool resulted in low nuclear GFP fluorescence, whereas bleach-ing apical NiGFP had no significant effect (Figures 4I and 4J).These results confirmed that receptors located along the lateralpIIa-pIIb interface produced NICD in pIIa, whereas signaling byapical Notch was not detectable by this approach.

Photo-Convertible Notch Confirms that Lateral NotchProduces NICDNext, to positively mark selective pools of Notch, we used thegreen-to-red photo-convertible fluorescent protein mMaple3[50] and generated a knockin NimMaple3 receptor by CRISPR-mediated HR (Figure 5A; mMaple3 was introduced at the sameposition as GFP in NiGFP). NimMaple3 flies were viable and fertile,with no detectable phenotype, indicating that the NimMaple3receptor was functional. Importantly, the maturation kinetics ofmMaple3 was fast enough to detect the green fluorescence ofNimMaple3 at the cell cortex in the absence of photo-conversion(Figure 5B) and NimMaple3 was efficiently photo-convertedby UV light (Figures 5B–5C’). Additionally, photo-convertedNimMaple3 was detected in the pIIa nucleus at t30 (Figures5D–5E’’): using raw intensity values to calculate pIIa/pIIb ratios,we found a weak but statistically significant difference uponUV illumination (Figure 5F). Because a small number of photo-converted NICD molecules might be present in the volume ofthe pIIa nucleus, only few pixels might actually measure photo-converted NimMaple3, whereas most pixels measured noise.We therefore applied increasing threshold values (to selectfrom 50% to 5% of the brightest pixels of the pIIa nucleus) andcalculated pIIa/pIIb ratios using thresholded images (see STARMethods). This image processing led to increasing pIIa/pIIb

AD

E

C

B C’

Figure 3. NICD Originated from RecentlySynthetized Receptors(A and B) Structure (A) and properties (B) of

NiGFP4Cherry5 (adapted from [38]). GFP (green)

and Cherry (red) are intracellular. Upon synthesis of

Notch, GFP, and Cherry are in a dark immature

state. NiGFP4Cherry5 molecules are more likely to

become first GFP-bright and Cherry-dark due to the

faster maturation of GFP relative to Cherry, before

being both GFP- and Cherry-bright. GFP-bright-

only molecules were detected at the cell surface

and in early endosomes, whereas Cherry-bright

receptors were detected into acidic intracellular

compartments (where GFP is quenched at low pH).

(C and C’) Snapshots showing the fluorescence of

NiGFP4Cherry5 at t30 (GFP, green and Cherry, red).

The pIIa and pIIb nuclei, marked with iRFP670nls

(not shown), are outlined. Nuclear NiGFP4Cherry5

was detected in pIIa only by GFP fluorescence.

(D and E) Quantification of the GFP (D) and

Cherry (E) fluorescence produced by nuclear

NiGFP4Cherry5 (n = 24). NICD came from a pool

of GFP-bright and Cherry-dark receptors. Data

are represented as mean ± SEM.




127

A

C

E

G I JH

E’ F

C’ D

B B’

B”

Figure 4. Tracing the Origin of NICD Using Photo-Bleaching(A) Principle of Notch photo-tracking using photo-bleaching. A loss of nuclear fluorescence upon the photo-bleaching of a specific pool of NiGFP (lateral pool,

greyed) would indicate that photo-bleached receptors contribute to NICD production.

(B–B’’) Selective photo-bleaching of the pIIa-pIIb interface (shown here using CadGFP, green; nuclei marked with iRFP670nls, red) in the notum. The position of

the apical interface (yellow dotted box in B) did not coincide with the lateral membrane (B’) due to a posterior tilt of the pIIa-pIIb interface (B’’; white dotted lines

show the angle of the pIIa-pIIb interface relative to the apical-basal axis).

(C–F) Effect of selective photo-bleaching (apical pool, C and C’; lateral pool, E and E’) on nuclear NiGFP fluorescence (D and F). The apical and lateral fluo-

rescence profiles of NiGFP at t30 (control, green) were specifically changed upon repeated photo-bleaching. Apical photo-bleaching efficiently reduced the

fluorescence of apical NiGFP (C, blue) but had no effect on basal NiGFP (C’, blue profile) and vice versa (E and E’; brown profiles). Photo-bleaching the lateral pool

of NiGFP specifically reduced the level of nuclear NiGFP fluorescence in pIIa at t30 (F; n = 19). By contrast, photo-bleaching the lateral pool of NiGFP had no

significant effect (D; n = 21). In (C), (C’), (E), and (E’), data are represented as mean ± SEM.

(G) Live imaging of sensory cells in pupal legs. Pupae aremounted legs up (top). Pupal legs appeared as tubular epithelia at the body surface so that pIIa-pIIb cells

located laterally could be imaged on side views (bottom).

(legend continued on next page)




128

ratios (Figure 5F), indicative of higher signal-to-noise ratios.Thus, photo-converted nuclear Notch was detectable in pIIa.Additionally, applying this image processing to the GFP andCherry signals of NiGFP4Cherry5 increased signal-to-noiseratios for GFP, but not Cherry (Figures S4B and S4B’). Thisconfirmed that Cherry-fluorescent Notch did not contribute toNICD production.

We then photo-converted the apical and basal pools ofNimMaple3 in pIIa-pIIb cell pairs in pupal legs. Repeatedphoto-conversion of the apical pool had a minor effect on thepIIa/pIIb ratio relative to the no UV illumination control (Figures5H, S4C–S4C’’, S4E, and S4E’), indicating that apical receptorshad only a minor contribution to the production of NICD. Bycontrast, photo-conversion of the lateral pool led to a clear in-crease of the pIIa/pIIb ratio at t30 (Figures 5G, 5H, and S4D–S4E’). We conclude that NICD was produced in pIIa mostlyfrom a pool of receptors located at (or close to) the lateral inter-

face at t15-t25. Thus, using two different photo-tracking ap-proaches, we found that Notch receptors present basal to themidbody along the pIIa-pIIb interface at any time between t15and t25 contributed to the accumulation of NICD at t30, whereasreceptors located apical to the midbody did not significantlycontribute to NICD production. We therefore propose that Notchreceptor activation mostly occurs on the pIIa side of the pIIa-pIIbinterface and that only a specific subset of Notch receptorsregulates the pIIa-pIIb fate decision.

DISCUSSION

Several methods are currently available to monitor in vivo thesignaling activity of Notch by measuring the level and/or activityof NICD [32, 51, 52]. By contrast, in vivo reporters for ligand-re-ceptor interaction, conformational change of Notch in responseto mechanical force, and S2 cleavage of Notch are lacking.

(H) Distribution of NiGFP (green) and BazCherry (magenta) along the pIIa-pIIb interface in pupal legs of living pupae. The lateral pool of NiGFP was clearly

observed at t15.

(I and J) Nuclear NiGFP fluorescence at t30 following selective photo-bleaching in pupal legs. Bleaching the apical pool of NiGFP had no significant effect

(I; n = 19), whereas bleaching the lateral pool specifically reduced nuclear NiGFP fluorescence in pIIa (J; n = 21).

n.s., not significant; ***p < 0.001. See also Figure S3.

A F

H

B

D

E

G G’ G”

E’ E”

D’ D”

B’

C

C’

Figure 5. Photo-Tracking Notch Using Photo-Convertible Receptors(A) Structure of NimMaple3 (as in Figure 3A). mMaple3 was inserted at position 2388 by CRISPR-mediated HR.

(B–C’) NimMaple3 was detected at the cell cortex prior to photo-conversion (B; green channel) and was efficiently photo-converted (C’, red channel).

(D–E’’) Photo-converted NimMaple3 (red/false colors) was detected in pIIa nuclei (SensGFP, green) at t30 in pupal legs. The red fluorescence emitted by

NimMaple3 was measured before (D–D’’) and after photo-conversion (E–E’’). Raw (D’ and E’) and processed signals (D’’ and E’’; using a threshold selecting the

5% brightest pixels of pIIa) are shown. Note the strong auto-fluorescence signal from the pre-cuticle (asterisk in D and E).

(F) Plot showing the pIIa/pIIb ratio values of nuclear photo-converted NimMaple3 for different thresholds (indicated as the percentage of brightest pixels in the pIIa

nucleus). Raw data (threshold value 100; lateral photo-conversion versus control) displayed statistically significant differences. Image processing efficiently

extracted the signal of nuclear photo-converted NimMaple3, as shown by the increasing difference between ratio values measured before (!405 nm) and after

photo-conversion (+405 nm; n = 6).

(G–G’’) Photo-converted NimMaple3 (red/false colors) was detected in pIIa nuclei (SensGFP, green) at t30 in pupal legs following repeated UV illumination of the

lateral domain at t15, t20, and t25.

(H) Plot showing the pIIa/pIIb ratio values of photo-converted NimMaple3 (as in F) in control cells (no photo-conversion, black) and upon photo-conversion of the

apical (red) and lateral (blue) pools. Higher levels of photo-converted NICDwere detected in pIIa upon photo-conversion of the lateral versus apical pool of Notch

(n = 25 for each condition).

In (F)–(H), data are represented as mean ± SEM. *p < 0.05, **p < 0.01, and ***p < 0.001. See also Figure S4.




129

Consequently, the subcellular location of Notch receptor activa-tion in vivo and the relative contribution of the different pools ofNotch to signaling remain unknown. Here, we developed twocomplementary fluorescent-based approaches to track for thefirst timewhere NICD comes from.We showed that Notch recep-tors present basal to the midbody along the pIIa-pIIb interfacecontributed to the accumulation of NICD, whereas receptorslocated apical to the midbody did not significantly contribute toNICD production. To our knowledge, this study provides the firstin vivo analysis of ligand-dependent Notch receptor activation atthe cell surface. Moreover, the photo-bleaching and photo-con-version approaches used here should be broadly applicable inmodel organisms that can be genetically engineered and easilyimaged.Other sites of Notch activation had previously been proposed in

pIIa. In one model, based on the specific requirements for Arp2/3and WASp activities for both Notch signaling and actin organiza-tion [53, 54], Dl at apical microvilli in pIIb would activate Notchlocated apically in pIIa [53]. However, loss of Arp2/3 activity alsodisruptedcorticalactinalong thebasalpIIa-pIIb interface [53], sug-gesting that regulation of the actin cytoskeleton at this location,rather than at microvilli, may be key for receptor activation. In asecond model, Dl-Notch signaling was proposed to occur at thenew apical pIIa-pIIb junction [13]. This model was largely basedon thedetectionofNotchat this location.Our study, however, indi-cated that this pool of Notch did not significantly contribute to theproduction of NICD in pIIa. In a third model, Notch activation wasproposed tooccur inspecificSara-positiveendosomes inpIIa [55].Whereas the possible contribution of these endosomes to NICDproduction could not be directly addressed by photo-tracking,two lines of evidence suggest that their contribution can only beminor. First, live imaging of Notch failed to detect this pool [32,38], indicating that this pool represents a minor fraction of Notchin pIIa. Second, symmetric partitioning of Sara endosomes didnot affect the pIIa-pIIb decision, indicating that this proposedpool is not essential for fate asymmetry [56]. Finally, the nature ofthe mechanical force acting on Notch at the limiting membraneof the Sara-positive endosomes remains to be addressed. In sum-mary, all available data are fully consistentwith our conclusion thatreceptor activation occurs mostly basal to the midbody.Whereas our experiments identified the signaling pool of

Notch along the pIIa-pIIb, they did not, however, addresswhether S3 cleavage takes place at the cell surface or intracellu-larly following endocytosis. Indeed, the photo-tracking approachused here did not inform whether the activation of Notch byDelta, i.e., s2 cleavage, is followed by S3 cleavage at the samelocation or whether S2-processed Notch is internalized to befurther processed in signaling endosomes [57]. We note, how-ever, that the accumulation of lateral Notch observed in Psnmutant cells is consistent with S3 cleavage taking place, at leastin part, at the cell surface.Our work also sheds new light on the general mechanism

whereby Notch signaling is specifically restricted to sister cellswithin a lineage. In several tissues, including the gut [58, 59],lung [7], and CNS [3], Notch regulates intra-lineage decisionsbetween sister cells soon after mitosis. Here, we propose thatNotch-mediated intra-lineage decisions are directly linked to di-vision. Indeed, we suggest that ligands and receptors localize tothe lateral membranes that separate the two sister cells at cyto-

kinesis so that Dl-Notch signaling is primarily restricted to sistercells. Thus, neighboring cells—belonging to other cell lineages—would not interfere with intra-lineage fate decisions. Our dataindicating that Neur-dependent activation of Notch by Dl pre-dominantly occurs along the pIIa-pIIb lateral interface, basal tothe midbody during cytokinesis, fully support this model. Also,the observation that core components of the secretory machin-ery, e.g., Sec15, are specifically required for Notch signaling inthe context of intra-lineage decisions [60] is also consistentwith this view. Thus, targeting both receptors and ligands alongthe newly formed interface during cytokinesis provides anelegant mechanism to restrict signaling between sister cells,thereby ensuring that intra-lineage signaling regulates intra-line-age fate decision. Because Notch generates fate diversity withinneural lineages in both vertebrates and invertebrates, this mech-anism of intra-lineage signaling may be conserved.

STAR+METHODS

Detailed methods are provided in the online version of this paperand include the following:

d KEY RESOURCES TABLEd CONTACT FOR REAGENT AND RESOURCE SHARINGd EXPERIMENTAL MODEL AND SUBJECT DETAILS

B Flies and CRISPR-mediated HRB Genotypes

d METHOD DETAILSB Immunostaining and live imagingB Photo-bleaching and photo-conversion

d QUANTIFICATION AND STATISTICAL ANALYSIS

SUPPLEMENTAL INFORMATION

Supplemental Information includes four figures and can be found with this

article online at http://dx.doi.org/10.1016/j.cub.2017.06.030.

AUTHOR CONTRIBUTIONS

M.T. performed imaging experiments and data analysis. K.M. performed mo-

lecular biology experiments. M.T. and F.S. designed the experiments and

wrote the paper.

ACKNOWLEDGMENTS

We thank H. Bellen, D. Kiehart, M. Rand, the Bloomington Drosophila Stock

Center, the Developmental Studies Hybridoma Bank (DSHB), and Flybase

for flies, antibodies, and other resources. We thank L. Couturier and V. Roca

for embryo injection, L. Bally-Cuif and R. Levayer for critical reading, and

all lab members for discussion. This work was supported by grant nos.

ARC-PGA120140200771 and ANR-10-LABX-0073. M.T. received a fellowship

from the UPMC.

Received: May 6, 2017

Revised: June 7, 2017

Accepted: June 12, 2017

Published: July 20, 2017

REFERENCES

1. Bertet, C., Li, X., Erclik, T., Cavey, M., Wells, B., and Desplan, C. (2014).

Temporal patterning of neuroblasts controls Notch-mediated cell survival

through regulation of Hid or Reaper. Cell 158, 1173–1186.




130

2. Bivik, C., MacDonald, R.B., Gunnar, E., Mazouni, K., Schweisguth, F., and

Thor, S. (2016). Control of neural daughter cell proliferation by multi-level

Notch/Su(H)/E(spl)-HLH signaling. PLoS Genet. 12, e1005984.

3. San-Juan, B.P., and Baonza, A. (2011). The bHLH factor deadpan is a

direct target of Notch signaling and regulates neuroblast self-renewal in

Drosophila. Dev. Biol. 352, 70–82.

4. Perdigoto, C.N., Gervais, L., Overstreet, E., Fischer, J., Guichet, A., and

Schweisguth, F. (2008). Overexpression of partner of numb induces asym-

metric distribution of the PI4P 5-Kinase Skittles in mitotic sensory organ

precursor cells in Drosophila. PLoS ONE 3, e3072.

5. Ohlstein, B., and Spradling, A. (2007). Multipotent Drosophila intestinal

stem cells specify daughter cell fates by differential notch signaling.

Science 315, 988–992.

6. Dong, Z., Yang, N., Yeo, S.Y., Chitnis, A., and Guo, S. (2012). Intralineage

directional Notch signaling regulates self-renewal and differentiation of

asymmetrically dividing radial glia. Neuron 74, 65–78.

7. Pardo-Saganta, A., Tata, P.R., Law, B.M., Saez, B., Chow, R.Dz., Prabhu,

M., Gridley, T., and Rajagopal, J. (2015). Parent stem cells can serve as

niches for their daughter cells. Nature 523, 597–601.

8. Kopan, R., and Ilagan, M.X. (2009). The canonical Notch signaling

pathway: unfolding the activation mechanism. Cell 137, 216–233.

9. Gordon, W.R., Vardar-Ulu, D., Histen, G., Sanchez-Irizarry, C., Aster, J.C.,

andBlacklow, S.C. (2007). Structural basis for autoinhibition of Notch. Nat.

Struct. Mol. Biol. 14, 295–300.

10. Tiyanont, K., Wales, T.E., Aste-Amezaga, M., Aster, J.C., Engen, J.R., and

Blacklow, S.C. (2011). Evidence for increased exposure of the Notch1

metalloprotease cleavage site upon conversion to an activated conforma-

tion. Structure 19, 546–554.

11. Sasaki, N., Sasamura, T., Ishikawa, H.O., Kanai, M., Ueda, R., Saigo, K.,

and Matsuno, K. (2007). Polarized exocytosis and transcytosis of Notch

during its apical localization in Drosophila epithelial cells. Genes Cells

12, 89–103.

12. Mizuhara, E., Nakatani, T., Minaki, Y., Sakamoto, Y., Ono, Y., and Takai, Y.

(2005). MAGI1 recruits Dll1 to cadherin-based adherens junctions and sta-

bilizes it on the cell surface. J. Biol. Chem. 280, 26499–26507.

13. Benhra, N., Lallet, S., Cotton, M., Le Bras, S., Dussert, A., and Le Borgne,

R. (2011). AP-1 controls the trafficking of Notch and Sanpodo toward

E-cadherin junctions in sensory organ precursors. Curr. Biol. 21, 87–95.

14. P!erez-Gomez, R., Slovakova, J., Rives-Quinto, N., Krejci, A., and

Carmena, A. (2013). A Serrate-Notch-Canoe complex mediates essential

interactions between glia and neuroepithelial cells during Drosophila optic

lobe development. J. Cell Sci. 126, 4873–4884.

15. Banda, E., McKinsey, A., Germain, N., Carter, J., Anderson, N.C., and

Grabel, L. (2015). Cell polarity and neurogenesis in embryonic stem cell-

derived neural rosettes. Stem Cells Dev. 24, 1022–1033.

16. Hatakeyama, J.,Wakamatsu,Y.,Nagafuchi, A., Kageyama,R., Shigemoto,

R., and Shimamura, K. (2014). Cadherin-based adhesions in the apical

endfoot are required for active Notch signaling to control neurogenesis in

vertebrates. Development 141, 1671–1682.

17. Sanders, P.G., Munoz-Descalzo, S., Balayo, T., Wirtz-Peitz, F., Hayward,

P., and Arias, A.M. (2009). Ligand-independent traffic of Notch buffers

activated Armadillo in Drosophila. PLoS Biol. 7, e1000169.

18. Clark, B.S., Cui, S., Miesfeld, J.B., Klezovitch, O., Vasioukhin, V., and Link,

B.A. (2012). Loss of Llgl1 in retinal neuroepithelia reveals links between

apical domain size, Notch activity and neurogenesis. Development 139,

1599–1610.

19. Rodrıguez-Fraticelli, A.E., Bagwell, J., Bosch-Fortea, M., Boncompain, G.,

Reglero-Real, N., Garcıa-Leon, M.J., Andr!es, G., Toribio, M.L., Alonso,

M.A., Millan, J., et al. (2015). Developmental regulation of apical endocy-

tosis controls epithelial patterning in vertebrate tubular organs. Nat. Cell

Biol. 17, 241–250.

20. Miyamoto, Y., Sakane, F., and Hashimoto, K. (2015). N-cadherin-based

adherens junction regulates the maintenance, proliferation, and differenti-

ation of neural progenitor cells during development. Cell Adhes. Migr. 9,

183–192.

21. Lopez-Schier, H., and St Johnston, D. (2001). Delta signaling from the

germ line controls the proliferation and differentiation of the somatic follicle

cells during Drosophila oogenesis. Genes Dev. 15, 1393–1405.

22. Del Bene, F., Wehman, A.M., Link, B.A., and Baier, H. (2008). Regulation of

neurogenesis by interkinetic nuclear migration through an apical-basal

notch gradient. Cell 134, 1055–1065.

23. Cohen, M., Georgiou, M., Stevenson, N.L., Miodownik, M., and Baum, B.

(2010). Dynamic filopodia transmit intermittent Delta-Notch signaling to

drive pattern refinement during lateral inhibition. Dev. Cell 19, 78–89.

24. Khait, I., Orsher, Y., Golan, O., Binshtok, U., Gordon-Bar, N., Amir-

Zilberstein, L., and Sprinzak, D. (2016). Quantitative analysis of Delta-

like 1 membrane dynamics elucidates the role of contact geometry on

Notch signaling. Cell Rep. 14, 225–233.

25. Schweisguth, F. (2015). Asymmetric cell division in the Drosophila bristle

lineage: from the polarization of sensory organ precursor cells to Notch-

mediated binary fate decision. Wiley Interdiscip. Rev. Dev. Biol. 4,

299–309.

26. Corson, F., Couturier, L., Rouault, H., Mazouni, K., and Schweisguth, F.

(2017). Self-organized Notch dynamics generate stereotyped sensory or-

gan patterns in Drosophila. Science 356, eaai7407.

27. Gho, M., Bellaıche, Y., and Schweisguth, F. (1999). Revisiting the

Drosophila microchaete lineage: a novel intrinsically asymmetric cell divi-

sion generates a glial cell. Development 126, 3573–3584.

28. Gho, M., and Schweisguth, F. (1998). Frizzled signalling controls orienta-

tion of asymmetric sense organ precursor cell divisions in Drosophila.

Nature 393, 178–181.

29. Rhyu, M.S., Jan, L.Y., and Jan, Y.N. (1994). Asymmetric distribution of

numb protein during division of the sensory organ precursor cell confers

distinct fates to daughter cells. Cell 76, 477–491.

30. Weinmaster, G., and Fischer, J.A. (2011). Notch ligand ubiquitylation: what

is it good for? Dev. Cell 21, 134–144.

31. Le Borgne, R., and Schweisguth, F. (2003). Unequal segregation of

Neuralized biases Notch activation during asymmetric cell division. Dev.

Cell 5, 139–148.

32. Couturier, L., Vodovar, N., and Schweisguth, F. (2012). Endocytosis by

Numb breaks Notch symmetry at cytokinesis. Nat. Cell Biol. 14, 131–139.

33. Hartenstein, V., and Posakony, J.W. (1990). A dual function of the Notch

gene in Drosophila sensillum development. Dev. Biol. 142, 13–30.

34. Cotton, M., Benhra, N., and Le Borgne, R. (2013). Numb inhibits the recy-

cling of Sanpodo in Drosophila sensory organ precursor. Curr. Biol. 23,

581–587.

35. Johnson, S.A., Zitserman, D., and Roegiers, F. (2016). Numb regulates

the balance between Notch recycling and late-endosome targeting in

Drosophila neural progenitor cells. Mol. Biol. Cell 27, 2857–2866.

36. Couturier, L., Mazouni, K., and Schweisguth, F. (2013). Numb localizes at

endosomes and controls the endosomal sorting of notch after asymmetric

division in Drosophila. Curr. Biol. 23, 588–593.

37. Berdnik, D., Torok, T., Gonzalez-Gaitan, M., and Knoblich, J.A. (2002). The

endocytic protein alpha-Adaptin is required for numb-mediated asym-

metric cell division in Drosophila. Dev. Cell 3, 221–231.

38. Couturier, L., Trylinski, M., Mazouni, K., Darnet, L., and Schweisguth, F.

(2014). A fluorescent tagging approach in Drosophila reveals late endoso-

mal trafficking of Notch and Sanpodo. J. Cell Biol. 207, 351–363.

39. Bailey, A.M., and Posakony, J.W. (1995). Suppressor of hairless directly

activates transcription of enhancer of split complex genes in response

to Notch receptor activity. Genes Dev. 9, 2609–2622.

40. Herszterg, S., Pinheiro, D., and Bellaıche, Y. (2014). A multicellular view of

cytokinesis in epithelial tissue. Trends Cell Biol. 24, 285–293.

41. Founounou, N., Loyer, N., and Le Borgne, R. (2013). Septins regulate the

contractility of the actomyosin ring to enable adherens junction remodel-

ing during cytokinesis of epithelial cells. Dev. Cell 24, 242–255.




131

42. Giagtzoglou, N., Yamamoto, S., Zitserman, D., Graves, H.K., Schulze,

K.L., Wang, H., Klein, H., Roegiers, F., and Bellen, H.J. (2012). dEHBP1

controls exocytosis and recycling of Delta during asymmetric divisions.

J. Cell Biol. 196, 65–83.

43. Benhra, N., Vignaux, F., Dussert, A., Schweisguth, F., and Le Borgne, R.

(2010). Neuralized promotes basal to apical transcytosis of delta in epithe-

lial cells. Mol. Biol. Cell 21, 2078–2086.

44. Perez-Mockus, G., Roca, V., Mazouni, K., and Schweisguth, F. (2017).

Neuralized regulates Crumbs endocytosis and epithelium morphogenesis

via specific Stardust isoforms. J. Cell Biol. 216, 1405–1420.

45. Bardin, A.J., and Schweisguth, F. (2006). Bearded family members inhibit

Neuralized-mediated endocytosis and signaling activity of Delta in

Drosophila. Dev. Cell 10, 245–255.

46. Lai, E.C., Bodner, R., Kavaler, J., Freschi, G., and Posakony, J.W. (2000).

Antagonism of notch signaling activity by members of a novel protein fam-

ily encoded by the bearded and enhancer of split gene complexes.

Development 127, 291–306.

47. Leviten, M.W., Lai, E.C., and Posakony, J.W. (1997). The Drosophila gene

Bearded encodes a novel small protein and shares 30 UTR sequence mo-

tifs with multiple Enhancer of split complex genes. Development 124,

4039–4051.

48. Kneen, M., Farinas, J., Li, Y., and Verkman, A.S. (1998). Green fluorescent

protein as a noninvasive intracellular pH indicator. Biophys. J. 74, 1591–

1599.

49. Khmelinskii, A., Keller, P.J., Bartosik, A., Meurer, M., Barry, J.D., Mardin,

B.R., Kaufmann, A., Trautmann, S., Wachsmuth, M., Pereira, G., et al.

(2012). Tandem fluorescent protein timers for in vivo analysis of protein dy-

namics. Nat. Biotechnol. 30, 708–714.

50. Wang, S., Moffitt, J.R., Dempsey, G.T., Xie, X.S., and Zhuang, X. (2014).

Characterization and development of photoactivatable fluorescent

proteins for single-molecule-based superresolution imaging. Proc. Natl.

Acad. Sci. USA 111, 8452–8457.

51. Housden, B.E., Li, J., and Bray, S.J. (2014). Visualizing Notch signaling

in vivo in Drosophila tissues. Methods Mol. Biol. 1187, 101–113.

52. Ilagan, M.X., Lim, S., Fulbright, M., Piwnica-Worms, D., and Kopan, R.

(2011). Real-time imaging of notch activation with a luciferase comple-

mentation-based reporter. Sci. Signal. 4, rs7.

53. Rajan, A., Tien, A.C., Haueter, C.M., Schulze, K.L., and Bellen, H.J. (2009).

The Arp2/3 complex and WASp are required for apical trafficking of Delta

into microvilli during cell fate specification of sensory organ precursors.

Nat. Cell Biol. 11, 815–824.

54. Ben-Yaacov, S., Le Borgne, R., Abramson, I., Schweisguth, F., and

Schejter, E.D. (2001). Wasp, the Drosophila Wiskott-Aldrich syndrome

gene homologue, is required for cell fate decisions mediated by Notch

signaling. J. Cell Biol. 152, 1–13.

55. Coumailleau, F., Furthauer, M., Knoblich, J.A., and Gonzalez-Gaitan, M.

(2009). Directional Delta and Notch trafficking in Sara endosomes during

asymmetric cell division. Nature 458, 1051–1055.

56. Loub!ery, S., Seum, C., Moraleda, A., Daeden, A., Furthauer, M., and

Gonzalez-Gaitan, M. (2014). Uninflatable and Notch control the targeting

of Sara endosomes during asymmetric division. Curr. Biol. 24, 2142–2148.

57. Vaccari, T., Lu, H., Kanwar, R., Fortini, M.E., and Bilder, D. (2008).

Endosomal entry regulates Notch receptor activation in Drosophila mela-

nogaster. J. Cell Biol. 180, 755–762.

58. Ohlstein, B., and Spradling, A. (2006). The adult Drosophila posterior

midgut is maintained by pluripotent stem cells. Nature 439, 470–474.

59. Perdigoto, C.N., Schweisguth, F., and Bardin, A.J. (2011). Distinct levels of

Notch activity for commitment and terminal differentiation of stem cells in

the adult fly intestine. Development 138, 4585–4595.

60. Jafar-Nejad, H., Andrews, H.K., Acar, M., Bayat, V., Wirtz-Peitz, F., Mehta,

S.Q., Knoblich, J.A., and Bellen, H.J. (2005). Sec15, a component of the

exocyst, promotes Notch signaling during the asymmetric division of

Drosophila sensory organ precursors. Dev. Cell 9, 351–363.

61. Delwig, A., Bland, C., Beem-Miller, M., Kimberly, P., and Rand, M.D.

(2006). Endocytosis-independent mechanisms of Delta ligand proteolysis.

Exp. Cell Res. 312, 1345–1360.

62. Kiehart, D.P., and Feghali, R. (1986). Cytoplasmic myosin from Drosophila

melanogaster. J. Cell Biol. 103, 1517–1525.

63. Huang, J., Zhou, W., Dong, W., Watson, A.M., and Hong, Y. (2009). From

the cover: directed, efficient, and versatile modifications of the Drosophila

genome by genomic engineering. Proc. Natl. Acad. Sci. USA 106, 8284–

8289.

64. Bosveld, F., Bonnet, I., Guirao, B., Tlili, S., Wang, Z., Petitalot, A.,

Marchand, R., Bardet, P.L., Marcq, P., Graner, F., and Bellaıche, Y.

(2012). Mechanical control of morphogenesis by Fat/Dachsous/Four-

jointed planar cell polarity pathway. Science 336, 724–727.

65. Chanet, S., Vodovar, N., Mayau, V., and Schweisguth, F. (2009). Genome

engineering-based analysis of Bearded family genes reveals both func-

tional redundancy and a nonessential function in lateral inhibition in

Drosophila. Genetics 182, 1101–1108.

66. Couturier, L., and Schweisguth, F. (2014). Antibody uptake assay and

in vivo imaging to study intracellular trafficking of Notch and Delta in

Drosophila. Methods Mol. Biol. 1187, 79–86.




132

STAR+METHODS

KEY RESOURCES TABLE

CONTACT FOR REAGENT AND RESOURCE SHARING

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, FrancoisSchweisguth ([email protected])

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Flies and CRISPR-mediated HRFlies were kept at 25!C, unless noted otherwise. The following flies and transgeneswere used in this study:DlGFP, a GFP knock-in lineproduced by CRISPR-HR (GFP at amino acid position 765 in the intracellular domain; the 3xP3-RFPmarked was still present 30 of theDl gene in this chromosome) [26]; NiGFP4Cherry5, a BAC transgenic line encoding a dual-tagged version of Notch [38]; NeurGFP,two BAC transgenic lines encoding a GFP-tagged version of Neur integrated at positions 22A3 (VK0037) and 99F8 (VK0020) [44];Cad-GFP, a GFP knock-in allele of shotgun [63]; Baz-Cherry, a transgene expressing mCherry-tagged Baz under the control of anubiquitous promoter [64]; Dlg1-GFP (BL-50859); Cherry-MyoII, a transgene expressing Cherry-tagged MyoII under the control ofthe sqh promoter (BL-59024); neur-iRFP670nls, a transgene expressing the far-red fluorescent protein iRFP670 fused to a nuclearlocalization sequence (nls) under the control of the neur regulatory sequences [38]; SensGFP, a BAC transgenic line encodingGFP-tagged Senseless (Sens; BL-38666); other flies were: Ubx-flp; UAS-BrdR [44]; UAS-dsRNA-numb (from R. Ueda); N55e11;Psn143; neurPGal4 [27]; pnr-Gal4 and tub-Gal80ts (BL-7019). FLP/FRT clones were detected using loss of nuclear RFP. To studythe effect of BrdR overexpression, neurPGal4 UAS-BrdR larvae were grown at 20!C to prevent ectopic and lethal expression ofBrdR and staged 0h APF pupae were switched to 29!C to increase Gal4 activity.

