ORP-Mediated ER Contact with Endocytic Sites Facilitates Actin Polymerization

ORP-Mediated ER Contact with Endocytic Sites Facilitates Actin Polymerization Javier Encinar del Dedo1, Fatima-Zahra Idrissi1, Isabel María Fernandez-Golbano1, Patricia Garcia1,2, Elena Rebollo1, Marek K. Krzyzanowski1,3, Helga Grötsch1,4 and Maria Isabel Geli1,5,* 1 Institute for Molecular Biology of Barcelona (CSIC), Baldiri Reixac 15, 08028 Barcelona, Spain 2 Present address: IDIBELL, Avenida de la Granvia de l'Hospitalet 199-203, 08090 L'Hospitalet de Llobregat, Spain 3 Present address: ETH Zürich, Otto-Stern-Weg 3, 8093 Zürich, Switzerland 4 Present address: EMBL Heidelberg, Meyerhofstraße 1, 69117 Heidelberg, Germany 5 Lead Contact ∗ Correspondence: [email protected] SUMMARY Oxysterol binding protein-related proteins (ORPs) are conserved lipid binding polypeptides, enriched at ER contacts sites. ORPs promote non-vesicular lipid transport and work as lipid sensors in the context of many cellular tasks, but the determinants of their distinct localization and function are not understood. Here, we demonstrate that the yeast endocytic invaginations associate with the ER and that this association specifically requires the ORPs Osh2 and Osh3, which bridge the endocytic myosin-I Myo5 to the ER integral-membrane VAMP-associated protein (VAP) Scs2. Disruption of the ER contact with endocytic sites using ORP, VAP, myosin-I, or reticulon mutants delays and weakens actin polymerization and interferes with vesicle scission. Finally, we provide evidence suggesting that ORP-dependent sterol transfer facilitates actin polymerization at endocytic sites. INTRODUCTION Oxysterol binding protein-related proteins (ORPs) are conserved sterol and phosphoglyceride binding proteins (Raychaudhuri and Prinz, 2010), which work as non-vesicular transport carriers and lipid sensors in the context of many cellular functions (Raychaudhuri and Prinz, 2010). As ORPs can transport sterols, which mislocalize or accumulate in neurological and cardiovascular diseases, altered ORP expression often accompanies disease onset and progression (Ngo et al., 2010). ORPs are enriched at contacts between the ER and other cellular compartments. Most ORPs bear FFAT (two phenylalanines in an acidic tract) motifs, which interact with the N-terminal MSP (major sperm protein) domain of the ER integral-membrane proteins VAMP-associated protein A (VAPA) and VAPB in mammals, and Scs2 and Scs22 in yeast (Loewen et al., 2003). Through their capacity to bind VAPs and lipids, ORPs link the ER to other cellular membranes (Raychaudhuri and Prinz, 2010). However, the membrane binding determinants that confer specificity to the ORP recruitment and function are not understood. In yeast, seven genes (OSH1 to OSH7) encode ORPs, which share at least a common essential function (Beh et al., 2001). Nevertheless, evidence for functional specialization exists. Thus, Osh4 localizes to the ER-Golgi contacts (Fain et al., 2007; Li et al., 2002) and Osh1 is targeted to the nuclear-vacuolar junction (Levine and Munro, 2001), whereas Osh5 localizes to the yeast bud neck and the vacuolar membrane (Yofe et al., 2016). Other ORPs (Osh2, Osh3, Osh6, and Osh7) are recruited to the cortical ER (cER) (Schulz et al., 2009), which in yeast appears in close apposition to the plasma membrane (PM) (West et al., 2011). Whether the cortical ORPs are merely redundant is unknown, but the observation that Osh2 and Osh3 have divergent N-terminal domains, and Osh6 and Osh7 transport phosphatidylserine (PS) rather than sterols (Maeda et al., 2013), also suggests specialization. Endocytic uptake in yeast and mammals implies the orchestrated recruitment of a number of functional modules (Figure 1A) (Kaksonen et al., 2003, 2005; Taylor et al., 2011). Upon assembly of the clathrin coat (Idrissi et al., 2012; Kukulski et al., 2012), membrane curvature emerges to produce an invagination that grows into a tubule of up to 200 nm in length, which is capped by the endocytic coat (Idrissi et al., 2012; Kukulski et al., 2012). As the tubule elongates, the neck constricts (Idrissi et al., 2012) to finally produce a vesicle (Avinoam et al., 2015; Kukulski et al., 2011, 2012). Actin polymerization is coupled to the emergence of membrane curvature and is essential for tubule elongation, which can be monitored as a slow inward movement of the endocytic coat by fluorescence microscopy (Idrissi et al., 2012; Kaksonen et al., 2003; Kukulski et al., 2012). Actin polymerization is mainly ignited by Las17/N-WASP (Galletta et al., 2008; Sun et al., 2006), which sits on the surface of the invagination tip and neck (Idrissi et al., 2008), and Myo5/myosin-I (Galletta et al., 2008; Sun et al., 2006), which accumulates where the tubule joints the PM (Idrissi et al., 2008). What exactly prompts actin polymerization, and which are the determinants organizing the actomyosin vectorial forces, are far from known. We here demonstrate that the endocytic sites associate with cER rims at the onset of actin polymerization and that in the absence of observable cER the endocytic coat can assemble, but initiation of actin polymerization and vesicle scission are impaired. We also found that Osh2 and Osh3 mediate binding of the endocytic myosin-I Myo5 to the ER integral-membrane VAP Scs2, and that disruption of the Myosin-I/ORP/VAP link alters the cER contact with endocytic sites. Finally, we provide evidence suggesting that ORP-dependent transport of sterols facilitates actin polymerization at endocytic sites. RESULTS Endocytic Invaginations Associate with cER Rims at the Onset of Actin Polymerization Previous work has demonstrated that the increasing length of endocytic invaginations can be used to scale time on quantitative immunoelectron microscopy studies defining the position of endocytic proteins (Idrissi and Geli, 2014), thereby allowing description of their dynamics with a resolution of a few nanometers (Femández-Golbano et al., 2014; Idrissi et al., 2008). Comparing PM invaginations labeled for different endocytic