Developmental Cell

Short Article

A Dynamic Interface between Vacuolesand Mitochondria in YeastYael Elbaz-Alon,1 Eden Rosenfeld-Gur,2 Vera Shinder,3 Anthony H. Futerman,2 Tamar Geiger,4 and Maya Schuldiner1,*1Department of Molecular Genetics, Weizmann Institute of Science, Rehovot 7610001, Israel2Department of Biological Chemistry, Weizmann Institute of Science, Rehovot 7610001, Israel3Electron Microscopy Unit, Weizmann Institute of Science, Rehovot 7610001, Israel4Department of Human Molecular Genetics and Biochemistry, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv 6997801, Israel

*Correspondence: maya.schuldiner@weizmann.ac.il

http://dx.doi.org/10.1016/j.devcel.2014.06.007

SUMMARY

Cellular life depends on continuous transport oflipids and small molecules between mitochondriaand the endomembrane system. Recently, endo-plasmic reticulum-mitochondrial encounter struc-ture (ERMES) was identified as an important yetnonessential contact for such transport. Using ahigh-content screen in yeast, we found a contactsite, marked by Vam6/Vps39, between vacuoles(the yeast lysosomal compartment) and mitochon-dria, named vCLAMP (vacuole and mitochondriapatch). vCLAMP is enriched with ion and amino-acid transporters and has a role in lipid relay betweenthe endomembrane system and mitochondria. Criti-cally, we show that mitochondria are dependent onhaving one of two contact sites, ERMES or vCLAMP.The absence of one causes expansion of theother, and elimination of both is lethal. Identificationof vCLAMP adds to our ability to understand thecomplexity of interorganellar crosstalk.

INTRODUCTION

Mitochondria generate the majority of cellular energy and house

enzymes required for the synthesis, breakdown, and intercon-

version of various species of amino acids, lipids, iron/sulfur clus-

ters, and other small molecules. Due to their diverse functions

and essential roles in cellular metabolism, mitochondria serve

as hubs for signaling in events such as growth, differentiation,

or cell death. Loss of optimal mitochondrial activity is therefore,

not surprisingly, implicated in a growing number of human dis-

eases as well as in aging (Bratic and Larsson, 2013; Costa and

Scorrano, 2012; Griffiths, 2012; Vives-Bauza and Przedborski,

2011; Yu et al., 2012).

The central tasks of mitochondria in cells necessitate constant

communication and transport of small molecules with other

organelles. However, mitochondria are not connected to the

endomembrane system via the vesicular pathway. Instead,

mitochondria have been shown to communicate with the endo-

membrane system by virtue of a membrane contact site (MCS)

where membranes of mitochondria come into close proximity to

membranes of the endoplasmic reticulum (ER). This MCS, also

De

termed mitochondria-associated-membranes (Achleitner et al.,

1999; Vance, 1990), enables ions and lipids to be rapidly

transported in a nonvesicular manner (Elbaz and Schuldiner,

2011; Levine and Loewen, 2006; Tatsuta et al., 2014). Under-

standing the molecular machineries creating and regulating

this MCS has been the arena of intense investigations in the

past decade.

In yeast, the molecular identity of the ER-mitochondria teth-

ering complex was recently uncovered and is mediated by a

four-protein complex termed the ER-mitochondria encounter

structure (ERMES) (Kornmann et al., 2009). One of its hypothe-

sized functions was to enable phospholipid transport (Kopec

et al., 2010; Kornmann and Walter, 2010) required for building

mitochondrial membranes as well as for the three-step bio-

synthetic pathway of aminoglycerophospholipids. Therefore, it

was of great surprise when the loss of ERMES subunits had

very little effect on cellular levels of aminoglycerophospholipids

(Kornmann et al., 2009; Nguyen et al., 2012; Voss et al., 2012).

Hence, it became clear that alternate routes of phospholipid

transport must exist in the cell, and uncovering them should

shed light on novel modes of communication between mito-

chondria and the endomembrane system. We report here our

findings of an MCS between mitochondria and vacuoles (the

yeast lysosomal compartment) and its functional significance

in lipid transport between the endomembrane system and

mitochondria.

RESULTS

A High-Content Screen Uncovers that Vam6/Vps39Influences the Number of MCSs between the ER andMitochondriaWe set out to uncover alternate paths for communication be-

tween mitochondria and the endomembrane system using a

systematic genetic screening approach. We reasoned that the

extent of ERMES mediated contact sites must be regulated so

that loss of a parallel pathway would result in an increase in

ERMES mediated contacts. Therefore, we screened for genetic

backgrounds yielding a pronounced increase in the amount of

ERMES foci per cell (Figure 1A). Using the synthetic genetic

array (SGA) methodology (Cohen and Schuldiner, 2011; Tong

and Boone, 2006; Tong et al., 2001), we created a library in

which each strain harbors a green fluorescent protein (GFP)-

tagged ERMES subunit (Mdm34-GFP) on the background of

a single gene mutation (Breslow et al., 2008; Giaever et al.,

velopmental Cell 30, 95–102, July 14, 2014 ª2014 Elsevier Inc. 95

A

B

Figure 1. A Genome-wide Screen to Un-

cover Mutants Affecting the Number of

ERMES-Mediated Mitochondria-ER Con-

tact Sites

(A) Schematic representation of the screen flow. 1.