REAGENT or RESOURCE SOURCE IDENTIFIER

Antibodies

Goat polyclonal anti-GFP Abcam ab6673; RRID: AB_305643

Rat monoclonal anti-DeltaICD (Intracellular Domain) [61] N/A

Rabbit polyclonal anti-MyoII [62] N/A

Experimental Models: Organisms/Strains

D. melanogaster: NiGFP CRISPR knock-in, with 3xP3-RFP

marker: w NiGFP

This study N/A

D. melanogaster: NimMaple3 CRISPR knock-in: w NimMaple3 This study N/A

D. melanogaster: dual-tagged Notch receptor encoded in a

BAC transgene: w ; ; PBac[y+-attP-9A-NiGFP4Cherry5]

VK00019

[38] N/A

D. melanogaster: DlGFP CRISPR knock-in line, with

3xP3-RFP marker: w ; ; DlGFP

[26] N/A

D. melanogaster: NeurGFP GFP-tagged BAC-encoded

transgenic lines: w; PB[y+ attP-3B NeurGFP] 22A3

and w;; PB[y+ attP-9A NeurGFP] 99F8

[44] N/A

D. melanogaster: nuclear fluorescent marker expressed

under the control of SOP-specific regulatory elements from

the neur gene, in a transgenic line: w neur-iRFP670nls

[38] N/A

D. melanogaster: mutant (stabilized) version of Brd, an

inhibitor of Neur activity, expressed under the control

of UAS: w ; ; UAS-BrdR

[44] N/A

D. melanogaster: dsRNA of numb, under the control of

UAS in a transgenic line: w ; ;UAS-dsRNA-numb

[32]; Japanese National

Institute of Genetics

3779R-3

Oligonucleotides

gRNA50 sequence used for CRISPR-mediated HR at

Notch locus: GAACTGGTGTCGTACGGACT

This study N/A

gRNA30 sequence used for CRISPR-mediated HR at

Notch locus: GTATTTATATATAGCATGTG

This study N/A

e1 Current Biology 27, 1–9.e1–e3, August 7, 2017



133

The BAC encodingGFP-E(spl)m8-HLHwas generated from the attB-P[acman]-Ap BAC encoding thewhole E(spl)-C [65]. This BACwasmodified using recombineeringmediated gap-repair to introduce GFP at the N terminus, with a GVG linker. Recombined regionswere verified by sequencing prior to phiC31-mediated integration at the M{3xP3-RFP.attP}ZH-51D site. The transgenic GFP-E(spl)m8-HLH BAC rescued the lethality associated with a deletion of the E(spl)-C.We used CRIPSR-mediated HR to tag intracellularly the endogenous N gene with GFP and Maple3. The NiGFP and NimMaple3

receptors resulted from the insertion of sfGFP and mMaple3, respectively, at amino acid position 2388 in the intracellular domainof Notch (exactly as in the BAC-encoded NiGFP that we generated earlier; see [32]). CRISPR-mediated GFP knock-in was achievedby injecting three plasmids, a donor template with two gRNAs, into Cas9-expressing embryos. The following gRNAs were selectedusing the Optimal Target Finder tool (http://flycrispr.molbio.wisc.edu/tools):

gRNA50: GAACTGGTGTCGTACGGACTgRNA30: GTATTTATATATAGCATGTG

Oligonucleotides were cloned into pU6-BbsI-chiRNA (Addgene #45946) as described in http://www.addgene.org/crispr/OConnor-Giles/. Donor templates for HR were produced by BAC recombineering in E. coli using a Notch BAC [32]. The PAM se-quences targeted by the gRNAs were mutated in the donor templates to avoid their Cas9-mediated cleavage. The 3xP3-RFPmarkerflanked by loxP sites was produced by gene synthesis and inserted 613 nucleotides downstream of the stop codon of Notch. The flyoptimized sequence of mMaple3 [50] flanked by 80nt-long homology arms was obtained by gene synthesis. Left and right homologyarms for CRISPR-mediated HR were 1 kb long. Additional cloning details will be provided upon request. A mix of donor template(300 ng/ml) and gRNA plasmids (150 ng/ml) was injected into 900-1,200 embryos from the PBac{vas-Cas9}VK00027 stock (BL-51324). Correct recombination events were checked by PCR. These new knock-in alleles of N were viable and fertile and adult fliesdisplayed no obvious developmental defects, indicating that these modified N receptors were active. The alleles used in this studystill carried the 3xP3-RFP marker located 30 of the N gene. These flies are noted NiGFP and NiMaple3.

GenotypesFigure 1

(A-B) NiGFP neur-iRFP670nls / Y(C) w ; M[p3xP3-RFP, GFP-E(spl)m8-HLH]51D neur-iRFP670nls / M[p3xP3-RFP, GFP-E(spl)m8-HLH]51D(D-D’’) Dlg1-GFP / + ; neur-iRFP670nls / + ; ubi-Baz-mCherry / +(F-G’’) NiGFP neur-iRFP670nls / Y ; ; pnrGal4 UAS-dsRNA-Numb / +(H-H’’) NiGFP Ubx-FLP neur-iRFP670nls / Y ; ; FRT82B e Psn143 / FRT82B ubi-RFPnls

Figure 2

(A-A’’) w ; Cad-GFP / neur-iRFP670nls(B-B’’) w ; neur-iRFP670nls / + ; DlGFP / DlGFP

(C-C’’) w ; neur-iRFP670nls / + ; DlGFP / DlGFP

(C-C’’) w ; neur-iRFP670nls / tub-Gal80ts ; DlGFP neurPGal4 / DlGFP UAS-BrdR

(E-E’’) w ; PB[y+ attP-3B NeurGFP] 22A3] neur-iRFP670nls / PB[y+ attP-3B NeurGFP] 22A3] ; PB[y+ attP-9A NeurGFP] 99F8 / +

Figure 3

(C-E) N55e11 / Y ; neur-iRFP670nls / + ; PBac[y+-attP-9A-NiGFP4Cherry5]VK00019 / +

Figure 4

(B-F, I-J) NiGFP neur-iRFP670nls / Y(H) NiGFP / Y ; neur-iRFP670nls / + ; ubi-Baz-mCherry / +

Figure 5

(B-C’) NiMaple3 / Y(D-H) NiMaple3 / Y ; SensGFP / +

METHOD DETAILS

Immunostaining and live imagingPupaewere collected at puparium formation (0h APF, hours After Puparium Formation) and processed for staining and live imaging at16h30 APF for SOP analysis in the notum (at 25!C; or 14h APF at 29!C) and at 15h APF for SOPs of the pupal legs.

Current Biology 27, 1–9.e1–e3, August 7, 2017 e2



134

Staged pupae were dissected according to standard procedures. The following antibodies were used: GFP (goat, 1:1000, Abcam),DlICD (rat, 1:1000, gift from M. Rand) [61] and MyoII (rabbit, 1:2000, gift from D. Kiehart) [62]. Secondary antibodies (Jackson Immu-noResearch Laboratories) were used at 1:1000 dilution.

Live pupae were staged, mounted and imaged as described in [66]. Image acquisitions were performed at 20 ± 2!C, using a laserscanning confocal microscope (LSM780; Zeiss) with a 63x (Plan APO, N.A. 1.4 DIC M27) objective. In all experiments, t = 0 was themetaphase-anaphase transition.

Tomeasure the apical and lateral pools of GFP-tagged Notch aswell as GFP-tagged NICD, we quantifiedGFP signals using large zstacks (Dz = 0.5 mm) encompassing the whole epithelium. To avoid bleaching, these cells were imaged at a single time point. Bycontrast, nuclear GFP was measured over time in NiGFP4Cherry5 pupae in small z stacks centered at the nucleus level (from t10to t40). Nuclei were identified using the SOP-specific iRFP670nls marker. ROIs delimiting the nuclei were drawn manually. GFP sig-nals were measured on 2 images (Dz = 1 mm; sum projection). Similarly, Notch levels at the apical and lateral pIIa-pIIb interface werequantified on maximal projection of 2 images (Dz = 1 mm) using the ‘‘auc’’ function under R software applied to intensity profilesmeasured with ROIs of fixed dimensions that were manually set across the pIIa-pIIb interface. The same approach was applied tomeasure DlGFP and NeurGFP levels.

Photo-bleaching and photo-conversionPhoto-bleaching of NiGFPwas performed using the Zeiss LSM780 FRAPmodule with a 405nmdiode laser (30mW) atmaximal power(no iteration, pixel dwell = 3.21 ms). ROIs corresponding to the apical or lateral area of the pIIa-pIIb interface were bleached at t15, t20and t25. Nuclear and interface GFP levels were quantified at t30. In both notum and legs, apical ROIs were defined using the NiGFPsignal at t15 whereas lateral ROIs were defined based on the position of pIIa/pIIb nuclei.

mMaple3-tagged molecules were photo-converted using a 405nm diode laser (30mW) at 1.8% power (40 iterations, pixel dwell:1.58 msec). In experiments involving photo-conversion of mMaple3-tagged Notch at the cell cortex, an initial pre-bleach step (561 nmlaser, 20mW at 30%) was performed at t10 on ROIs covering the pIIa and pIIb nuclei to decrease background signals in the acqui-sition channel for photo-converted MimMaple3. Repeated photo-activation was performed at t10, t15, t20 and t25, and nuclear sig-nals were acquired at t30 using two-slice z stacks (Dz = 0.5 mm) centered on pIIa and pIIb nuclei. In initial experiments, we found thatthe laser used for photo-converted NimMaple3 also excited the iRFP670nls far-red marker used to identify SOP. This prevented usfrom measuring photo-converted NimMaple3 in nuclei marked by iRFP670nls. We therefore used SensGFP as a SOP-specificmarker. However, since SensGFP remained partly cytoplasmic after mitosis, the apical pIIa-pIIb interface could not be identifiedin the notum.

Images were then processed using Fiji scripts. First, ROIs delimiting the pIIa and pIIb nuclei were automatically drawn and photo-converted NimMaple3 signals were measured on 2 images (Dz = 1 mm; sum projection). We then applied an intensity threshold thatwas calculated so as to select a given percentage of pixels in pIIa (from 100% to 5%). This threshold value was then subtracted to allpixel values in pIIa and pIIb nuclei (to substract noise) and pixels with positive intensity values were kept to calculate a mean pIIa/pIIbratio. The same imaging processing was applied to the no UV-illumination controls as well as to the NiGFP4Cherry5 signals.

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical details (n, number of cells, and p values) aswell as dispersionmeasures (standard error to themean, s.e.m.) are given in thefigures and figure legends. Statistical significance was tested by t tests or Wilcoxon tests depending on the normal distribution of thedata sample as determined using a Shapiro test. Statistical significancewas represented as follow: * p < 0.05, ** p < 0.01, *** p < 0.001

e3 Current Biology 27, 1–9.e1–e3, August 7, 2017



135

Current Biology, Volume 27

Supplemental Information

Intra-lineage Fate Decisions Involve Activation

of Notch Receptors Basal to the Midbody

in Drosophila Sensory Organ Precursor Cells

Mateusz Trylinski, Khalil Mazouni, and François Schweisguth


136

Figure S1: Localization of NiGFP relative to the midbody (related to Figure 1)

A-A’’) NiGFP (green; white in right panels) transiently accumulated at the apical pIIa-pIIb interface in living

pupae at t15 (A) and along the lateral interface (arrows in A’,A’’). The apical pIIa-pIIb junction was marked

with BazCherry (magenta; pIIa and pIIb nuclei were marked with iRFP670nls, red). NiGFP co-localized with

BazCherry into dots along the lateral pIIa-pIIb interface (A’).

B-B’’) The apical and lateral pools of NiGFP (green) were defined relative to the position of the midbody,

marked here by MyoIICherry (magenta, arrow in B’ left) in living pupae at t15. The nuclear iRFP670nls marker

(red/white) also weakly accumulated at the midbody (B’, right). This marker was used to define the apical and

lateral pools of NiGFP.

NiGFP iRFP670nlsMyoIICherryNiGFP iRFP670nlsBazCherry

A

A’

A’’

B

B’

B’’

apical

basal

z section

apical

midbody

basal

NiGFP iRFP670nls


137

Figure S2: Distribution of DlGFP and NeurGFP (related to Figure 2)

A,B) Quantification of nuclear NiGFP in the pIIa and pIIb nuclei during cytokinesis (t0, metaphase-anaphase

transition) in wild-type (green; same data as shown in Figure 1B), numbRNAi (A, red; n=22) and neur>BrdR (Brd

o/e in B, blue; n=12) pupae. Ectopic NICD was detected in the anterior ‘pIIb’ cell upon loss of numb activity

(A). Note that NICD levels were lower in pIIb than in pIIa, possibly reflecting the bias in Dl signaling resulting

from the unequal segregation of Neur [S1]. Conversely, NICD was not detected in the posterior ‘pIIa’ cell upon

inhibition of Neur (B). Data are represented as mean ± SEM.

C,C’) Quantification of the apical (C) and lateral (C’) pools of DlGFP in wild-type (green; n≥15) and upon

overexpression of BrdR (Brd o/e, blue; n≥21). Very low DlGFP levels were detected in wild-type cells. Upon

inhibition of Neur by BrdR, elevated DlGFP levels were measured at the lateral pIIa-pIIb interface (C’). Data are

represented as mean ± SEM.

D,D’) Quantification of NeurGFP levels at the apical (D) and lateral (D’) contact regions (n≥6). Decreasing Neur

levels were detected at the lateral pIIa-pIIb interface (D’). Neur eventually localized at the apical cortex (D).

Data are represented as mean ± SEM.

5 10 15 20 25 30

6080

100

120

140

160

5 10 15 20 25 30

4050

6070

80

time (min)

time (min)

D

D’

wild-type

wild-type

5 10 15 20 25 30

050

100

150

200

5 10 15 20 25 30

050

100

150

200

time (min)

time (min)

C

C’

wild-type

wild-type

Brd o/e

Brd o/e

apical pool - DlGFP apical pool - NeurGFP

lateral pool - DlGFP lateral pool - NeurGFP

10 15 20 25 30

6070

8090

100

●

●

●

●●

●

●

● ●

●

●● ●

●●

●● ● ●

●

pIIa wild-typepIIb wild-type

pIIa Brd o/e pIIb Brd o/e

time (min)

GFP

inte

nsity

(a.u

.)

10 15 20 25 30

6070

8090

100

●

●

●

●●

●

●

● ●

●

●

●

● ●●

●●

●

●●

time (min)

GFP

inte

nsity

(a.u

.)

pIIa wild-type

pIIb wild-typepIIb numbRNAi

pIIa numbRNAiA

B nuclear Notch

nuclear Notch

GFP

inte

nsity

(a.u

.)G

FP in

tens

ity (a

.u.)

GFP

inte

nsity

(a.u

.)G

FP in

tens

ity (a

.u.)


138

Figure S3: Photo-bleaching of NiGFP in the pupal notum (related to Figure 4)

A-D’) Time series of repeated photobleaching of NiGFP (green) at t15 (A,C), t20 (A’,C’) and t25 (A’’,C’’)

followed by quantification of NiGFP levels in the nucleus and along the pIIa-pIIb interface at t30 (B,D: surface

views; B’,D’: views taken at the level of the nuclei; iRFP670nls, red). The apical (A-B’) and basal (C-D’) pools

of NiGFP were photobleached in separate experiments (pre, pre-bleaching views; post, post-bleaching views).

Photo-bleached area are shown as enlarged insets (fluorescence signals are color-coded). These showed both a

loss of fluorescence signals upon bleaching and fluorescence recovery.

E-H) Photo-bleaching the apical pool of NiGFP only once at t15 led to reduced fluorescence signal at the apical

(E) but not lateral pIIa-pIIb interface (E’) at t30 (blue, fluorescence levels after photo-bleaching; green, no

photo-bleaching control). It also had no significant effect on nuclear NiGFP levels in pIIa at t30 (F; n≥24). By

contrast, photo-bleaching once the lateral pool at t15 led to low nuclear NiGFP levels in pIIa at t30 (H; n≥18).

The lateral NiGFP fluorescence signal was fully recovered at t30 (G’). Also, no change in fluorescence was seen

for apical NiGFP (G). Thus, a fraction of Notch receptors located along the lateral cortex at t15 appeared to

contribute to the production of NICD present in the nucleus at t30. Data are represented as mean ± SEM.

*** p <0.001

A A’ A’’ B B’

C C’ C’’ D D’

pre post pre post pre post

pre post pre post pre postt15 t20 t25

t15 t20 t25

t30

t30

apical

apical

basal

basal

0 5 10 15 20 25

100

150

200

250

300

0 10 20 30 40

4060

80100

120

0 5 10 15 20 25

100

150

200

250

300

0 10 20 30 40

4060

80100

120

4060

80100

120

4060

80100

120

E E’

G G’

x position (pixels)

x position (pixels)

x position (pixels)

x position (pixels)

controlbleaching

controlbleaching

controlbleaching

controlbleaching

F

H***

apical pool lateral pool nuclear Notch

bleaching: - + - +pIIa pIIb

bleaching: - + - +pIIa pIIb

pIIb pIIa pIIb pIIa

pIIb pIIa pIIb pIIa

apical pool lateral pool nuclear NotchG

FP in

tens

ity (a

.u.)

GFP

inte

nsity

(a.u

.)

GFP

inte

nsity

(a.u

.)G

FP in

tens

ity (a

.u.)

GFP

inte

nsity

(a.u

.)G

FP in

tens

ity (a

.u.)

n.s.

NiGFPiRFP670nls


139


140

Figure S4: Photo-conversion of NimMaple3 and data analysis (related to Figure 5)

A,A’) Nuclear NiGFP levels in pIIa and pIIb cells at t15 and t30 in pupal nota (A) and legs (A’). Nuclear NiGFP

accumulation levels in pIIa were similar in both epithelia.

B,B’) Plots showing the pIIa/pIIb ratio measured at t30 for nuclear NiGFP4Cherry5 (same data as in Figure

3D,E). The GFP (B) and Cherry (B’) fluorescence signals of NiGFP4Cherry5 were treated as described for

photo-converted NimMaple3 (Figure 5). Increasing pIIa/pIIb ratios were seen for GFP upon higher threshold

values, indicating that a small proportion of nuclear pixels contributed to the GFP signal in pIIa (B). By contrast,

no Cherry-fluorescent nuclear NiGFP4Cherry5 signal was detected following the same image processing (B’).

C-D’’) Photoconversion of the apical (C-C’’) and lateral pools (D-D’’) of NimMaple3 (red/false colors) along

the pIIa-pIIb interface in pupal legs (UV illumination in boxed areas). The pIIa and pIIb nuclei were marked

using SensGFP (green in C,D). A faint signal was detected in the boxed area following UV illumination (C’,D’:

pre-UV illumination; C’’,D’’: post-UV illumination).

E,E’) Plots showing the nuclear signal measured for photo-converted NimMaple3 in pIIa (E) and pIIb (E’) at t30

using a high threshold value (allowing to retain the 5% brightest pixels in the pIIa nucleus; correspond to the

data plotted as pIIa/pIIb ratios in Figure 5H). Background signals were measured in the no-UV control (black).

UV-illumination of the apical contact region (red) led to a weak but significant signal increase in pIIa (red in E)

but not pIIb (red in E’). A stronger increase was seen upon UV-illumination of the lateral pIIa-pIIb interface

(blue). Thus, selection of the brightest pixels indicated that only a fraction of the pixels detected the photo-

converted NimMaple3 signal in the pIIa nucleus. This in turn showed that receptors located along the lateral

pIIa-pIIb interface contributed to NICD production in pIIa whereas receptors located at the apical contact area

might contribute to a lesser extent.

*** p <0.001


141

Genotypes shown in Supplemental Figures

Figure S1

A-A’’) NiGFP / Y ; neur-iRFPnls / + ; ubi-Baz-mCherry / +

B-B’’) NiGFP neur-iRFPnls / Y ; ; sqh-mCherry-MyoII / +

Figure S2

A) related to Figure 1F-G’’

B) NiGFP neur-iRFPnls / Y ; tub-Gal80ts/+ ; neurPGal4 / UAS-BrdR

C-C’) related to Figure 2C-C’’ (green curves) and Figure 2D-D’’ (blue curves)

D-D’) related to Figure 2E-E’’

Figure S3

A-H) NiGFP neur-iRFPnls / Y

Figure S4

A) related to Figure 1B

A’) NiGFP neur-iRFPnls / Y

B-B’) related to Figure 3D-E

C-D’’) NiMaple3 / Y ; SensGFP / +

E-E’) related to Figure 5H

Supplemental References

S1. Le Borgne, R., and Schweisguth, F. (2003). Unequal segregation of Neuralized biases Notch activation during asymmetric cell division. Dev Cell 5, 139-148.

142

Chapter 5

Paper 2: The Arp2/3 complex couples cytokinesis to Notch activation in the

Drosophila sensory organ lineage

Chapter 5. Results. The Arp2/3 complex couples cytokinesis to Notch activation in the Drosophila sensory organ lineage

143

The Arp2/3 complex couples cytokinesis to Notch activation in the Drosophila sensory

organ lineage

Mateusz Trylinski1,2,3 and François Schweisguth2,3,4

1. Sorbonne Université, Collège Doctoral, F-75005, France

2. Institut Pasteur, Paris, F-75015, France

3. CNRS, UMR 3738, Paris, F-75015, France

4. corresponding author ([email protected])


144

Summary Notch signaling directs fate decisions between daughter cells in numerous lineages across the

animal kingdom. How Notch-dependent fate acquisition is coordinated with lineage

progression is currently not known. In Drosophila, following asymmetric cell division of the

sensory organ precursor cell (SOP), activation of Notch receptors at cytokinesis is necessary

to correctly establish cell identities. However, the molecular basis for this interplay remains

undetermined. Here, we show that the Arp2/3 complex acts as molecular pivot between

cytokinesis and Notch activation. By quantifying three distinct readouts at cytokinesis

(contact expansion, actin polymerization and Notch activity), we were able to show that

Arp2/3 integrates two inputs to elicit two distinct functions: Rac-SCAR signaling to expand

the interface contact area and WASp signaling to promote Notch activation. In addition, we

demonstrated that WASp-dependent Arp2/3 activity was required to promote Delta

endocytosis at cytokinesis in the signal-sending cell, thereby activating Notch receptors in the

second daughter cell. Thus, given that the requirement of Arp2/3 and WASp in Notch

signaling appears to be restricted to lineages, we propose that cytokinesis-dependent

recruitment of Arp2/3 at the newly formed interface permits WASp-dependent Arp2/3-

mediated Dl endocytosis at the newly-formed interface to ensure Notch activation onset at

cytokinesis.


145

Introduction Notch signaling is an evolutionary conserved pathway regulating fate decisions in lineages

among Metazoans. Notch receptors are transmembrane proteins made of an extracellular

domain that binds to ligands (Delta (Dl) and Serrate (Ser) in Drosophila) harbored at the

surface of a neighboring cell and a membrane-tethered intracellular domain released upon

receptor activation (Kovall et al., 2017). Following receptor-ligand binding, a pulling force

exerted on the ligand-receptor complex by ligand endocytosis unfolds a buried cleavage site

on the receptor extracellular domain (Langridge and Struhl, 2017; Lovendahl et al., 2018).

The receptor then undergoes a sequence of proteolytic cleavages leading to the release of

Notch intracellular domain (NICD) in the cytosol. Finally, NICD translocates to the nucleus

and binds to a CSL protein (CBF-1, Suppressor of Hairless, Lag-2) to direct target gene

expression (Bray and Gomez-Lamarca, 2018).

In the context of lineages where Notch-dependent fate decisions are mediated between

the daughter cells, fate specification is constrained by the division rate. In other words, all

Notch target genes required for fate acquisition must be expressed prior to the next division

round. As Notch transcriptional responses are characterized by time-demanding feedforward

interactions (Housden et al., 2013), this raises the question whether cells adopt a specific

strategy to coordinate fate acquisition with lineage progression.

To address this question, we used as a model the Drosophila sensory organ (SO)

lineage, a stereotyped lineage of the fly peripheral nervous system (Hartenstein and Posakony,

1989). In the notum, a monolayered epithelium covering Drosophila back, sensory organ

precursor cells (SOPs) are first singled out among proneural clusters through lateral inhibition

(Cabrera, 1990; Heitzler and Simpson, 1991; Simpson, 1990). Later on, they undergo a finite

number of stereotyped asymmetric cell divisions (ACDs) where binary fate choices are

regulated by directional Notch signaling (Gho et al., 1999; Guo et al., 1996; Hartenstein and

Posakony, 1990). SOPs first divide along the anterior-posterior axis to produce a posterior

pIIa cell where Notch is activated and an anterior pIIb cell where Notch is turned off (Figure

1a-b’) (Schweisguth, 2015). Notch activity is biased between the daughter cells by the

unequal segregation in pIIb of two cell fate determinants, Neuralized (Neur) and Numb (Le

Borgne and Schweisguth, 2003; Rhyu et al., 1994). Neur is an E3 ubiquitin ligase promoting

Delta (Dl, the key ligand in this context) endocytosis and thereby increasing the signal-


146

sending activity of pIIb (Le Borgne and Schweisguth, 2003). Numb, on the other hand,

inhibits Notch recycling towards the cell surface and promotes receptor targeting towards

degradation in pIIb, hence avoiding ectopic activation of Notch receptors in pIIb (Cotton et

al., 2013; Couturier et al., 2013, 2014).

Cells composing the SO lineage divide approximately every 2 hours (Gho et al.,

1999). In particular, following SOP division, pIIb enters mitosis within 120 minutes. As it is

the single signaling source for pIIa, this implies that the pIIa fate must be specified within this

time window. Moreover, a functional study probing the temporal requirement of Notch

activity in pIIa provided evidence that Notch receptors must be activated in pIIa within a 30-

minute time window after SOP division to ensure robust fate specification (Remaud et al.,

2008). When activation was delayed by one hour, only a few normal lineages were recovered.

Consistently, we recently showed by monitoring GFP-tagged Notch receptors in vivo that

NICD starts accumulating in pIIa nucleus at cytokinesis following telophase completion

(Couturier et al., 2012; Trylinski et al., 2017). Collectively, these observations point towards

an interplay between cytokinesis and Notch activation that coordinates fate acquisition with

lineage progression.

Mechanistically, this interplay might rely on the Arp2/3 complex, an actin regulator

polymerizing branched actin network (Pollard, 2007). Arp2/3 is a multiprotein complex

composed by the two subunits Arp2 and Arp3 that initiate daughter filament elongation and

five structural subunits Arpc1-5 that mediate binding to the mother filament. By itself, Arp2/3

is weakly active and requires activation by a nucleating promoting factor (NPF) (Pollard,

2007; Rotty et al., 2013). In the SO lineage, Arp3 and the NPF WASp (Wiskott-Aldrich

syndrome protein) are required for Notch-dependent fate acquisition (Ben-Yaacov et al.,

2001; Rajan et al., 2009). The pending model for Arp2/3 and WASp functions proposes that

their molecular activities are required in pIIb to traffic Dl from recycling endosomes towards

the apical side of the pIIa-pIIb interface (Rajan et al., 2009). However, Dl pulse-chase assays

supporting this model reported Dl apical localization after a one-hour chase, which does not

fit with Notch activation at cytokinesis (Couturier et al., 2012; Remaud et al., 2008; Trylinski

et al., 2017). In addition, we recently demonstrated that Notch receptors are activated at the

lateral side of the pIIa-pIIb interface, and not the apical one (Trylinski et al., 2017), which

directly contradicts the aforementioned model. Therefore, we decided to reinvestigate in this

study Arp2/3 function at cytokinesis during SOP division in light of two hypotheses. First,

given that Arp2/3 regulates contact expansion during cytokinesis (Herszterg et al., 2013) and


147

that Notch signaling is a juxtacrine pathway whose activity can scale with the contact area

(Khait et al., 2016; Shaya et al., 2017), we hypothesized that Arp2/3 and WASp are required

to establish and expand the pIIa-pIIb lateral contact to ensure high levels of Notch activity at

cytokinesis. Second, knowing that Arp2/3 and WASp can mediate endocytosis in many

systems (Kaksonen and Roux, 2018), we speculated that their activities might be required to

promote Dl endocytosis in pIIb at cytokinesis.

Results Arp2/3 activity is required for Notch activation in a cell-autonomous manner

To uncover Arp2/3 function at cytokinesis, we first wanted to test whether Arp2/3 activity is

indeed required for Notch activation in pIIa at cytokinesis. To do so, we used a previously

characterized GFP-tagged knock-in allele of Notch (NiGFP) (Trylinski et al., 2017) and

quantified GFP intensities in pIIa and pIIb nuclei as a proxy for Notch activity (Figure 1d-f).

Control cells displayed an initial strong activation step at cytokinesis (t = 10-30 min, t = 0 min

corresponding to metaphase-anaphase transition) before reaching a plateau that stabilized over

time (Figure 1f, green curves). In control pIIb cells, consistently with Notch being turned off,

no GFP fluorescence could be detected over the studied time course. In absence of Arp3, the

initial bulk of activation occurring at cytokinesis in pIIa cells was lost (Figure 1f, orange

curves). Only some residual activation was detected at later time points (t = 30-60 min), but

as loss of Arp3 triggers pIIa-to-pIIb fate changes (Rajan et al., 2009), we assumed that these

activity levels were insufficient to specify the pIIa fate. In addition, aggregates of NiGFP

molecules could be observed in the lateral side of the pIIa-pIIb interface of arp3 cells, but not

in control cells (Figure 1d-e). Since the GFP is intracellularly inserted in the protein sequence,

such aggregates might reflect the accumulation of unprocessed Notch molecules at the cell

surface, as reported in pIIa-pIIb pairs lacking gamma-secretase activity (Trylinski et al.,

2017). Thus, Arp2/3 activity is required to elicit high Notch activity levels at cytokinesis.

Next, we wondered whether this loss of Notch activity at cytokinesis was due to loss

of Arp2/3 activity in the SOP progeny or in surrounding epidermal cells. Indeed, cytokinesis

in epithelia can be viewed as a multicellular process where immediate neighbors play a

crucial role in setting the geometry of the adhesive contact between the daughter cells

(Founounou et al., 2013; Herszterg et al., 2013). In brief, during cytokinetic ring constriction,

myosin activity in immediate neighbors helps in juxtaposing the membranes of the dividing


148

cell and thereby facilitates formation and expansion of the new adhesive contact. Hence,

Arp2/3 activity in neighboring epidermal cells might theoretically contribute to set the contact

geometry in dividing SOPs. Then, in accordance with the first hypothesis, loss of Arp3 in

immediate neighbors might affect the pIIa-pIIb contact area, and consequently Notch activity

levels. To test this idea, we compared late pIIa/pIIb nuclear ratios (when residual Notch

activity is detected) of arp3 cells whose interface was surrounded either by two, one or no

mutant cell (Figure S1). If Arp2/3 activity is required in immediate neighbors, the genotype of

the neighbors should affect the values of the pIIa/pIIb ratios, with decreased values when

SOPs are surrounded by arp3 epidermal cells. As no difference could be observed between

the different conditions (Figure S1a), we concluded that Arp2/3 activity is required in a cell-

autonomous manner for Notch activation at cytokinesis.

Arp2/3 activity is required for contact formation and expansion at cytokinesis

Given that Arp3 is required in an autonomous manner in pIIa-pIIb pairs for Notch activation

and that Arp2/3 regulates cell-autonomously contact expansion at cytokinesis in epidermal

cells of the notum (Herszterg et al., 2013), we next sought to test whether Arp2/3 activity was

also setting the geometry of the lateral contact between pIIa and pIIb.

