University of Groningen

Intrinsic and extrinsic regulators of stem cell function in normal and malignant hematopoiesisCapala, Marta

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2015

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Capala, M. (2015). Intrinsic and extrinsic regulators of stem cell function in normal and malignanthematopoiesis. [Groningen]: University of Groningen.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 29-06-2020

CHAPTER 2 // IMAGING

HEMATOPOIETIC

STEM CELL DIVISION:

DETERMINING THE

SYMMETRY AND

ROLE OF RAC

PROTEINS

Marta E. Capala1, M. Pratt1, Edo Vellenga1, and Jan Jacob Schuringa1*.

1Department of Experimental Hematology, Cancer Research Center Groningen (CRCG), University Medical Center Groningen, University of Groningen, Hanzeplein 1, 9700RB Groningen, the Netherlands.

(manuscript in preparation)

ABSTRACT

Hematopoietic stem cells (HSCs) have the potential to both self-renew and give rise to more differentiated progeny that recon-stitute all blood lineages. This dual charac-teristic may originate from the two types of cell division that HSCs undergo: symmet-ric and asymmetric cell division. In sever-al stem cell models, interactions with the microenvironment dictate the positioning of centrosomes and consequently symme-try of cell division. However, cooperation between cell-extrinsic signals and cell-in-trinsic regulators is most likely necessary to determine the type of cell division that an HSC undergoes. Therefore, proteins such as RAC GTPases that translate mi-croenvironmental signals into intracellu-lar responses may be important regulators of HSC cell division. While our insights into the mechanisms that regulate stem cell fate are largely based on experiments performed on heterogeneous populations of stem cells, live cell imaging of sin-gle dividing HSCs is necessary to obtain

a better understanding of the mechanisms governing the symmetry of HSC division. Within this study we developed techniques that allow imaging the symmetry of an in- dividual human hematopoietic cell in real time. By generating imaging tools in he-matopoietic cell lines, applying them in human hematopoietic stem and progenitor cells (HSPCs) and developing methods of quantitative analysis, we were able to es-tablish a model in which we could observe and quantify the symmetry of HSPC cell divisions. Using these tools we observed specific localization patterns of the two RAC proteins, RAC1 and RAC2 in resting and dividing human hematopoietic cells and noted that RAC activity is necessary in the process of cell division in those cells. Although in this study we focused on RAC proteins, the methods that we de-veloped can be used to assess the effects of various cell-intrinsic and cell-extrinsic factors on the symmetry and progress of HSPC cell division.

IMAGING HEMATOPOIETIC STEM CELL DIVISION

2

// 33introduction

INTRODUCTION

Stem cells (SCs), and among them hema-topoietic stem cells (HSCs), are character-ized by their ability to both self-renew and give rise to more differentiated progeny that sustains tissue regeneration through-out life [1–3]. It has been proposed that this dual characteristic of SCs originates from different types of cell division that they undergo [4]. Asymmetric cell division would give rise to one daughter stem cell and one more differentiated progenitor cell, thus maintaining the stem cell pool and at the same time sustaining produc-tion of mature cells [5,6]. Symmetric cell division would result either in generation of two daughter stem cells and therefore SC expansion, or in two progenitor cells leading to SC exhaustion (Fig. 1A). These

different types of cell division are most likely not exclusive, but might co-exist within the stem cell pool in a balance that can change depending e.g. on the regen-erative requirements of the tissue [4,7–9].

Traditionally, two opposing views on the regulation of the symmetry of stem cell division have been proposed: the cell in-trinsic, and the cell extrinsic one [5,10–14] (Fig. 1B). In the cell intrinsic model, deter-minants of the symmetry of stem cell divi-sion are present within the cell and their distribution during cell division determines the type of division and the fate of the daughter cells. Among those factors Brat and Prospero in Drosophila neural stem cells and mammalian Inscuteable in devel-oping retina have been shown to influence

1

2

3

A.

B.

DETERMINANTS OF (A)SYMMETRY:1) intrinsic

- asymmetric segregation ofproteins/organelles/DNA

2) extrinsic- interactions with microenvironment

influence position of the spindle3) stochastic

Figure 1. Capala et al. 2015

asymmetric symmetric expansion symmetric exhaustion

Figure 1. Types and regulation of stem cell (SC) cell division. (A) Schematic representation of different types of SC division, in relation to their stroma. (B) Schematic overview of possible deter-minants of the symmetry of SC division: 1) cell intrinsic that segregate asymmetrically into daughter cells; 2) cell extrinsic influences from the microenvironment; 3) stochastic.

34 // introduction

the orientation of the mitotic spindle and therefore the symmetry of SC division [15,16]. However, stem cells reside within a specialized microenviromnent and can be accurately described only in relation to their niche [17–22]. Accordingly, the cell extrinsic model proposes that microenvi-ronmental signals can dictate the type of cell division that the stem cell undergoes as well as the fate of daughter cells. Ad-herent junctions formed by the cell-to-cell contacts between SC and their niche have been shown to influence the symmetry of SC division [14,23,24] In Drosophila male germline stem cells (GSCs), interaction be-tween a stem cell and the hub cell created an anchoring platform for Apc2 proteins that in turn oriented the mitotic spindle perpendicular to the hub. Such a position-ing of the mitotic spindle resulted in an asymmetric cell division. Consequently, the daughter cell that remained in contact with the niche retained stem cell proper-ties, while the daughter cell that moved away from the niche underwent differenti-ation [14]. This suggests that co-operation between cell extrinsic and cell intrinsic factors is required to determine the sym-metry of SC cell division [1,25] (Fig. 1B). In-terestingly, some of the intrinsic stemness factors are also closely involved in other aspects of SC-niche interactions. Cdc42 is a member of the family of Rho GTPases and acts as a molecular switch trans-mitting microenvironmental cues to the downstream cellular signaling pathways [26–29]. At the same time, asymmetric distribution of Cdc42 has been associated with a young HSC phenotype [30]. Anoth-er member of Rho family, Rac, has been shown to regulate asymmetric cell division in mammalian oocytes. Halet and Carroll have shown that polarized localization of active GTP-bound Rac regulated the

stability and anchoring of meiotic spin-dle to the cortex, which in turn enabled asymmetric cell division [31]. A similar function for Rac-GTP has been described in Drosophila female GSCs [32]. Whether Rac proteins have a role in regulating the symmetry of cell division also in other SC types is currently unclear.