proteins (Idrissi et al., 2008, 2012), we noticed many associated with cER rims (Figures 1A-1D and S1A). The yeast cER has an unequivocal appearance as an electrontranslucent zone of 20-65 nm in close apposition to the PM (West et al., 2011) (Figures 1B [arrowheads], S1A, and S1B). A possible association between PM invaginations and the cER in yeast was observed previously (Buser and Drubin, 2013). However, this work did not provide evidence discarding an accidental association due to the high abundance of cER. To discard this possibility, we analyzed whether such association evolved as the invaginations grew. We quantified the percentage of short (invagination length [IL]<50 nm), medium (50 nm≤IL<100 nm), or long (100 nm≤IL) invaginations falling into four categories: (1) not associated with cER (with no cER closer than 70 nm from the invagination) (no ER); (2) cER-sheltered (ER sh); (3) associated with one cER rim (ER 1r); or (4) flanked by two cER rims (ER 2r) (Figure 1B). We found striking differences as a function of the IL. Whereas 80% of the short invaginations appeared either not associated with cER or sheltered (Figures 1C, 1D, and S1A), more than 70% of the longest invaginations appeared associated with cER rims, being more than 30% flanked by cER (Figures 1B-1D and S1A). The data clearly pointed to an increased association of the matured endocytic invaginations with cER rims, discarding that the association was accidental (Figure 1C). Membrane curvature emergence during endocytic budding in yeast is coupled to initiation of actin polymerization (Idrissi et al., 2012; Kukulski et al., 2012). Consistently, labeling for markers of the actin module, such as Abp1, barely appear associated with invaginations shorter than 40 nm (Idrissi et al., 2012). Therefore, the results suggested that association of the endocytic sites with cER rims might occur at the onset of actin polymerization. To further investigate at which point endocytic patches became connected to cER rims, we imaged Sec61-mCherry, a subunit of the ER translocon, in combination with different GFP-tagged proteins, sequentially arriving at endocytic sites (Figure 1A): the eps15 homolog Ede1 (Stimpson et al., 2009), the coat component Sla1, and the actin binding protein Abp1 (Kaksonen et al., 2003) (Figures 1A, 2, and S2). Since the cER covers most PM in wild-type (WT) yeast, we also engineered an rtn1Δ reticulon mutant with the same fluorescent proteins. In the reticulon mutants the cER tubules collapse to a lamina, leaving extended surfaces of cER-uncovered PM (De Craene et al., 2006; Voeltz et al., 2006; West et al., 2011). Therefore, cER-free and -sheltered regions can be better defined. Analysis of the association of the different cortical patches with cER rims in WT cells already indicated significant differences between the early and coat components versus the actin marker Abp1. Abp1 was more associated with cER rims, regardless of whether equatorial planes, cortical planes, or 3D reconstructions of yeast were analyzed (Figures 2 and S2; Movies S1 and S2). A nonsignificant association of cortical Ede1 patches with cER rims was reported previously (Stradalova et al., 2012). The tighter association between Abp1 and the cER rims could be clearly visualized in WT cells in 3D projections (Figure 2C and Movie S2). The differences could be better appreciated in the rtn1Δ mutant (Figures 2A and 2B; Movie S1). Thus, whereas 70% of the Abp1 patches were closely associated with cER edges, only 30% of the Ede1 or Sla1 foci accumulated at these points in the mutant (Figures 2A and 2B). In WT cells, the cERfree regions are often beyond the resolution limits and, therefore, most Sla1 and Ede1 patches were recorded as assembling on cER-sheltered regions or at cER rims (Figures 2A, 2B, and S2). Interestingly, Ede1 and Sla1 patches assembled at cER-sheltered or -free regions in the rtn1Δ mutant (Figures 2A and 2B; Movie S1). In contrast, Abp1 patches hardly ever appeared on cER-depleted regions (Figures 2A and 2B; Movie S1). Consistent with the electron microscopy (EM) data, the liveimaging results indicated that endocytic patches become associated with cER rims at the onset of actin polymerization. Disruption of cER Contact with Endocytic Sites Impairs Actin Polymerization and Vesicle Scission The observation that Abp1 patches were hardly found at cER-free regions in the rtn1Δ mutant suggested that cER contact might be required to initiate actin polymerization and, therefore, to stabilize membrane curvature and promote tubule elongation. Tubule elongation can be monitored by live-cell fluorescence microscopy as a slow inward movement of the coat, which is evident in kymographs perpendicular to the cortex (Idrissi et al., 2012; Kaksonen et al., 2003; Kukulski et al., 2012). Consistent with this view, we observed a striking correlation between the ability of Sla1 to move inward and its association with the cER (Figures 3A-3C and S3; Movie S3). This correlation was obvious in a time projection, where most Sla1 patches moving inward in WTs appeared as tails, whereas the immobile patches in the cER-free regions of rtn1Δ cells appeared as dots (Figure 3A). More than 60% of the Sla1 patches assembled at cER-depleted regions in rtn1Δ cells failed to initiate the inward movement, exhibiting a flat trajectory in kymograph representations, whereas about 80% of the patches associated with cER rims did initiate the inward movement in the same strain background (Figures 3B and S3). The strong correlation between the endocytic defect and the absence of cER in the same strain discarded the proposal that bulk alterations of the PM secondarily altered actin polymerization. Triple-labeling experiments confirmed that endocytic coats assembled at ER-free regions did not or barely recruit Abp1, whereas those assembled at cER rims did, in the same rtn1Δ background (Figure 3C and Movie S3). In WTs, most Sla1 patches appeared at regions covered by very dynamic cER and often acquired Abp1, and moved into the cytosol when the cER