To do so, we used the FLP/FRT system to generate mosaic epithelia and selected

control or arp3 SOPs surrounded by the negative clone marker PH-ChFP (humain pleckstrin

membrane domain tagged with ChFP) (Figure 1g-h’, t = 0 min). PH-ChFP labeling

membranes, we could monitor the intercalation of immediate neighbors in between pIIa and

pIIb cells. In absence of this negative marker, due to the diffraction-limited resolution of

conventional fluorescent microscopy, loss of contact between pIIa and pIIb could not be

directly resolved with a GFP-tagged basolateral membrane marker. Therefore, this strategy

allowed us to unambiguously determine the edges of lateral contacts. Consistently with

reports pointing out that Arp2/3-mediated actin polymerization is not required during mitosis

progression (Herszterg et al., 2013; Rosa et al., 2015), arp3 mutant cells did not exhibit major

defects or delays from late anaphase till telophase completion (Figure 1g, h). However, at

cytokinesis, lateral membranes of immediate neighbors remained longer in between arp3

pIIa-pIIb pairs, leading to the formation of markedly reduced contact surface areas (Figure 1h,

h’). At t = 10 minutes, almost no contact could be measured between arp3 pIIa and pIIb cells

(Figure 1h-k). Later on, lateral contact expanded but remained significantly smaller compared

to control cells (Figure 1i). More specifically, these defects were mostly due to a failure in


149

pushing on the immediate neighbors, since decreased contact areas were due to reduced

lateral contact width, but not axial length (Figure 1j-k). Therefore, Arp2/3 activity is required

at cytokinesis to generate large contact surfaces at the lateral pIIa-pIIb interface. In addition,

we found that this functional requirement was consistent with Arp3 recruitment and

localization at the lateral side of the pIIa-pIIb interface at cytokinesis (Figure 1c-c’).

Lastly, Arp2/3 was shown to regulate the formation of an actin-rich structure (ARS)

found at the pIIa-pIIb interface, although this structure was not linked to cytokinesis (Rajan et

al., 2009). In epidermal cells, Arp2/3 activity expands contact by the polymerization of an

actin wave at cytokinesis (Herszterg et al., 2013). Thus, we wanted to know whether the ARS

and the actin polymerization wave correspond to the same structure appearing at cytokinesis,

and whether its formation depends on Arp2/3 activity. To do so, we followed actin dynamics

during SOP division using the F-actin marker utrophin-GFP (utro-GFP, utrophin actin-

binding-domain fused to GFP) and observed in control cells a transient actin polymerization

wave occurring around the midbody after cytokinetic ring closure (Figures 1a-b’, S2a). At

later time points, actin lateral levels decreased (Fig1a-a’, S2a, c) and the ARS was restricted

to apical side of the pIIa-pIIb interface (Figure 1a). As the shape of this actin network

resembled the ARS, we concluded that these two terminologies describe the same structure.

The term ARS will be kept for the rest of this study. In absence of Arp2/3 activity, as

expected from published results (Herszterg et al., 2013; Rajan et al., 2009), the ARS was not

detected at the lateral side of the pIIa-pIIb interface (Figure S1b, c’). Actin levels were close

to background fluorescence (Figure S1c’), suggesting that loss of Arp3 fully prevented the

polymerization of any actin network at this location after telophase completion.

Taken together, these results indicate that Arp2/3 activity regulates lateral contact

expansion in SOPs via the polymerization of an actin branched network.

WASp activity is not required for contact formation and expansion at cytokinesis

WASp role in specifying fates in the SO lineage was shown to be dependent on its Arp2/3

activating domain (Tal et al., 2002). Therefore, we hypothesized that pIIa-pIIb pairs lacking

WASp should display similar phenotypes to arp3 ones. First, Notch activity was identically

affected in WASp pIIa-pIIb pairs (of note, WASp corresponds in this study to a

WASp3/Df(3R)6210 deficient background): the initial bulk of activation was lost while some

residual activity was similarly detected at later time points (Figure 2a-c). Similar to arp3 cells,


150

NIGFP aggregates were found at the lateral side of the pIIa-pIIb interface (Figure 2b),

presumably reflecting the accumulation of unprocessed receptors.

To probe WASp role in lateral contact expansion, we first analyzed contact expansion

in WASp3 pIIa-pIIb pairs surrounded by epidermal cells expressing the Dlg-RFP membrane

marker (Figure 2e-e’). To our surprise, WASp3 pIIa-pIIb pairs did not exhibit any defects in

contact expansion (Figure 2e-h). Membranes of immediate neighbors did not remain between

pIIa and pIIb lateral membranes, and contact surface areas expanded similarly to control cells

(Figure 2e-f). The same experiments are currently carried out in the WASp deficient

background using the UAS/Gal4/Gal80 system to mark specifically membranes of epidermal

cells. Therefore, our preliminary data strongly indicate that WASp is not required for the

Arp2/3-mediated contact expansion. In line with this conclusion, formation of the ARS was

not affected by the absence of WASp (Figure S3). The actin polymerization wave was

detected at the same time points as control cells and with the exact same intensity levels

(Figure S3b’).

Nonetheless, using a GFP-trap line, we could detect WASp at the lateral pIIa-pIIb

contact at cytokinesis, albeit at low levels (Figure 2d-d’’), implying that the protein is present

but does not promote Arp2/3-dependent contact expansion and ARS formation at the pIIa-

pIIb interface. Given that WASp function in specifying fates in the SO lineage depends on

this Arp2/3 activating domain (Tal et al., 2002), it suggests that WASp-dependent Arp2/3

activity regulates another mechanism required for Notch activation. These puzzling results

raised the three following questions: (1) is Arp2/3 function in lateral contact formation and

expansion required for Notch activation at cytokinesis?, (2) is the ARS required for Notch

activation?, and (3) what is the cellular process regulated by WASp through Arp2/3 and

required for Notch activation? The next sections will address each of them.

WAVE and Rac activities are required for contact formation and expansion at

cytokinesis, but not to elicit high Notch activity levels

The Arp2/3 complex can be activated by multiple NPFs to regulate context-specific cellular

processes (Alekhina et al., 2017; Rotty et al., 2013). We therefore reasoned that other

upstream activators than WASp might activate Arp2/3 to direct contact expansion. In

neighboring epidermal cells, polymerization of the Arp2/3-dependent actin wave was shown

to rely on Rac activity (Herszterg et al., 2013). We thus examined whether this dependency


151

was conserved in SOPs, and more importantly whether Rac activity was required for Notch

activation.

In flies, three different genes encode for Rac activity: rac1, rac2 and mtl (Hakeda-

Suzuki et al., 2002). We analyzed contact expansion, ARS formation and Notch activity in a

Rac1 Rac2 homozygous null Mtl heterozygous null (termed hereafter rac) background where

some Rac activity can still be mediated by Mtl (Figure 3) and in SO lineages where the Rac1

dominant negative form Rac1N17 (referred to as Rac1DN hereafter) was co-expressed by the

SOP-specific driver neurPGal4 to interfere with Rac signaling together with the F-actin marker

LifeAct-GFP to visualize actin dynamics and cell contours (Figure S4). Under both

conditions, contact expansion and ARS formation were affected at cytokinesis but with

different penetrance (Figure 3e-h, S4a-e, S5). In the rac background, similar to arp3 pIIa-pIIb

interfaces, neighboring cells were still largely intervening at t = 10 min (Figure 3e), thereby

generating strongly reduced contact areas (Figure 3f-h). Later on, rac lateral interfaces

appeared to expand normally and presented similar geometric properties to control interfaces

from t = 30 min onwards (Figure 3f-h). In Rac1DN pIIa-pIIb pairs, lateral contact areas were

also markedly reduced at t = 10 min (Figure S4c-d). However, later on, contact expansion

appeared to be more affected than in rac pIIa-pIIb interfaces with smaller lateral interface

widths (Figure S4c-d). These defects were reminiscent of the ones observed in absence of

Arp3 (Figure 1j-k). Consistently, formation of the ARS was strongly affected under both

conditions, albeit to a lesser extent in the rac background (Figure S4e, S5) (Of note, the

remaining copy of Mtl in the rac background likely explains the milder defects reported

compared to Rac1N17 pIIa-pIIb pairs). Thus, our results support that Rac-dependent activation

of Arp2/3 at cytokinesis to expand the newly-forming contact is conserved in SOPs. Finally, a

GFP-tagged Rac1 transgene localized similarly to Arp3GFP along the lateral pIIa-pIIb

interface at cytokinesis (Figure 3d-d’), reinforcing the idea of a local activation of Arp2/3 by

Rac1 to mediate contact expansion through an actin polymerization wave (i.e., the ARS).

To our surprise, despite the defects we observed in lateral contact establishment and

unlike in arp3 pIIa cells, Notch signaling was only mildly affected in rac and Rac1DN pIIa

cells (Figure 3a-c, S4f-h). At early time points (t = 10-30 min), activity was indeed

diminished but later recovered at comparable levels with control cells (t = 60 min). Therefore,

expansion of the lateral pIIa-pIIb contact regulated by Rac-mediated Arp2/3 activity is not

required to elicit high levels of Notch activity at cytokinesis. The fact that the initial decrease

in Notch activity correlates with the delay in contact expansion in rac and Rac1DN


152

backgrounds might suggest that the establishment of the lateral contact is required for Notch

activity. Alternatively, in addition to promote Arp2/3-dependent contact expansion, Rac

activity might be specifically required in early time points (10 to 20 minutes) to stimulate

Notch activity.

To deepen our understanding on Arp2/3 regulation at the pIIa-pIIb interface, we

sought to identify the NPF involved in Arp2/3-mediated contact expansion. Indeed, the

Arp2/3 complex is not a direct target of any Rac GTPases, letting the connection between the

two effectors undetermined. In addition, as Rac GTPases have pleiotropic functions in cells,

the phenotypes we reported in rac and Rac1DN genetic backgrounds might be caused by other

Rac targets. Based on the literature, we chose to focus on SCAR/WAVE (suppressor of cyclic

AMP repressor/WASp-family verpolin-homologous protein), a WASP family member that is

found in cells within the WAVE regulatory complex (WRC). Importantly, SCAR is

maintained in an inactive state within the WRC (Campellone and Welch, 2010) and requires

Rac-dependent activation to stimulate Arp2/3 activity (Eden et al., 2002). Lastly, given that

the WRC was found to mediate contact expansion through Arp2/3 in multiple cell types

(Alekhina et al., 2017; Del Signore et al., 2018), we speculated that SCAR was the missing

link between Rac signaling and Arp2/3 activity during lateral contact expansion. Accordingly,

ARS formation was impaired in SCAR pIIa-pIIb pairs (Figure S6). Contact expansion in

SCAR cells is yet to be determined but appears to be affected in SCAR cells expressing utro-

GFP. Indeed, at early time points, pIIa and pIIb cell cortices were found separated, implying

the immediate neighboring cells were still intervening (Figure S6a, t = 10 min). In addition,

consistently with our results obtained in rac and Rac1DN backgrounds, Notch activity was

only mildly affected at early time points in absence of SCAR (Figure 4a-c). Of note, as Notch

signaling levels were less affected in SCAR than in rac and Rac1DN pIIa cells, it suggests that

Rac activity might stimulate Notch activation at early time points.

We therefore conclude that (1) the Rac-SCAR pathway, but not WASp, regulates

Arp2/3-mediated contact expansion in dividing SOPs, (2) this process does not play a key role

in triggering high levels of Notch activity at cytokinesis and (3) contrary to what was

proposed formerly (Rajan et al., 2009), the ARS as structure is not required to organize the

subcellular site for Notch activation. As a consequence, only the activation of Arp2/3 by

WASp appears to be required for strong Notch activation at cytokinesis.


153

WASp and Arp2/3 promote Delta endocytosis at cytokinesis

Given that the activation of Arp2/3 by the Rac-SCAR pathway does not explain the

requirement of Arp3 for Notch activation in the SO lineage, we sought to test our second

hypothesis where WASp and Arp2/3 promote Dl endocytosis at the pIIa-pIIb interface in pIIb,

thereby generating the pulling force required for activation of Notch receptors in pIIa. Rajan

et al. reported that Arp2/3 was not regulating Dl endocytosis in SOPs (Rajan et al., 2009),

albeit endocytosis assays were performed at random time points after cytokinesis and were

not quantified. Therefore, whether Arp2/3 and WASp regulate at cytokinesis the endocytosis

of a specific pool of Dl molecules located at the pIIa-pIIb interface remains an open question.

To test our hypothesis, we first analyzed Dl distribution with a GFP-tagged CRISPR

knock-in allele of Dl (DlGFP, (Corson et al., 2017)) in arp3 and WASp backgrounds (Figure

5a-c). In control cells, accordingly with our published data, DlGFP could not be detected at

the pIIa-pIIb interface (Figure 5a) (Trylinski et al., 2017). Noteworthy, this absence of

fluorescence is due to Neur activity instead of the absence of the protein (Trylinski et al.,

2017). Indeed, where Neur activity is impaired in SOPs, DlGFP can be detected in vivo at the

lateral pIIa-pIIb contact, thereby implying that DlGFP is endocytosed in a Neur-dependent

manner prior to GFP maturation (Trylinski et al., 2017). In arp3 and WASp cells, conversely,

DlGFP aggregates were detected at cytokinesis at the pIIa-pIIb lateral contact, likely

reflecting a defect in Dl endocytosis at this specific location (Figure 5b-c). These aggregates

further suggest that Dl addressing towards the interface is not affected in absence of WASp or

Arp2/3 activity.

Then, to directly test whether Dl endocytosis was impaired in arp3 and WASp pIIb

cells, we designed an antibody uptake assay at cytokinesis to probe Dl endocytosis (Methods,

Figure 4d-d’’) in the time window of Notch initial activation step (Figure 1f, green curves,

10-30 min). As loss of Arp3 and WASp strongly affects this first phase of Notch activity, we

expected that WASp and Arp2/3 activities would regulate Dl endocytosis within this period.

Accordingly, we observed a significant decrease of Dl-positive vesicles in arp3 and WASp

pIIb cells after performing our assay (Figure 5e-i). Moreover, the decrease in vesicle number

(Figure 5h) was not compensated by an increased amount of Dl molecules per endosomes or

an increased size of Dl-positive endosomes: total amounts of endocytosed Dl per cell were

also significantly decreased in arp3 and WASp cells (Figure 5i) and endosome volumes and

numbers of Dl molecules per endosome were similar among the different conditions (Figure

S7a-b). Therefore, the decrease in Dl-positive endosome numbers in arp3 and WASp pIIb


154

cells does reflect an impairment in Dl endocytosis. This demonstrates that Arp2/3 and WASp

activities do regulate Dl endocytosis in pIIb at cytokinesis.

It is noteworthy that Dl endocytosis was not abolished, but only decreased in arp3 and

WASp pIIb cells. This implies that Arp2/3 and WASp are not absolutely required for all Dl

endocytic events, but only for a subset of them. Indeed, some residual Notch can still be

detected at later time points in arp3 and WASp pIIa cells (Figure 1f, orange curves and 2c, red

curves), implying that some Notch receptors are still processed in pIIa cells, but at a much

slower rate. Thus, Arp2/3 and WASp have a quantitative effect on Notch signaling rather than

a qualitative one. Furthermore, as Arp3 and WASp are detected at the pIIa-pIIb interface

when Notch signaling starts and that DlGFP aggregates are detected at this location at

cytokinesis in arp3 and WASp backgrounds, we propose that WASp and Arp2/3 function

during cytokinesis is to promote and enhance Dl endocytosis at the pIIa-pIIb interface on the

pIIb side to activate Notch receptors in pIIa.

Collectively, our results favor a model where the Arp2/3 is initially recruited to the

newly-forming interface to expand contact in a Rac-SCAR-dependent manner. In the same

time, WASp-mediated Arp2/3 activation promotes and enhances Dl endocytosis in pIIb,

thereby ensuring high Notch activity levels in pIIa at cytokinesis.

Discussion

In cell lineages, coordinating fate acquisition with lineage progression is key to ensure

maintenance and diversification of cellular identities. In the fly SO lineage, activation of

Notch receptors at cytokinesis emerged as a potential mechanism mediating this interplay

(Couturier et al., 2012; Remaud et al., 2008; Trylinski et al., 2017), although its molecular

details were not determined. Here, by combining live-imaging, quantitative approaches and

genetic manipulations, we find that, during SOP division, the Arp2/3 complex acts as a

molecular pivot between the molecular mechanisms of cytokinesis and Notch activation to

select the pIIa fate in a timely manner. On one hand, as in the epidermal cells of the notum

(Herszterg et al., 2013), Arp2/3 integrates Rac-SCAR signaling to regulate expansion of the

lateral pIIa-pIIb contact (Figure 1, 3-4, S4-6). Counterintuitively, even though Notch activity

was shown to be dependent on contact area in in vitro systems (Shaya et al., 2017), impairing

Arp2/3-mediated contact expansion affects only mildly Notch activity levels (Figure 3-4, S4).

Thus, the conserved function of Arp2/3 in cytokinesis does not explain Arp3 requirement in


155

Notch signaling. On the other hand, we reported in this study that Arp2/3 activation by WASp

enhances Dl endocytosis at the pIIa-pIIb interface and consequently triggers high Notch

activity levels in pIIa at cytokinesis (Figure 5). As these two functions of Arp2/3 are elicited

at the same time, we speculate that the recruitment of Arp2/3 complexes by the cytokinesis

machinery at the pIIa-pIIb interface is hijacked by WASp to enhance locally Dl endocytosis.

In support of this model, WASp and Arp2/3 appear to affect Notch signaling only in

lineages. In the pioneering study characterizing WASp in flies, authors analyzed WASp

pharates (i.e. adults that did not hatch) and the only Notch-related phenotypes they reported

were found in the SO lineage (Ben-Yaacov et al., 2001). Other Notch-dependent cytokinesis-

independent processes, including wing margin formation, leg segmentation or wing vein

specification seemed unaffected. Further analyses during embryogenesis revealed that WASp

was also required in neural lineages to mediate Notch-dependent fate decisions (Ben-Yaacov

et al., 2001), confirming its lineage-specific requirement. In the same line, impairing Arp2/3

activity did not disturb Notch signaling in wing margin formation (Legent et al., 2012) nor

wing vein specification (Gohl et al., 2010). Finally, given that we were able to analyze

isolated SOPs in arp3 and WASp backgrounds, it indicates that lateral inhibition does not

depend either on WASp-dependent Arp2/3 activity. Hence, these context-specific

requirements of WASp and Arp2/3 point towards a general view where their role in

enhancing Dl endocytosis at cytokinesis is conserved among Notch-dependent lineages.

Mechanistically, this specific requirement in lineages implies that Arp2/3 and WASp

are not necessary to mediate Dl endocytosis nor Notch activation. Instead, they rather

potentiate Notch signaling by enhancing Dl endocytosis. Indeed, in arp3 and WASp pIIa cells,

low Notch activity was still detected at later time points (Figure 1f, orange curves and 2c, red

curves) and in their siblings, Dl was still endocytosed, albeit with a lower efficiency (Figure

5h-i). Accordingly, WASp phenotype was found to be rescued when one copy of Hairless, a

Notch antagonist that binds to CSL, was removed (Ben-Yaacov et al., 2001). Therefore, when

the inhibitory threshold is lowered in the WASp background, low Notch activity levels are

sufficient to specify the pIIa fate. This reinforces the idea that Arp2/3 and WASp exert a

quantitative effect on Notch activity, which translates from the developmental viewpoint into

a qualitative output.

Molecularly, our results suggest that WASp and Arp2/3 regulate a specific subclass of

Dl endocytic events which efficiently activate Notch receptors. Recent studies provided

experimental evidence that WASp-dependent Arp2/3-mediated actin polymerization


156

generates forces that are required for membrane invagination of the newly-forming endocytic

vesicle (Mund et al., 2018; Picco et al., 2018). In the context of Notch signaling, forces

generated by WASp and Arp2/3 during Dl endocytosis might exactly meet the requirements

of the mechanostranduction model (Lovendahl et al., 2018), although this remains to be

experimentally tested. Alternatively, as forces generated by endocytosis appear to be variable

(Meloty-Kapella et al., 2012), one might speculate that WASP-Arp2/3-dependent Dl

endocytosis exerts pulling forces in a more reproducible manner than other endocytic

pathways, thereby increasing the probability of Notch activation per endocytic event. Finally,

given that Notch activation in the SO lineage is independent of the Notch-related endocytic

adaptor Epsin (Wang and Struhl, 2004), we propose that the WASp-Arp2/3 pathway is an

alternative mechanism to the Epsin-dependent pathway (Langridge and Struhl, 2017) in

generating forces through ligand endocytosis. This also implicates that different actin

nucleators can drive Dl endocytosis depending on the context.

To conclude, we propose that the Arp2/3 coupling between cytokinesis and Notch

activation might be a conserved mechanism among phyla in neural lineages. During

mammalian corticogenesis, both Arp2/3 and Notch activities are required to maintain radial

glial cell fate and to diversify cell fates (Bigas and Porcheri, 2018; Bultje et al., 2009; Dong et

al., 2012; Wang et al., 2016). In addition, N-WASp is generally required in mammalian

neurogenesis (Jain et al., 2014; Liebau et al., 2011), although the connection with Notch

signaling and intralineage fate decisions remains to be explored. More broadly, given that

Notch signaling also regulates intralineage decisions in multiple tissues, including the thymus

(Radtke et al., 2004), the lung (Pardo-Saganta et al., 2015), the intestine (Ohlstein and

Spradling, 2006), the epidermis (Moriyama et al., 2006) and muscles (Mayeuf-Louchart et al.,

2014), one might speculate that the coupling between cytokinesis and Notch activation is used

in other contexts than neural lineages. Finally, this coupling might have dramatic effects in

cancer lineages where Notch activity is dysregulated. Given that Myc, an oncogenic gene, is a

direct target of Notch signaling in leukemia (Sanchez-Martin and Ferrando, 2017), synergy

between Notch-dependent Myc expression and Notch activation at cytokinesis might be a

potent mechanism of cancer cell proliferation.


157

Material and Methods Flies

Flies were grown at 25°C unless otherwise stated. The following lines were used in this study:

NiGFP, a GFP knock-in line produced by CRISPR-Homology Recombination (HR) where

GFP is inserted at amino acid position 2388 (Trylinski et al., 2017), DlGFP, a GFP knock-in

line produced by CRISPR-Homology Recombination (HR) where GFP is inserted at amino

acid position 765 (Corson et al., 2017), arp3515FC (BL-39727), WASp3 (BL-39725),

Df(3R)Exel6210 (BL-7688), Dlg-GFP (BL-50859), SCARD37 (BL-8754), Rac1J11 Rac2D

FRT2A MtlD (BL6678), UAS-Rac1N17 (BL-6292), Resille-GFP (Buszczak et al., 2006), UAS-

Arp3GFP (BL-39721), UAS-CAAX-mCherry (BL-59021), PH-ChFP (Herszterg et al., 2013),

Dlg-RFP (Pinheiro et al., 2017), utro-GFP (Rauzi et al., 2010), LifeAct-GFP (BL-35554),

LifeAct-Ruby (BL35545) Cherry-MyoII (BL-59024), neur-iRFP670nls (Couturier et al.,

2014), neur-PH-iRFP670 (PLC-g fused to iRFP670 under the control of neur regulatory

sequences), ubxFLP, neurPGal4, pnrGal4 and tub-Gal80ts (BL-7017 and BL-7019).

In Rac1DN experiments, expression of the UAS transgene by neurPGal4 was made

conditional with tub-Gal80ts to avoid embryonic lethality. At 0hAPF, pupae were collected

and grown at 29°C.

To express the membrane marker CAAX-mCherry specifically in epidermal cells, its

tissue-wide expression driven by pnr-Gal4 was inhibited in SOPs by neur-Gal80.

Additionally, to prevent expression of CAAX-mCherry prior to SOP selection, the system

was made conditional with tub-Gal80ts. 0hAPF pupae were kept for 17h at 18°C to inhibit

Gal4 activity during SOP selection, then shifted at 29°C for 7h to allow expression of CAAX-

mCherry prior to SOP divisions.

Live-imaging and quantifications

Pupae were collected at puparium formation (0hAPF, hours After Puparium Formation) and

prepared for live-imaging at 16h30 APF, unless otherwise stated. Pupae were dissected and

mounted according to standard procedures (Couturier and Schweisguth, 2014). Live-imaging

was performed at 20±2°C with a scanning confocal microscope (LSM780, Zeiss) with a 63x

Plan APO N.A. 1.4 DIC M27 objective. The t=0-minute time point corresponds to the SOP

metaphase-anaphase transition.


158

All parameters were quantified using z-stacks (Dz=0.5µm) covering cells from apical

to basal. Notch activity dynamics were determined by measuring the GFP fluorescence in pIIa

and pIIb nuclei, as described in (Trylinski et al., 2017). pIIa-pIIb pairs were imaged at single

time points at prevent photobleaching Specifically, in this study, raw values were pondered by

GFP signals measured at 10 minutes (i.e. when signaling is about to start) to buffer variability

between pupae. Actin levels were quantified on 2-image projections (Dz=1µm) corresponding

to the region just below the midbody. Intensity profiles were obtained after positioning

manually an ROI spanning the pIIa-pIIb interface. Images of the same pIIa-pIIb pairs were

acquired at 10, 20 and 30 minutes. Lateral surfaces were quantified in SOPs surrounded by

epidermal cells expressing a membrane marker. First, lateral interfaces were segmented semi-

automatically slice per slice based on the basolateral marker used in a given experiment.

Edges of the contact were determined using the fluorescent signal emitted by the neighboring

epidermal cells. Segmented interfaces were subsequently processed using the “3D ImageJ

Suite” on Fiji to reconstruct a 3D object and determine its surface area. Segmented interfaces

were further used to determine contact width and contact axial length with homemade Fiji

scripts. Raw data are presented. Images of the same pIIa-pIIb pairs were acquired at 10, 20,

30 and 60 minutes. In Rac1DN cells, edges of interfaces were determined using the weak actin

patches detected at lateral tricellular junctions. Contact width corresponds here to the distance

between these two vertices. Axial length was determined based on the distance between the

midbody and the most-basal actin signal.

Movies were acquired with a 1-minute (Fig1a-b’, 3d) or a 10-second frame rate

(Fig1c-c’, g-h’, 3d-d’, S2a-b, S3a, S4a-b, S5a, S6a) and a Dz=0.5µm step. Images were

furthered filtered with a Gaussian blur (s=1) for presentation.

Dl endocytosis assay

SOPs undergoing the metaphase-anaphase transition were first selected in living pupae within

a 10-minute window to stay in the timing of cytokinesis. After the selection step, pupae were

removed from the microscope and dissected for an antibody uptake assay according to

standard procedures (Couturier and Schweisguth, 2014). Right after being dissected, the

explanted notum was incubated at 25°C for 15 minutes in Schneider medium supplemented

with an antibody recognizing Dl ectodomain (mouse, 594.9B concentrate, 1:10, DHSB).

Antibody uptake was blocked by a 30-minute fixation step (paraformaldehyde 4%). The

tissue was subsequently stained with a goat anti-GFP (1:1000, Abcam) to enhance the


159

basolateral Dlg-GFP signal. Secondary antibodies (Jackson ImmunoResearch Laboratories)

used in this study were diluted at 1:1000.

Initially selected SOPs, which divided during the uptake assay, were tracked back in

fixed tissues based on their position in the SOP pattern and on somatic clone shapes. Dl-

positive vesicles were automatically detected for quantification using the Spot Detector plugin

under Icy (de Chaumont et al., 2012).

References

Alekhina, O., Burstein, E., and Billadeau, D.D. (2017). Cellular functions of WASP family proteins at a glance. J. Cell Sci. 130, 2235–2241.

Ben-Yaacov, S., Le Borgne, R., Abramson, I., Schweisguth, F., and Schejter, E.D. (2001). Wasp, the Drosophila Wiskott-Aldrich Syndrome Gene Homologue, Is Required for Cell Fate Decisions Mediated by Notch Signaling. J. Cell Biol. 152, 1–14.

Bigas, A., and Porcheri, C. (2018). Notch and Stem Cells. In Molecular Mechanisms of Notch Signaling, T. Borggrefe, and B.D. Giaimo, eds. (Cham: Springer International Publishing), pp. 235–263.

Bray, S.J., and Gomez-Lamarca, M. (2018). Notch after cleavage. Curr. Opin. Cell Biol. 51, 103–109.

Bultje, R.S., Castaneda-Castellanos, D.R., Jan, L.Y., Jan, Y.-N., Kriegstein, A.R., and Shi, S.-H. (2009). Mammalian Par3 Regulates Progenitor Cell Asymmetric Division via Notch Signaling in the Developing Neocortex. Neuron 63, 189–202.

Buszczak, M., Paterno, S., Lighthouse, D., Bachman, J., Planck, J., Owen, S., Skora, A.D., Nystul, T.G., Ohlstein, B., Allen, A., et al. (2006). The Carnegie Protein Trap Library: A Versatile Tool for Drosophila Developmental Studies. Genetics 175, 1505–1531.

Cabrera, C.V. (1990). Lateral inhibition and cell fate during neurogenesis in Drosophila: the interactions between scute, Notch and Delta. Dev. Camb. Engl. 110, 733–742.

Campellone, K.G., and Welch, M.D. (2010). A nucleator arms race: cellular control of actin assembly. Nat. Rev. Mol. Cell Biol. 11, 237–251.

de Chaumont, F., Dallongeville, S., Chenouard, N., Hervé, N., Pop, S., Provoost, T., Meas-Yedid, V., Pankajakshan, P., Lecomte, T., Le Montagner, Y., et al. (2012). Icy: an open bioimage informatics platform for extended reproducible research. Nat. Methods 9, 690–696.


160

Corson, F., Couturier, L., Rouault, H., Mazouni, K., and Schweisguth, F. (2017). Self-organized Notch dynamics generate stereotyped sensory organ patterns in Drosophila. Science 356.

Cotton, M., Benhra, N., and Le Borgne, R. (2013). Numb Inhibits the Recycling of Sanpodo in Drosophila Sensory Organ Precursor. Curr. Biol. 23, 581–587.

Couturier, L., and Schweisguth, F. (2014). Antibody Uptake Assay and In Vivo Imaging to Study Intracellular Trafficking of Notch and Delta in Drosophila. In Notch Signaling, H.J. Bellen, and S. Yamamoto, eds. (New York, NY: Springer New York), pp. 79–86.

Couturier, L., Vodovar, N., and Schweisguth, F. (2012). Endocytosis by Numb breaks Notch symmetry at cytokinesis. Nat. Cell Biol. 14, 131–139.

Couturier, L., Mazouni, K., and Schweisguth, F. (2013). Numb Localizes at Endosomes and Controls the Endosomal Sorting of Notch after Asymmetric Division in Drosophila. Curr. Biol. 23, 588–593.

Couturier, L., Trylinski, M., Mazouni, K., Darnet, L., and Schweisguth, F. (2014). A fluorescent tagging approach in Drosophila reveals late endosomal trafficking of Notch and Sanpodo. J. Cell Biol. 207, 351–363.

Del Signore, S.J., Cilla, R., and Hatini, V. (2018). The WAVE Regulatory Complex and Branched F-Actin Counterbalance Contractile Force to Control Cell Shape and Packing in the Drosophila Eye. Dev. Cell 44, 471-483.e4.

Dong, Z., Yang, N., Yeo, S.-Y., Chitnis, A., and Guo, S. (2012). Intralineage Directional Notch Signaling Regulates Self-Renewal and Differentiation of Asymmetrically Dividing Radial Glia. Neuron 74, 65–78.

Eden, S., Rohatgi, R., Podtelejnikov, A.V., Mann, M., and Kirschner, M.W. (2002). Mechanism of regulation of WAVE1-induced actin nucleation by Rac1 and Nck. Nature 418, 790–793.

Founounou, N., Loyer, N., and Le Borgne, R. (2013). Septins Regulate the Contractility of the Actomyosin Ring to Enable Adherens Junction Remodeling during Cytokinesis of Epithelial Cells. Dev. Cell 24, 242–255.

Gho, M., Bellaïche, Y., and Schweisguth, F. (1999). Revisiting the Drosophila microchaete lineage: a novel intrinsically asymmetric cell division generates a glial cell. Dev. Camb. Engl. 126, 3573–3584.

Gohl, C., Banovic, D., Grevelhörster, A., and Bogdan, S. (2010). WAVE Forms Hetero- and Homo-oligomeric Complexes at Integrin Junctions in Drosophila Visualized by Bimolecular Fluorescence Complementation. J. Biol. Chem. 285, 40171–40179.

Guo, M., Jan, L.Y., and Jan, Y.N. (1996). Control of Daughter Cell Fates during Asymmetric Division: Interaction of Numb and Notch. Neuron 17, 27–41.


161

Hakeda-Suzuki, S., Ng, J., Tzu, J., Dietzl, G., Sun, Y., Harms, M., Nardine, T., Luo, L., and Dickson, B.J. (2002). Rac function and regulation during Drosophila development. Nature 416, 438–442.

Hartenstein, V., and Posakony, J.W. (1989). Development of adult sensilla on the wing and notum of Drosophila melanogaster. Dev. Camb. Engl. 107, 389–405.

Hartenstein, V., and Posakony, J.W. (1990). A dual function of the Notch gene in Drosophila sensillum development. Dev. Biol. 142, 13–30.

Heitzler, P., and Simpson, P. (1991). The choice of cell fate in the epidermis of Drosophila. Cell 64, 1083–1092.