Although most of the insight into the mechanisms governing the symmetry of SC cell division comes from studies in the Drosophila model, there is evidence that symmetric and asymmetric divisions oc-cur also in hematopoietic system of higher vertebrae [33,34]. In vitro studies utilizing the Notch reporter as a marker of stemness showed that the different types of cell divi-sion co-exist in murine HSCs and that the balance between symmetric and asymmet-ric cell divisions can be affected by both microenvironmental signals as well as oncogenes [33]. Nup98-HoxA9 and BMI1 overexpression has been shown to alter the balance in HSC cell division, leading to an increase in symmetric cell divisions and expansion of HSCs [33,35,36]. In the human system, current protocols of HSC purification result in a mixed population of stem and progenitor cells (HSPCs). Mea-suring an average of such a heterogeneous group, as most experimental approaches do, has the pitfall of omitting small, but relevant subgroups. Methods of live cell imaging at the single cell level developed in recent years can ameliorate this prob-lem and are therefore highly useful in the field of stem cell research [37–40]. Employ-ing these methods for HSCs could lead to a better understanding of the mechanisms governing the symmetry of cell division.

Here, we describe the development of techniques that allow imaging the symme-try of individual hematopoietic cell in real time. Through generation of the imaging

IMAGING HEMATOPOIETIC STEM CELL DIVISION

2

// 35primary cell isolation and culture conditions

tools in hematopoietic cell lines, apply-ing them in human hematopoietic stem and progenitor cells (HSPCs) and finally through developing methods of quantitative analysis, we were able to establish a model in which we could observe and quantify the symmetry of HSPC cell division.

MATERIALS AND METHODS

PRIMARY CELL ISOLATION AND CULTURE CONDITIONS

Neonatal cord blood (CB) was obtained from healthy full-term pregnancies after informed consent in accordance with the Declaration of Helsinki form the obstet-rics departments of the University Medi-cal Centre Groningen (UMCG) and Mar-tini Hospital Groningen, Groningen, The Netherlands. All protocols were approved by the Medical Ethical Committee of the UMCG. After separation of mononuclear cells with lymphocyte separation medi-um (PAA Laboratories, Coble, Germany), CD34+ cells were isolated using magneti-cally activated cell sorting (MACS) CD34 progenitor kit (Miltenyi Biotech, Am-sterdam, The Netherlands). For the MS5 co-culture experiments cells were grown in Gartner’s medium consisting of α-mod-ified essential medium (α–MEM; Fisher Scientific Europe, Emergo, The Nether-lands) supplemented with 12,5% heat-inac-tivated fetal calf serum (Lonza, Leusden, The Netherlands), 12,5% heat-inactivat-ed horse serum (Invitrogen, Breda, The Netherlands), 1% penicillin and strepto-mycin, 2 mM glutamine (all from PAA Laboratories), 57,2 µM β-mercaptoethanol (Merck Sharp & Dohme BV, Haarlem, The Netherlands) and 1 µM hydrocortisone (Sigma-Aldrich Chemie B.V., Zwijndrecht,

The Netherlands). For the imaging experi-ments, Gartner’s medium was supplement-ed with 100 ng/mL stem cell factor (SCF), FLT3 Ligand (Flt3L; both from Amgen, Thousand Oaks, USA) and thrombopoie-tin (TPO; Kirin, Tokyo, Japan).

CELL LINES AND CULTURE CONDITIONS

293T embryonic kidney cells (ACC-635 DSMZ), PG13 packaging cells (ATCC CRL-10686) and HeLa cells (ACC-57 DSMZ) were grown in DMEM medium with 200 mM glutamine (BioWhittaker) supplemented with 10% FSC and 1% pen-icillin and streptomycin. K562 myelog-enous leukemia cells (ACC-10, DSMZ) and TF-1 erythroleukemic cells (ACC-334, DSMZ) were grown in RPMI medium with 200 mM glutamine (BioWhittaker) supple-mented with 10% FCS, and 1% penicillin and streptomycin, and for TF-1 cells with 5 ng/mL granulocyte-macrophage colony stimulating factor (GM-CSF; Genetics In-stitute, Cambridge, MA, USA). MS5 mu-rine stromal cells (ACC-441, DSMZ) were grown in αMEM with 200 mM glutamine (BioWhittaker) supplemented with 10% FCS and 1% penicillin and streptomycin.

RETRO- AND LENTIVIRUS GENERATION AND TRANSDUCTION

A C-terminal mCherry in-frame fusion with α-tubulin was cloned into the MSCV vector and stable PG13 producer cell lines were generated and used as published pre-viously [41]. Supernatants from the PG13 cells were harvested after 8-12 hours of incubation in HPGM and passed through

36 // materials and methods

0.45-mm filters before the retroviral trans-duction rounds (Sigma-Aldrich). Three rounds of transduction were performed on retronectin-coated 24-well plates in the presence of 4 µg/mL polybrene (Sigma- -Aldrich). With the last round of transduc-tion, lentiviral transduction with the GFP-RAC1 or GFP-RAC2 constructs described below was performed.

N-terminal GFP in-frame fusion con-structs of RAC1, RAC2 and C-terminal GFP in-frame fusion construct of γ-tubulin were cloned into the pRRL vector. 293T embry-onic kidney cells were transfected using Fu-GENE6 (Roche, Almere, The Netherlands) with 3 µg pCMV Δ8.91, 0,7 µg VSV-G, and 3 µg of vector constructs (pRRL-GFP-RAC1 (GFP-RAC1), pRRL-GFP-RAC2 (GFP-RAC2), or pRRL-γ-tubulin-GFP (γ-tubulin-GFP)). After 24 hours medium was changed to HPGM and after 12 hours supernatant containing lentiviral particles was har-vested and either stored at -80°C or used fresh for transduction of target cells. TF-1 cells, HeLa cells or isolated CD34+ CB cells that were pre-stimulated for 12 hours were subjected to 1 round of transduction with lentiviral particles in the presence of prestimulation cytokines and 4 µg/mL polybrene (Sigma) on retronectin-coated 24-well plates (Takara, Tokyo, Japan). Af-ter transduction GFP-positive or mCherry- and GFP-double-positive HeLa and TF-1 cells were sorted on a MoFlo sorter (Dako Cytomation). For CB cells, staining for CD34-PE and CD38-APC was performed and GFP+CD34+CD38- cells were sorted.