A strain harboring a GFP-tagged ERMES sub-

unit (Mdm34) was crossed against yeast KO and

DAmP collections, and a library was created so

that each strain harbors a tagged ERMES on the

background of a single mutation. 2. Live cells were

imaged using an automated image acquisition

system. 3. Images were manually inspected for an

increase in the number of ERMES foci.

(B) Fluorescence images of the increase in

Mdm34-GFP foci in deletion of either vps39

or vam7 relative to a WT background. Scale

bars, 5 mm.

See also Figure S1.

Developmental Cell

A Dynamic Mitochondria/Vacuole Interface

2002), covering the entire yeast genome. Following the auto-

mated imaging of all strains generated (Breker et al., 2013),

we manually examined the resulting images to uncover strains

with the phenotype of choice (Figure 1A). The screen resulted in

identification of over 100 different genetic backgrounds in

which the Mdm34-GFP signal was altered (Table S1 available

online). Strains that altered Mdm34-GFP morphology included

deletion strains of ERMES components themselves that

caused loss of the junction altogether, reduction in proteasomal

subunits that caused an increase in protein levels, and loss

of a large repertoire of mitochondrial proteins that caused

a decrease in the intensity of foci. We focused on the four

mutant backgrounds that displayed elevated number of foci

per cell: Ddnm1, Dfis1 (Figure S1A), Dvam6/vps39 (hereinafter

referred to as vps39), and Dvam7 (Figures 1B; Figures S1B

and S1C).

Dnm1 and Fis1 are essential components of the mitochon-

drial fission machinery (Bleazard et al., 1999; Mozdy et al.,

96 Developmental Cell 30, 95–102, July 14, 2014 ª2014 Elsevier Inc.

2000). Interestingly, it has recently been

shown that mitochondrial/ER contact

sites mark the sites for fission to occur

(Friedman et al., 2011; Murley et al.,

2013). We therefore reasoned that

the increase in ERMES foci on these

backgrounds is most probably a result

of an indirect effect on mitochondrial

morphology or due to crosstalk with the

ERMES complex.

Vps39 and Vam7 have been previ-

ously characterized as components of

the vacuolar fusion process (Price et al.,

2000; Stroupe et al., 2006). These hits

were surprising because no apparent

direct link was reported between vacu-

olar morphology and ER-mitochondria

connections or function. It has, how-

ever, previously been reported that the

absence of Vps39 leads to impaired

respiration capacity (Merz and Wester-

mann, 2009). Therefore, we decided to

examine the connection between the absence of Vps39 to the

increase in ERMES foci.

Vps39 Localizes to Contact Sites betweenMitochondriaand Vacuoles Termed vCLAMPsTo see if Vps39 has any spatial connection to mitochondria, we

tagged Vps39 with an N-terminal GFP (hereinafter referred to as

GFP-Vps39), since tagging on the N terminus was previously

shown to retain the function of the protein (Angers and Merz,

2009). Following GFP-Vps39 localization by fluorescence micro-

scopy, we found that it was highly enriched in defined patches on

the vacuolar membrane, which colocalized with a mitochondrial

marker (a matrix-targeted blue fluorescent protein; MTS-BFP)

(Figure 2A). We therefore attempted to uncover whether such

an overlap between vacuoles andmitochondria implies an unde-

scribed MCS.

Electron microscopy demonstrated that indeed there exists

a tight interface between the vacuolar membrane and the

A

B

Figure 2. Vps39 Marks a Contact Site between Mitochondria and

Vacuoles

(A) GFP-tagged Vps39 is enriched in defined patches on the vacuolar mem-

brane that colocalize with mitochondria. Scale bars, 5 mm.

(B) Immunoelectron microscopy analysis of a GFP-Vps39 strain, using anti-

bodies against GFP, demonstrates that Vps39 accumulates specifically at the

mitochondria-vacuole MCS that we termed vCLAMP. Distribution analysis of

gold particles (n = 400): 64% vCLAMP; 15% vacuole; 10% mitochondria; 7%

cytosol; 4% nucleus. V, vacuole; M, mitochondria.

Developmental Cell

A Dynamic Mitochondria/Vacuole Interface

mitochondrial outer membrane. Similarly to what was observed

by fluorescence microscopy, this MCS was labeled by GFP-

Vps39. The membranes of the two organelles maintained a

separate identity and were not fused, as would be expected

from a bona fide MCS (Figure 2B). We therefore referred to this

MCS as the vCLAMP (vacuole and mitochondria patch). This

MCS was discovered and characterized in parallel from studies

on Vps39 (see Honscher et al., 2014 [this issue ofDevelopmental

Cell]).

Vps39 has been studied in depth for many years as part of the

homotypic fusion and vacuole protein sorting (HOPS) tethering

complex (Stroupe et al., 2006). In order to start and characterize

its function at the vCLAMP, we decided to identify proteins inter-

acting with GFP-Vps39 specifically at the contact site. To do

that, we isolated cellular fractions enriched for mitochondria

and performed pull-downs with an anti-GFP antibody. We

reasoned that this should enrich for proteins that lie at the inter-

face of the two organelles. Enriched proteins were identified by

mass spectrometry analysis. (The complete list of significantly

enriched proteins is given in Table S2.) The strongest inter-

actions observed were among Vps39 and the additional five

HOPS complex proteins (Figure S2A), as would be expected

from the established role of Vps39. Interestingly, in addition to

these known interactions, we found nine small molecule trans-

porters that were enriched in our samples in a statistically signif-

icant manner (false discovery rate [FDR] = 0.05; Table S3). We

tagged all nine transporters with a GFP tag in order to verify their

De

exact cellular localization. Five of the eight vacuolar proteins

were not homogenously distributed over the vacuolar membrane

but rather localized to specific patches (Figure S2B). Further co-

localization with a mitochondrial marker showed that GFP-Mnr2

and GFP-Pho91 partially colocalized with mitochondria. (Figures

S2C and S2D). This could suggest that the vCLAMP, marked by

GFP-Vps39, could serve as a hub for interorganelle transport of

small molecules.