opened like a sliding door (Figures 3C and S3; Movie S3). The few Sla1 patches failing to initiate the inward movement in WT cells (Around 5%, Figure 3B, WT flat) also corresponded to patches that did not contact the ER (Figure S3, WT ER free), suggesting that ER contact was required to properly initiate actin polymerization not only in rtn1Δ cells but also in the WT. Analysis of the behavior of Sla1 in rtn1Δ cells also uncovered a defect in membrane fission, which is reflected by a v-shape in the Sla1-GFP kymographs, caused by the retraction of the endocytic coat back to the PM (Kaksonen et al., 2005; Toshima et al., 2006) (Figures 3B [v-shape] and S3). About 20% of the Sla1 patches had a v-shape phenotype in the rtn1Δ strain, compared with 4% in WTs (Figure 3B). The defects in vesicles scission in the rtn1Δ strain were mostly associated to cERsheltered or -free regions, but rarely observed at cER rims, where the Sla1 patches mostly showed WT dynamics (Figure 3B). In contrast to the actin polymerization defects, the v-shape phenotype in rtn1Δ cells was more frequent at cER-sheltered regions compared with cER-free areas (Figure 3B). This was expected, since a strong defect in actin polymerization in cER-free areas would mask a scission defect. In contrast to endocytic coats at cER-free regions, Sla2 patches assembled at cERsheltered regions in rtn1Δ cells and still recruited Abp1, but many failed to progress and retracted (Figures 3C and 3D; Movie S3), probably due to the altered properties of the cER in the mutant (De Craene et al., 2006; Voeltz et al., 2006; West et al., 2011). Indeed, the cER in the rtn1Δ mutant often failed to open under the endocytic coat (Figure 3C and Movie S3). Quantitative analysis of PM Sla1-HA-immunolabeled invaginations on ultrathin sections of rtn1Δ yeast confirmed this observation. cER-sheltered intermediate and long invaginations (IL>50 nm) accumulated in the reticulon mutant, as compared with the WT, indicating that actin polymerization and membrane invagination occurred, although the cER acted as a barrier that precluded elongation and, possibly, fission (Figures 3D-3F). Osh2 and Osh3 Link Myo5 to the ER Integral-Membrane Protein Scs2 To investigate which endocytic proteins were involved in establishing cER contact with endocytic sites, we compared the position of cER rims associated with 120 randomly chosen endocytic invaginations with that of 120 gold particles labeling each of 16 different endocytic proteins, previously described in Idrissi et al. (2012). To do so, we defined the ER relative position (ERRP) (Figure 4A) and statistically compared it with the Immunogold relative position (GRP) for the different proteins (Idrissi et al., 2008) (Figures 4A and 4B). The ER or gold relative position (ERRP or GRP) is the distance to the basal PM of the element of interest (ERDPM or GDPM) divided by the length of the associated invagination (IL). Elements associated with the invagination tip have relative position (RP) values close to 1, whereas those localized at the PM have RP values near 0, irrespective of the IL. The data showed maximum coincidence of the cER rim position with Myo5 (Figure 4B). Indeed, Immunogold particles decorating Myo5-HA were reported between the invagination and the cER rim (Figure 4C). The average distance of the ER rims to the associated invaginations (ERDI) was 48±1.2 nm (Figures 4A and 4B), compatible with a molecular link of two or three proteins. Genome-wide studies indicate that Myo5 might interact with Osh2 (Tonikian et al., 2009), a cortical ORP bearing an FFAT motif that binds Scs2 and Scs22 (Stefan et al., 2011) (Figure 4D). Two-hybrid assays confirmed that Myo5 interacted with Osh2 and also Osh3, but not other ORPs (Figure 4E, left panel). The data also indicated that the N-terminal region lacking the oxysterol binding proteinrelated domain (ORD) mediated the interaction with Myo5 (Figure 4E, left panel). Coimmunoprecipitations confirmed that Myo5 interacted with Osh2 and Osh3 (Figure 4F), but not with Osh1, which shares more than 60% sequence identity to Osh2. Reciprocally, Osh2 was not pulled down with the other endocytic nucleation-promoting factor (NPF) Las17 under the same experimental conditions (Figure S4A), despite previous data pointing to a possible interaction between the yeast N-WASP and the ORP (Michelot et al., 2010). The ORP/Myo5 interaction occurred mainly with proteins recruited to heavy membranes, since Osh2 and Myo5 did not coimmunoprecipitate from cytosolic or microsomal fractions (Figure S4B). Fluorescence microscopy confirmed that Myo5 and Osh2 and Osh3 transiently co-localized at cortical sites (Figure 4G and Movie S4). Interestingly, the ORPs did not concentrate on the endocytic patch progressively, as other endocytic proteins do, but rather dynamically approached as preformed foci, eventually came into contact with Myo5 and then either moved away or became faint (Movie S4). Analysis of the Osh2 intensity relative to Myo5 indicated that Osh2 arrived slightly after the Myo5 signal started to emerge, and the Myo5 signal continued to increase upon Osh2 contact (Figure S4C). Two-hybrid assays showed that Osh2 and Osh3 interacted with different Myo5 domains (Figure 4E, middle panel). As predicted, the Myo5-Osh2 interaction required the Myo5 SH3 domain and an Osh2 polyproline motif (PPPVP), absent in the Osh1 N-terminal domain (Figures 4D and 4E). In contrast, the Osh3 PH (pleckstrin homology) domain bound the Myo5 lipid binding TH1 (tail homology 1) domain (Figure 4E). A direct interaction between Myo5 and the ORPs was confirmed by pull-down assays using purified components (Figure S4D). To investigate whether Osh2 and Osh3 mediated binding of Myo5 to the ER Scs2 (Figure 5A), we asked whether Protein A-Myo5 (PA-Myo5) co-immunoprecipitated with Scs2-HA in an ORPdependent manner. As shown in Figures 5B and S5A, PA-Myo5 specifically co-precipitated with Scs2HA. The interaction was not due to massive co-precipitation of ER and PM, since the PM glycosylphosphatidylinositol-anchored