FLOW CYTOMETRY ANALYSIS AND SORTING

All fluorescence-activated cell sort-er (FACS) analyses were performed on

a FACScalibur (Becton-Dickinson [BD], Alpen a/d Rijn, the Netherlands) and data were analyzed using WinList 3D (Verity Software House, Topsham, USA). Cells were sorted on a MoFlo sorter. Antibodies: CD34-PE and CD38-APC were obtained from BD.

FLUORESCENT MICROSCOPY AND TIME-LAPSE IMAGING

Transduced TF-1 cells were plated into retronectin-coated glass bottom chamber slides (Nagle Nunc International, Naper-ville, Il, USA) either for time-lapse imag-ing as described below, or for high-resolu-tion confocal imaging. For the co-localiza-tion of γ-tubulin-GFP and RAC, transduc-ed TF-1 cells were cytospun and fixated in 4% paraformaldehyde, and staining with RAC antibody (610650; BD) was performed according to standard procotols. Images were acquired with a Leica SP2 AOBS confocal microscope, 63x objective and 3D reconstructions were generated using Imaris software, version 6.4.2 (Bitplane, Zurich, Switzerland). Transduced HeLa cells were plated on a retronectin-coated slides with adhesive square-shaped areas (CYTOO) according to manufacturer’s in-structions and cell spreading on micropat-terns was assessed on a Leica DMIL in-verted phase microscope. Subsequently, images were acquired with a Leica SP2 AOBS confocal microscope, 63x objective. GFP+CD34+CD38- CB cells were sorted and plated on MS5 in glass bottom chamber slides in Gartner’s medium supplemented with 100 ng/mL SCF, Flt3L and TPO. Cells were and analyzed every 5 minutes for 16 hours using a Solamere Nipkow Con-focal Live Cell Imaging system (based on Leica DM IRE2 Inverted microscope; Leica)

IMAGING HEMATOPOIETIC STEM CELL DIVISION

2

// 37statistical analysis

under the control of InVivo software (Me-dia Cybernetics, Silver Springs, USA). For the imaging of GFP-RAC1 and GFP-RAC2, a 63x objective was used and 20 confocal z-stacks were acquired at each timepoint. Images were analyzed with ImageJ soft-ware (freeware, developed by NIH). For the imaging of γ-tubulin-GFP, a 40x objec-tive was used and 10 confocal stacks were acquired. Image analysis and 3D recon-structions were performed using Imaris software.

STATISTICAL ANALYSIS

All values are expressed as means ± stan-dard deviation (SD). Student’s t test was used for all other comparisons. Differenc-es were considered statistically significant at p<0,05.

RESULTS

RAC1 AND RAC2 SHOW DISTINCT SUBCELLULAR LOCALIZATION IN RESTING AND DIVIDING HEMATOPOIETIC CELLS

RAC GTPases have been implicated in regulating the symmetry of cell division in Drosophila GSC and mammalian oo-cytes [31,32]. In the hematopoietic system, two members of that family are expressed (RAC1 and RAC2) and they have been shown to induce cytoskeleton rearrange-ments in response to microenviromental signals [42–44]. We questioned what the distribution pattern of RAC GTPases in human hematopoietic cells during cell di-vision would be and whether their local-

ization could influence the symmetry of HSC division. To allow imaging of living cells, we generated lentiviral constructs containing fluorescent-tagged RAC1 and RAC2, as well as α- and γ-tubulin (Fig. 2A). In the case of GFP-tagged RAC constructs, GFP was fused fused to the N-terminal ends of the proteins to ensure that the tag would not affect their functionality since the C-terminal region contains the polybasic motif that undergoes posttrans-lational modifications and it responsible for the targeting of RAC proteins to cell membranes. Those lentiviral constructs were then used to stably transduce hema-topoietic TF-1 cells that have a propensity to adhere and spread on retronectin-coat-ed plastic or glass surfaces. This ability of TF-1 cells allowed us to culture them in retronectin-coated chamber slides and perform time-lapse imaging of living cells. We observed a distinct localization pattern of the two RAC proteins, with RAC1 en-riched at the plasma membrane and RAC2 present in the cytoplasm (Fig. 2B). Impor-tantly, TF-1 cells plated on retronectin were still able to divide normally. During cell di-vision, formerly spread cells were round-ing up but remained attached to the sur-face and relatively immobile. This allowed us to acquire high-resolution confocal stacks of dividing TF-1 cells express-ing GFP-RAC1 or -RAC2 and α-tubulin- -mCherry, which were then used to create 3D reconstructions. In a 3D representa-tion of a dividing cell it was clearly visible that also during cell division RAC1 was located in the plasma membrane of the cell, while RAC2 was present in the cy-toplasm (Fig. 2C and Supplementary mov-ies 1 and 2). However, we did not observe co-localization of RAC proteins with the microtubules of the mitotic spindle, visual-ized by α-tubulin-mCherry. Subsequently,

38 // results

A B

C

GFP-RAC1

still fromtime-lapsemovie

GFP-RAC2

still fromtime-lapsemovie

Figure 2. Capala et al 2015.

GFP-RAC1α-tubulin-mCherry

TF-1

GFP-RAC2α-tubulin-mCherry

TF-1

DAPI �-Tubulin-GFP �-RAC1/2

D

merge

SFFV GFP RAC1 PRE

SFFV GFP RAC2 PRE

3'LTR mCherry 5'LTRα-tubulin

SFFV GFP PREγ-tubulin

Figure 2. RAC1 and RAC2 display different subcellular localization patterns in hematopoietic cell lines. (A) Schematic representation of fluorescent-tagged protein constructs used in the imaging experiments. (B) TF-1 cells expressing GFP-RAC1 or GFP-RAC2 were plated in retronectin-coated chamber slides and time-lapse imaging was performed for 16hrs. Stills from the time-lapse movies are shown. (C) TF-1 cells expressing GFP-RAC1 or GFP-RAC2 and α-tubulin-mCherry were plated in ret-ronectin-coated chamber slides and high-resolution confocal imaging was performed. Acquired con-focal stacks were used to generate 3D reconstructions of dividing TF1 cells that visualize the localization of GFP-RAC1 or –RAC2 (green channel) and α-tubulin-mCherry (red channel). (D) K562 cells express-ing γ-tubulin-GFP were cytospun, fixated in 4% formaldehyde and stained with anti-RAC1/2 antibody. DAPI was used to stain DNA. Blue channel – DAPI, green channel - γ-tubulin-GFP, red channel – RAC1/2.

since both our RAC and γ-tubulin con-structs were GFP-tagged, we performed immunostaining using anti-RAC antibody (recognizing both RAC1 and RAC2) on γ-tubulin-GFP-expressing TF-1 cells. In-terestingly, co-localization of both signals

was seen, indicating possible interaction of RAC proteins with centrosome structures (Fig. 2D) in non-dividing cells.