The Two Contact Sites of Mitochondria with theEndomembrane System Are in Dynamic Equilibrium andServe as Mutual Backup PathwaysOur original hypothesis for uncovering the vCLAMP was that

optimal communication of mitochondria and the endomembrane

system requires coregulation with ERMES. Indeed, our screen

showed that loss of vCLAMP caused an increase in the number

of ERMES mediated contact sites (Figure 1B and Figures S1A

and S1B). Therefore, we wanted to explore the effect of deleting

ERMES on the vCLAMP. Loss of ERMES (caused by deletion of

mdm34 [Figures 3A and S3A] or any of the other three complex

members [Figure 3C]) caused a dramatic expansion of the

vCLAMP interface. In fact, in this genetic background, vCLAMP

completely surrounded mitochondria, despite the fact that the

cells retained normal morphology (Figure S3C). Electron micro-

scopy demonstrated that the contact site expanded to the entire

circumference of eachmitochondrion (Figures 3B and S3B). This

finding demonstrates that ERMES and vCLAMP are tightly core-

gulated, which would ensure that mitochondria does not detach

from the endomembrane system and, as a consequence, no

reduction in endomembrane/mitochondrial transport occurs in

the absence of one MCS.

The tight interplay between ERMES and vCLAMP would

indeed ensure that, in the absence of one structure, the increase

in the other would serve as a ‘‘backup’’ path for small-molecule

transport. This predicts that disturbing both MCSs should, in

fact, cause a dramatic cellular phenotype. To obtain double

mutants, we crossed a Dvps39 strain with a strain harboring

a deletion in each of the ERMES subunits. Diploid yeast were

sporulated, and the haploid progeny from single meiosis events

were dissected. We found that, in all cases, the combination of

both mutations was lethal (Figure S4A), indicating that both

MCSs cannot be simultaneously disturbed. Thus, as would

be expected, mitochondria must be in contact with at least

one organelle in the endomembrane system to maintain proper

cellular functions.

Loss of Both Mitochondrial Contact Sites CausesDefects in Phospholipid TransportPhospholipid transport between the endomembrane system and

mitochondria is essential for the three-step enzymatic pathway

generating aminoglycerophospholipids (seemodel in Figure 4H).

In this pathway, phosphatidylserine (PtdSer) that is synthesized

in the ER is transported to mitochondria to be converted to

phosphatidylethanolamine (PtdEtn) by the inner mitochondrial

membrane enzyme, PtdSer decarboxylase 1 (Psd1). Some of

the PtdEtn is then transported back to the ER, where it can serve

as a source for generation of phosphatidylcholine (PtdCho)

(Birner and Daum, 2003; Daum et al., 1998). Although �5%

of PtdEtn is created in the late endomembrane system by the

velopmental Cell 30, 95–102, July 14, 2014 ª2014 Elsevier Inc. 97

A

B

C

Figure 3. vCLAMP and ERMES Are in Dynamic Equilibrium

(A) In a strain deleted for mdm34, mitochondria (visualized by MTS-RFP)

become large and spherical, and the GFP-Vps39 marked vCLAMP expands to

surround mitochondria. Scale bars, 5 mm.

Developmental Cell

A Dynamic Mitochondria/Vacuole Interface

98 Developmental Cell 30, 95–102, July 14, 2014 ª2014 Elsevier Inc.

paralogous enzyme Psd2 (Trotter and Voelker, 1995), the mito-

chondrial Psd1 is the major enzyme, and so a large lipid flux

between the ER and mitochondria should exist. If indeed

vCLAMP and ERMES are coregulated to sustain constant con-

tact area between the endomembrane system and mitochon-

dria, then this would explain the lack of phenotype on PtdEtn

levels caused by losing ERMES alone (Kornmann et al., 2009;

Nguyen et al., 2012; Voss et al., 2012). It would also predict

that concomitant loss of both MCSs would result in a more dra-

matic decrease in the phospholipids that require such transport

events. To measure this effect, we created a conditional strain

that is deleted for vps39 and expresses MDM34 under the con-

trol of a repressible GalS promoter. We eliminated additional

pathways for de novo PtdEtn synthesis by growing cells in

synthetic media depleted for ethanolamine, so as to limit the

Kennedy pathway (Daum et al., 1998), and by deleting the late

endomembrane localized PtdSer decarboxylase, Psd2 (Trotter

and Voelker, 1995). Under these conditions, the only existing

pathway for de novo PtdEtn synthesis involves themitochondrial

Psd1 and the obligation of transporting ER synthesized PtdSer

into mitochondria.

The phospholipid composition of the triple mutant compared

to control cells was examined after growth for 24 hr in glucose

(to repress GalSp-MDM34 expression), at which point the cells

are still alive (Figure S4B) but mitochondrial shape is already

affected (Figure S4C), indicating a depletion in Mdm34 levels.