proteins Gas1 and Las17 were not found in the Scs2-HA precipitate and PA-Myo5 failed to co-precipitate with the ER protein Rtn1-HA (Figure S5A). In agreement with the hypothesis that the ORPs specifically mediated binding of Myo5 to VAPs, the Myo5/Scs2 interaction was significantly reduced in the absence of Osh2, and nearly abolished in the absence of Osh2 and Osh3 (Figure 5B). Furthermore, mutation of the Osh2 PPPVP or FFAT motifs, which specifically prevented interaction with Myo5 and Scs2, respectively, altered the Myo5/Scs2 interaction, similar to depletion of Osh2 (Figure 5C). In addition, a point mutation in the Myo5 SH3 domain, predicted to impair binding to the polyp motifs (Myo5-W1123S), also affected the interaction with Scs2-HA to a similar extent (Figure S5B). Osh2 and Osh3 Functionally Link the cER to Endocytic Sites To further investigate whether the ORPs linked the cER to endocytic sites, we performed triplelabeling experiments in cells expressing Sec61-mCherry, Osh2-YFP, and Abp1-CFP. Consistent with our hypothesis, Osh2-YFP was often found in an intermediate position between the cER rims (Sec61mCherry) and the endocytic patches (Abp1-CFP), partially overlapping with both (Figures 5D and S5C). Time-lapse movies evidenced a very dynamic interaction between the cER rims bearing Osh2 and the Abp1-CFP patches (Figure S5C), with Abp1 appearing after arrival of Osh2 (Figure S5D). To refine these data, we analyzed the position of Osh2-HA and Osh3-HA, relative to the cER rims and the associated PM invaginations, using quantitative EM. Calculation of the average GRP for 36 gold particles labeling the ORPs demonstrated a striking coincidence with the position of Myo5 and the cER rim, but not with that of Sla1 (Figures 5E-5G). The average distance to the invagination of the gold particles decorating the ORPs was 35.4±5.3 nm, further demonstrating their intermediate localization between the cER rim and Myo5 (Figure 5G). To directly investigate whether Osh2 and Osh3 were required to establish functionally relevant contacts, we analyzed the effect of depleting the ORPs on the association of Sla1-GFP with cER rims in an rtn1Δ strain expressing Sec61-mCherry. The percentage of Sla1-GFP patches associated with cER rims progressively decreased upon depletion of Osh2 or Osh2 and Osh3 (Figures 5H, 5I, and S5E). In addition, the functionality of the endocytic patches still associated with cER rims was compromised upon depletion of the ORPs (Figure 5J). Next, we analyzed whether ORP and VAP mutants exhibited phenotypes that recapitulated those observed in rtn1Δ cells. The lifespan of Sla1 and Abp1 patches was expanded in the osh2Δosh3Δ and scs2Δscs22Δ strains, similar to the rtn1Δ mutant, with a considerable fraction of Sla1 patches initiating the inward movement later, or failing to so (Figures 6A-6C). Double-labeling experiments using Abp1-mCherry and Sla1-GFP demonstrated that actin polymerization at endocytic sites was delayed and weak in the osh2Δosh3Δ and scs2Δscs22Δ strains, similar to rtn1Δ cells (Figure 6D). In addition to the actin polymerization defect, a considerable number of v-shaped phenotypes was recorded in ORP and VAP mutants, which were more abundant for Abp1 than for Sla1 (Figures 6A and 6C). The data suggested that ORPand VAP-mediated ER contact to endocytic sites might have a dual function, first facilitating actin polymerization and second, promoting vesicle scission. The defects observed in the osh2Δosh3Δ mutant were unlikely a consequence of gross alteration of lipid or organelle homeostasis, since these cells show minor changes in the phosphoinositide content (Stefan et al., 2011) and do not exhibit major sterol-sensitive phenotypes (Beh et al., 2001), while transport to the vacuole and secretion is unaffected (Beh and Rine, 2004). Furthermore, we observed that the Sla1-GFP dynamics was unaltered in cells lacking Osh4 or Osh1, which have stronger growth defects compared with the osh2Δosh3Δ strain (Beh et al., 2001) (Figures S6A-S6C). No significant endocytic defects were observed in the scs2Δ or scs22Δ single mutants, suggesting strict functional redundancy (Figures S6A-S6C). In contrast, the phenotypes of the ORP single mutants indicated a certain specialization. Thus, even though depletion of either Osh2 or Osh3 caused a mild fission defect (Figures S6A-S6C), only depletion of Osh2 extended the lifespan of cortical Sla1-GFP patches, similar to the double mutant. The data suggested that Osh2 and Osh3 shared a function in vesicle scission but that Osh2 played a predominant role facilitating actin polymerization. The Osh2 Sterol Transport Domain Facilitates Actin Polymerization at Endocytic Sites Osh2 has the capacity to transfer sterols in vitro (Schulz et al., 2009), and sterols are required to sustain actin polymerization in the context of different cellular functions (Chadda et al., 2007; Kwik et al., 2003; Tsai et al., 2014). Furthermore, mutations affecting the sterol biosynthesis show endocytic defects in yeast (Heese-Peck et al., 2002; Munn et al., 1999), and we observed that treatment of cells with the sterol-sequestering drug filipin delayed and weakened actin polymerization at endocytic sites (Figures 7A-7C). In light of these results, we considered that the primary molecular function of the Myosin-I/Osh2/VAP link facilitating actin polymerization might be related to the localized transfer of sterols. We found that overexpression of Myo5 caused the appearance of ectopic membrane domains, which could be stained with filipin under non-permeabilizing conditions (Figure 7D). We decided to use this phenotype as a readout for the possible sterol transfer capacity of the myosin-I/ORP/VAP complex. The filipin/GFP-Myo5 co-localization completely disappeared in the scs2Δscs22Δ and osh2Δosh3Δ mutants and could not be recovered in the osh2Δosh3Δ background, by the Osh2PPPVP* and FFAT mutants, with impaired binding to Myo5 and Scs2, respectively (Figure 7D). Overexpression of Myo5-W1123S also failed to sustain the formation of filipin-stainable