IMAGING HEMATOPOIETIC STEM CELL DIVISION

2

// 39time-lapse imaging of living hspcs reveals specific localization patterns of rac1 and rac2

GFP-RAC1

top

bottom

CB CD34+/CD38- GFP-RAC2

top

bottom

CB CD34+/CD38-

Figure 3. Capala et al 2015.

A

B

CD34 CB+ GFP CD34 38 CB+ + -

GFP-RAC1/RAC2lentivirus

MS5-coated chamberslide

top slide

bottom slide

20µM

top

bottom

time-lapsemovie

Figure 3. RAC1 and RAC2 localize differently in resting and dividing human HSPCs. CD34+ CB cells were transduced with GFP-RAC1 or GFP-RAC2 constructs and GFP+ CD34+ CD38- cells were sorted and plated on MS5-coated chamber slides for confocal time-lapse imaging. Stills from time-lapse movies are shown as a composite of 20 confocal z-stacks (experimental setup shown in A). (B) Localization of GFP-RAC1 and GFP-RAC2 in human HSPCs plated on stroma, shown as a still form the time-lapse movie.

TIME-LAPSE IMAGING OF LIV-ING HSPCs REVEALS SPECIFIC LOCALIZATION PATTERNS OF RAC1 AND RAC2

Next, we wanted to visualize the local-ization of RAC1 and RAC2 in human HSC-enriched population, defined as lin-

CD34+CD38- cells. Since we were inter-ested in both resting and dividing cells,

we optimized the experimental protocols to ensure that the imaged cells remained alive and in good condition for the du-ration of the experiment. To this end, we pre-seeded MS5 cells in an imaging chamber and allowed them to reach con-fluency to create a feeder layer for hema-topoietic cells. CD34+CB cells were then transduced with lentiviral constructs con-taining GFP-RAC1 or -RAC2 and GFP+C-

40 // results

D34+CD38- cells were sorted and plated on the MS5-coated chamber slides in a cyto-kine-rich medium. Cells were allowed to recover overnight and imaging was per-formed on the following day. Exposure time, number of z-stacks acquired and ac-quisition frequency were set to achieve the highest possible quality of images while minimizing phototoxicity (Fig. 3A). This optimized experimental setup resulted in acquisition of several cell divisions with-in the HSC-enriched population. We could observe that, similar to TF-1 cells, RAC1 was enriched in the plasma membrane of resting HSCs and remained there also during cell division. In contrast, RAC2 was present in the cytoplasm and its distribu-tion was more dynamic as the cell division progressed. However, no clear asymmetry was observed in the localization of either RAC1 or RAC2 in the dividing cells, or in their distribution between the two daugh-ter cells (Fig. 3B and Supplementary mov-ies 3 and 4).

RETRONECTIN-COATED PATTERNS INDUCE CELL SPREADING AND (A)SYMMETRIC ORGANIZATION OF CELLULAR STRUCTURES

It has been previously shown that contacts between a cell and the components of the extracellular matrix (ECM) affect the or-ganization of cytoskeleton and can guide the symmetry of cell division [45]. This concept has been used to develop com-mercially available glass slides on which micrometer-level sized patterns are spot-ted. Those micropatterns can be coated with components of the ECM, and used to guide the symmetry of cell division. Con-

sequently, depending on the shape of the pattern, cells are forced to divide in ei-ther a symmetric or asymmetric manner (Fig. 4A). Therefore, we wondered whether upon imposing different symmetry of cell division by different micropatterns dis-tribution of RAC1 and RAC2 during cell division would also change. To setup the experiment, we first used slides spotted with symmetric micropatterns and coat-ed them with retronectin to enable bind-ing and spreading of TF-1 hematopoietic cells. Although we have previously seen TF-1 cells adhering readily to retronectin, we could not detect any spreading of those cells on the micropatterns, regardless of their size. Therefore, we proceeded to use the adherent HeLa cell line in which we stably expressed either GFP-RAC1 or GFP-RAC2. Those cells were able to ad-here and spread on the retronectin-coat-ed patterns, allowing for precise confocal imaging of the localization of RAC1 and RAC2 (Fig. 4B). Similarly to hematopoiet-ic cell lines and primary HSPCs, also in HeLa cells RAC1 was strongly enriched in the plasma membrane. RAC2 was present in the cytoplasm, with clear perinuclear localization (Fig. 4C). However, despite imaging several adherent cells, we did not observe any cell divisions occurring on the micropatterns, deeming this system not suitable for our experiments.

GFP-LABELLED γ-TUBULIN CAN BE USED TO TRACK CENTROSOME LOCALIZATION AND DISTINGUISH DIFFERENT TYPES OF CELL DIVISION

Since we were not able to setup and ex-perimental system in which a certain type

IMAGING HEMATOPOIETIC STEM CELL DIVISION

2

// 41gfp-labelled γ-tubulin can be used to track centrosome localization and distinguish different types of cell division.