Thin-layer chromatography (TLC) of phospholipids extracted

from whole cells (Figure 4A, independent quadruplicate quanti-

fied in Figure 4B) or enriched mitochondrial samples (Figure 4C,

independent triplicate quantified in Figure 4D) demonstrated that

themutant has up to 40%decrease in the amounts of PtdEtn and

cardiolipin (CL) (CL observed only in the enriched mitochondrial

samples) concomitantly with near doubling in phosphatidylinosi-

tol (PtdIns) levels. Specifically, the decrease in PtdEtn could only

be seen in this condition, whereas in the single mutants (Dvps39

orGalSp-MDM34 in glucose; data not shown) or after short-term

repression of GalSp-MDM34 in the triple mutant (Figures S4D

and S4E), this reduction was not seen. This demonstrates that

only the concomitant loss of bothMCSs causes a halt in shuttling

of PtdSer. More generally, PtdEtn, CL, and PtdIns are all synthe-

sized from a common precursor, phosphatidic acid (PA); how-

ever, PtdEtn and CL are synthesized in mitochondria, whereas

PtdIns is synthesized in the ER. Thus, accumulation of PtdIns

and reduction of CL most probably also reflect a defect in PA

transport from ER to mitochondria (Figure 4H).

In order to monitor the direct conversion of PtdSer to PtdEtn

and, consequently, to PtdCho, wemetabolically labeled the cells

using 3H-serine. The radioactive serine was incorporated in the

ER to PtdSer, and, following phospholipid extraction and sepa-

ration by TLC, we could follow its fate in control versus mutant

cells. In the mutant cells, where vCLAMP is absent and ERMES

is gradually depleted, there is a massive accumulation of PtdSer

(B) Electron microscopy analysis shows that, upon deletion of ERMES, the

vCLAMP is enlarged to surround mitochondria (visualized by immunogold

label against Por1). Cristae are marked by white arrows.

(C) Deletion of each of the ERMES subunits yields a similar effect on vCLAMP

dynamics.

See also Figure S3.

A B C D

E F G

H I

Figure 4. The vCLAMP Serves as a Route for Phospholipid Transport

(A) TLC of phospholipids extracted from whole cells of control (WT) and a mutant disrupted for both ERMES and vCLAMP. PA, phosphatidic acid; PI, phos-

phatidylinositol; PS, phosphatidylserine; PC, phosphatidylcholine; PE, phosphatidylethanolamine.

(B) Quantitation of four biologically independent repeats of whole cell extracts. Bars represent means ± SD. The disruption of both ERMES and vCLAMP causes a

reduction in PE levels, although Psd1 is present in mitochondria.

(C) TLC of phospholipids levels frommitochondrial (Mito) enriched fractions of control (WT) and a mutant disrupted for both ERMES and vCLAMP. Abbreviations

for lipids are described in (A).

(D) Quantitation of three biologically independent repeats of mitochondrial enriched fractions. Bars represent means ± SD.

(E) TLC of phospholipids extracted from whole cells following a 2 hr pulse of 3H-serine to control (WT) and a mutant disrupted for both ERMES and vCLAMP.

(F)Quantitationof fourbiologically independent repeats.Bars representmeans+SD.UpondisruptionofbothvCLAMPandERMEScells, phospholipid transport both

into and out of mitochondria is impaired. As a result, high levels of de novo synthesized PtdSer accumulate in the ER, while very low levels of PtdCho are generated.

(G) Autoradiograph of TLC of phospholipids extracted from whole cells of control (WT) and a mutant strain (Dvps39, GalSp-MDM34, Dpsd2) after 24 hr growth in

glucose-containing media. Cells were labeled with 3H-serine for indicated time points, followed by 1 mM cold L-serine addition and phospholipid extraction.

(H) Phospholipid biosynthesis pathways involve enzymes in both the ER and mitochondria, thus necessitating transport of substrates between both compart-

ments. Etn, ethanolamine; Cho, choline; CDP-DAG, cytidine diphosphate diacylglycerol; DAG, diacylglycerol.

(I) A schematic model representing the tight coregulation of ERMES and vCLAMP. Loss of one causes an enlargement of the other, and loss of both is lethal.

See also Figure S4.

Developmental Cell

A Dynamic Mitochondria/Vacuole Interface

Developmental Cell 30, 95–102, July 14, 2014 ª2014 Elsevier Inc. 99

Developmental Cell

A Dynamic Mitochondria/Vacuole Interface

(PtdSer content mutant/wild-type = 1.75), suggesting that as

both contacts between mitochondria and the endomembrane

system are gradually collapsing, more and more PtdSer, which

cannot be transported to mitochondria, is accumulating in the

ER (Figure 4E and independent quadruplicate quantified in Fig-

ure 4F; see also Figure 4G). Quantification of the various lipid

species demonstrates a large difference in the amounts of de

novo synthesized PtdCho. While PtdCho accumulates in control

cells, it is hardly synthesized in the mutant. This difference prob-

ably reflects the inability of PtdEtn that was synthesized in

the mitochondria to be transported back to the ER for further

processing. Taken together, the results of steady-state and

de novo phospholipid amounts demonstrate the role of both

ERMES and vCLAMP in phospholipid transport. More impor-

tantly, they highlight how well coordinated these back-up path-

ways are, as only elimination of both yields a dramatic change

in cellular phospholipid levels.