domains, consistent with the view that this myosin-I activity required the interaction with Osh2 (Figure S7C). Expression of an Osh2 C-terminal truncation lacking the ORD (Osh2-ORDΔ), which transfers sterols in vitro (Schulz et al., 2009), failed to sustain the appearance of filipin-stainable domains (Figure 7D). Substitution of the Osh2 ORD by that of Osh4, also known to transfer sterols (Schulz et al., 2009), recovered the filipin staining (Figure 7D). However, substitution by the Osh6 ORD, which transfers PS (Maeda et al., 2013), failed to do so (Figure 7D). Altogether, the data indicated that myosins-I could induce the formation of sterol-enriched domains at the PM and that this capacity required the Myosin-I/Osh2/VAP link and the sterol transfer activity of Osh2. The capacity of the different Osh2 mutants to restore the Myo5-induced filipin-stainable domains strongly correlated with their capacity to restore WT kinetics of the Sla1-GFP patches in the osh2Δ background (Figures 7E and 7F). Likewise, the Myo5-W1123S mutant expressed as the only source of myosin-I failed to sustain WT Sla1-YFP kinetics, even when it was artificially hooked to Las17 to guarantee its recruitment to endocytic sites (Figures S7D-S7G). The data suggested that the Myosin-I/Osh2/VAP complex facilitates actin polymerization at endocytic sites and that this activity probably required a sterol transfer activity of the complex. Interestingly, an Osh3 mutant lacking the ORD fully complemented the v-shape phenotype of an osh3Δ strain (Figures S7H-S7K), suggesting that, unlike the role of Osh2 promoting actin polymerization, the endocytic function of Osh3 did not rely on its lipid transfer capacity. DISCUSSION The present work demonstrates that endocytic coats can assemble at the PM in the presence or absence of cER, but associate with cER rims at the onset of actin polymerization. A non-accidental association of the cER with clathrin-coated pits at the PM was previously demonstrated in mammalian cells (Fox et al., 2013), indicating a possible conserved function of the cER in clathrinmediated endocytosis. We now identified a molecular link between the endocytic profiles and the cER, which implicates two ORPs, the VAPs and the myosins-I. Disruption of this molecular link by different means delayed the onset of actin polymerization and installed scission defects. Together with the timing of arrival of the cER at endocytic sites, the data strongly suggest that its primary role is facilitating actin polymerization and, as a consequence, membrane invagination. The lack of cER leads to a delay and weakening of actin polymerization rather than a complete inhibition. We favor the hypothesis that the actin-dependent recruitment of disassembly factors might still occur, but that actin polymerization might be insufficient to support invagination in the absence of cER contact. Alternatively or in addition, the lack of certain lipids at endocytic sites might impact on the stability of the endocytic coat, which could secondarily impact on actin polymerization. With respect to the scission defects, we cannot rule out at this point that these are a consequence of a defective endocytic actin network. However, our data suggest that proper initiation of actin polymerization and scission might be dissectible functions. Thus, we observe that depletion of Osh3 caused ORD-independent scission defects without apparently altering actin polymerization, and endocytic coats assembled at cER-sheltered regions in the rtn1Δ mutant still recruited Abp1 but failed to undergo fission. Together with the observation that longer invaginations tend to be flanked by cER rims, the data suggested that the ER might assist vesicle scission after embracing the elongated invaginations, as has been proposed for mitochondria or endosomes (Friedman et al., 2011; Rowland et al., 2014). Finally, we provide evidence suggesting that the molecular function of the Myo5/ORP/VAP link facilitating actin polymerization relies mostly on Osh2 and might implicate its sterol transfer capacity. We showed that deletion of its ORD, which has sterol transfer capacity in vitro (Schulz et al., 2009), installed actin polymerization defects similar to those caused by complete depletion of Osh2, and that this defect could be fully complemented by the ORD of Osh4, which also transfers sterols (Schulz et al., 2009), but not by that of Osh6, which transfers PS (Maeda et al., 2013). In parallel, we demonstrated that overexpression of Myo5 can induce the formation of membrane domains stainable with the sterol-sequestering drug filipin, and that this capacity required the Osh2 ORD and could be complemented by the ORD of Osh4 but not Osh6. Sterols control actin polymerization in the context of many cellular functions, including endocytic uptake via the GEEC pathway in mammals (Chadda et al., 2007; Kwik et al., 2003; Tsai et al., 2014). Based on our results and previous publications, we favor the hypothesis that sterol transport by Osh2 might organize phosphoglyceride-enriched microdomains (Kwik et al., 2003) or activate calcium channels (Dopico et al., 2012), which in turn activate or concentrate NPFs. However, we cannot completely rule out that Osh2 plays other functions, which secondarily impact on the initiation of actin polymerization. This work unveils the previously unnoticed contribution of myosins-I to the establishment of specialized contacts between the ER and the PM, which might promote transport of sterols to trigger localized actin polymerization. Myo5 is the yeast homolog of human Myo1E, also involved in endocytic uptake (Krendel et al., 2007) and cell migration (Jin et al., 2014; Tanimura et al., 2016). Myo1E promotes breast cancer malignancy (Ouderkirk-Pecone et al., 2016) and its misfunction causes a severe renal disease (Mele et al., 2011). Similar to endocytosis, cell migration requires assembly of sterol-enriched microdomains (Vasanji et al., 2004), suggesting that a primary molecular function of myosins-I might rely on their capacity to establish