GFP-RAC1

GFP-RAC2

top bottom

top bottom

Hela

Hela

BA

C

Figure 4. Capala et al 2015.

asymmertic

symmertic

Figure 4. Retronectin-coated micropattern induce spreading and symmetric intracellular or-ganization in HeLa cells. (A) Representation of asymmetric and symmetric micropatterns (left pan-els) and the cell spreading they induce (right panels) and organization of micropatterns on the cov-erslip. (B) HeLa cells were plated on the coverslip spotted with retronectin-coated micropatterns and allowed to adhere. Overview of a section of the slide acquired with the inverted phase microscope is shown. (C) GFP-RAC1 and GFP-RAC2-expressing HeLa cells were plated as described in panel B on symmetric micropatterns and confocal imaging of adherent cells was performed. Localization of GFP-RAC1 and GFP-RAC2 is shown as a composite of 5 confocal z-stacks.

was to observe a high number of cells and acquire images of as many cell divi-sions as possible, rather than observing details of subcellular organization. We observed that γ-tubulin-GFP was a clear marker of cell division, as the appearance of two centrosomes shortly preceded the beginning of mitosis and their segrega-tion into the daughter cells could be easily followed. Moreover, positioning of the two centrosomes could be used to distinguish two different types of cell division. In the symmetric cell division, two centrosomes were positioned in the same z-plane and both daughter cells remained in contact with stroma. In the asymmetric cell divi-

of cell division would be imposed, we fo-cused on determining whether symmetric and asymmetric cell divisions could natu-rally be observed in human HSC-enriched populations. As described above, we used a co-culture system to ensure optimal cul-ture conditions and register a high number of cell divisions. The GFP-labelled γ-tubu-lin construct was tested in TF-1 cell line (Supplementary movie 7) and then sta-bly introduced into CD34+ CB cells, after which CD34+38- cells were sorted to en-rich for HSCs and plated in stroma-coat-ed chamber slides. For the imaging of the co-cultures a relatively small magnifica-tion of 40x was used, since the objective

42 // results

Figure 5. Capala et al 2015.

sym

met

ric

asym

met

ric0

100

200

300

400

min

ute

sin

mito

sis

A

B

symmetric

asymmetric

C

sym

met

ric

asym

met

ric0

10

20

30

nu

mb

ero

fd

ivis

ions

glass coverslipcoated with stromal cells

z-axis

top

0 M�

15 M�

glass coverslipcoated with stromal cells

z-axis

top

0 M�

15 M�

Figure 5. γ-tubulin-GFP can be used to track HSPC cell division and to determine the sym-metry of cell division. (A) CD34+ CB cells were transduced with γ-tubulin-GFP construct and GFP+ CD34+ CD38- cells were sorted and plated on MS5-coated chamber slides for confocal time-lapse imaging. Confocal z-stacks were then used for the 3D reconstructions of dividing cells. An exam-ple of a symmetric and asymmetric cell division is shown with schematic representations on the left panels. (B) Overall quantification of the symmetric and asymmetric cell divisions registered in 3 independent imaging experiments. (C) Duration time of cell divisions registered in an experiment as described in panel A were analyzed and the two types of cell divisions were compared. Division time is represented as the number of minutes the individual cell spent in mitosis.

sion, centrosomes were located in differ-ent z-stacks and while one daughter cell remained attached, the other one was no more in contact with stromal cells after the division (Fig. 5A and Supplementary movie 5 and 6). In the three experiments

analyzed, similar number of symmetric and asymmetric cell divisions was ob-served (Fig. 5B). Moreover, we have an-alyzed the duration of acquired cell divi-sions and compared the two types. Sym-metric cell division took on average longer

IMAGING HEMATOPOIETIC STEM CELL DIVISION

2

// 43inhibition of rac activity in human hematopoietic cells results in a prolonged mitosis duration time

to complete (although not statistically sig-nificant), and larger variability in the du-ration of cell division was observed in this division type (Fig. 5C). Overall, this indi-cates that imaging GFP-labelled γ-tubulin is useful not only to detect the beginning of cell division, but also to determine its properties such as geometrical symmetry and duration time.

INHIBITION OF RAC ACTIVITY IN HUMAN HEMATOPOIETIC CELLS RESULTS IN A PROLONGED MITOSIS DURATION TIME

Having developed the tools to monitor progress of cell division in human hema-topoietic cells, we then wondered wheth-er interfering with intracellular signaling pathways would result in changes that could be registered using those tools. In-hibition of RAC activity with a small mol-ecule NSC23766 presented an interesting target, given the role of RAC proteins in

mediating the interactions between hema-topoietic stem cells and their niche, as well as in regulating cytoskeleton rearrange-ments [42,44]. TF-1 cell line stably express-ing α-tubulin-mCherry was used to enable robust imaging of cell divisions during an overnight time-lapse experiment. Indeed, the first experiment performed yielded acquisition of a large number of cell divi-sions. Analysis of all the cell divisions ob-tained for the NSC-treated as well as un-treated cells revealed that RAC inhibition resulted in a largely decreased frequency of cell division. Moreover, the duration of mitosis was significantly extended in NCS-treated cells (Fig. 6 and Supplemen-tary movie 8 and 9. The two other ex-periments performed showed the same trend, although in those cases a much lower number of cell divisions was ac-quired, highlighting the risk of phototox-icity during prolonged time-lapse imaging. Taken together, this data shows that the imaging setup we developed can be use-ful to identify and analyze changes in the

untre

ated

NSC

-trea

ted

0

100

200

300

400

500

min

ute

sin

mitosis

untre

ated

NSC-tr

eate

d0

100

200

300

400

min

ute

sin

mitosis

Experiment 1 Experiment 3Experiment 2

untre

ated

NSC

-trea

ted

0

100

200

300

min

ute

sin

mito

sis

*** *

Figure 6. Capala et al 2015.

Figure 6. RAC activity is required for the progress of cell division in hematopoietic cells. TF1 cells expressing α-tubulin-mCherry were plated on MS5-coated 2 chamber slides for confocal time-lapse imaging. At the beginning of the imaging experiment RAC inhibitor was added to one chamber, while the other chamber remained untreated. Cells were then followed for 16hrs. Duration time of the registered cell division was analyzed and compared between NSC-treated and untreated cells. The results of 3 independent experiments are shown. * P<0.05, *** P<0.001.

44 // discussion

frequency and progression of mitosis in pimary human HSCs grown on bone mar-row stromal cells. Furthermore, we con-clude that inhibiting RAC activity results in a decreased number of cell divisions taht take longer to complete.