DISCUSSION

Functional cooperation between mitochondria and other organ-

elles is essential. Since mitochondria is not connected to the

endomembrane system through vesicular trafficking, the ER-

mitochondria MCS serves as an important interorganellar inter-

action, allowing for fast and direct transport of phospholipids

and Ca+2. We show here that, in yeast, mitochondria maintain

a secondMCS, vCLAMP, linking it to the endomembrane system

through the vacuole. Since vacuoles are linked to the ER—both

directly, through the nuclear vacuolar junction (NVJ), and indi-

rectly, through vesicular traffic—this MCS can serve as a bypass

for flow of information and nutrients between the ER and mito-

chondria. Indeed, we show that ERMES and vCLAMP are core-

gulated and that vCLAMP serves as a backup for ERMES, as

only in the case where both are missing do yeast suffer from dra-

matic alternations in phospholipid levels and death.

Loss of both mitochondrial MCSs is lethal, and although a dra-

matic reduction in PtdEtn is observed when there is concomitant

loss of ERMES and vCLAMP, it is possible that there exist addi-

tional essential functions for the two MCSs. In yeast, where the

major ion stores (such as calcium, iron, and copper) reside

in vacuoles, an additional essential function may be ion trans-

port. Condition-specific direct traffic between endosomes and

mitochondria in reticulocytes have been shown to facilitate iron

transfer (Sheftel et al., 2007). Recently, it has been reported

that mitochondria and melanosomes, pigment-synthesizing

lysosome-like organelles, establish physical contacts. The prox-

imities observed between both organelles correlated with mela-

nosome biogenesis andmaturation (Daniele et al., 2014). Hence,

the need for sustaining a contact with later endomembrane

organelles may be conserved from yeast to higher eukaryotes.

More broadly, a better understanding of vCLAMP resident pro-

teins must be obtained before the full extent of crosstalk

between mitochondria and the endomembrane system can be

determined.

Previously described MCSs have been shown to be dynamic

and respond to the needs of the cell. For example, the NVJ en-

larges upon entry to stationary growth phase (Pan et al., 2000),

the formation of new plasma membrane-ER MCSs is induced

upon Ca2+ store depletion (Wu et al., 2006), and a late endo-

100 Developmental Cell 30, 95–102, July 14, 2014 ª2014 Elsevier Inc

some/lysosome-ER MCS forms solely under conditions of

increased cholesterol levels (Du et al., 2011). vCLAMP is dy-

namic too, in response to both loss of ERMES and changes in

carbon source (see Honscher et al., 2014). It is this dynamic

and condition-specific nature that can most probably account

for the fact that the vCLAMP has remained elusive for so long.

The dynamic nature of vCLAMP and its apparent regulation

through metabolic cues (see Honscher et al., 2014) marks

yet another link between intracellular organelle positioning,

morphology, and degree of contact as determinants in the

cell’s adaptation to changing metabolic requirements (Liesa

and Shirihai, 2013) and cell fate (Csordas et al., 2006).

To summarize, we demonstrate an MCS between vacuoles

and mitochondria. This contact site, which we term vCLAMP,

is in dynamic equilibrium with the ERMES mediated junction be-

tween mitochondria and the ER and works in parallel to enable

the shuttling of small molecules (such as lipids) between the en-

domembrane system and mitochondria (model in Figure 4I).

More generally, identification of the vCLAMP demonstrates

the true complexity of interorganellar crosstalk: not only must

an interface between two organelles be formed, but bypass sys-

tems must also be in place, and their size should be coregulated

to serve the changing needs of the cell. Future efforts should

be directed to elucidate the nature of the vCLAMP tether. Our

proteomic analysis did not yield a definite answer as to how

Vps39 is tethered to the mitochondria; whether through an

adaptor protein or a yet-to-be-discoveredmitochondrial partner.

Future studies aimed at understanding the coregulation of the

vCLAMP and ERMES should provide novel insights into what

the cells sense to ensure optimal MCS surface area. Having

visual markers and the genetic capacity to alter junction size

should now give a platform for in depth investigations of this

new cellular organization module.

EXPERIMENTAL PROCEDURES

Strains and Plasmids

Strains created in this study are listed in a table in the Supplemental Experi-

mental Procedures. All yeast strains in this study are based on the BY4741

laboratory strains (Brachmann et al., 1998). Genetic manipulations were per-

formed using the Li-acetate, polyethylene glycol, single-stranded DNAmethod

for transforming yeast strains (Gietz and Woods, 2006) using integration plas-

mids described elsewhere (Janke et al., 2004; Longtine et al., 1998).

For mitochondria staining, we used either an MTS-red fluorescent protein

(RFP) plasmid (kindly provided by Jodi Nunnari) or an MTS-BFP plasmid

(kindly provided by Christian Ungermann and Benedikt Westermann).

SGA and High-Content Screening

The SGA technique was used to efficiently introduce an ERMES marker (GFP-

taggedMdm34) into systematic yeast deletion and hypomorphic allele libraries

(Breslow et al., 2008; Giaever et al., 2002). SGA was performed as described

elsewhere (Cohen and Schuldiner, 2011; Tong and Boone, 2006; Tong et al.,

2001). Microscopic screening was performed using an automated microscopy

setup as described elsewhere (Cohen and Schuldiner, 2011). Further informa-

tion is present in the Supplemental Experimental Procedures.

Manual Fluorescence Microscopy

Imaging was performed using an Olympus IX71 microscope controlled by the

DeltaVision SoftWoRx 3.5.1 software with a 603 or 1003 oil lens. Images were

captured by a Phoetometrics Coolsnap HQ camera with excitation at 490/

20 nm and emission at 528/38 nm (for GFP); excitation at 555/28 nm and emis-

sion at 617/73 nm (for mCherry/RFP); or excitation at 402 nm and emission at

.