transient ER/PM contacts to promote sterol transport and actin polymerization. Altogether, we describe a transient link between the ER and the endocytic sites and define its nature and function. Despite most ER contacts with other cellular membranes being rather stable (Phillips and Voeltz, 2016), a role for transient ER contacts assisting the formation or scission of transport intermediates is starting to emerge (Caldieri et al., 2017; Dong et al., 2016). MATERIALS AND METHODS Strains and Growth Conditions Yeast strains used are listed in the Key Resources Table. GFP, mCherry, YFP and HA tags were fused at the C-terminus of each protein by homologous recombination in the genome as described previously (Wach et al., 1997). Genome edited strains behaved as WTs. Strains without plasmids were grown in complete yeast peptone dextrose and strains with plasmids were selected on synthetic dextrose complete (SDC) lacking the appropriate nutrient (Dulic et al., 1991) at 25∘C. Transformation of yeast was accomplished by the lithium acetate method (Ito et al., 1983). DNA Techniques and Plasmid Construction DNA manipulations were performed as described (Sambrook et al., 1989). Enzymes for molecular biology were obtained from New England Biolabs. PCRs were performed with a Vent polymerase (New England Biolabs) and a TRIO-thermoblock (Biometra GmbH). All plasmids used in this study bear AMP resistance for selection in E. coli and are listed in the Key Resources Table, where the yeast KEY RESOURCES TABLE REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies Peroxidase Anti-Peroxidase Sigma-Aldrich Cat# P1291, RRID:AB_1079562 Anti-HA-Peroxidase High Affinity (3F10) Roche Cat# 12013819001, RRID:AB_390917 Anti-GAS1 Muniz et al. (2001) N/A c-Myc Antibody, HRP conjugate (9E10) Thermo Fisher Scientific Cat# MA1-980-HRP, RRID:AB_2537628 Anti-GFP (JL-8) Clontech Cat# 632380, RRID:AB_10013427 Anti-HA Rat monoclonal antibody (3F10) Roche Cat# 11867423001, RRID:AB_10094468 EM Goat anti-Rat IgG Jackson Cat# 112-205-143, RRID:AB_2338271 Chemicals, Peptides, and Recombinant Proteins Peptone BD Difco 211677 Bacto yeast extract BD Difco 288620 Stericup Filter Sigma-Aldrich Z660493-12EA Sorbitol Sigma-Aldrich 240850100G Formaldehyde 16% MetOH free, ultra-pure EM grade Polysciences 18814 Glutaraldehyde 25% EM grade Polysciences 01909 Sodium Metaperiodate (NaIO4) Sigma-Aldrich S1878-25G Ammonium Chloride (NH4Cl) Sigma-Aldrich 254134-25G Ethanol PA ACS 99,8% Quimivita 414607 LR White Embedding Media Medium Polysciences 17411 Gelatin Capsules Polysciences 23313 BSA Sigma-Aldrich A9085 5-Bromo-4-chloro-3-indolyl β-D-galactopyranoside Sigma-Aldrich B4252-100MG Monoclonal Anti-HA Agarose Sigma-Aldrich A2095-1ML IgG Sepharose 6FastFlow GE Healthcare 17-0969-01 GFP-Trap ChromoTek gta-20 Glutathione Sepharose™ 4B GE Healthcare 17-0756-01 Filipin III from Streptomyces filipinensis Sigma-Aldrich F4767-1MG Critical Commercial Assays Mini-Protean TGX Gels BIO-RAD 456-1096 Experimental Models: Organisms/Strains Yeast strains Mata his3 leu2 met15 ura3 Euroscarf SCMIG381 Mata his3 leu2 ura3 trp1 bar1−1 EDE1−3HA::TRP1 Idrissi et al., 2012 SCMIG1213 Mata his3 leu2 ura3 trp1 bar1−1 SYP1−3HA::TRP1 Idrissi et al., 2012 SCMIG1212 Mata his3 leu2 ura3 trp1 bar1−1 ENT1−3HA::TRP1 Idrissi et al., 2012 SCMIG1214 ENT2−3HA::HIS3MX Mata his3 leu2 met15 ura3 bar1Δ::URA3 SLA1−3HA::HIS3MX Idrissi et al., 2012 SCMIG1218 Mata his3 leu2 ura3 trp1 bar1−1 PAN1−3HA::TRP1 Idrissi et al., 2008 SCMIG946 Mata his3 leu2 met15 ura3 bar1Δ::URA3 BZZ1−3HA::HIS3MX Idrissi et al., 2012 SCMIG1216 Mata his3Δ ura3Δ leu2Δ met15Δ bar1Δ::URA3 BBC1−3HA::HIS3MX Idrissi et al., 2008 SCMIG903 Mata his3Δ ura3Δ leu2Δ met15Δ bar1Δ::URA3 MYO5−3HA::HIS3MX Idrissi et al., 2008 SCMIG994 Mata his3Δ ura3Δ leu2Δ met15Δ bar1Δ::URA3 ABP1−3HA::HIS3MX Idrissi et al., 2008 SCMIG723 Mata his3 leu2 met15 ura3 bar1Δ::URA3 CAP1−3HA::HIS3MX Idrissi et al., 2012 SCMIG1217 Mata his3 leu2 trp1 ura3 bar1 SJL2−HA::TRP1 Fernández-Golbano et al., 2014 SCMIG1262 Mata his3 leu2 trp1 ura3 bar1 sla2Δ::HIS3 Idrissi et al., 2012 SCMIG669 Mata lys2 ade2 ura3 leu2 vrp1Δ::URA3 bar1 Idrissi et al., 2012 SCMIG79 Mata his3 leu2 met15 ura3 bar1Δ::URA3 ARC15−3HA::HIS3MX Idrissi et al., 2012 SCMIG1215 Mata his3 leu2 met15 ura3 bar1Δ::URA3 SAC6−3HA::HIS3MX Idrissi et al., 2012 SCMIG907 Mata his3 ura3 leu2 met15 bar1Δ::URA3 RVS167−3HA::HIS3MX Idrissi et al., 2008 SCMIG995 Mata his3 leu2 ura3 met15 trp1Δ::URA3 SEC61−mCherry::HIS3MX SLA1GFP−::TRP1 This study SCMIG1198 Mata his3 leu2 ura3 met15 trp1Δ::URA3 SEC61−mCherry::HIS3MX ABP1−GFP::TRP1 This study SCMIG1206 Mata his3 leu2 ura3 met15 trp1Δ::URA3 SEC61−mCherry::HIS3MX SLA1GFP−::TRP1 rt n1Δ::KMX This study SCMIG1208 Mata his3 leu2 ura3 met15 trp1Δ::URA3 SEC61−mCherry::HIS3MX ABP1−GFP::TRP1 rt n1Δ::KMX This study SCMIG1210 Mata his3 leu2 ura3 met15 trp1Δ::URA3 SEC61−mCherry::HIS3MX EDE1−GFP::TRPMX This study SCMIG1481 Mata his3 leu2 ura3 met15 trp1Δ::URA3 SEC61−mCherry::HIS3MX rtn1Δ::KMX EDE1− GFP::TRPMX This study SCMIG1482 Mata his3 leu2 ura3 met15 trp1Δ::URA3 SEC61−mCherry::HIS3MX rtn1Δ::KMX This study SCMIG1195 Mata his3 leu2 lys2 ura3 myo5Δ::KMX Euroscarf SCMIG136 Mato his3 leu2 lys2 ura3 myo5Δ::KMX OSH2−HA::HISMX This study SCMIG1483 Mato his3 leu2 lys2 ura3 myo5Δ::KMX OSH3−HA::HISMX This study SCMIG1484 Mato his3 leu2 lys2 ura3 myo5Δ::KMX OSH1−HA::HISMX This study SCMIG1485 Mata his3 leu2 lys2 ura3 myo5Δ::KMX OSH2−mCherry::HISMX This study SCMIG1486 REAGENT or RESOURCE SOURCE IDENTIFIER Mato his3 leu2 lys2 ura3 myo5Δ::KMX SCS2−HA::HISMX This study SCMIG1487 Mata his3 leu2 lys2 ura3 myo5Δ::KMX RTN1−HA::HISMX This study SCMIG1488 Mato osh2Δ::KMX ura3 leu2 his3 met15, myo5Δ::KMX SCS2−HA::HISMX This study SCMIG1489 Mata his3 leu2 met15 ura3 osh2Δ::KMX osh3Δ::KMX SCS2−HA::HISMX myo5Δ::LEU2 This study SCMIG1490 Mata his3 leu2 ura3 met15 trp1Δ::URA3 SEC61−mCherry::HIS3MX SLA1−GFP::TRP1 rt n1Δ::KMX osh2Δ::LEU2 This study SCMIG1491 Mata his3 leu2 met15 ura3 EDE1−GFP::HISMX This study SCMIG1492 Mata his3 leu2 met15 ura3 osh2Δ::KMX osh3Δ::KMX EDE1GFP−::HISMX