DISCUSSION

RAC PROTEINS AS REGULATORS OF THE SYMMETRY OF HSC CELL DIVISION

RAC GTPases act as molecular switches that in response to microenvironmental stimuli activate various effector proteins, resulting in for instance cytoskeleton re-arrangements in HSCs [42–44]. Since the role of RAC proteins in regulating the symmetry of division of other stem cell types has been described, they presented as potentially interesting candidates for determining the symmetry of HSC cell division. Murine knock-out models have shown that despite the very high amino-acid sequence homology, RAC1 and RAC2 have non-redundant functions in hemato-poietic cells [42]. This specificity of func-tion could be explained at least in part by the differences in the C-termini of the two proteins that lead to their distinct subcel-lular distribution in murine neutrophils [46,47]. Here, we showed that RAC1 and RAC2 displayed different subcellular lo-calization in human hematopoietic cell lines, as well as in HSC-enriched human CD34+38- CB cells and that this distinct lo-calization pattern was also apparent in di- viding cells. However, no asymmetry in dis- tribution of RAC1 or RAC2 before or during cell division was observed, and the level

of RAC protein that was passed on to the two daughter cells was usually comparable.

Although RAC2 was present in the cytoplasm of dividing cells, no co-localiza-tion with the mitotic spindle could be seen. Interestingly, when TF-1 cells expressing γ-tubulin-GFP were immunostained with anti-RAC antibody, a clear co-localization of the two proteins was observed. In pro-teomic experiments that we have previously performed, several proteins involved in reg-ulation of mitosis were identified in RAC2- associated complexes, including Aurora Kinase B (Capala et al. 2015, in press). The Aurora kinases are highly conserved group of proteins that regulate chromosomal alignment and segregation during cell di-vision. Aurora B specifically controls chro-mosome condensation, the spindle check-point and cytokinesis as a member of the chromosome passenger complex [48]. It is therefore possible that RAC proteins are involved in regulating HSC division in an indirect manner, by activating key players of the mitotic machinery, responsible for the organization of the mitotic spindle and timely progression of cell division. Impor-tantly, when we blocked RAC activity in TF-1 cells, we observed not only fewer cell divisions, but also a significantly prolonged division time. This indicated that the mi-totic machinery was disturbed, although the exact mechanisms remain to be de-termined. Whether blocking RAC activity could also influence the symmetry of HSC division requires further investigation.

IMAGING HEMATOPOIETIC STEM CELL DIVISION

2

// 45use of retronectin-coated micropatterns for enforcing symmetric or asymmetric cell division

USE OF RETRONECTIN-COATED MICROPATTERNS FOR ENFORCING SYMMETRIC OR ASYMMETRIC CELL DIVISION

Since we did not observe asymmetry in the distribution of RAC proteins during or af-ter cell division, we wondered whether by enforcing a particular subcellular organi-zation and consequently cell division type we could affect RAC localization. We coat-ed commercially available CYTOO chips with retronectin that we previously used for inducing spreading of TF-1 cells, yet were not able to stimulate them to attach to the micropatterns. It is possible that TF-1 cells, being naturally a non-adherent cell line, require a larger retronectin-coat-ed surface for adhering. In agreement with that, TF-1 cells were able to spread on the control area of a CYTOO chip, where adherent surface was not limited to the micropatterns. Although adherent HeLa cell lines that we subsequently used were readily attaching and spreading on the retronectin-coated micropatterns, we were not able to observe any cell divisions despite imaging several individual cells. It is possible that the combined stress ap-plied by spreading on a micropattern and by prolonged imaging resulted in a block in cell cycle progression in the observed cells making this system unsuitable for our purpose.

DEFINING THE SYMMETRY OF HSC CELL DIVISION – CHALLENGES AND OPPORTUNITIES

In various types of stem cells, specific po-sitioning of the mitotic spindle in respect

to the supporting stromal cells has been shown to result in a defined symmetry of cell division [14,16,25]. Therefore, we an-alyzed centrosome positioning, visualized by γ-tubulin-GFP, to determine the sym-metry of HSC cell divisions and observed both symmetric and asymmetric types of cell division. However, monitoring a suf-ficient number of cell divisions in these types of experiments poses several tech-nical challenges since cells divide infre-quently and prolonged culture periods are therefore necessary. While setting up the time-lapse imaging of RAC1/RAC2 local-ization in human HSPCs, we were able to find the optimal culture conditions for prolonged imaging experiments. By keep-ing the HSPCs on stroma and in Gartner’s medium, cells remained in good condition and the addition of cytokines increased the division rate. Moreover, presence of stromal cells allowed modeling the inter-actions between HSCs and their niche, which we then used for determining the symmetry of HSC division. Despite using optimal culture conditions, phototoxicity remained an issue during long imaging experiments [49]. Our experience high-lights the necessity to find balance be-tween obtaining the best images possible and preventing toxic effects from occur-ring in observed cells. Therefore, in the experiments aiming at determining the symmetry of cell division, we limited the number of confocal z-stacks we acquired for each timepoint to 10. Although the 3D reconstructions obtained in that way were not detailed, they were sufficiently accu-rate to distinguish position of the two cen-trosomes, while limiting the exposure time to the minimum. Interestingly, primary HSPCs seemed more resistant to photo-toxicity than TF-1 cell lines that suffered significantly in two out of three experi-

46 // acknowledgements

ments performed. One reason for that may be the protective effect of stroma, used in the experiments with primary material but not with cell lines. Moreover, we ob-served more pronounced cytotoxic effects in cells expressing higher levels of GFP. Since transduction efficiency achieved in cell lines is in general higher than in pri-mary cells, higher overexpression of GFP-tagged constructs is obtained, making transduced cells susceptible to phototoxic-ity. This further stresses the need to select for those cells in which levels of fluoren-cent-tagged protein are within the range of endogenous protein for all the imaging experiments.

It is postulated that in the normal steady-state hematopoiesis asymmetric cell divisions would be the dominant kind as they allow both maintenance of HSC pool and generation of mature cells [4]. However, in our experiments we observed a compara-ble numbers of symmetric and asymmetric cell division. This may be explained by the culture conditions in which the HSCs were kept as culturing HSCs in vitro requires using high concentration of cytokines for their survival and proliferation. These con-ditions may therefore promote symmetric cell division and therefore expansion of HSCs [7,8]. It must be therefore noted that any in vitro imaging experiment provides information on what resembles stress he-matopoiesis, rather than on the normal, steady state division of HSCs. Although we were able to visualize two types of HSCs cell division of different geometrical sym-metry, it reminds to be verified whether that (a)symmetry corresponds to different fate of the daughter cells. Previous studies on mouse HSCs have used Notch reporter assays to distinguish stem cells from more differentiated progenitors and in that way determine the rate of symmetric versus

asymmetric cell divisions in a population [33]. However, determinants of stemness in human HSCs are not as well described as in the murine system and therefore re-porter assays are not readily available. In an alternative approach, micromanipu-lation techniques were used to physically separate daughter cells after division and assess their multilineage potential in a col-ony-forming assay [35]. Employing single cell assay for monitoring the geometry of stem cell division and combining it with a functional readout of the LTC-IC poten-tial of daughter cell would therefore be the optimal approach for studying the symme-try of HSC division.