Developmental Cell

A Dynamic Mitochondria/Vacuole Interface

457 nm (for BFP). Images were transferred to Adobe Photoshop CS3 for slight

contrast and brightness adjustments.

Electron Microscopy

For immunoelectron microscopy, cells were fixed in 4% paraformaldehyde

with 0.1% glutaraldehyde in 0.1 M cacodylate buffer (pH, 7.4) for 1 hr at

room temperature and kept at 4�C during 1–2 days. The samples were

soaked overnight in 2.3 M sucrose and rapidly frozen in liquid nitrogen.

Frozen ultrathin (70–90 nm) sections were cut with a diamond knife

at �120�C on a Leica EM UC6 ultramicrotome. The sections were collected

on 200-mesh formvar-coated nickel grids. Sections were blocked by a block-

ing solution containing 1% BSA, 0.1% glycine, 0.1% gelatin, and 1% Tween-

20. Immunolabeling was performed using rabbit polyclonal anti-GFP antibody

(ab6556, 1:100; Abcam) during 1.5–2 hr at room temperature followed by goat

anti-rabbit immunoglobulin G coupled to 10 nm (or 15 nm) gold particles

(1:20 dilution) for 30 min at room temperature. Contrasting and embedding

were performed as described elsewhere (Tokuyasu, 1986). The embedded

sections were scanned and digitally viewed on a Tecnai Spirit transmission

electron microscope (FEI) at 120 kV using a CCD Eagle camera with TIA soft-

ware (FEI).

Phospholipid Analysis

Phospholipids were extracted (Bligh and Dyer, 1959) from mitochondria-en-

riched fractions (Daum et al., 1982) or from whole-cell homogenates (Folch

et al., 1957) and analyzed by TLC using chloroform:acetone:methanol:acetic

acid:water (50:20:10:15:5, v:v:v:v:v) as the developing solvent. Lipids were

visualized using copper sulfate (6.25 mM in 9.4 ml phosphoric acid) and

heated at 110�C for 30–60 min, identified using lipid standards (Sigma,

Avanti Polar Lipids), and quantified using ImageGauge Ver 4.0 software

(FUJIFILM).

For measuring de novo phospholipid synthesis, we metabolically labeled

cells with 3H-serine as follows: cells were grown to early logarithmic phase

and pelleted, and 15 optical density units were resuspended in 25 ml syn-

thetic media (S-dextrose) supplemented with 0.5 mM ethanolamine and

1 mg/ml myriocin. 3H-serine (1 mCi/ml) was then added to a final concentra-

tion of 2 mCi/ml. Cells continued growth at 30�C until pulse was stopped by

the addition of 1 mM cold L-serine at various time points (as indicated in

the figures). Phospholipids were extracted (Folch et al., 1957) and separated

by TLC using chloroform:methanol:acetic acid:water (25:15:4:2, v/v/v) as the

developing solvent. [3H]-labeled lipids were visualized using a phosphori-

maging screen (Fuji) and quantified using ImageGauge Ver 4.0 software

(FUJIFILM).

Interaction Proteomics

Pull-down of GFP-Vps39 was performed from mitochondria-enriched

preparations followed by liquid chromatography-tandem mass spectrometry

analysis. Detailed information is present in the Supplemental Experimental

Procedures.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures,

four figures, and three tables and can be found with this article online at

http://dx.doi.org/10.1016/j.devcel.2014.06.007.

ACKNOWLEDGMENTS

We thank Jodi Nunnari for sharing the MTS-RFP plasmid and Christian Unger-

mann and Benedikt Westermann for the mBFP plasmid. We thank Oren Elbaz

for immense help with the graphic design, Michal Breker and Silvia Chuartz-

man for technical assistance with the screening system, and Shai Fuchs for

technical assistance with the flow cytometer. We thank Oren Schuldiner, Tslil

Ast, and Shai Fuchs for critical reading of the manuscript. The electron micro-

scopy studies were conducted at the Irving and Cherna Moskowitz Center for

Nano and Bio-Nano Imaging at the Weizmann Institute of Science. This study

was funded by a European Research Council Starting Grant (260395). Y.E.-A.

was the recipient of the Weizmann Institute Dean’s postdoctoral fellowship.

M.S. is an awardee of the EMBO Young Investigator Program, in conjunction

Dev

with the Israel Ministry of Science, and a recipient of Marie Curie International

Reintegration Grant 239224 of the European Union. A.H.F. is the Joseph

Meyerhoff Professor of Biochemistry at the Weizmann Institute of Science.

T.G. is supported by the Israel Science Foundation.

Received: October 15, 2013

Revised: April 14, 2014

Accepted: June 9, 2014

Published: July 14, 2014

REFERENCES

Achleitner, G., Gaigg, B., Krasser, A., Kainersdorfer, E., Kohlwein, S.D.,

Perktold, A., Zellnig, G., and Daum, G. (1999). Association between the endo-

plasmic reticulum andmitochondria of yeast facilitates interorganelle transport

of phospholipids through membrane contact. Eur. J. Biochem. 264, 545–553.

Angers, C.G., and Merz, A.J. (2009). HOPS interacts with Apl5 at the vacuole

membrane and is required for consumption of AP-3 transport vesicles. Mol.

Biol. Cell 20, 4563–4574.

Birner, R., and Daum,G. (2003). Biogenesis and cellular dynamics of aminogly-

cerophospholipids. Int. Rev. Cytol. 225, 273–323.