Herszterg, S., Leibfried, A., Bosveld, F., Martin, C., and Bellaiche, Y. (2013). Interplay between the Dividing Cell and Its Neighbors Regulates Adherens Junction Formation during Cytokinesis in Epithelial Tissue. Dev. Cell 24, 256–270.

Housden, B.E., Fu, A.Q., Krejci, A., Bernard, F., Fischer, B., Tavaré, S., Russell, S., and Bray, S.J. (2013). Transcriptional Dynamics Elicited by a Short Pulse of Notch Activation Involves Feed-Forward Regulation by E(spl)/Hes Genes. PLoS Genet. 9, e1003162.

Jain, N., Lim, L.W., Tan, W.T., George, B., Makeyev, E., and Thanabalu, T. (2014). Conditional N-WASP knockout in mouse brain implicates actin cytoskeleton regulation in hydrocephalus pathology. Exp. Neurol. 254, 29–40.

Kaksonen, M., and Roux, A. (2018). Mechanisms of clathrin-mediated endocytosis. Nat. Rev. Mol. Cell Biol. 19, 313–326.

Khait, I., Orsher, Y., Golan, O., Binshtok, U., Gordon-Bar, N., Amir-Zilberstein, L., and Sprinzak, D. (2016). Quantitative Analysis of Delta-like 1 Membrane Dynamics Elucidates the Role of Contact Geometry on Notch Signaling. Cell Rep. 14, 225–233.

Kovall, R.A., Gebelein, B., Sprinzak, D., and Kopan, R. (2017). The Canonical Notch Signaling Pathway: Structural and Biochemical Insights into Shape, Sugar, and Force. Dev. Cell 41, 228–241.

Langridge, P.D., and Struhl, G. (2017). Epsin-Dependent Ligand Endocytosis Activates Notch by Force. Cell 171, 1383-1396.e12.

Le Borgne, R., and Schweisguth, F. (2003). Unequal segregation of Neuralized biases Notch activation during asymmetric cell division. Dev. Cell 5, 139–148.

Legent, K., Steinhauer, J., Richard, M., and Treisman, J.E. (2012). A Screen for X-Linked Mutations Affecting Drosophila Photoreceptor Differentiation Identifies Casein Kinase 1α as an Essential Negative Regulator of Wingless Signaling. Genetics 190, 601–616.

Liebau, S., Steinestel, J., Linta, L., Kleger, A., Storch, A., Schoen, M., Steinestel, K., Proepper, C., Bockmann, J., Schmeisser, M.J., et al. (2011). An SK3 Channel/nWASP/Abi-1 Complex Is Involved in Early Neurogenesis. PLoS ONE 6, e18148.


162

Lovendahl, K.N., Blacklow, S.C., and Gordon, W.R. (2018). The Molecular Mechanism of Notch Activation. In Molecular Mechanisms of Notch Signaling, T. Borggrefe, and B.D. Giaimo, eds. (Cham: Springer International Publishing), pp. 47–58.

Mayeuf-Louchart, A., Lagha, M., Danckaert, A., Rocancourt, D., Relaix, F., Vincent, S.D., and Buckingham, M. (2014). Notch regulation of myogenic versus endothelial fates of cells that migrate from the somite to the limb. Proc. Natl. Acad. Sci. 111, 8844–8849.

Meloty-Kapella, L., Shergill, B., Kuon, J., Botvinick, E., and Weinmaster, G. (2012). Notch Ligand Endocytosis Generates Mechanical Pulling Force Dependent on Dynamin, Epsins, and Actin. Dev. Cell 22, 1299–1312.

Moriyama, M., Osawa, M., Mak, S.-S., Ohtsuka, T., Yamamoto, N., Han, H., Delmas, V., Kageyama, R., Beermann, F., Larue, L., et al. (2006). Notch signaling via Hes1 transcription factor maintains survival of melanoblasts and melanocyte stem cells. J. Cell Biol. 173, 333–339.

Mund, M., van der Beek, J.A., Deschamps, J., Dmitrieff, S., Hoess, P., Monster, J.L., Picco, A., Nédélec, F., Kaksonen, M., and Ries, J. (2018). Systematic Nanoscale Analysis of Endocytosis Links Efficient Vesicle Formation to Patterned Actin Nucleation. Cell 174, 884-896.e17.

Ohlstein, B., and Spradling, A. (2006). The adult Drosophila posterior midgut is maintained by pluripotent stem cells. Nature 439, 470–474.

Pardo-Saganta, A., Tata, P.R., Law, B.M., Saez, B., Chow, R.D.-W., Prabhu, M., Gridley, T., and Rajagopal, J. (2015). Parent stem cells can serve as niches for their daughter cells. Nature 523, 597–601.

Picco, A., Kukulski, W., Manenschijn, H.E., Specht, T., Briggs, J.A.G., and Kaksonen, M. (2018). The contributions of the actin machinery to endocytic membrane bending and vesicle formation. Mol. Biol. Cell 29, 1346–1358.

Pinheiro, D., Hannezo, E., Herszterg, S., Bosveld, F., Gaugue, I., Balakireva, M., Wang, Z., Cristo, I., Rigaud, S.U., Markova, O., et al. (2017). Transmission of cytokinesis forces via E-cadherin dilution and actomyosin flows. Nature 545, 103–107.

Pollard, T.D. (2007). Regulation of Actin Filament Assembly by Arp2/3 Complex and Formins. Annu. Rev. Biophys. Biomol. Struct. 36, 451–477.

Radtke, F., Wilson, A., and MacDonald, H.R. (2004). Notch signaling in T- and B-cell development. Curr. Opin. Immunol. 16, 174–179.

Rajan, A., Tien, A.-C., Haueter, C.M., Schulze, K.L., and Bellen, H.J. (2009). The Arp2/3 complex and WASp are required for apical trafficking of Delta into microvilli during cell fate specification of sensory organ precursors. Nat. Cell Biol. 11, 815–824.

Rauzi, M., Lenne, P.-F., and Lecuit, T. (2010). Planar polarized actomyosin contractile flows control epithelial junction remodelling. Nature 468, 1110–1114.


163

Remaud, S., Audibert, A., and Gho, M. (2008). S-Phase Favours Notch Cell Responsiveness in the Drosophila Bristle Lineage. PLoS ONE 3, e3646.

Rhyu, M.S., Jan, L.Y., and Jan, Y.N. (1994). Asymmetric distribution of numb protein during division of the sensory organ precursor cell confers distinct fates to daughter cells. Cell 76, 477–491.

Rosa, A., Vlassaks, E., Pichaud, F., and Baum, B. (2015). Ect2/Pbl Acts via Rho and Polarity Proteins to Direct the Assembly of an Isotropic Actomyosin Cortex upon Mitotic Entry. Dev. Cell 32, 604–616.

Rotty, J.D., Wu, C., and Bear, J.E. (2013). New insights into the regulation and cellular functions of the ARP2/3 complex. Nat. Rev. Mol. Cell Biol. 14, 7–12.

Sanchez-Martin, M., and Ferrando, A. (2017). The NOTCH1-MYC highway toward T-cell acute lymphoblastic leukemia. Blood 129, 1124–1133.

Schweisguth, F. (2015). Asymmetric cell division in the Drosophila bristle lineage: from the polarization of sensory organ precursor cells to Notch-mediated binary fate decision. Wiley Interdiscip. Rev. Dev. Biol. 4, 299–309.

Shaya, O., Binshtok, U., Hersch, M., Rivkin, D., Weinreb, S., Amir-Zilberstein, L., Khamaisi, B., Oppenheim, O., Desai, R.A., Goodyear, R.J., et al. (2017). Cell-Cell Contact Area Affects Notch Signaling and Notch-Dependent Patterning. Dev. Cell 40, 505-511.e6.

Simpson, P. (1990). Lateral inhibition and the development of the sensory bristles of the adult peripheral nervous system of Drosophila. Dev. Camb. Engl. 109, 509–519.

Tal, T., Vaizel-Ohayon, D., and Schejter, E.D. (2002). Conserved Interactions with Cytoskeletal but Not Signaling Elements Are an Essential Aspect of Drosophila Wasp Function. Dev. Biol. 243, 260–271.

Trylinski, M., Mazouni, K., and Schweisguth, F. (2017). Intra-lineage Fate Decisions Involve Activation of Notch Receptors Basal to the Midbody in Drosophila Sensory Organ Precursor Cells. Curr. Biol. 27, 2239-2247.e3.

Wang, W., and Struhl, G. (2004). Drosophila Epsin mediates a select endocytic pathway that DSL ligands must enter to activate Notch. Dev. Camb. Engl. 131, 5367–5380.

Wang, P.-S., Chou, F.-S., Ramachandran, S., Xia, S., Chen, H.-Y., Guo, F., Suraneni, P., Maher, B.J., and Li, R. (2016). Crucial roles of the Arp2/3 complex during mammalian corticogenesis. Development 143, 2741–2752.


164

Figures and Legend

Figure 1: The Arp2/3 complex is required for Notch activation and lateral contact

expansion at cytokinesis

a-b’) Time-lapse images of an SOP expressing utro-GFP, MyoII-Cherry (a-a’) and PH-

iRFP670 (b-b’). t = 0 minute (min) corresponds to the metaphase-anaphase transition. Lateral

(a, b) and orthogonal views (a’, b’) are shown. Anterior is to the left, posterior to the right. In

orthogonal views, apical is to the top, basal to the bottom. Stars in a’, t = 10 min and t = 20

min highlight the apically positioned midbody and the formation of the actin rich structure

(ARS).

c-c’) Time-lapse images of an SOP expressing H2B-RFP and Arp3-GFP. Lateral (c) and

orthogonal (c’) views are shown. Arp3-GFP accumulated at the membrane during lateral

membrane ingression and contact expansion (arrow at t = 5 min, compare with Fig1a-b’, t = 5


165

min). Later on, Arp3-GFP signal is restricted above and below the midbody (stars at t = 10

min and t = 20 min).

d-e’) Snapshots showing the nuclear accumulation of NiGFP in control (d-d’), but not in arp3

pIIa cells (e-e’) at t = 30 min. Clones were negatively marked by the PH-ChFP marker and

outlined by the orange dotted line. pIIa and pIIb nuclei, outlined in d’ and e’, were identified

using iRFP670nls. NiGFP molecules accumulated at the lateral pIIa-pIIb interface in absence

of Arp3 (orange arrow, e).

f) Quantification of NiGFP nuclear levels in control (green curves) and arp3 (orange curves)

backgrounds. High Notch activity levels were not observed in arp3 pIIa cells (t = 10-30

minutes, light orange curve). Weak Notch activity could be detected at later time points (t =

30-60min, light orange curve). Standard errors to the mean are shown (as in all the following

graphs). n≥20 for each time point.

g-h’) Time-lapse images of control (g-g’) or arp3 (h-h’) SOPs expressing Dlg-GFP and

iRFP670nls. Neighboring cells were marked by PH-ChFP (magenta in c, d, grey in c’, d’). In

arp3 pIIa-pIIb pairs, membranes of neighboring cells remained longer in between the lateral

pIIa-pIIb interface (arrows, d’).

i-k) Quantification of the lateral contact geometric properties. Surface area (i), interface width

(j) and interface axial length (k) were markedly reduced in arp3 pIIa-pIIb pairs (orange

curves) compared to control pairs (green curves). Note that arp3 pIIa-pIIb interfaces

preferentially expanded along the z-axis (compare t = 60 min, j and k). n=8 (resp. 15) for

control (resp. arp3) cells.


166

Figure 2: WASp is required for Notch activation at cytokinesis, but not for contact

expansion

a-b’) Snapshots showing the nuclear accumulation of NiGFP in control (a-a’), but not in

WASp pIIa cells (e-e’) at t = 30 min. Clones were negatively marked by the PH-ChFP marker

and outlined by the orange dotted line. pIIa and pIIb nuclei, outlined in a’ and b’, were

identified using iRFP670nls. NiGFP molecules accumulated at the lateral pIIa-pIIb interface

in absence of WASp (orange arrow, e). Images in a, a’ are the same as Fig1d-d’.

c) Quantification of NiGFP nuclear levels in control (green curves from Fig1f) and WASp (red

curves) backgrounds. High Notch activity levels were not observed in WASp pIIa cells (t =

10-30 minutes, red curve). Weak Notch activity could be detected at later time points (t = 30-

60min, red curve). n≥20 for each time point.

d-d’) Snapshots showing weak WASp-GFP (green in c, grey in c’) lateral localization

(arrows) at the lateral pIIa-pIIb interface at cytokinesis.

e-e’) Time-lapse images of WASp3 (e-e’) SOPs expressing Dlg-GFP and iRFP670nls.

Neighboring cells were marked by Dlg-RFP (magenta in e, grey in e’). In WASp3 pIIa-pIIb

pairs, membranes of neighboring cells did not intervene longer in between pIIa and pIIb.

f-h) Quantification of the lateral contact geometric properties. Surface area (f), interface width

(g) and interface axial length (h) were not affected in WASp3 pIIa-pIIb pairs (red curves)

compared to control pairs (green curves, same as in Fig1i-k). n=3 for WASp3 cells.


167

Figure 3: Rac activity regulates contact expansion at cytokinesis, but affects only mildly

Notch activity

a-b’) Snapshots showing the nuclear accumulation of NiGFP in control (a-a’) and in rac pIIa

cells (e-e’) at t = 30 min. GFP levels appears slightly weaker in the rac pIIa cell. Clones were

negatively marked by the PH-ChFP (a) or His-RFP (b) marker and outlined by the orange

dotted line. pIIa and pIIb nuclei, outlined in a’ and b’, were identified using iRFP670nls.

c) Quantification of NiGFP nuclear levels in control (green curves from Fig1f) and rac (light

blue curves) backgrounds. Notch activity was diminished at early time points in rac pIIa cells

(t = 10-30 minutes, light blue curve) but reached similar levels to control pIIa cells at later

time points (t = 30-60min, green and light blue curves). n≥15 for each time point.

d-d’) Time-lapse images of an SOP expressing LifeAct-Ruby and Rac1-GFP (green in d, grey

in d’). Rac1-GFP accumulated at the membrane during lateral membrane ingression and

contact expansion similarly to Arp3GFP (arrows, compare with Fig1a-c’).

e-e’) Time-lapse images of WASp3 (e-e’) SOPs expressing Dlg-GFP and iRFP670nls.

Neighboring cells were marked by PH-ChFP (magenta in e, grey in e’). In rac pIIa-pIIb pairs,

membranes of neighboring cells remained longer in between the lateral pIIa-pIIb interface

(arrows, e’).

f-h) Quantification of the lateral contact geometric properties. Surface area (f), interface width

(g) and interface axial length (h) were markedly reduced in rac pIIa-pIIb pairs (red curves)

compared to control pairs at t = 10 min (green curves, same as in Fig1i-k). Later on, lateral

appeared to expand similarly to control cells. n=5 for rac cells.


168

Figure 4: SCAR is not required for Notch activation at cytokinesis

a-b’) Snapshots showing the nuclear accumulation of NiGFP in control (a-a’) and in SCAR

pIIa cells (e-e’) at t = 30 min. GFP levels appears slightly weaker in the SCAR pIIa cell.

Clones were negatively marked by the PH-ChFP (a) or RFPnls (b) marker and outlined by the

orange dotted line. pIIa and pIIb nuclei, outlined in a’ and b’, were identified using

iRFP670nls.

c) Quantification of NiGFP nuclear levels in control (green curves from Fig1f) and SCAR

(purple curves) backgrounds. Notch activity was only weakly diminished at early time points

in SCAR pIIa cells (t = 10-30 minutes, light blue curve) and reached similar levels to control

pIIa cells at later time points (t = 30-60min, green and light blue curves). n≥16 for each time

point.

10 20 30 40 50 60

1.0

1.2

1.4

1.6

1.8

●

●

●

●

●

● ● ●●

●

●

●

●●

● ●

NiGFP PH-ChFP/RFPnls (neighbor) iRFP670nls

NiGFP

a

30 minlateral

ctrl

pIIb pIIa

pIIb

pIIa

a’

SCARb

b’

c

time (min)

Notch activity

pIIa ctrl

pIIa SCAR

pIIb ctrl

pIIb SCAR

Fig4

norm

alized G

FP

inte

nsity (

a.u

.)


169

05

1015

01000

2000

3000

DlgGFP PH-ChFP (neighbor) iRFP670nlsendo-Dl

e fctrl arp3 WASpg

lateral

d’

d

d’’

DlGFP PH-ChFP (neighbor) iRFP670nls

Fig5

a bctrl arp3 WASpc

ctrl

Number of Dl-vesicles in pIIb Total intensity of endo-Dl per pIIb cell

arp3 WASp ctrl arp3 WASp

*** *** *** ***

h i

in vivo fixed* **

*

*

*

(1) selection of dividing SOPs in vivo (10 min)

(2) dissection and incubationwith anti-Dl (15 min)

(3) Imaging divided SOPsin fixed tissues

anti-DlECD

uptake assay


170

Figure 5: Arp2/3 and WASp regulate Dl endocytosis at cytokinesis

a-c) Snapshots showing the lateral distribution of DlGFP in control (a), arp3 (b) and WASp (c)

pIIa-pIIb pairs at t ≥ 20 min. DlGFP accumulated laterally in arp3 and WASp backgrounds.

PH-ChFP marked epidermal neighbors in b.

d-d’’) Dl endocytosis assay at cytokinesis. Principle of the experiment (d) (adapted from

Besson et al., 2015 and Le Borgne and Schwesiguth, 2003). Snapshots of the same tissue area

before (d’) and after (d’’) the uptake assay. Stars highlight monitored SOPs that divided

during the experiment.

e-g) Snapshots of monitored SOPs during the endocytosis assay in control (e), arp3 (f) and

WASp (g) backgrounds. Bright magenta puncta represent vesicles containing endocytosed Dl

molecules (endo-Dl). DlGFP (green) marks the lateral membranes. Weak magenta signal in e

comes from PH-ChFP expressed by epidermal cells.

h-i) Quantifications of the Dl endocytosis assay. The number Dl-positive vesicles (h) and the

integrated intensity of endo-Dl per cell (i) were decreased in arp3 (orange) and WASp (red)

pIIb cells compared to control (green). *** p<0.001 (Student t-test). n≥16.


171

Supplementary Figures and Legends

Figure S1: The Arp2/3 complex affects Notch activity in a cell-autonomous manner

a) Quantifications of pIIa/pIIb ratios at t = 60 minutes in arp3 pIIa-pIIb pairs whose interface

contacted two arp3 epidermal cells (orange), one arp3 epidermal cell (dark orange), no arp3

cell (brown) and in control pIIa-pIIb pairs (green). No statistical difference was found

between the different arp3 conditions. n≥9.

b-d) Snapshots of the different arp3 conditions. Stars highlight non-mutant arp3 neighbors

(either arp3+/- or arp3+/+). Clones are negatively marked by PH-ChFP and outlined by an

orange dotted line.


172

Figure S2: The Arp2/3 complex is required for the formation of the ARS at cytokinesis

a-b) Time-lapse images showing actin dynamics using utro-GFP in control (a) or arp3 (b)

SOPs. The lateral tail of the ARS is only detected in control pIIa-pIIb pairs (arrow, compare

with Fig1a-a’).

c-c’) Actin intensity profiles in control (c) and arp3 (c’) pIIa-pIIb pairs at cytokinesis. Three

time points were quantified: t = 10 (light colors), 20 (intermediate colors) and 30 (dark colors)

minutes. Note the absence of ARS formation in arp3 pIIa-pIIb interfaces (t = 10 min, light

orange curve). n≥18.

a

b

0 min 5 min 10 min 20 min 30 minlateral

ctrl

Arp3

utroGFP PH-ChFP (neighbor) iRFP670nls

pIIbpIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

FigS2

0 5 10 15 20 25 30

0200

400

600

800

1000

1200

1400

0 5 10 15 20 25 30

0200

400

600

800

1000

1200

1400

Actin intensity profile

Control arp310 min

20 min

30 min

10 min

20 min

30 min

c c’

GF

P in

ten

sity (

a.u

.)

x position (pix) x position (pix)


173

Figure S3: WASP is not required for the formation of the ARS at cytokinesis

a) Time-lapse images showing actin dynamics using utro-GFP in WASp SOPs. The formation

of the ARS appeared unaffected in WASp pIIa-pIIb interfaces (arrow, compare with Fig1a-a’).

b-b’) Actin intensity profiles in control (b, curves from FigS1c) and WASp (b’) pIIa-pIIb pairs

at cytokinesis. Three time points were quantified: t = 10 (light colors), 20 (intermediate

colors) and 30 (dark colors) minutes. n≥18.

0 5 10 15 20 25 30

0200

400

600

800

1000

1200

1400

a


WASputroGFP iRFP670nls

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

FigS3

0 5 10 15 20 25 30

0200

400

600

800

1000

1200

1400


Control WASp10 min20 min30 min

10 min20 min30 min

b b’

GFP

inte

nsity

(a.u

.)



174

Figure S4: Disturbing Rac activity with Rac1DN impairs contact expansion and mildly

affects Notch activity at cytokinesis

a-b) Time-lapse images showing actin dynamics using LifeAct-GFP in control and Rac1DN

SOPs. Both constructs were driven using the SOP-specific neurPGal4 driver. The formation

of the ARS could only be detected in control pIIa-pIIb pairs (arrow).

c-d) Quantification of the lateral contact geometric properties. Interface width (c) and axial

length are strongly reduced in Rac1DN pIIa-pIIb pairs (blue curves) compared to control ones

10 20 30 40 50 60

02

46

810

12

●

●

●

●

●

●

●

●

0 5 10 15 20 25 30

01000

2000

3000

4000

0 5 10 15 20 25 30

01000

2000

3000

4000

10 20 30 40 50 60

23

45

67

●

●

●

●

●

●●

●

a

b

c d e

ctrl

Rac1DN

LifeActGFP

0 min 5 min 10 min 20 min 30 min

time (min) time (min) x position (pix) x position (pix)

Interface lateral width Interface axial lengthActin intensity profile

control Rac1DNctrlRac1DN

ctrl 10 minRac1DN

30 min20 min 20 min

10 min

30 min

10 20 30 40 50 60

1.0

1.2

1.4

1.6

1.8

●

●

●

●

●●

●●

●

●

●

●

●●

●

●

pIIa ctrl

pIIa Rac1DN

pIIb ctrl pIIb Rac1DN

Notch activity

time (min)

NiGFP

NiGFP

30 min

ctrl Rac1DNiRFP670nls

pIIbpIIa pIIb pIIa

f g

f’ g’

h

lateral

pIIbpIIa

pIIb

pIIa

FigS4

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

leng

th (µ

m)

leng

th (µ

m)


175

(green curves). Note that similarly to arp3 cells (Fig1j-k), contact expanded preferentially

along the z-axis.

e) Actin intensity profiles in control (green curves) and Rac1DN pIIa-pIIb pairs (blue curves).

Three time points were quantified: t = 10 (light colors), 20 (intermediate colors) and 30 (dark

colors) minutes. n≥24. Note that at t = 10 min, Rac1DN pIIa and pIIb actin cortices could be

resolved, indicating that the two cells were physically separated (light blue curve).

f-g’) Snapshots showing the nuclear accumulation of NiGFP in control (f-f’) and in Rac1DN

pIIa cells (g-g’) at t = 30 min. GFP levels appears slightly weaker in the Rac1DN pIIa cell. pIIa

and pIIb nuclei, outlined in f’ and g’, were identified using iRFP670nls.

h) Quantification of NiGFP nuclear levels in control (green curves) and Rac1DN (blue curves)

backgrounds. Notch activity was diminished at early time points in Rac1DN pIIa cells (t = 10-

30 minutes, light blue curve) but reached similar levels to control pIIa cells at later time

points (t = 30-60min, green and light blue curves). n≥18 for each time point.


176

Figure S5: Rac activity is required for the formation of the ARS at cytokinesis

a) Time-lapse images showing actin dynamics using utro-GFP in rac SOPs. The formation of

the ARS was not detected in rac pIIa-pIIb interfaces (arrow, compare with Fig1a-a’ and

FigS1a).

b-b’) Actin intensity profiles in control (b, curves from FigS1c) and rac (b’) pIIa-pIIb pairs at

cytokinesis. Three time points were quantified: t = 10 (light colors), 20 (intermediate colors)

and 30 (dark colors) minutes. n≥11.

0 5 10 15 20 25 30

0200

400

600

800

1000

1200

1400

a


racutroGFP iRFP670nls

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

FigS5

0 5 10 15 20 25 30

0200

400

600

800

1000

1200

1400


Control rac10 min

20 min

30 min

10 min

20 min

30 min

PH-ChFP (neighbor)

b b’

GF

P in

ten

sity (

a.u

.)



177

Figure S6: SCAR is required for the formation of the ARS at cytokinesis

a) Time-lapse images showing actin dynamics using utro-GFP in SCAR SOPs. The formation

of the ARS was not detected in SCAR pIIa-pIIb interfaces (arrow, compare with Fig1a-a’ and

FigS1a). Note that pIIa and pIIb

b-b’) Actin intensity profiles in control (b, curves from FigS1c) and SCAR (b’) pIIa-pIIb pairs

at cytokinesis. Three time points were quantified: t = 10 (light colors), 20 (intermediate

colors) and 30 (dark colors) minutes. n≥18. Note that at t = 10 min, SCAR pIIa and pIIb actin

cortices could be resolved with utro-GFP, indicating that the two cells were physically

separated and likely that the pIIa-pIIb lateral contact was affected.

0 5 10 15 20 25 30

0200

400

600

800

1000

1200

1400

a


SCARutroGFP iRFP670nls

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

pIIb

pIIa

FigS6

0 5 10 15 20 25 30

0200

400

600

800

1000

1200

1400


Control SCAR10 min20 min30 min

10 min20 min30 min

RFPnls (neighbor)

b b’

GFP

inte

nsity

(a.u

.)



178

Figure S7: Decreased number in Dl-positive vesicles is not compensated by increased

numbers of Dl molecules per endosome or increased vesicle size

a) Boxplot showing average endo-Dl intensity per endosome in control (green), arp3 (orange)

and WASp (red) backgrounds. Levels are comparable, indicating that the same amounts of Dl

molecules are incorporated per vesicle between the different conditions. n≥16.

b) Boxplot showing volumes of Dl-positive vesicles in control (green), arp3 (orange) and

WASp (red) backgrounds. No significant change was observed between the different

conditions. n≥16.

0100

200

300

400

0.2

0.4

0.6

0.8

1.0

FigS7

ctrl

Intensity of Dl-vesicles in pIIb Volume of Dl-vesicles in pIIb

arp3 WASp ctrl arp3 WASp

a ben

do-D

l int

ensi

ty (a

.u.)

volu

me

(µm

3 )

179

180

Part III

Discussion and Perspectives

181

Chapter 6

Discussion

6.1. Apical VS lateral Notch signaling

The first axis of my thesis consisted in developing a new experimental strategy to identify

Notch activation site in vivo during SOP ACD. This work led to the conclusion that apical

Notch receptors, albeit easily detectable by fluorescence, do not contribute significantly to

signaling in pIIa. On the other hand, the significant pool located laterally along the pIIa-pIIb

interface is barely detectable in vivo. This conclusion raises several points that will be

discussed below. First, it appears that localization of Notch receptors does not, in this context,

reflect their function. However, such conclusion might confuse qualitative function and

biological significance. Second, it questions why Notch receptors are still addressed above the

midbody if they are not required for signaling. Third, why the lateral side of the pIIa-pIIb

interface is favored against the apical one lacks a molecular basis.

6.1.1. Notch receptors: is localization reflecting function?

Investigators often illustrate the tension between protein localization and function by referring

to the American school bus parabola. From a top view, yellow school buses are easily

detectable when concentrated at a parking lot, i.e. when they are inactive. Conversely, once

mobilized for school transport, they become diluted in the traffic and thereby barely

detectable. By analogy, accumulation of fluorophore-tagged proteins at subcellular sites might

lead to erroneous interpretations when the connection between localization and function is set

too rapidly.

Nonetheless, in the case of Notch receptors, the analogy itself might be misleading.

Indeed, it associates accumulation with inactivity and dilution/relocalization with

functionality. When transposed to the context of the SOP division, it suggests that apical

Notch receptors serve as an inactive reservoir for the lateral interface, or any signaling site.

However, photobleaching data (Chapter 4) show that the apical pool and the lateral pool are

Chapter 6. Discussion

182

largely independent since (1) photobleaching the apical pool does not change Notch levels at

the lateral interface and vice versa, and (2) photobleaching/photoconverting the apical pool

affects only weakly and non-significantly fluorescence levels in pIIa nucleus.

In fact, this weak and non-significant contribution might rather hint that Notch

receptors are activated apically (i.e. are functional), but in much lesser amounts compared to

the lateral interface. Following that idea, it would mean that Notch receptors are either

activated more efficiently laterally (discussed below) or in larger numbers at this location,

although they are weakly detected. One explanation to this contrast between Notch

distribution and contribution to signaling might reside in the size of the contact surfaces.

Because Notch receptors are accumulating on a reduced area at the level of AJs forming the

adhesive belt, they can be easily detected (in addition to the fact that they are closer to the

microscope objective). Conversely, as the lateral contact area is 10- to 20-time larger,

receptors would be diluted over a large surface and difficult to observe with conventional

fluorescent microscopy.

Therefore, this contrast between receptor distribution and function might finally be

due to the limitations of our optic systems. It would be intriguing to test in other epithelia

experiencing Notch signaling whether apical accumulation and lateral activation of Notch

receptors is a conserved feature or a cytokinesis-specific phenomenon.

6.1.2. Why an apical pool of Notch receptors in SOPs?

Then, if apical Notch receptors do not contribute significantly to signaling activity in pIIa,

why would cells spend energy to synthetize and traffic them towards the apical AJs? A

comparison with the neighboring epidermal cells reveals that Notch receptors are

physiologically addressed towards this subcellular location in the notum. As SOPs share both

a proneural and an epithelial identity, one might simply infer that apical addressing of Notch

receptors is their default trafficking route in a polarized epithelium.

Therefore, knowing whether Notch receptors are activated laterally or apically in

epithelia might provide valuable information concerning this paradox specific to SOPs.

Indeed, if lateral activation would appear to be specific to cell division, a tantalizing idea

might be that apical activation of Notch receptors is the common mechanism for Notch

signaling in epithelia, albeit with slow dynamics, while lateral activation specifically

occurring at cytokinesis would sustain rapidly high levels of Notch activity.

6.1. Apical VS lateral Notch signaling

183

Experimentally, this question could be addressed with a new set of nanobody-based

GFP-traps, termed GrabFP, which allows the control of transmembrane protein distribution

(Harmansa et al., 2017). These tools consist in the fusion of a lateral or apical membrane

tether with a nanobody recognizing the GFP. Thereby, GrabFPs can be used to force the

localization of GFP-tagged transmembrane proteins either to apical or lateral domains.

Applied to Notch, it could help in understanding the requirement of its polarized distribution

in epithelia.

6.1.3. The lateral contact: a simple matter of area?

Why signaling between pIIa and pIIb is mostly, if not completely, mediated laterally is not

determined in the here presented work. The simplest hypothesis would assume that the lateral

contact is favored due to its large area compared to apical AJs, in accordance with recent

reports (Khait et al., 2016; Shaya et al., 2017). Although this question was not experimentally

addressed in this thesis, reducing the lateral pIIa-pIIb contact area does not seem to strongly

impair Notch activity levels. In Rac1DN pIIa-pIIb pairs (Chapter 5), signaling levels reached

similar values to control cells even though the lateral area was much smaller. In SCAR cells,

Notch activity is even less affected and actin quantifications hint towards similar defects in

contact expansion. In fact, it rather appears that as soon as the lateral contact is set between

pIIa and pIIb, Notch signaling starts. A careful examination of the interplay between lateral

contact areas and signaling dynamics in diverse genetic backgrounds might help in

determining the nature of the correlation, if any, between these two parameters.

Alternatively, the explanation for lateral Notch activation might lie in the distribution

of Notch pathway components. Both Notch and Neur can be detected apically, but not Delta

either in fixed tissues or in vivo in absence of Neur activity (Chapter 4). Thereby, Notch

receptors cannot be activated apically, or only by a weak pool of Delta molecules that I could

not detect. Intriguingly, Delta localizes at the AJs in neighboring epidermal cells. Why it is

weakly detected at this location, even under favorable conditions (Brd overexpression), in

pIIa-pIIb interfaces is a mystery. Analysis of the components forming the exocyst, in

particular Sec15 (Jafar-Nejad et al., 2005), might be a promising direction to understand why

Delta would only be targeted laterally at cytokinesis. On the other side of the membrane,

considering that Notch receptors display a large array of glycosylation patterns in vivo

(Harvey et al., 2016), one might consider that differentially modified receptors enter distinct

trafficking routes and bind to Delta with different affinities.