In conclusion, although the experi-mental setup we established still requires functional validation, it provides a new ap-proach to study the symmetry of HSC cell division. Moreover, while giving an insight into the symmetry of division of an indi-vidual cell, it is robust enough to provide information on the changes in cell division on a population level, as seen in the exper-iment using RAC inhibitor. Therefore, this system could be useful for studying the ef-fect of other small molecule inhibitors, e.g. of Aurora Kinase B, growth factors or on-cogenes on the symmetry of cell division and ultimately unraveling the mechanism regulating this process in human HSCs.

ACKNOWLEDGEMENTS

We would like to acknowledge K. Sjollema and M. Meijer (UMIC, UMCG) for help with microscopy, and I. Leegte and M. Geugien for help with experiments and cre-ating reagents. This work was supported by a grant from The Netherlands Organiza-tion for Scientific Research (NWO-VIDI 91796312) to JJS.

IMAGING HEMATOPOIETIC STEM CELL DIVISION

2

// 47reference list

REFERENCE LIST1. Fuchs E, Tumbar T, Guasch

G (2004) Socializing with the Neighbors: Stem Cells and Their Niche. Cell 116: 769-778.

2. Osawa M, Hanada Ki, Hamada H, Nakauchi H (1996) Long-Term Lymphohematopoietic Reconstitution by a Single CD34-Low/Negative Hema-topoietic Stem Cell. Science 273: 242-245.

3. Watt FM, Hogan AB (2000) Out of Eden: Stem Cells and Their Niches. Science 287: 1427-1430.

4. Morrison SJ, Kimble J (2006) Asymmetric and symmetric stem-cell divisions in devel-opment and cancer. Nature 441: 1068-1074.

5. Betschinger J, Knoblich JA (2004) Dare to Be Different: Asymmetric Cell Division in Drosophila, C. elegans and Ver- tebrates. Current Biology 14: R674-R685.

6. Horvitz HR, Herskowitz I (1992) Mechanisms of asym-metric cell division: Two Bs or not two Bs, that is the ques-tion. Cell 68: 237-255.

7. Morrison SJ, Wright DE, Weissman IL (1997) Cyclo-phosphamide/granulocyte colony-stimulating factor induces hematopoietic stem cells to proliferate prior to mobilization. Proc Natl Acad Sci USA 94: 1908-1913.

8. Bodine DM, Seidel NE, Orlic D (1996) Bone marrow col-lected 14 days after in vivo ad-ministration of granulocyte colony-stimulating factor and stem cell factor to mice has 10-fold more repopulating ability than untreated bone marrow. Blood 88: 89-97.

9. Zhang R, Zhang Z, Zhang C, Zhang L, Robin A, Wang Y, Lu M, Chopp M (2004) Stroke Transiently Increas-es Subventricular Zone Cell Division from Asymmetric

to Symmetric and Increases Neuronal Differentiation in the Adult Rat. The Journal of Neuroscience 24: 5810-5815.

10. Ho AD, Wagner W (2007) The beauty of asymmetry: asymmetric divisions and self-renewal in the haema-topoietic system. Current Opinion in Hematology 14.

11. Clevers H (2005) Stem cells, asymmetric division and can-cer. Nat Genet 37: 1027-1028.

12. Doe CQ, Bowerman B (2001) Asymmetric cell division: fly neuroblast meets worm zy-gote. Current Opinion in Cell Biology 13: 68-75.

13. Yamashita Y, Fuller M (2005) Asymmetric Stem Cell Divi- sion and function of the Niche in the Drosophila Male Germ Line. Int J Hematol 82: 377-380.

14. Yamashita YM, Jones DL, Fuller MT (2003) Orienta-tion of asymmetric stem cell division by the APC tumor suppressor and centrosome. Science 301: 1547-1550.

15. Betschinger J, Mechtler K, Knoblich JA (2006) Asym-metric Segregation of the Tumor Suppressor Brat Reg-ulates Self-Renewal in Dro-sophila Neural Stem Cells. Cell 124: 1241-1253.

16. Žigman M, Cayouette M, Charalambous C, Schleiffer A, Hoeller O, Dunican D, McCudden CR, Firnberg N, Barres BA, Siderovski DP, Knoblich JA (2005) Mamma-lian Inscuteable Regulates Spindle Orientation and Cell Fate in the Developing Reti-na. Neuron 48: 539-545.

17. Schofield R (1978) The rela-tionship between the spleen colony-forming cell and the haemopoietic stem cell. Blood Cells 4: 7-25.

18. Kimble JE, White JG (1981) On the control of germ cell development in Caenorhab-

ditis elegans. Developmental Biology 81: 208-219.

19. Xie T, Spradling AC (1998) decapentaplegic Is Essential for the Maintenance and Division of Germline Stem Cells in the Drosophila Ova-ry. Cell 94: 251-260.

20. Kiger AA, Jones DL, Schulz C, Rogers MB, Fuller MT (2001) Stem Cell Self-Renew-al Specified by JAK-STAT Activation in Response to a Support Cell Cue. Science 294: 2542-2545.

21. Palmer TD, Willhoite AR, Gage FH (2000) Vascular niche for adult hippocampal neurogenesis. J Comp Neu-rol 425: 479-494.

22. Zhang J, Niu C, Ye L, Huang H, He X, Tong WG, Ross J, Haug J, Johnson T, Feng JQ, Harris S, Wiedemann LM, Mishina Y, Li L (2003) Iden-tification of the haematopoi-etic stem cell niche and con-trol of the niche size. Nature 425: 836-841.

23. Lu B, Roegiers F, Jan LY, Jan YN (2001) Adherens junc-tions inhibit asymmetric di-vision in the Drosophila ep- ithelium. Nature 409: 522-525.