Bleazard, W., McCaffery, J.M., King, E.J., Bale, S., Mozdy, A., Tieu, Q.,

Nunnari, J., and Shaw, J.M. (1999). The dynamin-related GTPase Dnm1 regu-

lates mitochondrial fission in yeast. Nat. Cell Biol. 1, 298–304.

Bligh, E.G., and Dyer, W.J. (1959). A rapid method of total lipid extraction and

purification. Can. J. Biochem. Physiol. 37, 911–917.

Brachmann, C.B., Davies, A., Cost, G.J., Caputo, E., Li, J., Hieter, P., and

Boeke, J.D. (1998). Designer deletion strains derived from Saccharomyces

cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated

gene disruption and other applications. Yeast 14, 115–132.

Bratic, A., and Larsson, N.G. (2013). The role of mitochondria in aging. J. Clin.

Invest. 123, 951–957.

Breker, M., Gymrek, M., and Schuldiner, M. (2013). A novel single-cell

screening platform reveals proteome plasticity during yeast stress responses.

J. Cell Biol. 200, 839–850.

Breslow, D.K., Cameron, D.M., Collins, S.R., Schuldiner, M., Stewart-Ornstein,

J., Newman, H.W., Braun, S., Madhani, H.D., Krogan, N.J., and Weissman,

J.S. (2008). A comprehensive strategy enabling high-resolution functional

analysis of the yeast genome. Nat. Methods 5, 711–718.

Cohen, Y., and Schuldiner, M. (2011). Advanced methods for high-throughput

microscopy screening of genetically modified yeast libraries. Methods Mol.

Biol. 781, 127–159.

Costa, V., and Scorrano, L. (2012). Shaping the role of mitochondria in the

pathogenesis of Huntington’s disease. EMBO J. 31, 1853–1864.

Csordas, G., Renken, C., Varnai, P., Walter, L., Weaver, D., Buttle, K.F., Balla,

T., Mannella, C.A., and Hajnoczky, G. (2006). Structural and functional features

and significance of the physical linkage between ER and mitochondria. J. Cell

Biol. 174, 915–921.

Daniele, T., Hurbain, I., Vago, R., Casari, G., Raposo, G., Tacchetti, C., and

Schiaffino, M.V. (2014). Mitochondria and melanosomes establish physical

contacts modulated by Mfn2 and involved in organelle biogenesis. Curr.

Biol. 24, 393–403.

Daum,G., Bohni, P.C., and Schatz, G. (1982). Import of proteins intomitochon-

dria. Cytochrome b2 and cytochrome c peroxidase are located in the inter-

membrane space of yeast mitochondria. J. Biol. Chem. 257, 13028–13033.

Daum, G., Lees, N.D., Bard, M., and Dickson, R. (1998). Biochemistry, cell

biology and molecular biology of lipids of Saccharomyces cerevisiae. Yeast

14, 1471–1510.

Du, X., Kumar, J., Ferguson, C., Schulz, T.A., Ong, Y.S., Hong, W., Prinz, W.A.,

Parton, R.G., Brown, A.J., and Yang, H. (2011). A role for oxysterol-binding

protein-related protein 5 in endosomal cholesterol trafficking. J. Cell Biol.

192, 121–135.

Elbaz, Y., and Schuldiner, M. (2011). Staying in touch: the molecular era of

organelle contact sites. Trends Biochem. Sci. 36, 616–623.

elopmental Cell 30, 95–102, July 14, 2014 ª2014 Elsevier Inc. 101

Developmental Cell

A Dynamic Mitochondria/Vacuole Interface

Folch, J., Lees, M., and Sloane Stanley, G.H. (1957). A simple method for the

isolation and purification of total lipides from animal tissues. J. Biol. Chem.

226, 497–509.

Friedman, J.R., Lackner, L.L., West, M., DiBenedetto, J.R., Nunnari, J., and

Voeltz, G.K. (2011). ER tubules mark sites of mitochondrial division. Science

334, 358–362.

Giaever, G., Chu, A.M., Ni, L., Connelly, C., Riles, L., Veronneau, S., Dow, S.,

Lucau-Danila, A., Anderson, K., Andre, B., et al. (2002). Functional profiling of

the Saccharomyces cerevisiae genome. Nature 418, 387–391.

Gietz, R.D., and Woods, R.A. (2006). Yeast transformation by the LiAc/SS

Carrier DNA/PEG method. Methods Mol. Biol. 313, 107–120.

Griffiths, E.J. (2012). Mitochondria and heart disease. Adv. Exp. Med. Biol.

942, 249–267.

Honscher, C., Mari, M., Auffarth, K., Bohnert, M., Griffith, J., Geerts, W., van

der Laan, M., Cabrera, M., Reggiori, F., and Ungermann, C. (2014). Cellular

metabolism regulates contact sites between vacuoles and mitochondria.

Dev. Cell 30, this issue, 86–94.

Janke, C., Magiera, M.M., Rathfelder, N., Taxis, C., Reber, S., Maekawa, H.,

Moreno-Borchart, A., Doenges, G., Schwob, E., Schiebel, E., and Knop, M.

(2004). A versatile toolbox for PCR-based tagging of yeast genes: new fluores-

cent proteins, more markers and promoter substitution cassettes. Yeast 21,

947–962.

Kopec, K.O., Alva, V., and Lupas, A.N. (2010). Homology of SMP domains to

the TULIP superfamily of lipid-binding proteins provides a structural basis

for lipid exchange between ER and mitochondria. Bioinformatics 26, 1927–

1931.