This study SCMIG1493 Mato his3 leu2 lys2 ura3 scs2Δ::KMX scs22Δ::KMX EDE1−GFP::HISMX This study SCMIG1494 Mata his3 leu2 met15 ura3 SLA1−GFP::HISMX This study SCMIG1495 Mata his3 leu2 met15 ura3 osh2Δ::KMX osh3Δ::KMX SLA1−GFP::HISMX This study SCMIG1496 Mata his3 leu2 lys2 ura3 scs2Δ::KMX scs22Δ::KMX SLA1−GFP::HISMX This study SCMIG1497 Mata his3 leu2 lys2 ura3 osh2Δ::KMX osh3Δ::KMX This study SCMIG1498 Mata his3 leu2 met15 ura3 scs2Δ::KMX scs22Δ::KMX This study SCMIG1499 Mata his3 ura3 leu2 met1 bar1::URA3 ABP1−GFP::HISMX This study SCMIG738 Mata his3 leu2 met15 ura3 osh2Δ::KMX osh3Δ::KMX ABP1−GFP::HISMX This study SCMIG1500 Mata his3 leu2 lys2 ura3 scs2Δ::KMX scs22Δ::KMX ABP1−GFP::HISMX This study SCMIG1501 Mata his3 leu2 trp1 ura3 bar1 H. Riezman SCMIG19 Mata his3 leu2 lys2 ura3 osh2Δ::KMX osh3Δ::KMX This study SCMIG1502 Mata osh2Δ::KMX ura3 leu2 his3 met15 SLA1−GFP::HISMX This study SCMIG1503 Mata osh3Δ::KMX ura3 leu2 his3 met15 SLA1−GFP::HISMX This study SCMIG1504 Mata scs2Δ::KMX ura3 leu2 his3 met15 SLA1−HA::HISMX This study SCMIG1505 Mata scs22Δ::KMX ura3 leu2 his3 met15 Sla1−GFP::HISMX This study SCMIG1506 Mata ura3 leu2 his3 met15 rtn1Δ::KMX SEC61−mCherry::HISMX This study SCMIG1507 Mata ura3 leu2 his3 met15 rtn1Δ::KMX SLA1−HA::HISMX This study SCMIG1508 Mata ura3 leu2 his3 met15 osh2Δ::KMX osh3Δ::KMX SLA1−HA::HISMX This study SCMIG1509 Mata ura3 leu2 his3 met15 rtn1Δ::KMX SLA1−GFP::HISMX This study SCMIG1510 Mata his3 leu2 trp1 ura3 bar1 Sec61−mCherry::HISMX This study SCMIG1519 Mata his3 leu2 met15 ura3 OSH3−GFP::HISMX osh2Δ (ydl019c)::KMX This study SCMIG1520 Mata osh2Δ::KMX ura3 leu2 his3 met15 Sec61−mCherry::HISMX This study SCMIG1521 Mata his3 leu2 met15 ura3 OSH2−HA::HISMX OSH3−HA::HISMX This study SCMIG1522 Mata his3 leu2 met15 ura3 osh2Δ::KMX osh3Δ::KMX Sec61−mCherry::HISMX rtn1Δ::U RA3 This study SCMIG1523 Mata his3 leu2 met15 ura3 Osh3.ORDΔ-HA:HISMX This study SCMIG1524 Mata osh1Δ::KMX ura3 leu2 his3 met15 Euroscarf SCMIG372 Mata osh3Δ::KMX ura3 leu2 his3 met15 Euroscarf SCMIG374 Mata osh4Δ::KMX ura3 leu2 his3 met15 Euroscarf SCMIG375 Mata his3 leu2 ura3 myo3Δ::KMX myo5Δ::KMX PMYO5::URA3 This study SCMIG1525 Mata his3 leu2 met15 ura3 OSH3−HA::HISMX This study SCMIG1526 Mata his3 leu2 met15 ura3 LAS17−GFP::HISMX This study SCMIG1527 Recombinant DNA CEN4 LEU2 SLA2−3HA Idrissi et al., 2008 p111SLA2−3HA CEN4 LEU2 TRP1 VRP1−3HA Geli et al., 2000 p111VPR1−3HA CEN4 LEU2 HIS3 SLA1−YFP This study p111SLA1−YFP CEN4 LEU2 HIS5 SLA2−YFP This study P111SLA2−YFP CEN4 URA3 KMXr ABP1−CFP This study p33ABP1−CFP CEN4 URA3 CFP−ABP1 This study p33CFP−ABP1 CEN4 URA3 HIS3 ABP1−mCherry This study p33ABP1−mCherry CEN4 LEU2 HIS3 ABP1−mCherry This study p33ABP1−mCherry ura3Δ::LEU2 CEN4 URA3 ProteinA−MYO5 Grosshans et al., 2006 p33ProteinA−MYO5 CEN4 LEU2 GFP−MYO5 Grotsch et al., 2010 p33GFP−MYO5 ura3Δ::LEU2 2μ URA3 MYC−MYO5 Fernández-Golbano et al., 2014 p195MYC−MYO5 2μ HIS3 LexA−myo5(aa757−1219) Geli et al., 2000 pLexA−myo5−tail 2μ HIS3 LexA−myo5(aa984−1219) Grotsch et al., 2010 pLexA−myo5−TH2.SH3.CA 2μ HIS3 LexA−myo5(aa1085−1219) Grotsch et al., 2010 pLexA−myo5−SH3.CA 2μ HIS3 LexA−myo5(aa1085−1219) Grotsch et al., 2010 pLexA−myo5−SH3.CA-W1123S 2μ HIS3 LexA−myo5(aa1142−1219) Grotsch et al., 2010 pLexA−myo5−CA 2μ HIS3 LexA−myo5(aa757−1181) Grotsch et al., 2010 pLexA−myo5−TH1.TH2.SH3 2μ HIS3 LexA−myo5(aa757−1091) Grotsch et al., 2010 pLexA−myo5−TH1.TH2 2μ HIS3 LexA−myo5(aa757−996) Grotsch et al., 2010 pLexA−myo5−TH1 2μ HIS3 LexA−myo5(aa984−1091) Grotsch et al., 2010 pLexA−myo5−TH2 2μ HIS3 LexA−myo5(aa1085−1181) Grotsch et al., 2010 pLexA−myo5−SH3 2μ HIS3 LexA−BICOID (D. melanogaster) Gyuris et al., 1993 pLexA−BICOID 2μ TRP1 B42−HA−VPR1 Geli et al., 2000 pB42−VRP1 2μ TRP1 B42−osh1(aa1−860) This study pB42−osh1−N 2μ TRP1 B42−osh2(aa1−860) This study pB42−osh2−N REAGENT or RESOURCE SOURCE IDENTIFIER 2μ TRP1 B42−osh2(aa1−860)−PPPVP::AAAA This study pB42−osh2−N−PPPVP∗ 2μ TRP1 B42−osh3(aa1−604) This study pB42−osh3−N 2μ TRP1 B42−OSH6 This study PJG4−5OSH6 2μ TRP1 B42−OSH7 This study PJG4−5OSH7 2μ TRP1 B42−ORD−osh2(aa840−1240) This study pJG4−5osh2−C 2μ TRP1 B42−ORD−osh3(aa605−997) This study pJG4−5osh3−C CEN4 LEU2 OSH2 This study p111OSH2 CEN4 LEU2 osh2−PPPVP-AAAVA This study p111osh2−PPPVP CEN4 LEU2 osh2−FYD-AAA This study p111osh2−FFAT CEN4 LEU2 HIS3 OSH2−HA This study P111OSH2−HA CEN4 LEU2 HIS3 osh2−ORDΔ-HA(aa1−897) This study p111osh2−ORDΔ−HA CEN4 LEU2 HIS3 osh2−PPPVP-AAAVA-HA This study p111osh2−PPPVP−HA CEN4 LEU2 HIS3 osh2−FYD-AAA-HA This study p111osh2−FFAT−HA CEN4 LEU2 HIS3 OSH2GFPThis study p111osh2−GFP CEN4 LEU2 HIS3 osh2−ORDΔ-GFP(aa1−897) This study p111osh2−ORDΔGFPCEN4 URA3 MYO5 Geli and Riezman, 1996 p33MYO5 CEN4 URA3 MYO5−mCherry Fernández-Golbano et al., 2014 p33MYO5−mCherry CEN4 URA3 ProteinA−LAS17 This study p33ProteinA−LAS17 CEN4 LEU2 HIS5 OSH2−YFP This study p111OSH2−YFP CEN4 LEU2 osh2−ORD4(aa1−880 Osh2, aa118−434 Osh4) This study p111osh2−ORD4 CEN4 LEU2 osh2−ORD6(aa1−916 Osh2, aa71−448 Osh6) This study p111osh2−ORD6 CEN4 LEU2 HIS5 osh2−ORD4−HA(aa1−880 Osh2, aa118−434 Osh4) This study p111osh2−ORD4−HA CEN4 LEU2 HIS5 osh2−ORD6−HA(aa1−916 Osh2, aa71−448 Osh6) This study p111osh2−ORD6−HA CEN4 URA3 myo5−W1123S This study p33myo5−W1123S CEN4 URA3 GFP−myo5−W1123S This study p33GFP−myo5−W1123S 2μ TRP1 osh3−GOLD(aa1−218) This study pJG4−5.osh3GOLD2μ TRP1 osh3−PH(aa219−354) This study pJG4−5.osh3−PH 2μ TRP1 osh3−L(aa335−603) This study pJG4−5.osh3−L 2μ TRP1 osh3−PH−L(aa219−603) This study pJG4−5.osh3−PH−L GST−osh3−PH(aa219−235) This study PGST−osh3−PH GST−osh2−L(PPPVP−AAAVA)(aa219−603) This study PGST−osh2−L(PPPVP∗) CEN4 URA3 MYO5−LAS17 This study p33MYO5−LAS17 CEN4 URA3 myo5−W1123S−LAS17 This study p33myo5−W1123S−LAS17 2μ URA3 myo5−W1123S This study p195−myo5−W1123S CEN4 URA3 HIS3 MYO5−LAS17GFPThis study P33MYO5−LAS17GFPCEN4 URA3 TRP1 myo5−W1123S−LAS17GFPThis study p33myo5−W1123S−LAS17GFP- Fig. 1. Endocytic Invaginations Associate with cER Rims (A) Sequential recruitment of endocytic proteins in yeast. Mammalian homologs are in parentheses. (B) Electron micrographs