24. Song X, Zhu CH, Doan C, Xie T (2002) Germline Stem Cells Anchored by Adherens Junctions in the Drosophila Ovary Niches. Science 296: 1855-1857.

25. Colombo K, Grill SW, Kimple RJ, Willard FS, Siderovski DP, Gonczy P (2003) Trans-lation of Polarity Cues into Asymmetric Spindle Posi-tioning in Caenorhabditis el-egans Embryos. Science 300: 1957-1961.

26. Cau J, Hall A (2005) Cdc42 controls the polarity of the actin and microtubule cy-toskeletons through two distinct signal transduction pathways. Journal of Cell Science 118: 2579-2587.

48 // reference list

27. Etienne-Manneville S (2004) Cdc42 - the centre of polar-ity. Journal of Cell Science 117: 1291-1300.

28. Carolina Florian M, Geiger H (2010) Concise Review: Polarity in Stem Cells, Dis-ease, and Aging. Stem Cells 28: 1623-1629.

29. Sinha S, Yang W (2008) Cel-lular signaling for activation of Rho GTPase Cdc42. Cel-lular Signalling 20: 1927-1934.

30. Florian M, Dorr K, Niebel A, Daria D, Schrezenmeier H, Rojewski M, Filippi MD, Hasenberg A, Gunzer M, Scharffetter-Kochanek K, Zheng Y, Geiger H (2012) Cdc42 Activity Regulates Hematopoietic Stem Cell Ag-ing and Rejuvenation. Cell Stem Cell 10: 520-530.

31. Halet G, Carroll J (2007) Rac Activity Is Polarized and Regulates Meiotic Spindle Stability and Anchoring in Mammalian Oocytes. Devel-opmental Cell 12: 309-317.

32. Lu W, Casanueva MO, Mahowald AP, Kato M, Lauterbach D, Ferguson EL (2012) Niche-Associated Ac-tivation of Rac Promotes the Asymmetric Division of Drosophila Female Germline Stem Cells. PLoS Biol 10: e1001357.

33. Wu M, Kwon HY, Rattis F, Blum J, Zhao C, Ashkenazi R, Jackson TL, Gaiano N, Oliver T, Reya T (2007) Im-aging Hematopoietic Pre-cursor Division in Real Time. Cell Stem Cell 1: 541-554.

34. Beckmann J, Scheitza S, Wernet P, Fischer JC, Giebel B (2007) Asymmetric cell division within the human hematopoietic stem and pro-genitor cell compartment: identification of asymmet-rically segregating proteins. Blood 109: 5494-5501.

35. Iwama A, Oguro H, Negishi M, Kato Y, Morita Y, Tsukui H, Ema H, Kamijo T,

Katoh-Fukui Y, Koseki H, van Lohuizen M, Nakauchi H (2004) Enhanced self-re-newal of hematopoietic stem cells mediated by the poly-comb gene product Bmi-1. Immunity 21: 843-851.

36. Rizo A, Dontje B, Vellenga E, de HG, Schuringa JJ (2008) Long-term maintenance of human hematopoietic stem/progenitor cells by expres-sion of BMI1. Blood 111: 2621-2630.

37. Rieger MA, Hoppe PS, Smejkal BM, Eitelhuber AC, Schroeder T (2009) Hema-topoietic cytokines can in-struct lineage choice. Sci-ence 325: 217-218.

38. Eilken HM, Nishikawa SI, Schroeder T (2009) Contin-uous single-cell imaging of blood generation from hae-mogenic endothelium. Na-ture 457: 896-900.

39. Costa MR, Ortega F, Brill MS, Beckervordersandforth R, Petrone C, Schroeder T, Goetz M, Berninger B (2011) Continuous live imaging of adult neural stem cell divi-sion and lineage progression in vitro. Development 138: 1057-1068.

40. Dykstra B, Ramunas J, Kent D, McCaffrey L, Szumsky E, Kelly L, Farn K, Blaylock A, Eaves C, Jervis E (2006) High-resolution video moni-toring of hematopoietic stem cells cultured in single-cell arrays identifies new fea-tures of self-renewal. Pro-ceedings of the National Academy of Sciences 103: 8185-8190.

41. Rizo A, Horton SJ, Olthof S, Dontje B, Ausema A, van Os R, van den Boom V, Vellenga E, de Haan G, Schuringa JJ (2010) BMI1 collaborates with BCR-ABL in leukemic transformation of human CD34+ cells. Blood 116: 4621-4630.

42. Cancelas JA, Lee AW, Prabhakar R, Stringer KF,

Zheng Y, Williams DA (2005) Rac GTPases differ-entially integrate signals regulating hematopoietic stem cell localization. Nat Med 11: 886-891.

43. Gu Y, Filippi MD, Cancelas JA, Siefring JE, Williams EP, Jasti AC, Harris CE, Lee AW, Prabhakar R, Atkinson SJ, Kwiatkowski DJ, Williams DA (2003) Hematopoietic cell regulation by Rac1 and Rac2 guanosine triphospha-tases. Science 302: 445-449.

44. Cancelas JA (2011) On how Rac controls hematopoietic stem cell activity. Transfu-sion 51 Suppl 4: 153S-159S.

45. Thery M, Racine V, Pepin A, Piel M, Chen Y, Sibarita JB, Bornens M (2005) The extracellular matrix guides the orientation of the cell di-vision axis. Nat Cell Biol 7: 947-953.

46. Filippi MD, Harris CE, Meller J, Gu Y, Zheng Y, Williams DA (2004) Localization of Rac2 via the C terminus and aspartic acid 150 specifies superoxide generation, actin polarity and chemotaxis in neutrophils. Nat Immunol 5: 744-751.

47. Tao W, Filippi MD, Bailey JR, Atkinson SJ, Connors B, Evan A, Williams DA (2002) The TRQQKRP motif locat-ed near the C-terminus of Rac2 is essential for Rac2 biologic functions and intra-cellular localization. Blood 100: 1679-1688.

48. Goldenson B, Crispino JD (2015) The aurora kinases in cell cycle and leukemia. On-cogene 34: 537-545. Review.

49. Schroeder T (2011) Long-term single-cell imaging of mammalian stem cells. Nat Meth 8: S30-S35.