184

Lastly, local molecular microenvironments might also explain why signaling is

preferentially mediated at the lateral contact. First, composition in polarity proteins is

completely different above and below the midbody. Given that Crumbs was proposed to

prevent Notch ligand-independent activation by direct binding (Nemetschke and Knust,

2016), one might hypothesize that other polarity can modulate signaling either by promoting

or inhibiting activation. Such additional layers of regulation are yet to be determined. Second,

the different domains of the interface might also determine how Notch and Delta distribute at

their surface. The idea that ligand-receptor clustering influences Notch activity has emerged

in the Notch field the last years (Nandagopal et al., 2018; Narui and Salaita, 2013), although

supporting experimental evidence is still lacking. This hypothesis assumes that concentrating

signaling molecules (receptor, ligand, endocytosis effector) in discrete points of a large

contact area translates into larger rates of NICD production. In the context of the pIIa-pIIb

interface, Notch clustering specifically occurring below the midbody could provide a

molecular basis for the high Notch activity levels detected in pIIa. Furthermore, given that

Notch interacts physically with E-Cad (Sasaki et al., 2007) and Spdo (O’Connor-Giles and

Skeath, 2003), distribution of these two transmembrane proteins could be key in organizing

Notch “signaling hubs”. Lastly, organization of the plasma membrane itself may determine

the lateral location of Notch activation site. Specific domains characterized by lipid

compositions were shown in many contexts to influence multiple signaling pathways,

including the famous lipid rafts (Santos and Preta, 2018; Sunshine and Iruela-Arispe, 2017).

In line with this idea, several reports showed an influence of membrane composition on

ligand ability to activate Notch receptors (Chillakuri et al., 2013; Hamel et al., 2010; Suckling

et al., 2017) Whether the pIIa-pIIb interface displays differential lipid compositions along the

apical-basal axis, and whether this correlates with Notch and Delta distributions, would be

exciting and not the least challenging to determine.

6.2. Interplay between cytokinesis and Notch signaling

The two axes of my theses point towards an interplay between cytokinesis and Notch

signaling: the first revealing the location of Notch activation site, the second providing a

molecular mechanism bridging the two processes. Nonetheless several questions


185

remain unaddressed, including the coordination between cytokinesis and exocytosis in the

context of Notch signaling, the molecular details of Arp2/3 dual activation mode at the pIIa-

pIIb interface and the role of Notch activation at cytokinesis in the SO lineage.

6.2.1. Contact formation and exocytosis

During cytokinesis, a new interface needs to be generated to resolve the daughter cells’ own

membranes. This process requires active exocytosis to provide membrane materials (Frémont

and Echard, 2018) and, in the context of an epithelium, presumably adhesive transmembrane

proteins to establish stable cell-cell contacts. In addition to these conserved requirements,

SOPs need to address signaling molecules (Notch, Delta, Neur…) to the newly-forming pIIa-

pIIb interface to ensure that signaling will occur after telophase completion.

Whether this additional requirement relies on a SOP-specific mechanism is currently

not known. One appealing candidate is Sec15, a member of the exocyst. Sec15 was shown to

be required for Notch signaling in the SO lineage, although its potential function at

cytokinesis has been overshadowed by the Delta recycling model (Jafar-Nejad et al., 2005).

Similar to Arp2/3 and WASp, Sec15 was not found to regulate other Notch-dependent

developmental processes and appears to be lineage-specific. Therefore, exploring its function

at cytokinesis using optogenetic-based tools to induce temporally-controlled perturbations,

thereby avoiding earlier pleiotropic effects, might help in understanding the interplay between

Notch signaling at cytokinesis and exocytosis in the context of lineages.

6.2.2. Dual mode of Arp2/3 activation

The model proposed for the interplay between cytokinesis and Notch activation implicates

that Arp2/3 complexes located at the pIIa-pIIb interface might either be activated by

SCAR/WAVE or by WASp. This implication in turn interrogates how at the molecular scale

the two signaling inputs are coordinated to efficiently mediate contact expansion and Delta

endocytosis.

If SCAR and WASp activated simultaneously Arp2/3 complexes at the pIIa-pIIb

interfaces, it could question how a competition between the two NPFs in binding Arp2/3

complexes leads to two cellular processes robustly reproducible throughout SOPs. The

simplest solution would consist in an excess of Arp2/3 molecules compared to the NPFs.


186

Thereby, activated SCAR or WASp would never saturate the total pool of available Arp2/3.

Alternatively, a balance between NPF concentration and affinity for Arp2/3 (for example, low

concentration is compensated by higher affinity) would fine-tune the differential mode of

Arp2/3 function.

On the other hand, SCAR- and WASp-dependent activations might occur sequentially

at cytokinesis. First, Arp2/3 complexes would be saturated by Rac-activated WRC complexes

and only employed to expand the newly-forming contact. Once this contact is stabilized by

adhesive molecules, such as E-Cad, Rac activity would be inhibited and WRC complexes

inactivated. Free Arp2/3 complexes would then be available to be bound by WASp molecules

and initiate Delta endocytosis.

Testing these different hypotheses would require developing an imaging setup with

high temporal and spatial resolutions to resolve NPF distributions along the pIIa-pIIb

interface at cytokinesis as well as to probe their activities using bio-sensors.

6.2.3. Is lineage progression hiding a “competence window”?

Analysis of Notch functional requirement after SOP division clearly showed that Notch

receptors need to be activated at cytokinesis to ensure pIIa fate acquisition prior to the next

division round (Remaud et al., 2008). The most parsimonious hypothesis would assume that

this requirement underlies the time required for the expression of all Notch target genes

specifying the pIIa fate.

During my thesis, I found that the Arp2/3 complex couples Notch activation with

cytokinesis to ensure both Notch activity onset and high activity levels following telophase

completion in pIIa (Chapter 5). Nonetheless, presuming that this mechanism is conserved at

each step of the SO lineage, the constraint of the lineage division rate stops being relevant for

the terminal divisions of pIIa and pIIIb. Indeed, arp3 lineages were reported to produce only

neurons, implying that pIIa-to-pIIb and sheath-to-neuron fate transformations occurred (Rajan

et al., 2009). Similar requirements were found with WASp (Ben-Yaacov et al., 2001).

Therefore, WASp-dependent Arp2/3 activity is also required for fate specification after pIIIb

division, even though the daughter cells will stay interphasic throughout animal life. In

addition, even though Notch is weakly activated in pIIa after division, integration over longer

time periods of this weak signaling activity is not sufficient to specify the sheath fate. This

suggests that additional regulatory mechanisms are in place in pIIIb daughter cells to lock the


187

neuronal fate in absence of strong Notch activity and thereby restrict temporally the response

to Notch signaling.

In light of this interpretation, one might consider that the pIIa-to-pIIb fate

transformation observed in arp3 and WASp SO lineages would also be caused by the

expression of genes that determine the pIIb fate and prevent the expression of pIIa-specific

genes. Thereby, the switch mediated by Notch signaling between these alternative fates would

be restricted in time by a so-called “competence window”. This potential restriction would

reinforce the necessity for Notch receptors to be activated at cytokinesis to ensure Notch

target expression in pIIa prior to the expression by default of “pIIb” genes.

Temporal profiling of gene expression in pIIa and pIIb cells would uncover whether

such transcriptional regulatory mechanisms exist and how they are organized to direct fate

acquisitions and to antagonize the opposite fate.

188

Chapter 7

Perspectives

7.1. A widely-applicable technique to identify Notch activation

site in vivo

Combination of photobleaching and photoconversion of fluorophore-tagged Notch receptors

provides a powerful technique to map Notch activation at the subcellular scale (Chapter 4).

As genome editing strategies have improved in efficiency over the past years following the

fast development of the CRISPR technology, tagging Notch genes at the endogenous locus

with the fluorescent protein of interest became more accessible and allows to monitor

signaling activities at endogenous levels. Advances in imaging systems also helps in precisely

quantifying Notch levels at the cell surface and in the nucleus under its cleaved form, a

prerequisite to apply the phototracking technique.

This section will briefly present applications of this technique to biological problems

related to Notch signaling.

7.1.1. Testing the contact area/Notch activity correlation

In vitro demonstration of the correlation between the contact area and Notch activity levels in

pairs of signal-sending and signal-receiving cells begs the question of its significance in vivo

(Shaya et al., 2017). In their experimental approach, Shaya et al. proceeded by reducing the

contacting interface to a single straight line and observed that Notch activity levels scaled

with the length of the line. Methodically, reducing the dimensionality of the interface (i.e.

from a surface to a line) was key to test such fundamental hypothesis. In vivo, as interfaces in

most biological systems can hardly be considered as mere lines, the distance seems quite

broad to extrapolate these findings. First and foremost, testing this correlation requires

knowing the location of Notch activation site.

To overcome this limitation, the phototracking technique offers a powerful

experimental strategy. Not only it identifies where Notch receptors are activated, but also

7.1. A widely-applicable technique to identify Notch activation site in vivo

189

provides Notch signaling dynamics. Combination with the geometric properties of interfaces

would reveal whether the correlation is valid or not in hypothetically any system allowing

decent live-imaging conditions.

7.1.2. Notch signaling site and regulatory mechanisms

The SO lineage provides a clear example of how untested assumptions, including Notch

apical activation and Delta recycling, have presumably misled the interpretations on the

cellular functions related to Notch signaling of multiple regulators. During my thesis,

identifying Notch activation site at the lateral side of the pIIa-pIIb interface further orientated

my work when reinvestigating Arp3 function during SOP ACD. More generally, all Notch

regulators identified in this experimental model and linked with one of the two

aforementioned assumptions (Benhra et al., 2010, 2011; Jafar-Nejad et al., 2005; Rajan et al.,

2009) should be, in principle, reexamined in light of lateral Notch activation at cytokinesis.

From a wider perspective, determining receptor activation site could be the starting

point in understanding the context-specific regulatory mechanisms governing and organizing

Notch signaling at the protein level. Upstream of the signaling site, trafficking pathways

dictate where receptors, ligands and other essential regulators are targeted. How these

pathways are regulated, for example by receptor or ligand post-translational modifications,

remains largely unknown. At the level of the signaling site, our understanding on how the

membrane microenvironment, including the lipid composition of the membrane bilayer, the

cortical cytoskeleton or neighboring transmembrane or membrane-associated proteins, affects

in vivo Notch activation is scarce.

Thereby, once Notch activation site is established, analysis of regulatory mechanisms

might spread out from and around this subcellular site. Lastly, such knowledge would be

particularly valuable in pathological contexts where common cellular mechanisms become

dysregulated and lead to aberrant Notch activity levels (Siebel and Lendahl, 2017).

Chapter 7. Perspectives.

190

7.2. Arp2/3-mediated Delta endocytosis: a molecular basis for the

mechanotransduction model

The mechanotransduction model for Notch activation necessitates a molecular mechanism

transmitting a pulling force to the receptor-ligand complex. Over the past years, ligand

endocytosis became the prevalent source of such forces (Lovendahl et al., 2018). Nonetheless,

the molecular machinery generating the forces during ligand endocytosis is not determined.

The endocytic adaptor Epsin seems to be required in many different systems (Langridge and

Struhl, 2017; Meloty-Kapella et al., 2012), albeit its molecular activity does not trigger

membrane bending. Actin also appears required (Meloty-Kapella et al., 2012), although actin

nucleators were not characterized.

Work from Chapter 5 provides evidence that WASp-dependent Arp2/3 activity

regulates a subset of Dl endocytic events at the pIIa-pIIb interface. Given that WASp-Arp2/3-

mediated actin polymerization at endocytic sites was recently demonstrated to efficently

induce membrane bending and endosome tubulation (Mund et al., 2018; Picco et al., 2018), it

becomes plausible that Arp2/3 is one of the actin nucleators generating the pulling force.

Implications of this hypothesis are discussed in the following subsections.

7.2.1. Towards an array of Delta endocytosis mechanisms

The fact that WASp and Arp2/3 appear to be required for Notch signaling only in lineages

implies that they are not essential components of Notch activation mechanism. Molecularly,

this implication means that neither Neur-mediated nor Mib1-mediated Delta endocytoses are

in general dependent on the WASp-Arp2/3 pathway. Consistently, in arp3 or WASp pIIb cells,

Delta was still endocytosed, albeit in lower amounts (Chapter 5). Conversely, Epsin is not

required in SO lineages to mediate Notch signaling (Wang and Struhl, 2004).

Collectively, these context-specific requirements of endocytosis regulators indicate

that Delta can enter different endocytic pathways to activate Notch receptors depending on

the cellular environment. Mechanistically, it asks whether all these endocytic pathways

activate Notch receptors with the same efficiencies. In arp3 and WASp pIIb cells, Delta was

still endocytosed while Notch activity was weak, but not abolished, in their siblings (Chapter

5). One interpretation might be that the remaining mechanism driving Delta endocytosis only

7.2. Arp2/3-mediated Delta endocytosis: a molecular basis for the mechanotransduction model

191

weakly activate Notch receptors. In the light of Notch activation, this WASp-Arp2/3-

independent Delta endocytosis would rather be inefficient in reproducibly generating pulling

forces within the 2-10 pN range.

Further investigations are needed to uncover context-specific molecular requirements

for Delta endocytosis in vivo. These findings could then be correlated with the studied

biological context to generate a comprehensive map linking Delta endocytosis pathways with

the features of the Notch-dependent process (inductive signaling, fate specification, time

constraints, location of the signaling site).

7.2.2. Probing the forces generated by WASp-Arp2/3-dependent Delta

endocytosis

SOPs are presumably to best model to probe the range of pulling forces generated by WASp-

Arp2/3-dependent Delta endocytosis. In vitro, one would need to find cell lines where most, if

not all, Delta endocytoses are dependent on WASp-dependent Arp2/3 activity. Such research

might reveal to be highly time consuming. In SOPs, on the other hand, this requirement is

already fulfilled.

Experimentally, forces could be estimated using the strategy based on the

mechanosensitive vWF A2 domain and developed by Langridge and Struhl (Langridge and

Struhl, 2017). Concretely, synthetic pairs of receptor-ligand could be specifically expressed in

SOPs using the neurPGal4 SOP driver where: (1) Notch ECD would be replaced by a vWF

A2 domain fused to a motif recognizing the ligand synthetic ECD (for instance GFP/nano-

GFP, FSH/FSHR) and (2) Notch ICD would be tagged with a fluorescent protein to follow the

activity of the synthetic pair in pIIa. As several A2 domain variants exist with different force

requirements for activation, forces generated by WASp-Arp2/3-dependent (control

background) or independent (arp3 or WASp background) Delta endocytosis at cytokinesis

could be estimated, as well as their efficiency in triggering activation of the chimeric Notch

receptor. Results obtained in SOPs could finally be compared with values found during Epsin-

dependent Delta endocytosis in the wing disc.

Finally, this method could be applied in many different Notch-dependent processes

where signal-sending and signal-receiving cells are clearly identified. By doing so, one could

build a comprehensive map with context-specific force requirements and test in a wide range

of in vivo contexts the mechanotransduction model.

192

193

References

Abella, J.V.G., Galloni, C., Pernier, J., Barry, D.J., Kjær, S., Carlier, M.-F., and Way, M. (2016). Isoform diversity in the Arp2/3 complex determines actin filament dynamics. Nat. Cell Biol. 18, 76–86.

Acar, M., Jafar-Nejad, H., Takeuchi, H., Rajan, A., Ibrani, D., Rana, N.A., Pan, H., Haltiwanger, R.S., and Bellen, H.J. (2008). Rumi Is a CAP10 Domain Glycosyltransferase that Modifies Notch and Is Required for Notch Signaling. Cell 132, 247–258.

Adam, J., Myat, A., Le Roux, I., Eddison, M., Henrique, D., Ish-Horowicz, D., and Lewis, J. (1998). Cell fate choices and the expression of Notch, Delta and Serrate homologues in the chick inner ear: parallels with Drosophila sense-organ development. Dev. Camb. Engl. 125, 4645–4654.

del Álamo, D., Rouault, H., and Schweisguth, F. (2011). Mechanism and Significance of cis-Inhibition in Notch Signalling. Curr. Biol. 21, R40–R47.

Alekhina, O., Burstein, E., and Billadeau, D.D. (2017). Cellular functions of WASP family proteins at a glance. J. Cell Sci. 130, 2235–2241.

Almeida, M.S., and Bray, S.J. (2005). Regulation of post-embryonic neuroblasts by Drosophila Grainyhead. Mech. Dev. 122, 1282–1293.

Andersen, P., Uosaki, H., Shenje, L.T., and Kwon, C. (2012). Non-canonical Notch signaling: emerging role and mechanism. Trends Cell Biol. 22, 257–265.

Aulehla, A., Wiegraebe, W., Baubet, V., Wahl, M.B., Deng, C., Taketo, M., Lewandoski, M., and Pourquié, O. (2008). A β-catenin gradient links the clock and wavefront systems in mouse embryo segmentation. Nat. Cell Biol. 10, 186–193.

Bailey, A.M., and Posakony, J.W. (1995). Suppressor of hairless directly activates transcription of enhancer of split complex genes in response to Notch receptor activity. Genes Dev. 9, 2609–2622.

Baker, N.E., and Yu, S.-Y. (1997). Proneural function of neurogenic genes in the developing Drosophila eye. Curr. Biol. 7, 122–132.

Baker, N., Mlodzik, M., and Rubin, G. (1990). Spacing differentiation in the developing Drosophila eye: a fibrinogen-related lateral inhibitor encoded by scabrous. Science 250, 1370–1377.

Baker, N.E., Yu, S., and Han, D. (1996). Evolution of proneural atonal expression during distinct regulatory phases in the developing Drosophila eye. Curr. Biol. 6, 1290–1302.

Banda, E., McKinsey, A., Germain, N., Carter, J., Anderson, N.C., and Grabel, L. (2015). Cell Polarity and Neurogenesis in Embryonic Stem Cell-Derived Neural Rosettes. Stem Cells Dev. 24, 1022–1033.

References

194

Banerjee, J.J., Aerne, B.L., Holder, M.V., Hauri, S., Gstaiger, M., and Tapon, N. (2017). Meru couples planar cell polarity with apical-basal polarity during asymmetric cell division. ELife 6.

Barad, O., Rosin, D., Hornstein, E., and Barkai, N. (2010). Error Minimization in Lateral Inhibition Circuits. Sci. Signal. 3, ra51–ra51.

Basant, A., and Glotzer, M. (2018). Spatiotemporal Regulation of RhoA during Cytokinesis. Curr. Biol. 28, R570–R580.

Bellaïche, Y., Gho, M., Kaltschmidt, J.A., Brand, A.H., and Schweisguth, F. (2001). Frizzled regulates localization of cell-fate determinants and mitotic spindle rotation during asymmetric cell division. Nat. Cell Biol. 3, 50–57.

Bellaïche, Y., Beaudoin-Massiani, O., Stuttem, I., and Schweisguth, F. (2004). The planar cell polarity protein Strabismus promotes Pins anterior localization during asymmetric division of sensory organ precursor cells in Drosophila. Dev. Camb. Engl. 131, 469–478.

Bellaıche, Y., Radovic, A., Woods, D.F., Hough, C.D., Parmentier, M.-L., O’Kane, C.J., Bryant, P.J., and Schweisguth, F. (2001). The Partner of Inscuteable/Discs-Large Complex Is Required to Establish Planar Polarity during Asymmetric Cell Division in Drosophila. Cell 106, 355–366.

Benedito, R., Roca, C., Sörensen, I., Adams, S., Gossler, A., Fruttiger, M., and Adams, R.H. (2009). The Notch Ligands Dll4 and Jagged1 Have Opposing Effects on Angiogenesis. Cell 137, 1124–1135.

Benhra, N., Vignaux, F., Dussert, A., Schweisguth, F., and Le Borgne, R. (2010). Neuralized Promotes Basal to Apical Transcytosis of Delta in Epithelial Cells. Mol. Biol. Cell 21, 2078–2086.

Benhra, N., Lallet, S., Cotton, M., Le Bras, S., Dussert, A., and Le Borgne, R. (2011). AP-1 Controls the Trafficking of Notch and Sanpodo toward E-Cadherin Junctions in Sensory Organ Precursors. Curr. Biol. 21, 87–95.

Benjamin, J.M., Kwiatkowski, A.V., Yang, C., Korobova, F., Pokutta, S., Svitkina, T., Weis, W.I., and Nelson, W.J. (2010). αE-catenin regulates actin dynamics independently of cadherin-mediated cell–cell adhesion. J. Cell Biol. 189, 339–352.

Ben-Yaacov, S., Le Borgne, R., Abramson, I., Schweisguth, F., and Schejter, E.D. (2001). Wasp, the Drosophila Wiskott-Aldrich Syndrome Gene Homologue, Is Required for Cell Fate Decisions Mediated by Notch Signaling. J. Cell Biol. 152, 1–14.

Berndt, N., Seib, E., Kim, S., Troost, T., Lyga, M., Langenbach, J., Haensch, S., Kalodimou, K., Delidakis, C., and Klein, T. (2017). Ubiquitylation-independent activation of Notch signalling by Delta. ELife 6.

Berry, L.W., Westlund, B., and Schedl, T. (1997). Germ-line tumor formation caused by activation of glp-1, a Caenorhabditis elegans member of the Notch family of receptors. Dev. Camb. Engl. 124, 925–936.

References

195

Besson, C., Bernard, F., Corson, F., Rouault, H., Reynaud, E., Keder, A., Mazouni, K., and Schweisguth, F. (2015). Planar Cell Polarity Breaks the Symmetry of PAR Protein Distribution prior to Mitosis in Drosophila Sensory Organ Precursor Cells. Curr. Biol. CB 25, 1104–1110.

Binshtok, U., and Sprinzak, D. (2018). Modeling the Notch Response. In Molecular Mechanisms of Notch Signaling, T. Borggrefe, and B.D. Giaimo, eds. (Cham: Springer International Publishing), pp. 79–98.

Blanchoin, L., Pollard, T.D., and Mullins, R.D. (2000). Interactions of ADF/cofilin, Arp2/3 complex, capping protein and profilin in remodeling of branched actin filament networks. Curr. Biol. CB 10, 1273–1282.

Bowman, S.K., Neumüller, R.A., Novatchkova, M., Du, Q., and Knoblich, J.A. (2006). The Drosophila NuMA Homolog Mud regulates spindle orientation in asymmetric cell division. Dev. Cell 10, 731–742.

Bowman, S.K., Rolland, V., Betschinger, J., Kinsey, K.A., Emery, G., and Knoblich, J.A. (2008). The Tumor Suppressors Brat and Numb Regulate Transit-Amplifying Neuroblast Lineages in Drosophila. Dev. Cell 14, 535–546.

Bozkulak, E.C., and Weinmaster, G. (2009). Selective Use of ADAM10 and ADAM17 in Activation of Notch1 Signaling. Mol. Cell. Biol. 29, 5679–5695.

Brand, M., Jarman, A.P., Jan, L.Y., and Jan, Y.N. (1993). asense is a Drosophila neural precursor gene and is capable of initiating sense organ formation. Dev. Camb. Engl. 119, 1–17.

Bray, S.J. (2016). Notch signalling in context. Nat. Rev. Mol. Cell Biol. 17, 722–735.

Bray, S., and Furriols, M. (2001). Notch pathway: Making sense of Suppressor of Hairless. Curr. Biol. 11, R217–R221.

Bray, S.J., and Gomez-Lamarca, M. (2018). Notch after cleavage. Curr. Opin. Cell Biol. 51, 103–109.

Brou, C., Logeat, F., Gupta, N., Bessia, C., LeBail, O., Doedens, J.R., Cumano, A., Roux, P., Black, R.A., and Israël, A. (2000). A Novel Proteolytic Cleavage Involved in Notch Signaling. Mol. Cell 5, 207–216.

Brückner, K., Perez, L., Clausen, H., and Cohen, S. (2000). Glycosyltransferase activity of Fringe modulates Notch–Delta interactions. Nature 406, 411–415.

Cabrera, C.V. (1990). Lateral inhibition and cell fate during neurogenesis in Drosophila: the interactions between scute, Notch and Delta. Dev. Camb. Engl. 110, 733–742.

Cagan, R.L., and Ready, D.F. (1989). Notch is required for successive cell decisions in the developing Drosophila retina. Genes Dev. 3, 1099–1112.

Cai, L., Holoweckyj, N., Schaller, M.D., and Bear, J.E. (2005). Phosphorylation of coronin 1B by protein kinase C regulates interaction with Arp2/3 and cell motility. J. Biol. Chem. 280, 31913–31923.

References

196

Cai, L., Makhov, A.M., Schafer, D.A., and Bear, J.E. (2008). Coronin 1B Antagonizes Cortactin and Remodels Arp2/3-Containing Actin Branches in Lamellipodia. Cell 134, 828–842.

Campellone, K.G., and Welch, M.D. (2010). A nucleator arms race: cellular control of actin assembly. Nat. Rev. Mol. Cell Biol. 11, 237–251.

Campellone, K.G., Webb, N.J., Znameroski, E.A., and Welch, M.D. (2008). WHAMM Is an Arp2/3 Complex Activator That Binds Microtubules and Functions in ER to Golgi Transport. Cell 134, 148–161.

Campos-Ortega, J.A. (1994). Cellular interactions in the developing nervous system of Drosophila. Cell 77, 969–975.

Caridi, C.P., D’Agostino, C., Ryu, T., Zapotoczny, G., Delabaere, L., Li, X., Khodaverdian, V.Y., Amaral, N., Lin, E., Rau, A.R., et al. (2018). Nuclear F-actin and myosins drive relocalization of heterochromatic breaks. Nature 559, 54–60.

Carrieri, F.A., and Dale, J.K. (2017). Turn It Down a Notch. Front. Cell Dev. Biol. 4.

Castro, B., Barolo, S., Bailey, A.M., and Posakony, J.W. (2005). Lateral inhibition in proneural clusters: cis-regulatory logic and default repression by Suppressor of Hairless. Dev. Camb. Engl. 132, 3333–3344.

Cave, J.W., Loh, F., Surpris, J.W., Xia, L., and Caudy, M.A. (2005). A DNA transcription code for cell-specific gene activation by notch signaling. Curr. Biol. CB 15, 94–104.

Cave, J.W., Xia, L., and Caudy, M. (2011). Differential regulation of transcription through distinct Suppressor of Hairless DNA binding site architectures during Notch signaling in proneural clusters. Mol. Cell. Biol. 31, 22–29.

Cavey, M., and Lecuit, T. (2009). Molecular Bases of Cell-Cell Junctions Stability and Dynamics. Cold Spring Harb. Perspect. Biol. 1, a002998–a002998.

de Celis, J.F., and Bray, S. (1997). Feed-back mechanisms affecting Notch activation at the dorsoventral boundary in the Drosophila wing. Dev. Camb. Engl. 124, 3241–3251.

Chastagner, P., Rubinstein, E., and Brou, C. (2017). Ligand-activated Notch undergoes DTX4-mediated ubiquitylation and bilateral endocytosis before ADAM10 processing. Sci. Signal. 10.

Chen, N., and Greenwald, I. (2004). The lateral signal for LIN-12/Notch in C. elegans vulval development comprises redundant secreted and transmembrane DSL proteins. Dev. Cell 6, 183–192.

Chen, B., Chou, H.-T., Brautigam, C.A., Xing, W., Yang, S., Henry, L., Doolittle, L.K., Walz, T., and Rosen, M.K. (2017). Rac1 GTPase activates the WAVE regulatory complex through two distinct binding sites. ELife 6.

Chen, H., Ko, G., Zatti, A., Di Giacomo, G., Liu, L., Raiteri, E., Perucco, E., Collesi, C., Min, W., Zeiss, C., et al. (2009). Embryonic arrest at midgestation and disruption of Notch

References

197

signaling produced by the absence of both epsin 1 and epsin 2 in mice. Proc. Natl. Acad. Sci. 106, 13838–13843.

Chen, J., Xu, N., Wang, C., Huang, P., Huang, H., Jin, Z., Yu, Z., Cai, T., Jiao, R., and Xi, R. (2018). Transient Scute activation via a self-stimulatory loop directs enteroendocrine cell pair specification from self-renewing intestinal stem cells. Nat. Cell Biol. 20, 152–161.

Chen, Z., Borek, D., Padrick, S.B., Gomez, T.S., Metlagel, Z., Ismail, A.M., Umetani, J., Billadeau, D.D., Otwinowski, Z., and Rosen, M.K. (2010). Structure and control of the actin regulatory WAVE complex. Nature 468, 533–538.

Chillakuri, C.R., Sheppard, D., Ilagan, M.X.G., Holt, L.R., Abbott, F., Liang, S., Kopan, R., Handford, P.A., and Lea, S.M. (2013). Structural Analysis Uncovers Lipid-Binding Properties of Notch Ligands. Cell Rep. 5, 861–867.

Chorev, D.S., Moscovitz, O., Geiger, B., and Sharon, M. (2014). Regulation of focal adhesion formation by a vinculin-Arp2/3 hybrid complex. Nat. Commun. 5.

Chou, F.-S., and Wang, P.-S. (2016). The Arp2/3 complex is essential at multiple stages of neural development. Neurogenesis 3, e1261653.

Chowdhury, F., Li, I.T.S., Ngo, T.T.M., Leslie, B.J., Kim, B.C., Sokoloski, J.E., Weiland, E., Wang, X., Chemla, Y.R., Lohman, T.M., et al. (2016). Defining Single Molecular Forces Required for Notch Activation Using Nano Yoyo. Nano Lett. 16, 3892–3897.

Cohen, M., Georgiou, M., Stevenson, N.L., Miodownik, M., and Baum, B. (2010). Dynamic Filopodia Transmit Intermittent Delta-Notch Signaling to Drive Pattern Refinement during Lateral Inhibition. Dev. Cell 19, 78–89.

Cordle, J., Johnson, S., Zi Yan Tay, J., Roversi, P., Wilkin, M.B., de Madrid, B.H., Shimizu, H., Jensen, S., Whiteman, P., Jin, B., et al. (2008). A conserved face of the Jagged/Serrate DSL domain is involved in Notch trans-activation and cis-inhibition. Nat. Struct. Mol. Biol. 15, 849–857.

Corson, F., Couturier, L., Rouault, H., Mazouni, K., and Schweisguth, F. (2017). Self-organized Notch dynamics generate stereotyped sensory organ patterns in Drosophila. Science 356.

Cotton, M., Benhra, N., and Le Borgne, R. (2013). Numb Inhibits the Recycling of Sanpodo in Drosophila Sensory Organ Precursor. Curr. Biol. 23, 581–587.

Coumailleau, F., Fürthauer, M., Knoblich, J.A., and González-Gaitán, M. (2009). Directional Delta and Notch trafficking in Sara endosomes during asymmetric cell division. Nature 458, 1051–1055.

Coutts, A.S., and La Thangue, N.B. (2015). Actin nucleation by WH2 domains at the autophagosome. Nat. Commun. 6.

Coutts, A.S., Weston, L., and La Thangue, N.B. (2009). A transcription co-factor integrates cell adhesion and motility with the p53 response. Proc. Natl. Acad. Sci. 106, 19872–19877.

References

198

Couturier, L., Vodovar, N., and Schweisguth, F. (2012). Endocytosis by Numb breaks Notch symmetry at cytokinesis. Nat. Cell Biol. 14, 131–139.

Couturier, L., Mazouni, K., and Schweisguth, F. (2013). Numb Localizes at Endosomes and Controls the Endosomal Sorting of Notch after Asymmetric Division in Drosophila. Curr. Biol. 23, 588–593.

Couturier, L., Trylinski, M., Mazouni, K., Darnet, L., and Schweisguth, F. (2014). A fluorescent tagging approach in Drosophila reveals late endosomal trafficking of Notch and Sanpodo. J. Cell Biol. 207, 351–363.

Cowan, C.R., and Hyman, A.A. (2004). Asymmetric cell division in C. elegans: cortical polarity and spindle positioning. Annu. Rev. Cell Dev. Biol. 20, 427–453.

Daeden, A., and Gonzalez-Gaitan, M. (2018). Endosomal Trafficking During Mitosis and Notch-Dependent Asymmetric Division. In Endocytosis and Signaling, C. Lamaze, and I. Prior, eds. (Cham: Springer International Publishing), pp. 301–329.