Kornmann, B., andWalter, P. (2010). ERMES-mediated ER-mitochondria con-

tacts: molecular hubs for the regulation of mitochondrial biology. J. Cell Sci.

123, 1389–1393.

Kornmann, B., Currie, E., Collins, S.R., Schuldiner, M., Nunnari, J., Weissman,

J.S., andWalter, P. (2009). An ER-mitochondria tethering complex revealed by

a synthetic biology screen. Science 325, 477–481.

Levine, T., and Loewen, C. (2006). Inter-organelle membrane contact sites:

through a glass, darkly. Curr. Opin. Cell Biol. 18, 371–378.

Liesa,M., and Shirihai, O.S. (2013). Mitochondrial dynamics in the regulation of

nutrient utilization and energy expenditure. Cell Metab. 17, 491–506.

Longtine, M.S., McKenzie, A., 3rd, Demarini, D.J., Shah, N.G., Wach, A.,

Brachat, A., Philippsen, P., and Pringle, J.R. (1998). Additional modules for

versatile and economical PCR-based gene deletion and modification in

Saccharomyces cerevisiae. Yeast 14, 953–961.

Merz, S., and Westermann, B. (2009). Genome-wide deletion mutant analysis

reveals genes required for respiratory growth, mitochondrial genome mainte-

nance and mitochondrial protein synthesis in Saccharomyces cerevisiae.

Genome Biol. 10, R95.

Mozdy, A.D., McCaffery, J.M., and Shaw, J.M. (2000). Dnm1p GTPase-medi-

ated mitochondrial fission is a multi-step process requiring the novel integral

membrane component Fis1p. J. Cell Biol. 151, 367–380.

102 Developmental Cell 30, 95–102, July 14, 2014 ª2014 Elsevier Inc

Murley, A., Lackner, L.L., Osman, C., West, M., Voeltz, G.K., Walter, P., and

Nunnari, J. (2013). ER-associated mitochondrial division links the distribution

of mitochondria and mitochondrial DNA in yeast. eLife 2, e00422.

Nguyen, T.T., Lewandowska, A., Choi, J.Y., Markgraf, D.F., Junker, M., Bilgin,

M., Ejsing, C.S., Voelker, D.R., Rapoport, T.A., and Shaw, J.M. (2012). Gem1

and ERMES do not directly affect phosphatidylserine transport from ER to

mitochondria or mitochondrial inheritance. Traffic 13, 880–890.

Pan, X., Roberts, P., Chen, Y., Kvam, E., Shulga, N., Huang, K., Lemmon, S.,

and Goldfarb, D.S. (2000). Nucleus-vacuole junctions in Saccharomyces

cerevisiae are formed through the direct interaction of Vac8p with Nvj1p.

Mol. Biol. Cell 11, 2445–2457.

Price, A., Wickner, W., and Ungermann, C. (2000). Proteins needed for vesicle

budding from the Golgi complex are also required for the docking step of

homotypic vacuole fusion. J. Cell Biol. 148, 1223–1229.

Sheftel, A.D., Zhang, A.S., Brown, C., Shirihai, O., and Ponka, P. (2007). Direct

interorganellar transfer of iron from endsome to mitochondrion. Blood 110,

125–132.

Stroupe, C., Collins, K.M., Fratti, R.A., and Wickner, W. (2006). Purification

of active HOPS complex reveals its affinities for phosphoinositides and the

SNARE Vam7p. EMBO J. 25, 1579–1589.

Tatsuta, T., Scharwey, M., and Langer, T. (2014). Mitochondrial lipid traf-

ficking. Trends Cell Biol. 24, 44–52.

Tokuyasu, K.T. (1986). Application of cryoultramicrotomy to immunocyto-

chemistry. J. Microsc. 143, 139–149.

Tong, A.H., and Boone, C. (2006). Synthetic genetic array analysis in

Saccharomyces cerevisiae. Methods Mol. Biol. 313, 171–192.

Tong, A.H., Evangelista, M., Parsons, A.B., Xu, H., Bader, G.D., Page, N.,

Robinson, M., Raghibizadeh, S., Hogue, C.W., Bussey, H., et al. (2001).

Systematic genetic analysis with ordered arrays of yeast deletion mutants.

Science 294, 2364–2368.

Trotter, P.J., and Voelker, D.R. (1995). Identification of a non-mitochondrial

phosphatidylserine decarboxylase activity (PSD2) in the yeast Saccharomyces

cerevisiae. J. Biol. Chem. 270, 6062–6070.

Vance, J.E. (1990). Phospholipid synthesis in a membrane fraction associated

with mitochondria. J. Biol. Chem. 265, 7248–7256.

Vives-Bauza, C., and Przedborski, S. (2011). Mitophagy: the latest problem for

Parkinson’s disease. Trends Mol. Med. 17, 158–165.

Voss, C., Lahiri, S., Young, B.P., Loewen, C.J., and Prinz, W.A. (2012).

ER-shaping proteins facilitate lipid exchange between the ER and mitochon-

dria in S. cerevisiae. J. Cell Sci. 125, 4791–4799.

Wu, M.M., Buchanan, J., Luik, R.M., and Lewis, R.S. (2006). Ca2+ store deple-

tion causes STIM1 to accumulate in ER regions closely associated with the

plasma membrane. J. Cell Biol. 174, 803–813.

Yu, E., Mercer, J., and Bennett, M. (2012). Mitochondria in vascular disease.

Cardiovasc. Res. 95, 173–182.

.