D’Amato, G., Luxán, G., and de la Pompa, J.L. (2016). Notch signalling in ventricular chamber development and cardiomyopathy. FEBS J. 283, 4223–4237.

Dang, I., Gorelik, R., Sousa-Blin, C., Derivery, E., Guérin, C., Linkner, J., Nemethova, M., Dumortier, J.G., Giger, F.A., Chipysheva, T.A., et al. (2013). Inhibitory signalling to the Arp2/3 complex steers cell migration. Nature 503, 281–284.

Daniel, E., Daudé, M., Kolotuev, I., Charish, K., Auld, V., and Le Borgne, R. (2018). Coordination of Septate Junctions Assembly and Completion of Cytokinesis in Proliferative Epithelial Tissues. Curr. Biol. 28, 1380-1391.e4.

Daskalaki, A., Shalaby, N.A., Kux, K., Tsoumpekos, G., Tsibidis, G.D., Muskavitch, M.A.T., and Delidakis, C. (2011). Distinct intracellular motifs of Delta mediate its ubiquitylation and activation by Mindbomb1 and Neuralized. J. Cell Biol. 195, 1017–1031.

De Strooper, B., Annaert, W., Cupers, P., Saftig, P., Craessaerts, K., Mumm, J.S., Schroeter, E.H., Schrijvers, V., Wolfe, M.S., Ray, W.J., et al. (1999). A presenilin-1-dependent γ-secretase-like protease mediates release of Notch intracellular domain. Nature 398, 518–522.

Deblandre, G.A., Lai, E.C., and Kintner, C. (2001). Xenopus Neuralized Is a Ubiquitin Ligase that Interacts with XDelta1 and Regulates Notch Signaling. Dev. Cell 1, 795–806.

Del Signore, S.J., Cilla, R., and Hatini, V. (2018). The WAVE Regulatory Complex and Branched F-Actin Counterbalance Contractile Force to Control Cell Shape and Packing in the Drosophila Eye. Dev. Cell 44, 471-483.e4.

Derivery, E., Sousa, C., Gautier, J.J., Lombard, B., Loew, D., and Gautreau, A. (2009). The Arp2/3 Activator WASH Controls the Fission of Endosomes through a Large Multiprotein Complex. Dev. Cell 17, 712–723.

Derivery, E., Seum, C., Daeden, A., Loubéry, S., Holtzer, L., Jülicher, F., and Gonzalez-Gaitan, M. (2015). Polarized endosome dynamics by spindle asymmetry during asymmetric cell division. Nature 528, 280–285.

References

199

Dix, C.L., Matthews, H.K., Uroz, M., McLaren, S., Wolf, L., Heatley, N., Win, Z., Almada, P., Henriques, R., Boutros, M., et al. (2018). The Role of Mitotic Cell-Substrate Adhesion Re-modeling in Animal Cell Division. Dev. Cell 45, 132-145.e3.

Doe, C.Q., and Spana, E.P. (1995). A collection of cortical crescents: asymmetric protein localization in CNS precursor cells. Neuron 15, 991–995.

Dong, Z., Yang, N., Yeo, S.-Y., Chitnis, A., and Guo, S. (2012). Intralineage Directional Notch Signaling Regulates Self-Renewal and Differentiation of Asymmetrically Dividing Radial Glia. Neuron 74, 65–78.

Donnelly, S.K., Weisswange, I., Zettl, M., and Way, M. (2013). WIP Provides an Essential Link between Nck and N-WASP during Arp2/3-Dependent Actin Polymerization. Curr. Biol. 23, 999–1006.

Dornier, E., Coumailleau, F., Ottavi, J.-F., Moretti, J., Boucheix, C., Mauduit, P., Schweisguth, F., and Rubinstein, E. (2012). TspanC8 tetraspanins regulate ADAM10/Kuzbanian trafficking and promote Notch activation in flies and mammals. J. Cell Biol. 199, 481–496.

Dye, C.A., Lee, J.K., Atkinson, R.C., Brewster, R., Han, P.L., and Bellen, H.J. (1998). The Drosophila sanpodo gene controls sibling cell fate and encodes a tropomodulin homolog, an actin/tropomyosin-associated protein. Dev. Camb. Engl. 125, 1845–1856.

Eden, S., Rohatgi, R., Podtelejnikov, A.V., Mann, M., and Kirschner, M.W. (2002). Mechanism of regulation of WAVE1-induced actin nucleation by Rac1 and Nck. Nature 418, 790–793.

Ehrlich, J.S., Hansen, M.D.H., and Nelson, W.J. (2002). Spatio-temporal regulation of Rac1 localization and lamellipodia dynamics during epithelial cell-cell adhesion. Dev. Cell 3, 259–270.

Elowitz, M.B., Santat, L.A., and Nandagopal, N. (2018). -activation in the Notch signaling pathway.

Emery, G., Hutterer, A., Berdnik, D., Mayer, B., Wirtz-Peitz, F., Gaitan, M.G., and Knoblich, J.A. (2005). Asymmetric Rab11 Endosomes Regulate Delta Recycling and Specify Cell Fate in the Drosophila Nervous System. Cell 122, 763–773.

Engl, W., Arasi, B., Yap, L.L., Thiery, J.P., and Viasnoff, V. (2014). Actin dynamics modulate mechanosensitive immobilization of E-cadherin at adherens junctions. Nat. Cell Biol. 16, 584–591.

Espinoza-Sanchez, S., Metskas, L.A., Chou, S.Z., Rhoades, E., and Pollard, T.D. (2018). Conformational changes in Arp2/3 complex induced by ATP, WASp-VCA, and actin filaments. Proc. Natl. Acad. Sci. 115, E8642–E8651.

Evans, I.R., Ghai, P.A., Urbančič, V., Tan, K.-L., and Wood, W. (2013). SCAR/WAVE-mediated processing of engulfed apoptotic corpses is essential for effective macrophage migration in Drosophila. Cell Death Differ. 20, 709–720.

References

200

Fang, J.S., Coon, B.G., Gillis, N., Chen, Z., Qiu, J., Chittenden, T.W., Burt, J.M., Schwartz, M.A., and Hirschi, K.K. (2017). Shear-induced Notch-Cx37-p27 axis arrests endothelial cell cycle to enable arterial specification. Nat. Commun. 8.

Fehon, R.G., Kooh, P.J., Rebay, I., Regan, C.L., Xu, T., Muskavitch, M.A.T., and Artavanis-Tsakonas, S. (1990). Molecular interactions between the protein products of the neurogenic loci Notch and Delta, two EGF-homologous genes in Drosophila. Cell 61, 523–534.

Fichelson, P., and Gho, M. (2003). The glial cell undergoes apoptosis in the microchaete lineage of Drosophila. Dev. Camb. Engl. 130, 123–133.

Fischer, A., and Gessler, M. (2007). Delta Notch and then? Protein interactions and proposed modes of repression by Hes and Hey bHLH factors. Nucleic Acids Res. 35, 4583–4596.

Fleming, R.J., Gu, Y., and Hukriede, N.A. (1997). Serrate-mediated activation of Notch is specifically blocked by the product of the gene fringe in the dorsal compartment of the Drosophila wing imaginal disc. Dev. Camb. Engl. 124, 2973–2981.

Fostier, M., Evans, D.A., Artavanis-Tsakonas, S., and Baron, M. (1998). Genetic characterization of the Drosophila melanogaster Suppressor of deltex gene: A regulator of notch signaling. Genetics 150, 1477–1485.

Founounou, N., Loyer, N., and Le Borgne, R. (2013). Septins Regulate the Contractility of the Actomyosin Ring to Enable Adherens Junction Remodeling during Cytokinesis of Epithelial Cells. Dev. Cell 24, 242–255.

Frankfort, B.J., and Mardon, G. (2002). R8 development in the Drosophila eye: a paradigm for neural selection and differentiation. Dev. Camb. Engl. 129, 1295–1306.

Frémont, S., and Echard, A. (2018). Membrane Traffic in the Late Steps of Cytokinesis. Curr. Biol. 28, R458–R470.

Fricke, R., Gohl, C., Dharmalingam, E., Grevelhörster, A., Zahedi, B., Harden, N., Kessels, M., Qualmann, B., and Bogdan, S. (2009). Drosophila Cip4/Toca-1 Integrates Membrane Trafficking and Actin Dynamics through WASP and SCAR/WAVE. Curr. Biol. 19, 1429–1437.

Furriols, M., and Bray, S. (2001). A model Notch response element detects Suppressor of Hairless-dependent molecular switch. Curr. Biol. CB 11, 60–64.

Gandhi, M., Smith, B.A., Bovellan, M., Paavilainen, V., Daugherty-Clarke, K., Gelles, J., Lappalainen, P., and Goode, B.L. (2010). GMF Is a Cofilin Homolog that Binds Arp2/3 Complex to Stimulate Filament Debranching and Inhibit Actin Nucleation. Curr. Biol. 20, 861–867.

Gautier, J.J., Lomakina, M.E., Bouslama-Oueghlani, L., Derivery, E., Beilinson, H., Faigle, W., Loew, D., Louvard, D., Echard, A., Alexandrova, A.Y., et al. (2011). Clathrin is required for Scar/Wave-mediated lamellipodium formation. J. Cell Sci. 124, 3414–3427.

Gazave, E., Lapébie, P., Richards, G.S., Brunet, F., Ereskovsky, A.V., Degnan, B.M., Borchiellini, C., Vervoort, M., and Renard, E. (2009). Origin and evolution of the Notch signalling pathway: an overview from eukaryotic genomes. BMC Evol. Biol. 9, 249.

References

201

Gazave, E., Lemaître, Q.I.B., and Balavoine, G. (2017). The Notch pathway in the annelid Platynereis : insights into chaetogenesis and neurogenesis processes. Open Biol. 7, 160242.

Georgiou, M., and Baum, B. (2010). Polarity proteins and Rho GTPases cooperate to spatially organise epithelial actin-based protrusions. J. Cell Sci. 123, 1089–1098.

Georgiou, M., Marinari, E., Burden, J., and Baum, B. (2008). Cdc42, Par6, and aPKC Regulate Arp2/3-Mediated Endocytosis to Control Local Adherens Junction Stability. Curr. Biol. 18, 1631–1638.

Gho, M., Bellaïche, Y., and Schweisguth, F. (1999). Revisiting the Drosophila microchaete lineage: a novel intrinsically asymmetric cell division generates a glial cell. Dev. Camb. Engl. 126, 3573–3584.

Giagtzoglou, N., Li, T., Yamamoto, S., and Bellen, H.J. (2013). Drosophila EHBP1 regulates Scabrous secretion during Notch-mediated lateral inhibition. J. Cell Sci. 126, 3686–3696.

Glittenberg, M., Pitsouli, C., Garvey, C., Delidakis, C., and Bray, S. (2006). Role of conserved intracellular motifs in Serrate signalling, cis-inhibition and endocytosis. EMBO J. 25, 4697–4706.

Glotzer, M. (2017). Cytokinesis in Metazoa and Fungi. Cold Spring Harb. Perspect. Biol. 9, a022343.

Go, M.J., and Artavanis-Tsakonas, S. (1998). A genetic screen for novel components of the notch signaling pathway during Drosophila bristle development. Genetics 150, 211–220.

Gohl, C., Banovic, D., Grevelhörster, A., and Bogdan, S. (2010). WAVE Forms Hetero- and Homo-oligomeric Complexes at Integrin Junctions in Drosophila Visualized by Bimolecular Fluorescence Complementation. J. Biol. Chem. 285, 40171–40179.

Goley, E.D., and Welch, M.D. (2006). The ARP2/3 complex: an actin nucleator comes of age. Nat. Rev. Mol. Cell Biol. 7, 713–726.

Gomez, T.S., and Billadeau, D.D. (2009). A FAM21-Containing WASH Complex Regulates Retromer-Dependent Sorting. Dev. Cell 17, 699–711.

Gomez-Lamarca, M.J., Snowdon, L.A., Seib, E., Klein, T., and Bray, S.J. (2015). Rme-8 depletion perturbs Notch recycling and predisposes to pathogenic signaling. J. Cell Biol. 210, 303–318.

Gomez-Lamarca, M.J., Falo-Sanjuan, J., Stojnic, R., Abdul Rehman, S., Muresan, L., Jones, M.L., Pillidge, Z., Cerda-Moya, G., Yuan, Z., Baloul, S., et al. (2018). Activation of the Notch Signaling Pathway In Vivo Elicits Changes in CSL Nuclear Dynamics. Dev. Cell.

Gómez-Skarmeta, J.L., Campuzano, S., and Modolell, J. (2003). Half a century of neural prepatterning: the story of a few bristles and many genes. Nat. Rev. Neurosci. 4, 587–598.

Gordon, W.R., Vardar-Ulu, D., Histen, G., Sanchez-Irizarry, C., Aster, J.C., and Blacklow, S.C. (2007). Structural basis for autoinhibition of Notch. Nat. Struct. Mol. Biol. 14, 295–300.

References

202

Gordon, W.R., Arnett, K.L., and Blacklow, S.C. (2008). The molecular logic of Notch signaling - a structural and biochemical perspective. J. Cell Sci. 121, 3109–3119.

Gordon, W.R., Vardar-Ulu, D., L’Heureux, S., Ashworth, T., Malecki, M.J., Sanchez-Irizarry, C., McArthur, D.G., Histen, G., Mitchell, J.L., Aster, J.C., et al. (2009a). Effects of S1 Cleavage on the Structure, Surface Export, and Signaling Activity of Human Notch1 and Notch2. PLoS ONE 4, e6613.

Gordon, W.R., Roy, M., Vardar-Ulu, D., Garfinkel, M., Mansour, M.R., Aster, J.C., and Blacklow, S.C. (2009b). Structure of the Notch1-negative regulatory region: implications for normal activation and pathogenic signaling in T-ALL. Blood 113, 4381–4390.

Gordon, W.R., Zimmerman, B., He, L., Miles, L.J., Huang, J., Tiyanont, K., McArthur, D.G., Aster, J.C., Perrimon, N., Loparo, J.J., et al. (2015). Mechanical Allostery: Evidence for a Force Requirement in the Proteolytic Activation of Notch. Dev. Cell 33, 729–736.

Gournier, H., Goley, E.D., Niederstrasser, H., Trinh, T., and Welch, M.D. (2001). Reconstitution of human Arp2/3 complex reveals critical roles of individual subunits in complex structure and activity. Mol. Cell 8, 1041–1052.

Groot, A.J., and Vooijs, M.A. (2012). The Role of Adams in Notch Signaling. In Notch Signaling in Embryology and Cancer, J. Reichrath, and S. Reichrath, eds. (New York, NY: Springer US), pp. 15–36.

Gu, Y., Hukriede, N.A., and Fleming, R.J. (1995). Serrate expression can functionally replace Delta activity during neuroblast segregation in the Drosophila embryo. Dev. Camb. Engl. 121, 855–865.

Gulino, A., Di Marcotullio, L., and Screpanti, I. (2010). The multiple functions of Numb. Exp. Cell Res. 316, 900–906.

Guo, B., McMillan, B.J., and Blacklow, S.C. (2016). Structure and function of the Mind bomb E3 ligase in the context of Notch signal transduction. Curr. Opin. Struct. Biol. 41, 38–45.

Guo, M., Jan, L.Y., and Jan, Y.N. (1996). Control of Daughter Cell Fates during Asymmetric Division: Interaction of Numb and Notch. Neuron 17, 27–41.

Haining, E.J., Yang, J., Bailey, R.L., Khan, K., Collier, R., Tsai, S., Watson, S.P., Frampton, J., Garcia, P., and Tomlinson, M.G. (2012). The TspanC8 subgroup of tetraspanins interacts with A disintegrin and metalloprotease 10 (ADAM10) and regulates its maturation and cell surface expression. J. Biol. Chem. 287, 39753–39765.

Hambleton, S., Valeyev, N.V., Muranyi, A., Knott, V., Werner, J.M., McMichael, A.J., Handford, P.A., and Downing, A.K. (2004). Structural and Functional Properties of the Human Notch-1 Ligand Binding Region. Structure 12, 2173–2183.

Hamel, S., Fantini, J., and Schweisguth, F. (2010). Notch ligand activity is modulated by glycosphingolipid membrane composition in Drosophila melanogaster. J. Cell Biol. 188, 581–594.

References

203

Han, Z., and Bodmer, R. (2003). Myogenic cells fates are antagonized by Notch only in asymmetric lineages of the Drosophila heart, with or without cell division. Dev. Camb. Engl. 130, 3039–3051.

Hansen, S.D., Kwiatkowski, A.V., Ouyang, C.-Y., Liu, H., Pokutta, S., Watkins, S.C., Volkmann, N., Hanein, D., Weis, W.I., Mullins, R.D., et al. (2013). αE-catenin actin-binding domain alters actin filament conformation and regulates binding of nucleation and disassembly factors. Mol. Biol. Cell 24, 3710–3720.

Hansson, E.M., Lanner, F., Das, D., Mutvei, A., Marklund, U., Ericson, J., Farnebo, F., Stumm, G., Stenmark, H., Andersson, E.R., et al. (2010). Control of Notch-ligand endocytosis by ligand-receptor interaction. J. Cell Sci. 123, 2931–2942.

Hao, Y.-H., Doyle, J.M., Ramanathan, S., Gomez, T.S., Jia, D., Xu, M., Chen, Z.J., Billadeau, D.D., Rosen, M.K., and Potts, P.R. (2013). Regulation of WASH-Dependent Actin Polymerization and Protein Trafficking by Ubiquitination. Cell 152, 1051–1064.

Harmansa, S., Alborelli, I., Bieli, D., Caussinus, E., and Affolter, M. (2017). A nanobody-based toolset to investigate the role of protein localization and dispersal in Drosophila. ELife 6.

Hartenstein, V., and Posakony, J.W. (1989). Development of adult sensilla on the wing and notum of Drosophila melanogaster. Dev. Camb. Engl. 107, 389–405.

Hartenstein, V., and Posakony, J.W. (1990). A dual function of the Notch gene in Drosophila sensillum development. Dev. Biol. 142, 13–30.

Hartenstein, V., and Wodarz, A. (2013). Initial neurogenesis in Drosophila: Neurogenesis in Drosophila. Wiley Interdiscip. Rev. Dev. Biol. 2, 701–721.

Harvey, B.M., and Haltiwanger, R.S. (2018). Regulation of Notch Function by O-Glycosylation. In Molecular Mechanisms of Notch Signaling, T. Borggrefe, and B.D. Giaimo, eds. (Cham: Springer International Publishing), pp. 59–78.

Harvey, B.M., Rana, N.A., Moss, H., Leonardi, J., Jafar-Nejad, H., and Haltiwanger, R.S. (2016). Mapping Sites of O -Glycosylation and Fringe Elongation on Drosophila Notch. J. Biol. Chem. 291, 16348–16360.

Hass, M.R., Liow, H., Chen, X., Sharma, A., Inoue, Y.U., Inoue, T., Reeb, A., Martens, A., Fulbright, M., Raju, S., et al. (2015). SpDamID: Marking DNA Bound by Protein Complexes Identifies Notch-Dimer Responsive Enhancers. Mol. Cell 59, 685–697.

Heitzler, P., and Simpson, P. (1991). The choice of cell fate in the epidermis of Drosophila. Cell 64, 1083–1092.

Helgeson, L.A., and Nolen, B.J. (2013). Mechanism of synergistic activation of Arp2/3 complex by cortactin and N-WASP. ELife 2.

Helwani, F.M., Kovacs, E.M., Paterson, A.D., Verma, S., Ali, R.G., Fanning, A.S., Weed, S.A., and Yap, A.S. (2004). Cortactin is necessary for E-cadherin-mediated contact formation and actin reorganization. J. Cell Biol. 164, 899–910.

References

204

Herszterg, S., Leibfried, A., Bosveld, F., Martin, C., and Bellaiche, Y. (2013). Interplay between the Dividing Cell and Its Neighbors Regulates Adherens Junction Formation during Cytokinesis in Epithelial Tissue. Dev. Cell 24, 256–270.

Herszterg, S., Pinheiro, D., and Bellaïche, Y. (2014). A multicellular view of cytokinesis in epithelial tissue. Trends Cell Biol. 24, 285–293.

Ho, D.M., Guruharsha, K.G., and Artavanis-Tsakonas, S. (2018). The Notch Interactome: Complexity in Signaling Circuitry. In Molecular Mechanisms of Notch Signaling, T. Borggrefe, and B.D. Giaimo, eds. (Cham: Springer International Publishing), pp. 125–140.

Ho, H.-Y.H., Rohatgi, R., Lebensohn, A.M., Le Ma, null, Li, J., Gygi, S.P., and Kirschner, M.W. (2004). Toca-1 mediates Cdc42-dependent actin nucleation by activating the N-WASP-WIP complex. Cell 118, 203–216.

Homem, C.C.F., Repic, M., and Knoblich, J.A. (2015). Proliferation control in neural stem and progenitor cells. Nat. Rev. Neurosci. 16, 647–659.

Hori, K., Fostier, M., Ito, M., Fuwa, T.J., Go, M.J., Okano, H., Baron, M., and Matsuno, K. (2004). Drosophila deltex mediates suppressor of Hairless-independent and late-endosomal activation of Notch signaling. Dev. Camb. Engl. 131, 5527–5537.

Hori, K., Sen, A., Kirchhausen, T., and Artavanis-Tsakonas, S. (2011). Synergy between the ESCRT-III complex and Deltex defines a ligand-independent Notch signal. J. Cell Biol. 195, 1005–1015.

Housden, B.E., and Perrimon, N. (2014). Spatial and temporal organization of signaling pathways. Trends Biochem. Sci. 39, 457–464.

Housden, B.E., Fu, A.Q., Krejci, A., Bernard, F., Fischer, B., Tavaré, S., Russell, S., and Bray, S.J. (2013). Transcriptional Dynamics Elicited by a Short Pulse of Notch Activation Involves Feed-Forward Regulation by E(spl)/Hes Genes. PLoS Genet. 9, e1003162.

Huang, H., and Kornberg, T.B. (2015). Myoblast cytonemes mediate Wg signaling from the wing imaginal disc and Delta-Notch signaling to the air sac primordium. ELife 4.

Huenniger, K., Krämer, A., Soom, M., Chang, I., Köhler, M., Depping, R., Kehlenbach, R.H., and Kaether, C. (2010). Notch1 signaling is mediated by importins alpha 3, 4, and 7. Cell. Mol. Life Sci. 67, 3187–3196.

Hunter, G.L., Hadjivasiliou, Z., Bonin, H., He, L., Perrimon, N., Charras, G., and Baum, B. (2016). Coordinated control of Notch/Delta signalling and cell cycle progression drives lateral inhibition-mediated tissue patterning. Development 143, 2305–2310.

Hutterer, A., and Knoblich, J.A. (2005). Numb and α-Adaptin regulate Sanpodo endocytosis to specify cell fate in Drosophila external sensory organs. EMBO Rep. 6, 836–842.

Imayoshi, I., Shimojo, H., Sakamoto, M., Ohtsuka, T., and Kageyama, R. (2013). Genetic visualization of notch signaling in mammalian neurogenesis. Cell. Mol. Life Sci. 70, 2045–2057.

References

205

Innocenti, M., Gerboth, S., Rottner, K., Lai, F.P.L., Hertzog, M., Stradal, T.E.B., Frittoli, E., Didry, D., Polo, S., Disanza, A., et al. (2005). Abi1 regulates the activity of N-WASP and WAVE in distinct actin-based processes. Nat. Cell Biol. 7, 969–976.

Ismail, A.M., Padrick, S.B., Chen, B., Umetani, J., and Rosen, M.K. (2009). The WAVE regulatory complex is inhibited. Nat. Struct. Mol. Biol. 16, 561–563.

Itoh, M., Kim, C.-H., Palardy, G., Oda, T., Jiang, Y.-J., Maust, D., Yeo, S.-Y., Lorick, K., Wright, G.J., Ariza-McNaughton, L., et al. (2003). Mind Bomb Is a Ubiquitin Ligase that Is Essential for Efficient Activation of Notch Signaling by Delta. Dev. Cell 4, 67–82.

Jafar-Nejad, H., Andrews, H.K., Acar, M., Bayat, V., Wirtz-Peitz, F., Mehta, S.Q., Knoblich, J.A., and Bellen, H.J. (2005). Sec15, a component of the exocyst, promotes notch signaling during the asymmetric division of Drosophila sensory organ precursors. Dev. Cell 9, 351–363.

Jain, N., Lim, L.W., Tan, W.T., George, B., Makeyev, E., and Thanabalu, T. (2014). Conditional N-WASP knockout in mouse brain implicates actin cytoskeleton regulation in hydrocephalus pathology. Exp. Neurol. 254, 29–40.

Jarman, A.P., Brand, M., Jan, L.Y., and Jan, Y.N. (1993). The regulation and function of the helix-loop-helix gene, asense, in Drosophila neural precursors. Dev. Camb. Engl. 119, 19–29.

Jarman, A.P., Grell, E.H., Ackerman, L., Jan, L.Y., and Jan, Y.N. (1994). atonal is the proneural gene for Drosophila photoreceptors. Nature 369, 398–400.

Jarman, A.P., Sun, Y., Jan, L.Y., and Jan, Y.N. (1995). Role of the proneural gene, atonal, in formation of Drosophila chordotonal organs and photoreceptors. Dev. Camb. Engl. 121, 2019–2030.

Jia, D., Gomez, T.S., Metlagel, Z., Umetani, J., Otwinowski, Z., Rosen, M.K., and Billadeau, D.D. (2010). WASH and WAVE actin regulators of the Wiskott-Aldrich syndrome protein (WASP) family are controlled by analogous structurally related complexes. Proc. Natl. Acad. Sci. 107, 10442–10447.

Johnston, C.A., Manning, L., Lu, M.S., Golub, O., Doe, C.Q., and Prehoda, K.E. (2013). Formin-mediated actin polymerization cooperates with Mushroom body defect (Mud)-Dynein during Frizzled-Dishevelled spindle orientation. J. Cell Sci. 126, 4436–4444.

de Joussineau, C., Soulé, J., Martin, M., Anguille, C., Montcourrier, P., and Alexandre, D. (2003). Delta-promoted filopodia mediate long-range lateral inhibition in Drosophila. Nature 426, 555–559.

Kaksonen, M., and Roux, A. (2018). Mechanisms of clathrin-mediated endocytosis. Nat. Rev. Mol. Cell Biol. 19, 313–326.

Kakuda, S., and Haltiwanger, R.S. (2017). Deciphering the Fringe-Mediated Notch Code: Identification of Activating and Inhibiting Sites Allowing Discrimination between Ligands. Dev. Cell 40, 193–201.

References

206

Kapus, A., and Janmey, P. (2013). Plasma Membrane-Cortical Cytoskeleton Interactions: A Cell Biology Approach with Biophysical Considerations. In Comprehensive Physiology, R. Terjung, ed. (Hoboken, NJ, USA: John Wiley & Sons, Inc.), p.

Kast, D.J., Zajac, A.L., Holzbaur, E.L.F., Ostap, E.M., and Dominguez, R. (2015). WHAMM Directs the Arp2/3 Complex to the ER for Autophagosome Biogenesis through an Actin Comet Tail Mechanism. Curr. Biol. 25, 1791–1797.

Kershaw, N.J., Church, N.L., Griffin, M.D.W., Luo, C.S., Adams, T.E., and Burgess, A.W. (2015). Notch ligand delta-like1: X-ray crystal structure and binding affinity. Biochem. J. 468, 159–166.

Khait, I., Orsher, Y., Golan, O., Binshtok, U., Gordon-Bar, N., Amir-Zilberstein, L., and Sprinzak, D. (2016). Quantitative Analysis of Delta-like 1 Membrane Dynamics Elucidates the Role of Contact Geometry on Notch Signaling. Cell Rep. 14, 225–233.

Khmelinskii, A., Keller, P.J., Bartosik, A., Meurer, M., Barry, J.D., Mardin, B.R., Kaufmann, A., Trautmann, S., Wachsmuth, M., Pereira, G., et al. (2012). Tandem fluorescent protein timers for in vivo analysis of protein dynamics. Nat. Biotechnol. 30, 708–714.

Kidd, S., and Lieber, T. (2002). Furin cleavage is not a requirement for Drosophila Notch function. Mech. Dev. 115, 41–51.

Kim, A.S., Kakalis, L.T., Abdul-Manan, N., Liu, G.A., and Rosen, M.K. (2000). Autoinhibition and activation mechanisms of the Wiskott-Aldrich syndrome protein. Nature 404, 151–158.

King, J.S., Veltman, D.M., Georgiou, M., Baum, B., and Insall, R.H. (2010). SCAR/WAVE is activated at mitosis and drives myosin-independent cytokinesis. J. Cell Sci. 123, 2246–2255.

Kitt, K.N., and Nelson, W.J. (2011). Rapid Suppression of Activated Rac1 by Cadherins and Nectins during De Novo Cell-Cell Adhesion. PLoS ONE 6, e17841.

Kiyota, T., Jono, H., Kuriyama, S., Hasegawa, K., Miyatani, S., and Kinoshita, T. (2001). X-Serrate-1 is involved in primary neurogenesis in Xenopus laevis in a complementary manner with X-Delta-1. Dev. Genes Evol. 211, 367–376.

Klueg, K.M., and Muskavitch, M.A. (1999). Ligand-receptor interactions and trans-endocytosis of Delta, Serrate and Notch: members of the Notch signalling pathway in Drosophila. J. Cell Sci. 112 ( Pt 19), 3289–3297.

Knoblich, J.A. (2008). Mechanisms of Asymmetric Stem Cell Division. Cell 132, 583–597.

Knoblich, J.A., Jan, L.Y., and Jan, Y.N. (1997). The N terminus of the Drosophila Numb protein directs membrane association and actin-dependent asymmetric localization. Proc. Natl. Acad. Sci. U. S. A. 94, 13005–13010.

Kolodziejczyk, A.A., Kim, J.K., Svensson, V., Marioni, J.C., and Teichmann, S.A. (2015). The Technology and Biology of Single-Cell RNA Sequencing. Mol. Cell 58, 610–620.

References

207

Koo, B.-K., Lim, H.-S., Song, R., Yoon, M.-J., Yoon, K.-J., Moon, J.-S., Kim, Y.-W., Kwon, M.-C., Yoo, K.-W., Kong, M.-P., et al. (2005a). Mind bomb 1 is essential for generating functional Notch ligands to activate Notch. Dev. Camb. Engl. 132, 3459–3470.

Koo, B.-K., Yoon, K.-J., Yoo, K.-W., Lim, H.-S., Song, R., So, J.-H., Kim, C.-H., and Kong, Y.-Y. (2005b). Mind bomb-2 is an E3 ligase for Notch ligand. J. Biol. Chem. 280, 22335–22342.

Koo, B.-K., Yoon, M.-J., Yoon, K.-J., Im, S.-K., Kim, Y.-Y., Kim, C.-H., Suh, P.-G., Jan, Y.N., and Kong, Y.-Y. (2007). An Obligatory Role of Mind Bomb-1 in Notch Signaling of Mammalian Development. PLoS ONE 2, e1221.

Kopan, R., and Ilagan, M.X.G. (2009). The Canonical Notch Signaling Pathway: Unfolding the Activation Mechanism. Cell 137, 216–233.

Kopan, R., Schroeter, E.H., Weintraub, H., and Nye, J.S. (1996). Signal transduction by activated mNotch: importance of proteolytic processing and its regulation by the extracellular domain. Proc. Natl. Acad. Sci. 93, 1683–1688.

Kovacs, E.M., Goodwin, M., Ali, R.G., Paterson, A.D., and Yap, A.S. (2002). Cadherin-directed actin assembly: E-cadherin physically associates with the Arp2/3 complex to direct actin assembly in nascent adhesive contacts. Curr. Biol. CB 12, 379–382.

Kovall, R.A., Gebelein, B., Sprinzak, D., and Kopan, R. (2017). The Canonical Notch Signaling Pathway: Structural and Biochemical Insights into Shape, Sugar, and Force. Dev. Cell 41, 228–241.

Krejci, A., Bernard, F., Housden, B.E., Collins, S., and Bray, S.J. (2009). Direct Response to Notch Activation: Signaling Crosstalk and Incoherent Logic. Sci. Signal. 2, ra1–ra1.

Kressmann, S., Campos, C., Castanon, I., Fürthauer, M., and González-Gaitán, M. (2015). Directional Notch trafficking in Sara endosomes during asymmetric cell division in the spinal cord. Nat. Cell Biol. 17, 333–339.

Kühbacher, A., Emmenlauer, M., Rämo, P., Kafai, N., Dehio, C., Cossart, P., and Pizarro-Cerdá, J. (2015). Genome-Wide siRNA Screen Identifies Complementary Signaling Pathways Involved in Listeria Infection and Reveals Different Actin Nucleation Mechanisms during Listeria Cell Invasion and Actin Comet Tail Formation. MBio 6.

Kumar, J.P. (2012). Building an ommatidium one cell at a time: Ommatidial Assembly in Drosophila. Dev. Dyn. 241, 136–149.

Lai, E.C., and Orgogozo, V. (2004). A hidden program in Drosophila peripheral neurogenesis revealed: fundamental principles underlying sensory organ diversity. Dev. Biol. 269, 1–17.

Lai, E.C., Deblandre, G.A., Kintner, C., and Rubin, G.M. (2001). Drosophila Neuralized Is a Ubiquitin Ligase that Promotes the Internalization and Degradation of Delta. Dev. Cell 1, 783–794.

Lai, E.C., Roegiers, F., Qin, X., Jan, Y.N., and Rubin, G.M. (2005a). The ubiquitin ligase Drosophila Mind bomb promotes Notch signaling by regulating the localization and activity of Serrate and Delta. Dev. Camb. Engl. 132, 2319–2332.

References

208

Lai, E.C., Roegiers, F., Qin, X., Jan, Y.N., and Rubin, G.M. (2005b). The ubiquitin ligase Drosophila Mind bomb promotes Notch signaling by regulating the localization and activity of Serrate and Delta. Dev. Camb. Engl. 132, 2319–2332.

Lake, R.J., Grimm, L.M., Veraksa, A., Banos, A., and Artavanis-Tsakonas, S. (2009). In Vivo Analysis of the Notch Receptor S1 Cleavage. PLoS ONE 4, e6728.

Lammich, S., Kojro, E., Postina, R., Gilbert, S., Pfeiffer, R., Jasionowski, M., Haass, C., and Fahrenholz, F. (1999). Constitutive and regulated alpha-secretase cleavage of Alzheimer’s amyloid precursor protein by a disintegrin metalloprotease. Proc. Natl. Acad. Sci. U. S. A. 96, 3922–3927.

Langridge, P.D., and Struhl, G. (2017). Epsin-Dependent Ligand Endocytosis Activates Notch by Force. Cell 171, 1383-1396.e12.

Le Borgne, R., and Schweisguth, F. (2003). Unequal segregation of Neuralized biases Notch activation during asymmetric cell division. Dev. Cell 5, 139–148.

Le Borgne, R., Remaud, S., Hamel, S., and Schweisguth, F. (2005). Two Distinct E3 Ubiquitin Ligases Have Complementary Functions in the Regulation of Delta and Serrate Signaling in Drosophila. PLoS Biol. 3, e96.

Le Bras, S., Loyer, N., and Le Borgne, R. (2011). The Multiple Facets of Ubiquitination in the Regulation of Notch Signaling Pathway. Traffic 12, 149–161.

Le Bras, S., Rondanino, C., Kriegel-Taki, G., Dussert, A., and Le Borgne, R. (2012). Genetic identification of intracellular trafficking regulators involved in Notch-dependent binary cell fate acquisition following asymmetric cell division. J. Cell Sci. 125, 4886–4901.

Lebensohn, A.M., and Kirschner, M.W. (2009). Activation of the WAVE complex by coincident signals controls actin assembly. Mol. Cell 36, 512–524.

LeBon, L., Lee, T.V., Sprinzak, D., Jafar-Nejad, H., and Elowitz, M.B. (2014). Fringe proteins modulate Notch-ligand cis and trans interactions to specify signaling states. ELife 3.

Lecourtois, M., and Schweisguth, F. (1998). Indirect evidence for Delta-dependent intracellular processing of Notch in Drosophila embryos. Curr. Biol. 8, 771–775.

Lee, E.C., Yu, S.Y., and Baker, N.E. (2000). The scabrous protein can act as an extracellular antagonist of notch signaling in the Drosophila wing. Curr. Biol. CB 10, 931–934.

Lee, T.V., Sethi, M.K., Leonardi, J., Rana, N.A., Buettner, F.F.R., Haltiwanger, R.S., Bakker, H., and Jafar-Nejad, H. (2013). Negative Regulation of Notch Signaling by Xylose. PLoS Genet. 9, e1003547.

Legent, K., Steinhauer, J., Richard, M., and Treisman, J.E. (2012). A Screen for X-Linked Mutations Affecting Drosophila Photoreceptor Differentiation Identifies Casein Kinase 1α as an Essential Negative Regulator of Wingless Signaling. Genetics 190, 601–616.

Lehmann, R., Dietrich, U., Jim�nez, F., and Campos-Ortega, J.A. (1981). Mutations of early neurogenesis inDrosophila. Wilhelm Rouxs Arch. Dev. Biol. 190, 226–229.

References

209

Lehmann, R., Jim�nez, F., Dietrich, U., and Campos-Ortega, J.A. (1983). On the phenotype and development of mutants of early neurogenesis inDrosophila melanogaster. Wilhelm Rouxs Arch. Dev. Biol. 192, 62–74.

Leibfried, A., Fricke, R., Morgan, M.J., Bogdan, S., and Bellaiche, Y. (2008). Drosophila Cip4 and WASp Define a Branch of the Cdc42-Par6-aPKC Pathway Regulating E-Cadherin Endocytosis. Curr. Biol. 18, 1639–1648.

Leitch, C.C., Lodh, S., Prieto-Echague, V., Badano, J.L., and Zaghloul, N.A. (2014). Basal body proteins regulate Notch signaling through endosomal trafficking. J. Cell Sci. 127, 2407–2419.

Lekomtsev, S., Su, K.-C., Pye, V.E., Blight, K., Sundaramoorthy, S., Takaki, T., Collinson, L.M., Cherepanov, P., Divecha, N., and Petronczki, M. (2012). Centralspindlin links the mitotic spindle to the plasma membrane during cytokinesis. Nature 492, 276–279.

Leung, D.W., and Rosen, M.K. (2005). The nucleotide switch in Cdc42 modulates coupling between the GTPase-binding and allosteric equilibria of Wiskott-Aldrich syndrome protein. Proc. Natl. Acad. Sci. U. S. A. 102, 5685–5690.

Lewellyn, E.B., Pedersen, R.T.A., Hong, J., Lu, R., Morrison, H.M., and Drubin, D.G. (2015). An Engineered Minimal WASP-Myosin Fusion Protein Reveals Essential Functions for Endocytosis. Dev. Cell 35, 281–294.

Li, Y., and Baker, N.E. (2001). Proneural enhancement by Notch overcomes Suppressor-of-Hairless repressor function in the developing Drosophila eye. Curr. Biol. CB 11, 330–338.

Li, H., Koo, Y., Mao, X., Sifuentes-Dominguez, L., Morris, L.L., Jia, D., Miyata, N., Faulkner, R.A., van Deursen, J.M., Vooijs, M., et al. (2015). Endosomal sorting of Notch receptors through COMMD9-dependent pathways modulates Notch signaling. J. Cell Biol. 211, 605–617.

Li, X., Erclik, T., Bertet, C., Chen, Z., Voutev, R., Venkatesh, S., Morante, J., Celik, A., and Desplan, C. (2013). Temporal patterning of Drosophila medulla neuroblasts controls neural fates. Nature 498, 456–462.

Li, Y., Fetchko, M., Lai, Z.-C., and Baker, N.E. (2003). Scabrous and Gp150 are endosomal proteins that regulate Notch activity. Dev. Camb. Engl. 130, 2819–2827.

Liebau, S., Steinestel, J., Linta, L., Kleger, A., Storch, A., Schoen, M., Steinestel, K., Proepper, C., Bockmann, J., Schmeisser, M.J., et al. (2011). An SK3 Channel/nWASP/Abi-1 Complex Is Involved in Early Neurogenesis. PLoS ONE 6, e18148.

Lieber, T., Kidd, S., and Young, M.W. (2002). kuzbanian-mediated cleavage of Drosophila Notch. Genes Dev. 16, 209–221.

Ligoxygakis, P., Yu, S.Y., Delidakis, C., and Baker, N.E. (1998). A subset of notch functions during Drosophila eye development require Su(H) and the E(spl) gene complex. Dev. Camb. Engl. 125, 2893–2900.

Liu, S., and Boulianne, G.L. (2017). The NHR domains of Neuralized and related proteins: Beyond Notch signalling. Cell. Signal. 29, 62–68.

References

210

Liu, R., Abreu-Blanco, M.T., Barry, K.C., Linardopoulou, E.V., Osborn, G.E., and Parkhurst, S.M. (2009). Wash functions downstream of Rho and links linear and branched actin nucleation factors. Development 136, 2849–2860.

Liu, T., Dai, A., Cao, Y., Zhang, R., Dong, M.-Q., and Wang, H.-W. (2017). Structural Insights of WHAMM’s Interaction with Microtubules by Cryo-EM. J. Mol. Biol. 429, 1352–1363.

Logeat, F., Bessia, C., Brou, C., LeBail, O., Jarriault, S., Seidah, N.G., and Israel, A. (1998). The Notch1 receptor is cleaved constitutively by a furin-like convertase. Proc. Natl. Acad. Sci. 95, 8108–8112.

Loria, A., Longhini, K.M., and Glotzer, M. (2012). The RhoGAP Domain of CYK-4 Has an Essential Role in RhoA Activation. Curr. Biol. 22, 213–219.

Loubéry, S., Daeden, A., Seum, C., Holtzer, L., Moraleda, A., Damond, N., Derivery, E., Schmidt, T., and Gonzalez-Gaitan, M. (2017). Sara phosphorylation state controls the dispatch of endosomes from the central spindle during asymmetric division. Nat. Commun. 8, 15285.

Lovendahl, K.N., Blacklow, S.C., and Gordon, W.R. (2018). The Molecular Mechanism of Notch Activation. In Molecular Mechanisms of Notch Signaling, T. Borggrefe, and B.D. Giaimo, eds. (Cham: Springer International Publishing), pp. 47–58.

Lu, B., Rothenberg, M., Jan, L.Y., and Jan, Y.N. (1998). Partner of Numb Colocalizes with Numb during Mitosis and Directs Numb Asymmetric Localization in Drosophila Neural and Muscle Progenitors. Cell 95, 225–235.

Lu, B., Ackerman, L., Jan, L.Y., and Jan, Y.N. (1999). Modes of protein movement that lead to the asymmetric localization of partner of Numb during Drosophila neuroblast division. Mol. Cell 4, 883–891.

Lubensky, D.K., Pennington, M.W., Shraiman, B.I., and Baker, N.E. (2011). A dynamical model of ommatidial crystal formation. Proc. Natl. Acad. Sci. 108, 11145–11150.

Luca, V.C., Jude, K.M., Pierce, N.W., Nachury, M.V., Fischer, S., and Garcia, K.C. (2015). Structural basis for Notch1 engagement of Delta-like 4. Science 347, 847–853.

Luca, V.C., Kim, B.C., Ge, C., Kakuda, S., Wu, D., Roein-Peikar, M., Haltiwanger, R.S., Zhu, C., Ha, T., and Garcia, K.C. (2017). Notch-Jagged complex structure implicates a catch bond in tuning ligand sensitivity. Science 355, 1320–1324.

Mack, J.J., Mosqueiro, T.S., Archer, B.J., Jones, W.M., Sunshine, H., Faas, G.C., Briot, A., Aragón, R.L., Su, T., Romay, M.C., et al. (2017). NOTCH1 is a mechanosensor in adult arteries. Nat. Commun. 8.

Maritzen, T., Zech, T., Schmidt, M.R., Krause, E., Machesky, L.M., and Haucke, V. (2012). Gadkin negatively regulates cell spreading and motility via sequestration of the actin-nucleating ARP2/3 complex. Proc. Natl. Acad. Sci. 109, 10382–10387.

References

211

Marshall, O.J., Southall, T.D., Cheetham, S.W., and Brand, A.H. (2016). Cell-type-specific profiling of protein–DNA interactions without cell isolation using targeted DamID with next-generation sequencing. Nat. Protoc. 11, 1586–1598.

Massarwa, R., Carmon, S., Shilo, B.-Z., and Schejter, E.D. (2007). WIP/WASp-based actin-polymerization machinery is essential for myoblast fusion in Drosophila. Dev. Cell 12, 557–569.

McGill, M.A., Dho, S.E., Weinmaster, G., and McGlade, C.J. (2009). Numb Regulates Post-endocytic Trafficking and Degradation of Notch1. J. Biol. Chem. 284, 26427–26438.

Meloty-Kapella, L., Shergill, B., Kuon, J., Botvinick, E., and Weinmaster, G. (2012). Notch Ligand Endocytosis Generates Mechanical Pulling Force Dependent on Dynamin, Epsins, and Actin. Dev. Cell 22, 1299–1312.

Micchelli, C.A., Rulifson, E.J., and Blair, S.S. (1997). The function and regulation of cut expression on the wing margin of Drosophila: Notch, Wingless and a dominant negative role for Delta and Serrate. Dev. Camb. Engl. 124, 1485–1495.

Michelot, A., Grassart, A., Okreglak, V., Costanzo, M., Boone, C., and Drubin, D.G. (2013). Actin Filament Elongation in Arp2/3-Derived Networks Is Controlled by Three Distinct Mechanisms. Dev. Cell 24, 182–195.

Mikami, S., Nakaura, M., Kawahara, A., Mizoguchi, T., and Itoh, M. (2015). Mindbomb 2 is dispensable for embryonic development and Notch signalling in zebrafish. Biol. Open 4, 1576–1582.

Miller, A.C., Lyons, E.L., and Herman, T.G. (2009). cis-Inhibition of Notch by Endogenous Delta Biases the Outcome of Lateral Inhibition. Curr. Biol. 19, 1378–1383.

Mizuhara, E., Nakatani, T., Minaki, Y., Sakamoto, Y., Ono, Y., and Takai, Y. (2005). MAGI1 Recruits Dll1 to Cadherin-based Adherens Junctions and Stabilizes It on the Cell Surface. J. Biol. Chem. 280, 26499–26507.

Moloney, D.J., Panin, V.M., Johnston, S.H., Chen, J., Shao, L., Wilson, R., Wang, Y., Stanley, P., Irvine, K.D., Haltiwanger, R.S., et al. (2000a). Fringe is a glycosyltransferase that modifies Notch. Nature 406, 369–375.

Moloney, D.J., Shair, L.H., Lu, F.M., Xia, J., Locke, R., Matta, K.L., and Haltiwanger, R.S. (2000b). Mammalian Notch1 is modified with two unusual forms of O-linked glycosylation found on epidermal growth factor-like modules. J. Biol. Chem. 275, 9604–9611.

Montagne, C., and Gonzalez-Gaitan, M. (2014). Sara endosomes and the asymmetric division of intestinal stem cells. Development 141, 2014–2023.

Morsut, L., Roybal, K.T., Xiong, X., Gordley, R.M., Coyle, S.M., Thomson, M., and Lim, W.A. (2016). Engineering Customized Cell Sensing and Response Behaviors Using Synthetic Notch Receptors. Cell 164, 780–791.

Mukherjee, S., Tucker-Burden, C., Zhang, C., Moberg, K., Read, R., Hadjipanayis, C., and Brat, D.J. (2016). Drosophila Brat and Human Ortholog TRIM3 Maintain Stem Cell

References

212

Equilibrium and Suppress Brain Tumorigenesis by Attenuating Notch Nuclear Transport. Cancer Res. 76, 2443–2452.

Mumm, J.S., Schroeter, E.H., Saxena, M.T., Griesemer, A., Tian, X., Pan, D.., Ray, W.J., and Kopan, R. (2000). A Ligand-Induced Extracellular Cleavage Regulates γ-Secretase-like Proteolytic Activation of Notch1. Mol. Cell 5, 197–206.

Mummery-Widmer, J.L., Yamazaki, M., Stoeger, T., Novatchkova, M., Bhalerao, S., Chen, D., Dietzl, G., Dickson, B.J., and Knoblich, J.A. (2009). Genome-wide analysis of Notch signalling in Drosophila by transgenic RNAi. Nature 458, 987–992.

Mund, M., van der Beek, J.A., Deschamps, J., Dmitrieff, S., Hoess, P., Monster, J.L., Picco, A., Nédélec, F., Kaksonen, M., and Ries, J. (2018). Systematic Nanoscale Analysis of Endocytosis Links Efficient Vesicle Formation to Patterned Actin Nucleation. Cell 174, 884-896.e17.

Nagel, B.M., Bechtold, M., Rodriguez, L.G., and Bogdan, S. (2017). Drosophila WASH is required for integrin-mediated cell adhesion, cell motility and lysosomal neutralization. J. Cell Sci. 130, 344–359.

Nandagopal, N., Santat, L.A., LeBon, L., Sprinzak, D., Bronner, M.E., and Elowitz, M.B. (2018). Dynamic Ligand Discrimination in the Notch Signaling Pathway. Cell 172, 869-880.e19.

Narui, Y., and Salaita, K. (2013). Membrane Tethered Delta Activates Notch and Reveals a Role for Spatio-Mechanical Regulation of the Signaling Pathway. Biophys. J. 105, 2655–2665.

Nemetschke, L., and Knust, E. (2016). Drosophila Crumbs prevents ectopic Notch activation in developing wings by inhibiting ligand-independent endocytosis. Dev. Camb. Engl. 143, 4543–4553.

Neto, H., Collins, L.L., and Gould, G.W. (2011). Vesicle trafficking and membrane remodelling in cytokinesis. Biochem. J. 437, 13–24.

Neujahr, R., Heizer, C., and Gerisch, G. (1997). Myosin II-independent processes in mitotic cells of Dictyostelium discoideum: redistribution of the nuclei, re-arrangement of the actin system and formation of the cleavage furrow. J. Cell Sci. 110 ( Pt 2), 123–137.

Nguyen, H.T., Voza, F., Ezzeddine, N., and Frasch, M. (2007). Drosophila mind bomb2 is required for maintaining muscle integrity and survival. J. Cell Biol. 179, 219–227.

O’Connor-Giles, K.M., and Skeath, J.B. (2003). Numb inhibits membrane localization of Sanpodo, a four-pass transmembrane protein, to promote asymmetric divisions in Drosophila. Dev. Cell 5, 231–243.

Oikawa, T., Yamaguchi, H., Itoh, T., Kato, M., Ijuin, T., Yamazaki, D., Suetsugu, S., and Takenawa, T. (2004). PtdIns(3,4,5)P3 binding is necessary for WAVE2-induced formation of lamellipodia. Nat. Cell Biol. 6, 420–426.

Okajima, T., and Irvine, K.D. (2002). Regulation of notch signaling by o-linked fucose. Cell 111, 893–904.

References

213

Okano, M., Matsuo, H., Nishimura, Y., Hozumi, K., Yoshioka, S., Tonoki, A., and Itoh, M. (2016). Mib1 modulates dynamin 2 recruitment via Snx18 to promote Dll1 endocytosis for efficient Notch signaling. Genes Cells 21, 425–441.

Oswald, F., Täuber, B., Dobner, T., Bourteele, S., Kostezka, U., Adler, G., Liptay, S., and Schmid, R.M. (2001). p300 acts as a transcriptional coactivator for mammalian Notch-1. Mol. Cell. Biol. 21, 7761–7774.

Overstreet, E., Chen, X., Wendland, B., and Fischer, J.A. (2003). Either part of a Drosophila epsin protein, divided after the ENTH domain, functions in endocytosis of delta in the developing eye. Curr. Biol. CB 13, 854–860.

Overstreet, E., Fitch, E., and Fischer, J.A. (2004). Fat facets and Liquid facets promote Delta endocytosis and Delta signaling in the signaling cells. Dev. Camb. Engl. 131, 5355–5366.

Palmer, W.H., and Deng, W.-M. (2015). Ligand-Independent Mechanisms of Notch Activity. Trends Cell Biol. 25, 697–707.

Palomero, T., Lim, W.K., Odom, D.T., Sulis, M.L., Real, P.J., Margolin, A., Barnes, K.C., O’Neil, J., Neuberg, D., Weng, A.P., et al. (2006). NOTCH1 directly regulates c-MYC and activates a feed-forward-loop transcriptional network promoting leukemic cell growth. Proc. Natl. Acad. Sci. 103, 18261–18266.

Panin, V.M., Papayannopoulos, V., Wilson, R., and Irvine, K.D. (1997). Fringe modulates Notch–ligand interactions. Nature 387, 908–912.

Park, M., Yaich, L.E., and Bodmer, R. (1998). Mesodermal cell fate decisions in Drosophila are under the control of the lineage genes numb, Notch, and sanpodo. Mech. Dev. 75, 117–126.

Parks, A.L., Klueg, K.M., Stout, J.R., and Muskavitch, M.A. (2000). Ligand endocytosis drives receptor dissociation and activation in the Notch pathway. Dev. Camb. Engl. 127, 1373–1385.

Pavlopoulos, E., Pitsouli, C., Klueg, K.M., Muskavitch, M.A.T., Moschonas, N.K., and Delidakis, C. (2001). neuralized Encodes a Peripheral Membrane Protein Involved in Delta Signaling and Endocytosis. Dev. Cell 1, 807–816.

Perez-Mockus, G., and Schweisguth, F. (2017). Cell Polarity and Notch Signaling: Linked by the E3 Ubiquitin Ligase Neuralized? BioEssays 39, 1700128.

Perez-Mockus, G., Mazouni, K., Roca, V., Corradi, G., Conte, V., and Schweisguth, F. (2017a). Spatial regulation of contractility by Neuralized and Bearded during furrow invagination in Drosophila. Nat. Commun. 8, 1594.

Perez-Mockus, G., Roca, V., Mazouni, K., and Schweisguth, F. (2017b). Neuralized regulates Crumbs endocytosis and epithelium morphogenesis via specific Stardust isoforms. J. Cell Biol. 216, 1405–1420.

Picco, A., Kukulski, W., Manenschijn, H.E., Specht, T., Briggs, J.A.G., and Kaksonen, M. (2018). The contributions of the actin machinery to endocytic membrane bending and vesicle formation. Mol. Biol. Cell 29, 1346–1358.

References

214

Piekny, A.J., and Maddox, A.S. (2010). The myriad roles of Anillin during cytokinesis. Semin. Cell Dev. Biol. 21, 881–891.

Pinheiro, D., Hannezo, E., Herszterg, S., Bosveld, F., Gaugue, I., Balakireva, M., Wang, Z., Cristo, I., Rigaud, S.U., Markova, O., et al. (2017). Transmission of cytokinesis forces via E-cadherin dilution and actomyosin flows. Nature 545, 103–107.

Pitsouli, C., and Delidakis, C. (2005). The interplay between DSL proteins and ubiquitin ligases in Notch signaling. Dev. Camb. Engl. 132, 4041–4050.

Pizarro-Cerdá, J., Chorev, D.S., Geiger, B., and Cossart, P. (2017). The Diverse Family of Arp2/3 Complexes. Trends Cell Biol. 27, 93–100.

Polacheck, W.J., Kutys, M.L., Yang, J., Eyckmans, J., Wu, Y., Vasavada, H., Hirschi, K.K., and Chen, C.S. (2017). A non-canonical Notch complex regulates adherens junctions and vascular barrier function. Nature.

Pollard, T.D. (2007). Regulation of Actin Filament Assembly by Arp2/3 Complex and Formins. Annu. Rev. Biophys. Biomol. Struct. 36, 451–477.

Poulson, D.F. (1937). Chromosomal Deficiencies and the Embryonic Development of Drosophila Melanogaster. Proc. Natl. Acad. Sci. U. S. A. 23, 133–137.

Powell, P.A., Wesley, C., Spencer, S., and Cagan, R.L. (2001). Scabrous complexes with Notch to mediate boundary formation. Nature 409, 626–630.

Preuße, K., Tveriakhina, L., Schuster-Gossler, K., Gaspar, C., Rosa, A.I., Henrique, D., Gossler, A., and Stauber, M. (2015). Context-Dependent Functional Divergence of the Notch Ligands DLL1 and DLL4 In Vivo. PLOS Genet. 11, e1005328.

Qiu, L., Joazeiro, C., Fang, N., Wang, H.-Y., Elly, C., Altman, Y., Fang, D., Hunter, T., and Liu, Y.-C. (2000). Recognition and Ubiquitination of Notch by Itch, a Hect-type E3 Ubiquitin Ligase. J. Biol. Chem. 275, 35734–35737.

Rajan, A., Tien, A.-C., Haueter, C.M., Schulze, K.L., and Bellen, H.J. (2009). The Arp2/3 complex and WASp are required for apical trafficking of Delta into microvilli during cell fate specification of sensory organ precursors. Nat. Cell Biol. 11, 815–824.

Ramkumar, N., and Baum, B. (2016). Coupling changes in cell shape to chromosome segregation. Nat. Rev. Mol. Cell Biol. 17, 511–521.

Rampal, R., Luther, K.B., and Haltiwanger, R.S. (2007). Notch signaling in normal and disease States: possible therapies related to glycosylation. Curr. Mol. Med. 7, 427–445.

Ready, D.F., Hanson, T.E., and Benzer, S. (1976). Development of the Drosophila retina, a neurocrystalline lattice. Dev. Biol. 53, 217–240.

Rebay, I., Fleming, R.J., Fehon, R.G., Cherbas, L., Cherbas, P., and Artavanis-Tsakonas, S. (1991). Specific EGF repeats of Notch mediate interactions with Delta and serrate: Implications for notch as a multifunctional receptor. Cell 67, 687–699.

References

215

Reiss, K., Maretzky, T., Ludwig, A., Tousseyn, T., de Strooper, B., Hartmann, D., and Saftig, P. (2005). ADAM10 cleavage of N-cadherin and regulation of cell-cell adhesion and beta-catenin nuclear signalling. EMBO J. 24, 742–752.

Remaud, S., Audibert, A., and Gho, M. (2008). S-Phase Favours Notch Cell Responsiveness in the Drosophila Bristle Lineage. PLoS ONE 3, e3646.

Ren, Q., Awasaki, T., Wang, Y.-C., Huang, Y.-F., and Lee, T. (2018). Lineage-guided Notch-dependent gliogenesis by Drosophila multi-potent progenitors. Dev. Camb. Engl. 145.

Renaud, O., and Simpson, P. (2001). scabrous Modifies Epithelial Cell Adhesion and Extends the Range of Lateral Signalling during Development of the Spaced Bristle Pattern in Drosophila. Dev. Biol. 240, 361–376.

Rhyu, M.S., Jan, L.Y., and Jan, Y.N. (1994). Asymmetric distribution of numb protein during division of the sensory organ precursor cell confers distinct fates to daughter cells. Cell 76, 477–491.

Richardson, B.E., Beckett, K., Nowak, S.J., and Baylies, M.K. (2007). SCAR/WAVE and Arp2/3 are crucial for cytoskeletal remodeling at the site of myoblast fusion. Dev. Camb. Engl. 134, 4357–4367.

Robinson, R.C., Turbedsky, K., Kaiser, D.A., Marchand, J.B., Higgs, H.N., Choe, S., and Pollard, T.D. (2001). Crystal structure of Arp2/3 complex. Science 294, 1679–1684.

Rocca, D.L., Martin, S., Jenkins, E.L., and Hanley, J.G. (2008). Inhibition of Arp2/3-mediated actin polymerization by PICK1 regulates neuronal morphology and AMPA receptor endocytosis. Nat. Cell Biol. 10, 259–271.

Rodrigues, N.T.L., Lekomtsev, S., Jananji, S., Kriston-Vizi, J., Hickson, G.R.X., and Baum, B. (2015). Kinetochore-localized PP1–Sds22 couples chromosome segregation to polar relaxation. Nature 524, 489–492.

Roegiers, F., Younger-Shepherd, S., Jan, L.Y., and Jan, Y.N. (2001a). Two types of asymmetric divisions in the Drosophila sensory organ precursor cell lineage. Nat. Cell Biol. 3, 58–67.

Roegiers, F., Younger-Shepherd, S., Jan, L.Y., and Jan, Y.N. (2001b). Bazooka is required for localization of determinants and controlling proliferation in the sensory organ precursor cell lineage in Drosophila. Proc. Natl. Acad. Sci. 98, 14469–14474.

Rotty, J.D., Wu, C., and Bear, J.E. (2013). New insights into the regulation and cellular functions of the ARP2/3 complex. Nat. Rev. Mol. Cell Biol. 14, 7–12.

Rouiller, I., Xu, X.-P., Amann, K.J., Egile, C., Nickell, S., Nicastro, D., Li, R., Pollard, T.D., Volkmann, N., and Hanein, D. (2008). The structural basis of actin filament branching by the Arp2/3 complex. J. Cell Biol. 180, 887–895.

Roy, S., Hsiung, F., and Kornberg, T.B. (2011). Specificity of Drosophila Cytonemes for Distinct Signaling Pathways. Science 332, 354–358.

References

216

Roy, S., Huang, H., Liu, S., and Kornberg, T.B. (2014). Cytoneme-Mediated Contact-Dependent Transport of the Drosophila Decapentaplegic Signaling Protein. Science 343, 1244624–1244624.

Sachan, N., Mishra, A.K., Mutsuddi, M., and Mukherjee, A. (2013). The Drosophila Importin-α3 Is Required for Nuclear Import of Notch In Vivo and It Displays Synergistic Effects with Notch Receptor on Cell Proliferation. PLoS ONE 8, e68247.

Sakata, T., Sakaguchi, H., Tsuda, L., Higashitani, A., Aigaki, T., Matsuno, K., and Hayashi, S. (2004). Drosophila Nedd4 Regulates Endocytosis of Notch and Suppresses Its Ligand-Independent Activation. Curr. Biol. 14, 2228–2236.

San-Juán, B.P., and Baonza, A. (2011). The bHLH factor deadpan is a direct target of Notch signaling and regulates neuroblast self-renewal in Drosophila. Dev. Biol. 352, 70–82.

Santos, A.L., and Preta, G. (2018). Lipids in the cell: organisation regulates function. Cell. Mol. Life Sci. 75, 1909–1927.

Sasaki, N., Sasamura, T., Ishikawa, H.O., Kanai, M., Ueda, R., Saigo, K., and Matsuno, K. (2007). Polarized exocytosis and transcytosis of Notch during its apical localization in Drosophila epithelial cells. Genes Cells Devoted Mol. Cell. Mech. 12, 89–103.

Schaefer, M., Shevchenko, A., Shevchenko, A., and Knoblich, J.A. (2000). A protein complex containing Inscuteable and the Gα-binding protein Pins orients asymmetric cell divisions in Drosophila. Curr. Biol. 10, 353–362.

Schäfer, G., Weber, S., Holz, A., Bogdan, S., Schumacher, S., Müller, A., Renkawitz-Pohl, R., and Onel, S.-F. (2007). The Wiskott-Aldrich syndrome protein (WASP) is essential for myoblast fusion in Drosophila. Dev. Biol. 304, 664–674.

Schechtman, A.M. (1937). LOCALIZED CORTICAL GROWTH AS THE IMMEDIATE CAUSE OF CELL DIVISION. Science 85, 222–223.

Schier, A.F., Neuhauss, S.C., Harvey, M., Malicki, J., Solnica-Krezel, L., Stainier, D.Y., Zwartkruis, F., Abdelilah, S., Stemple, D.L., Rangini, Z., et al. (1996). Mutations affecting the development of the embryonic zebrafish brain. Dev. Camb. Engl. 123, 165–178.

Schnute, B., Troost, T., and Klein, T. (2018). Endocytic Trafficking of the Notch Receptor. In Molecular Mechanisms of Notch Signaling, T. Borggrefe, and B.D. Giaimo, eds. (Cham: Springer International Publishing), pp. 99–122.

Scholey, J., Civelekoglu-Scholey, G., and Brust-Mascher, I. (2016). Anaphase B. Biology 5, 51.

Schrank, B.R., Aparicio, T., Li, Y., Chang, W., Chait, B.T., Gundersen, G.G., Gottesman, M.E., and Gautier, J. (2018). Nuclear ARP2/3 drives DNA break clustering for homology-directed repair. Nature 559, 61–66.

Schroeter, E.H., Kisslinger, J.A., and Kopan, R. (1998). Notch-1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature 393, 382–386.

References

217

Schweisguth, F. (2015). Asymmetric cell division in the Drosophila bristle lineage: from the polarization of sensory organ precursor cells to Notch-mediated binary fate decision. Wiley Interdiscip. Rev. Dev. Biol. 4, 299–309.

Seegar, T.C.M., Killingsworth, L.B., Saha, N., Meyer, P.A., Patra, D., Zimmerman, B., Janes, P.W., Rubinstein, E., Nikolov, D.B., Skiniotis, G., et al. (2017). Structural Basis for Regulated Proteolysis by the α-Secretase ADAM10. Cell 171, 1638-1648.e7.

Ségalen, M., Johnston, C.A., Martin, C.A., Dumortier, J.G., Prehoda, K.E., David, N.B., Doe, C.Q., and Bellaïche, Y. (2010). The Fz-Dsh Planar Cell Polarity Pathway Induces Oriented Cell Division via Mud/NuMA in Drosophila and Zebrafish. Dev. Cell 19, 740–752.

Seo, D., Southard, K.M., Kim, J., Lee, H.J., Farlow, J., Lee, J., Litt, D.B., Haas, T., Alivisatos, A.P., Cheon, J., et al. (2016). A Mechanogenetic Toolkit for Interrogating Cell Signaling in Space and Time. Cell 165, 1507–1518.

Seugnet, L., Simpson, P., and Haenlin, M. (1997). Requirement for Dynamin during Notch Signaling inDrosophilaNeurogenesis. Dev. Biol. 192, 585–598.

Severson, E., Arnett, K.L., Wang, H., Zang, C., Taing, L., Liu, H., Pear, W.S., Shirley Liu, X., Blacklow, S.C., and Aster, J.C. (2017). Genome-wide identification and characterization of Notch transcription complex–binding sequence-paired sites in leukemia cells. Sci. Signal. 10, eaag1598.

Shan, Z., Tu, Y., Yang, Y., Liu, Z., Zeng, M., Xu, H., Long, J., Zhang, M., Cai, Y., and Wen, W. (2018). Basal condensation of Numb and Pon complex via phase transition during Drosophila neuroblast asymmetric division. Nat. Commun. 9.

Shaya, O., Binshtok, U., Hersch, M., Rivkin, D., Weinreb, S., Amir-Zilberstein, L., Khamaisi, B., Oppenheim, O., Desai, R.A., Goodyear, R.J., et al. (2017). Cell-Cell Contact Area Affects Notch Signaling and Notch-Dependent Patterning. Dev. Cell 40, 505-511.e6.

Shen, Q.-T., Hsiue, P.P., Sindelar, C.V., Welch, M.D., Campellone, K.G., and Wang, H.-W. (2012). Structural insights into WHAMM-mediated cytoskeletal coordination during membrane remodeling. J. Cell Biol. 199, 111–124.

Shergill, B., Meloty-Kapella, L., Musse, A.A., Weinmaster, G., and Botvinick, E. (2012). Optical Tweezers Studies on Notch: Single-Molecule Interaction Strength Is Independent of Ligand Endocytosis. Dev. Cell 22, 1313–1320.

Shi, S., and Stanley, P. (2003). Protein O-fucosyltransferase 1 is an essential component of Notch signaling pathways. Proc. Natl. Acad. Sci. 100, 5234–5239.

Shimizu, H., Woodcock, S.A., Wilkin, M.B., Trubenová, B., Monk, N.A.M., and Baron, M. (2014). Compensatory Flux Changes within an Endocytic Trafficking Network Maintain Thermal Robustness of Notch Signaling. Cell 157, 1160–1174.

Shimizu, K., Chiba, S., Kumano, K., Hosoya, N., Takahashi, T., Kanda, Y., Hamada, Y., Yazaki, Y., and Hirai, H. (1999). Mouse Jagged1 Physically Interacts with Notch2 and Other Notch Receptors: ASSESSMENT BY QUANTITATIVE METHODS. J. Biol. Chem. 274, 32961–32969.

References

218

Siebel, C., and Lendahl, U. (2017). Notch Signaling in Development, Tissue Homeostasis, and Disease. Physiol. Rev. 97, 1235–1294.

Simonetti, B., and Cullen, P.J. (2019). Actin-dependent endosomal receptor recycling. Curr. Opin. Cell Biol. 56, 22–33.

Simpson, P. (1990). Lateral inhibition and the development of the sensory bristles of the adult peripheral nervous system of Drosophila. Dev. Camb. Engl. 109, 509–519.

Simpson, P. (2007). The stars and stripes of animal bodies: evolution of regulatory elements mediating pigment and bristle patterns in Drosophila. Trends Genet. 23, 350–358.

Skalska, L., Stojnic, R., Li, J., Fischer, B., Cerda-Moya, G., Sakai, H., Tajbakhsh, S., Russell, S., Adryan, B., and Bray, S.J. (2015). Chromatin signatures at Notch-regulated enhancers reveal large-scale changes in H3K56ac upon activation. EMBO J. 34, 1889–1904.

Skeath, J.B., and Carroll, S.B. (1992). Regulation of proneural gene expression and cell fate during neuroblast segregation in the Drosophila embryo. Dev. Camb. Engl. 114, 939–946.

Skeath, J.B., and Doe, C.Q. (1998). Sanpodo and Notch act in opposition to Numb to distinguish sibling neuron fates in the Drosophila CNS. Dev. Camb. Engl. 125, 1857–1865.

Skeath, J.B., Panganiban, G., Selegue, J., and Carroll, S.B. (1992). Gene regulation in two dimensions: the proneural achaete and scute genes are controlled by combinations of axis-patterning genes through a common intergenic control region. Genes Dev. 6, 2606–2619.

Smith, C.A., Lau, K.M., Rahmani, Z., Dho, S.E., Brothers, G., She, Y.M., Berry, D.M., Bonneil, E., Thibault, P., Schweisguth, F., et al. (2007). aPKC-mediated phosphorylation regulates asymmetric membrane localization of the cell fate determinant Numb. EMBO J. 26, 468–480.

Sokolova, O.S., Chemeris, A., Guo, S., Alioto, S.L., Gandhi, M., Padrick, S., Pechnikova, E., David, V., Gautreau, A., and Goode, B.L. (2017). Structural Basis of Arp2/3 Complex Inhibition by GMF, Coronin, and Arpin. J. Mol. Biol. 429, 237–248.

Song, Y., and Lu, B. (2011). Regulation of cell growth by Notch signaling and its differential requirement in normal vs. tumor-forming stem cells in Drosophila. Genes Dev. 25, 2644–2658.

Sorensen, E.B., and Conner, S.D. (2010). γ-Secretase-Dependent Cleavage Initiates Notch Signaling from the Plasma Membrane. Traffic 11, 1234–1245.

Sossey-Alaoui, K., Li, X., and Cowell, J.K. (2007). c-Abl-mediated phosphorylation of WAVE3 is required for lamellipodia formation and cell migration. J. Biol. Chem. 282, 26257–26265.

Southall, T.D., and Brand, A.H. (2009). Neural stem cell transcriptional networks highlight genes essential for nervous system development. EMBO J. 28, 3799–3807.

Southall, T.D., Gold, K.S., Egger, B., Davidson, C.M., Caygill, E.E., Marshall, O.J., and Brand, A.H. (2013). Cell-Type-Specific Profiling of Gene Expression and Chromatin Binding

References

219

without Cell Isolation: Assaying RNA Pol II Occupancy in Neural Stem Cells. Dev. Cell 26, 101–112.

Spana, E.P., and Doe, C.Q. (1996). Numb antagonizes Notch signaling to specify sibling neuron cell fates. Neuron 17, 21–26.

Spratford, C.M., and Kumar, J.P. (2014). Hedgehog and extramacrochaetae in the Drosophila eye: An irresistible force meets an immovable object. Fly (Austin) 8, 36–42.

Sprinzak, D., Lakhanpal, A., LeBon, L., Santat, L.A., Fontes, M.E., Anderson, G.A., Garcia-Ojalvo, J., and Elowitz, M.B. (2010). Cis-interactions between Notch and Delta generate mutually exclusive signalling states. Nature 465, 86–90.

Stevenson, V., Hudson, A., Cooley, L., and Theurkauf, W.E. (2002). Arp2/3-Dependent Psuedocleavage Furrow Assembly in Syncytial Drosophila Embryos. Curr. Biol. 12, 705–711.

Struhl, G., and Adachi, A. (1998). Nuclear Access and Action of Notch In Vivo. Cell 93, 649–660.

Struhl, G., and Greenwald, I. (1999). Presenilin is required for activity and nuclear access of Notch in Drosophila. Nature 398, 522–525.

Suckling, R.J., Korona, B., Whiteman, P., Chillakuri, C., Holt, L., Handford, P.A., and Lea, S.M. (2017). Structural and functional dissection of the interplay between lipid and Notch binding by human Notch ligands. EMBO J. e201796632.

Sun, Y., Jan, L.Y., and Jan, Y.N. (1998). Transcriptional regulation of atonal during development of the Drosophila peripheral nervous system. Dev. Camb. Engl. 125, 3731–3740.

Sun, Y., Leong, N.T., Jiang, T., Tangara, A., Darzacq, X., and Drubin, D.G. (2017). Switch-like Arp2/3 activation upon WASP and WIP recruitment to an apparent threshold level by multivalent linker proteins in vivo. ELife 6.

Sunshine, H., and Iruela-Arispe, M.L. (2017). Membrane lipids and cell signaling: Curr. Opin. Lipidol. 28, 408–413.

Swaney, K.F., and Li, R. (2016). Function and regulation of the Arp2/3 complex during cell migration in diverse environments. Curr. Opin. Cell Biol. 42, 63–72.

Swanson, C.I., Evans, N.C., and Barolo, S. (2010). Structural Rules and Complex Regulatory Circuitry Constrain Expression of a Notch- and EGFR-Regulated Eye Enhancer. Dev. Cell 18, 359–370.

Swanson, C.I., Schwimmer, D.B., and Barolo, S. (2011). Rapid Evolutionary Rewiring of a Structurally Constrained Eye Enhancer. Curr. Biol. 21, 1186–1196.

Tagami, S., Okochi, M., Yanagida, K., Ikuta, A., Fukumori, A., Matsumoto, N., Ishizuka-Katsura, Y., Nakayama, T., Itoh, N., Jiang, J., et al. (2008). Regulation of Notch Signaling by Dynamic Changes in the Precision of S3 Cleavage of Notch-1. Mol. Cell. Biol. 28, 165–176.

References

220

Takano, K., Takano, K., Toyooka, K., and Suetsugu, S. (2008). EFC/F-BAR proteins and the N-WASP-WIP complex induce membrane curvature-dependent actin polymerization. EMBO J. 27, 2817–2828.

Tal, T., Vaizel-Ohayon, D., and Schejter, E.D. (2002). Conserved Interactions with Cytoskeletal but Not Signaling Elements Are an Essential Aspect of Drosophila Wasp Function. Dev. Biol. 243, 260–271.

Terriente-Felix, A., Li, J., Collins, S., Mulligan, A., Reekie, I., Bernard, F., Krejci, A., and Bray, S. (2013). Notch cooperates with Lozenge/Runx to lock haemocytes into a differentiation programme. Dev. Camb. Engl. 140, 926–937.

van Tetering, G., van Diest, P., Verlaan, I., van der Wall, E., Kopan, R., and Vooijs, M. (2009). Metalloprotease ADAM10 is required for Notch1 site 2 cleavage. J. Biol. Chem. 284, 31018–31027.

Thieleke-Matos, C., Osório, D.S., Carvalho, A.X., and Morais-de-Sá, E. (2017). Emerging Mechanisms and Roles for Asymmetric Cytokinesis. In International Review of Cell and Molecular Biology, (Elsevier), pp. 297–345.

Tian, X., Hansen, D., Schedl, T., and Skeath, J.B. (2004). Epsin potentiates Notch pathway activity in Drosophila and C. elegans. Dev. Camb. Engl. 131, 5807–5815.

Tognon, E., Wollscheid, N., Cortese, K., Tacchetti, C., and Vaccari, T. (2014). ESCRT-0 Is Not Required for Ectopic Notch Activation and Tumor Suppression in Drosophila. PLoS ONE 9, e93987.

Tomasevic, N., Jia, Z., Russell, A., Fujii, T., Hartman, J.J., Clancy, S., Wang, M., Beraud, C., Wood, K.W., and Sakowicz, R. (2007). Differential regulation of WASP and N-WASP by Cdc42, Rac1, Nck, and PI(4,5)P2. Biochemistry 46, 3494–3502.

Torres, E., and Rosen, M.K. (2006). Protein-tyrosine kinase and GTPase signals cooperate to phosphorylate and activate Wiskott-Aldrich syndrome protein (WASP)/neuronal WASP. J. Biol. Chem. 281, 3513–3520.

Tran, D.T., Masedunskas, A., Weigert, R., and Ten Hagen, K.G. (2015). Arp2/3-mediated F-actin formation controls regulated exocytosis in vivo. Nat. Commun. 6.

Truman, J.W., Moats, W., Altman, J., Marin, E.C., and Williams, D.W. (2010). Role of Notch signaling in establishing the hemilineages of secondary neurons in Drosophila melanogaster. Development 137, 53–61.

Tyler, J.J., Allwood, E.G., and Ayscough, K.R. (2016). WASP family proteins, more than Arp2/3 activators. Biochem. Soc. Trans. 44, 1339–1345.

Uemura, T., Shepherd, S., Ackerman, L., Jan, L.Y., and Jan, Y.N. (1989). numb, a gene required in determination of cell fate during sensory organ formation in Drosophila embryos. Cell 58, 349–360.

Upadhyay, A., Kandachar, V., Zitserman, D., Tong, X., and Roegiers, F. (2013). Sanpodo controls sensory organ precursor fate by directing Notch trafficking and binding γ-secretase. J. Cell Biol. 201, 439–448.

References

221

Vaccari, T., Rusten, T.E., Menut, L., Nezis, I.P., Brech, A., Stenmark, H., and Bilder, D. (2009). Comparative analysis of ESCRT-I, ESCRT-II and ESCRT-III function in Drosophila by efficient isolation of ESCRT mutants. J. Cell Sci. 122, 2413–2423.

Volkmann, N., Amann, K.J., Stoilova-McPhie, S., Egile, C., Winter, D.C., Hazelwood, L., Heuser, J.E., Li, R., Pollard, T.D., and Hanein, D. (2001). Structure of Arp2/3 complex in its activated state and in actin filament branch junctions. Science 293, 2456–2459.

Wang, W., and Struhl, G. (2004). Drosophila Epsin mediates a select endocytic pathway that DSL ligands must enter to activate Notch. Dev. Camb. Engl. 131, 5367–5380.

Wang, W., and Struhl, G. (2005). Distinct roles for Mind bomb, Neuralized and Epsin in mediating DSL endocytosis and signaling in Drosophila. Dev. Camb. Engl. 132, 2883–2894.

Wang, X., and Ha, T. (2013). Defining Single Molecular Forces Required to Activate Integrin and Notch Signaling. Science 340, 991–994.

Wang, H., Somers, G.W., Bashirullah, A., Heberlein, U., Yu, F., and Chia, W. (2006). Aurora-A acts as a tumor suppressor and regulates self-renewal of Drosophila neuroblasts. Genes Dev. 20, 3453–3463.

Wang, H., Ouyang, Y., Somers, W.G., Chia, W., and Lu, B. (2007). Polo inhibits progenitor self-renewal and regulates Numb asymmetry by phosphorylating Pon. Nature 449, 96–100.

Wang, H., Zang, C., Taing, L., Arnett, K.L., Wong, Y.J., Pear, W.S., Blacklow, S.C., Liu, X.S., and Aster, J.C. (2014). NOTCH1-RBPJ complexes drive target gene expression through dynamic interactions with superenhancers. Proc. Natl. Acad. Sci. 111, 705–710.

Wang, Z., Bosveld, F., and Bellaïche, Y. (2018). Tricellular junction proteins promote disentanglement of daughter and neighbour cells during epithelial cytokinesis. J. Cell Sci. 131, jcs215764.

Weinmaster, G., and Fischer, J.A. (2011). Notch Ligand Ubiquitylation: What Is It Good For? Dev. Cell 21, 134–144.

Weisshuhn, P.C., Sheppard, D., Taylor, P., Whiteman, P., Lea, S.M., Handford, P.A., and Redfield, C. (2016). Non-Linear and Flexible Regions of the Human Notch1 Extracellular Domain Revealed by High-Resolution Structural Studies. Structure 24, 555–566.

Welch, M.D., Rosenblatt, J., Skoble, J., Portnoy, D.A., and Mitchison, T.J. (1998). Interaction of human Arp2/3 complex and the Listeria monocytogenes ActA protein in actin filament nucleation. Science 281, 105–108.

Weng, A.P., Ferrando, A.A., Lee, W., Morris, J.P., Silverman, L.B., Sanchez-Irizarry, C., Blacklow, S.C., Look, A.T., and Aster, J.C. (2004). Activating mutations of NOTCH1 in human T cell acute lymphoblastic leukemia. Science 306, 269–271.

Weng, A.P., Millholland, J.M., Yashiro-Ohtani, Y., Arcangeli, M.L., Lau, A., Wai, C., Del Bianco, C., Rodriguez, C.G., Sai, H., Tobias, J., et al. (2006). c-Myc is an important direct target of Notch1 in T-cell acute lymphoblastic leukemia/lymphoma. Genes Dev. 20, 2096–2109.

References

222

Wilkin, M.B., Carbery, A.-M., Fostier, M., Aslam, H., Mazaleyrat, S.L., Higgs, J., Myat, A., Evans, D.A.P., Cornell, M., and Baron, M. (2004). Regulation of Notch Endosomal Sorting and Signaling by Drosophila Nedd4 Family Proteins. Curr. Biol. 14, 2237–2244.

Windler, S.L., and Bilder, D. (2010). Endocytic Internalization Routes Required for Delta/Notch Signaling. Curr. Biol. 20, 538–543.

Wirtz-Peitz, F., Nishimura, T., and Knoblich, J.A. (2008). Linking Cell Cycle to Asymmetric Division: Aurora-A Phosphorylates the Par Complex to Regulate Numb Localization. Cell 135, 161–173.

Wolff, T., and Ready, D.F. (1991). The beginning of pattern formation in the Drosophila compound eye: the morphogenetic furrow and the second mitotic wave. Dev. Camb. Engl. 113, 841–850.

Xu, X., Choi, S.H., Hu, T., Tiyanont, K., Habets, R., Groot, A.J., Vooijs, M., Aster, J.C., Chopra, R., Fryer, C., et al. (2015). Insights into Autoregulation of Notch3 from Structural and Functional Studies of Its Negative Regulatory Region. Struct. Lond. Engl. 1993 23, 1227–1235.

Yamada, S., and Nelson, W.J. (2007). Localized zones of Rho and Rac activities drive initiation and expansion of epithelial cell-cell adhesion. J. Cell Biol. 178, 517–527.

Yamada, K., Fuwa, T.J., Ayukawa, T., Tanaka, T., Nakamura, A., Wilkin, M.B., Baron, M., and Matsuno, K. (2011). Roles of Drosophila Deltex in Notch receptor endocytic trafficking and activation: Deltex functions in Notch endocytosis. Genes Cells 16, 261–272.

Yamamoto, S., Charng, W.-L., Rana, N.A., Kakuda, S., Jaiswal, M., Bayat, V., Xiong, B., Zhang, K., Sandoval, H., David, G., et al. (2012). A Mutation in EGF Repeat-8 of Notch Discriminates Between Serrate/Jagged and Delta Family Ligands. Science 338, 1229–1232.

Yatim, A., Benne, C., Sobhian, B., Laurent-Chabalier, S., Deas, O., Judde, J.-G., Lelievre, J.-D., Levy, Y., and Benkirane, M. (2012). NOTCH1 Nuclear Interactome Reveals Key Regulators of Its Transcriptional Activity and Oncogenic Function. Mol. Cell 48, 445–458.

Yeaman, C., Grindstaff, K.K., Wright, J.R., and Nelson, W.J. (2001). Sec6/8 complexes on trans-Golgi network and plasma membrane regulate late stages of exocytosis in mammalian cells. J. Cell Biol. 155, 593–604.

Yeh, E., Zhou, L., Rudzik, N., and Boulianne, G.L. (2000). Neuralized functions cell autonomously to regulate Drosophila sense organ development. EMBO J. 19, 4827–4837.

Yeh, E., Dermer, M., Commisso, C., Zhou, L., McGlade, C.J., and Boulianne, G.L. (2001). Neuralized functions as an E3 ubiquitin ligase during Drosophila development. Curr. Biol. 11, 1675–1679.

Yonemura, Y., Futai, E., Yagishita, S., Kaether, C., and Ishiura, S. (2016). Specific combinations of presenilins and Aph1s affect the substrate specificity and activity of γ-secretase. Biochem. Biophys. Res. Commun. 478, 1751–1757.

References

223

Yousefian, J., Troost, T., Grawe, F., Sasamura, T., Fortini, M., and Klein, T. (2013). Dmon1 controls recruitment of Rab7 to maturing endosomes in Drosophila. J. Cell Sci. 126, 1583–1594.

Zacharioudaki, E., and Bray, S.J. (2014). Tools and methods for studying Notch signaling in Drosophila melanogaster. Methods 68, 173–182.

Zacharioudaki, E., Magadi, S.S., and Delidakis, C. (2012). bHLH-O proteins are crucial for Drosophila neuroblast self-renewal and mediate Notch-induced overproliferation. Dev. Camb. Engl. 139, 1258–1269.

Zallen, J.A., Cohen, Y., Hudson, A.M., Cooley, L., Wieschaus, E., and Schejter, E.D. (2002). SCAR is a primary regulator of Arp2/3-dependent morphological events in Drosophila. J. Cell Biol. 156, 689–701.

Zeng, C., Younger-Shepherd, S., Jan, L.Y., and Jan, Y.N. (1998). Delta and Serrate are redundant Notch ligands required for asymmetric cell divisions within the Drosophila sensory organ lineage. Genes Dev. 12, 1086–1091.

Zhang, Y., Yu, J.C., Jiang, T., Fernandez-Gonzalez, R., and Harris, T.J.C. (2018). Collision of Expanding Actin Caps with Actomyosin Borders for Cortical Bending and Mitotic Rounding in a Syncytium. Dev. Cell 45, 551-564.e4.

Zhong, W., Feder, J.N., Jiang, M.M., Jan, L.Y., and Jan, Y.N. (1996). Asymmetric localization of a mammalian numb homolog during mouse cortical neurogenesis. Neuron 17, 43–53.

Zhong, W., Jiang, M.M., Schonemann, M.D., Meneses, J.J., Pedersen, R.A., Jan, L.Y., and Jan, Y.N. (2000). Mouse numb is an essential gene involved in cortical neurogenesis. Proc. Natl. Acad. Sci. U. S. A. 97, 6844–6849.

Zhu, K., Shan, Z., Zhang, L., and Wen, W. (2016). Phospho-Pon Binding-Mediated Fine-Tuning of Plk1 Activity. Structure 24, 1110–1119.

Zuchero, J.B., Coutts, A.S., Quinlan, M.E., Thangue, N.B.L., and Mullins, R.D. (2009). p53-cofactor JMY is a multifunctional actin nucleation factor. Nat. Cell Biol. 11, 451–459.

Zuchero, J.B., Belin, B., and Mullins, R.D. (2012). Actin binding to WH2 domains regulates nuclear import of the multifunctional actin regulator JMY. Mol. Biol. Cell 23, 853–863.

224

List of figures

Figure 1. Overview of the steps leading to Notch activation. .................................................. 6 Figure 2. Structure and organization of Notch receptors and DSL ligands. .............................. 7 Figure 3. Structure of the Notch1-Dll4 complex in trans and cis conformations ...................... 8 Figure 4. Types of glycan modifications found in Notch EGF repeats. .................................. 11 Figure 5. A comprehensive and functional map of glycan modifications on Notch EGF repeats. ................................................................................................................................. 13 Figure 6. Mechanisms of receptor-ligand cis-inhibition. ....................................................... 14 Figure 7. Structure and force-dependent unfolding of the NRR............................................. 16 Figure 8. Overview of experimental strategies probing the force range required for Notch activation ............................................................................................................................. 18 Figure 9. Overview of mechanisms regulating NICD transcriptional activity ........................ 23 Figure 10. Notch trafficking from the cell surface to the lysosomes ...................................... 27 Figure 11. Ligand endocytosis as a molecular mechanism generating pulling forces during receptor activation. ............................................................................................................... 33 Figure 12. Neur and Mib1 functional domains ...................................................................... 39 Figure 13. Delta recycling model .......................................................................................... 43 Figure 14. Temporal profiling of Notch target gene expression ............................................. 46 Figure 15. Principle and application of the TaDa technique .................................................. 47 Figure 16. Notch-Gal4 chimeric receptors as a tool to follow Notch signaling dynamics ...... 49 Figure 17. Expression patterns of NRE-GFP and Cut, a Notch target gene............................ 50 Figure 18. Dynamics of Notch signaling using a NiGFP transgene. ...................................... 52 Figure 19. Types of gene expression profiles in response to an NICD stimulus ..................... 53 Figure 20. Dynamic encoding of Notch activity elicits different biological outputs ............... 56 Figure 21. Selection of neural progenitors during embryogenesis: from prepatterning genes to neuroblast ............................................................................................................................. 59 Figure 22. Diversity of sensory organs and lineages in Drosophila. ...................................... 61 Figure 23. Neural precursor selection occurs through gradual refinement of the cluster ........ 62 Figure 24. Patterning of the bristle rows in the notum combines prepatterning and self-organization ......................................................................................................................... 63 Figure 25. Combination between short-range and long-range signaling is a potent mechanism explaining lateral inhibition in large proneural clusters ......................................................... 64 Figure 26. Morphogenetic furrow progression and formation of ommatidial rows in the eye disc ...................................................................................................................................... 66 Figure 27. R8 selection and patterning in the eye disc........................................................... 67 Figure 28. The thoracic microchaete lineage ......................................................................... 72 Figure 29. Unequal segregation of Numb during SOP ACD in pIIb ...................................... 73 Figure 30. Polarization of polarity proteins prior to SOP division ensures unequal segregation of Numb and Neur and orientation of the mitotic spindle ...................................................... 74 Figure 31. Phosphorylation-based exclusion of Numb from the posterior pole prior to SOP division ................................................................................................................................ 75 Figure 32. Time-lapse images illustrating the relocalization of Numb from the anterior crescent to subapical endosomes in pIIb. .............................................................................. 77 Figure 33. Current model for Numb function in pIIb ........................................................... 78 Figure 34. Unequal segregation of Neur in pIIb during SOP ACD ....................................... 78 Figure 35. Correlation between contact area and Notch signaling activity ............................ 80 Figure 36. Notch localizes at the AJ level at cytokinesis in fixed tissues and in vivo ............ 81

List of Figures

225

Figure 37. Evolution of the Delta recycling model in the SO lineage .................................... 83 Figure 38. An actin rich structure containing microvilli forms at the pIIa-pIIb interface. ...... 84 Figure 39. Notch is detected laterally in vivo in numb and spdo backgrounds ....................... 86 Figure 40. Directionally trafficking Sara endosomes carrying Notch and Delta molecules are inherited by pIIa ................................................................................................................... 87 Figure 41. Notch activation at cytokinesis is required to specify the pIIa fate ....................... 89 Figure 42. Notch is activated at cytokinesis .......................................................................... 89 Figure 43. Cytokinetic ring positioning at anaphase onset ..................................................... 92 Figure 44. Overview of the molecular mechanisms at stake during anaphase elongation ....... 93 Figure 45. Midbody formation and abscission ...................................................................... 94 Figure 46. MyoII (red) activity in neighbors helps in juxtaposing the dividing cell membranes during cytokinetic ring constriction ...................................................................................... 95 Figure 47. Cell-autonomous Rac-dependent Arp2/3 activity ensures neighbor membrane withdrawal and contact expansion ........................................................................................ 97 Figure 48. Asymmetric furrowing in the notum caused by asymmetric MyoII distribution. .. 97 Figure 49. Two physically separated contact surfaces above and below the midbody at cytokinesis ........................................................................................................................... 98 Figure 50. The pIIa-pIIb interface is polarized along the apical-basal axis ............................ 99 Figure 51. Structure of the Apr2/3 complex under its active or inactive form ...................... 100 Figure 52. Diversity of Arp2/3 complexes found in cells .................................................... 102 Figure 53. Overview of cellular processes regulated by the Arp2/3 complex....................... 102 Figure 54. Summary of Arp2/3 interactors .......................................................................... 103 Figure 55. Domain structure of NPFs ................................................................................. 104 Figure 56. Activation modes of type I NPFs ....................................................................... 105 Figure 57. Arp2/3-mediated contact expansion as a zipper-like mechanism ........................ 109 Figure 58. WASp is required for correct bristle development in multiple epidermises......... 111 Figure 59. Decreasing Hairless activity is sufficient to rescue the WASp phenotype ............ 112 Figure 60. The neuronal fate is favored in absence of WASp in the RP2 lineage................. 113 Figure 61. Impairing Arp2/3 function does not impair Notch signaling during vein morphogenesis or wing margin formation........................................................................... 114 Figure 62. Alternative functions for the Arp2/3 complex at cytokinesis .............................. 116

Abstract

226

Abstract and Résumé Abstract Interplay between Notch signaling and cytokinesis in the Drosophila sensory organ

lineage

Notch signaling regulates fate specification in lineages among Metazoans. Although its

functional requirement is established, it remains unclear how Notch activity is coordinated

with lineage progression to ensure fate specification between each division round. To address

this question, I used the division of the Drosophila sensory organ precursor cell (SOP) as an

experimental model. During fly development, SOPs divide asymmetrically and generate a

pIIa cell where Notch is activated and a pIIb cell where Notch is turned off. Notch signaling is

mediated in an intralineage manner where pIIb serves as a signal source for pIIa. Following

SOP division, pIIb divides within two hours, thereby constraining pIIa fate acquisition within

this time window. In addition, activation of Notch receptors at cytokinesis was shown to be

required to specify the pIIa fate prior pIIb division. However, the molecular basis for the

Notch-cytokinesis interplay was not determined.

During my thesis, I first developed a strategy based on photobleaching and

photoconversion of fluorophore-tagged Notch receptors to determine Notch activation site

along the pIIa-pIIb interface. By doing so, I demonstrated that, in contrast with former

models, Notch receptors were activated at the lateral side of the pIIa-pIIb interface at

cytokinesis. Using live-imaging, I then provided evidence that the actin regulator Arp2/3 was

recruited to the lateral pIIa-pIIb contact during SOP division to expand the contact area and to

activate Notch receptors via Delta endocytosis. Thereby, Arp2/3 couples cytokinesis to Notch

activation following SOP division.

Keywords: Notch, Arp2/3, cytokinesis, sensory organ precursor cell, Drosophila, live-

imaging,

Résumé

227

Résumé Interaction entre la voie de signalisation Notch et la cytocinèse dans le lignage des soies

sensorielles chez la Drosophile

La voie Notch régule la spécification de destins cellulaires dans les lignages parmi les

Métazoaires. Bien sa nécessité fonctionnelle ait été établie, il est encore mal compris

comment l’activité de la voie se coordonne avec la progression du lignage pour assurer

l’acquisition des destins entre chaque division. Au cours de ma thèse, j’ai utilisé la division de

la cellule précurseur des soies sensorielles (SOP) chez la Drosophile comme modèle pour

comprendre cette coordination. Les SOPs se divisent de manière asymétrique au cours du

développement de la Drosophile et produisent une cellule pIIa, où Notch est activé, et une

cellule pIIb, où Notch est inhibé. Dans ce contexte, la signalisation se fait entre les cellules-

filles. Après la division de la SOP, pIIb se divise à son tour dans les deux heures qui suivent,

ce qui contraint l’acquisition de l’identité pIIa à cette fenêtre temporelle. De plus, des travaux

récents ont montré que les récepteurs Notch doivent être activés à la cytocinèse pour

déterminer l’identité pIIa. Cependant, la nature de l’interaction Notch-cytocinèse n’était pas

déterminée.

Au cours de ma thèse, j’ai d’abord développé une technique de photo-pistage de

récepteurs Notch marqués par des protéines fluorescentes pour déterminer le site de

signalisation le long de l’interface pIIa-pIIb, un prérequis pour comprendre l’interaction

Notch-cytocinèse. Par la suite, j’ai montré que le régulateur d’actine Arp2/3 est recruté

pendant la division de la SOP à l’interface pIIa-pIIb pour étendre le contact et pour activer les

récepteurs via l’endocytose de Delta. Ce faisant, Arp2/3 couple la cytocinèse à l’activation de

la voie Notch pendant la division de la SOP.

Mots-clés : Notch, Arp2/3, cytocinèse, cellule précurseur des soies sensorielles, Drosophile,

imagerie in vivo.

Date post:	20-Apr-2023
Category:	Documents
Upload:	khangminh22
View:	0 times
Download:	0 times

Interplay between Notch signaling and cytokinesis in the Drosophila ...

Documents