of yeast ultrathin sections showing endocytic invaginations decorated with Immunogold particles labeling Ede1-HA, either not associated with (no ER) or sheltered by cER (ER sh); associated with one (ER 1r) or two cER rims (ER 2r). Arrowheads indicate cER. Scale bar, 100 nm. (C) Average ±SEM frequency of short (IL<50 nm), medium (50 nm≤IL<100 nm), and long (IL≥100 nm) invaginations, associated with cER rims (upper panel) or flanked by cER (lower panel). Student's t test p values are indicated. Significant differences are in red. (D) Relative frequency of short (IL<50 nm), medium (50 nm≤IL<100 nm), and long (IL≥100 nm) invaginations labeled for the indicated proteins, either sheltered by cER (ER sh) or associated with one or two cER rims (ER 1r or ER 2r). At least 50 invaginations were analyzed per protein. See also Figure S1. Fig. 2. cER Contact with Endocytic Sites Occurs at the Onset of Actin Polymerization (A) Representative confocal fluorescence micrographs of WT or rtn1Δ cells expressing Sec61-mCherry and either Ede1-GFP, Sla1-GFP, or Abp1-GFP. The individual channels and the merge are shown. In the merge, Sec61-mCherry is red and GFP fusion is green. Arrowheads indicate endocytic patches: 1, associated with a cER rim; 2, assembled at a cER-free region; 3, assembled at a cER-sheltered region. Scale bars, 1μm. (B) Average ±SEM frequency of Ede1-GFP, Sla1-GFP, or Abp1-GFP patches associated with either cER-free (Fr) or sheltered (Sh) regions, or overlapping with cER rims (Ri). Student's t test p values are indicated for the cER rim association. Significant differences are in red. At least 150 patches in at least 15 cells were analyzed per 3 experiments. (C) 3D projections of deconvoluted confocal images of yeast expressing Abp1-GFP (green) and Sec61-mCherry (red). The merge is shown. See also Figure S2; Movies S1 and S2. Fig. 3. Disruption of cER Contact with Endocytic Sites Impairs Actin Polymerization and Scission (A) Maximum-intensity projections of confocal movies of WT or rtn1Δ cells expressing Sec61-mCherry and Sla1-GFP. A single focal plane was captured every 2 s for 2 min. Images of individual channels and the merge are shown. In the merge, Sec61-mCherry is red and Sla1-GFP is green. The square insets show an Sla1-GFP "tail" (undergoing inward movement) (WT) or a "spot" (immobile) (rtn1Δ). Arrows indicate Sla1-GFP patches assembled at cER free regions, which did not move inwards. Scale bar, 1μm. (B) Average ±SEM frequencies of normal, flat or v-shape Sla1-GFP kymographs (exemplified above) in WT cells or in cER-free (Fr) or -sheltered (Sh) regions, or overlapping with cER rims (Ri) in rtn1Δ cells. Student's t test p values are indicated for patches showing WT dynamics. Very significant differences are in red. At least 90 patches in at least 12 cells were analyzed per 3 experiments. (C) Time-lapse confocal fluorescence micrographs of cropped endocytic patches from WT and rtn1Δ cells expressing Sec61-mCherry, CFP-Abp1, and Sla2-YFP, either associated with cER rims or assembled in cER-sheltered or -free regions. Images of individual channels and the merge are shown. In the merge, Sec61-mCherry, CFP-Abp1, and Sla2-YFP are shown in red, cyan, and yellow, respectively. Enlarged images of the frames indicated with ′, ′′, and ′′′ are shown on top. (D) Electron micrographs of WT and rtn1Δ ultrathin sections of endocytic invaginations decorated with Immunogold particles labeling Sla1-HA. Scale bar, 100 nm. (E) Frequency of short (IL<50 nm), medium (50 nm≤IL<100 nm), and long (IL≥100 nm) invaginations in WT and rtn1Δ yeast for the complete population (left) or those sheltered by cER (right). At least 70 invaginations were analyzed per strain. (F) Average ±SEM invagination length (IL) of either the complete population of invaginations (left) or those sheltered by ER (right). Student's t test p values are indicated. Significant differences are in red. See also Figure S3 and Movie S3. Fig. 4. Myo5 Interacts with Osh2 and Osh3 (A) Left: diagram illustrating the parameters used to analyze the position of the cER and the Immunogold particles decorating endocytic proteins, relative to the associated invagination. ERRP, ER relative position; ERDPM, ER distance to the PM; ERDI, ER distance to the invagination; IL, invagination length; GRP, gold relative position; GDPM, gold distance to the PM. Right: graph representing the ERDPM versus the IL for 120 endocytic invaginations. The x axis represents the level of the PM. Red lines converging at the x-y axes crossing point define the same relative position (RP). 1.00 corresponds to the invagination tip (IT). (B) Left: graph representing the average ±SEM of the ERRP and GRPs for different endocytic proteins described in Idrissi et al. (2012). Student's t test p values between the ERRP and GRPs are indicated. Very significant differences are in red. n>120 gold particles per endocytic protein. Right: scheme representing the centroids of the positions for the ER rims and Myo5 and Sla1 Immunogold labeling, for an invagination of 150 nm in length. The centroid was calculated using the average ±SEM of the ERRP and ERDI and the parameters described in Idrissi et al. (2012). (C) Electron micrographs showing PM invaginations associated with cER rims, labeled with Immunogold particles decorating Myo5-HA. Scale bar, 100 nm. (D) Domain organization of the indicated ORPs and Myo5. See text for further details. (E) Myo5 versus ORP two-hybrid assays. Patches of yeast bearing the β-galactosidase reporter (pSH18-34) and plasmids expressing the indicated ORP fragments, fused to LexA, and the indicated MYO5 constructs fused to the B42 transcription activator, grown on X-Gal-containing plates. Blue indicates interaction. VRP1 was used as control for a positive MYO5 interaction and LexA alone or B42 fused to D. melanogaster BICOID as negative controls. (F) Immunoblots of immunoglobulin G-Sepharose (IP (PA)) or anti-HA agarose (IP (HA)) pulldowns of protein extracts from yeast expressing the indicated Protein A (PA) or HA-tagged proteins (+), or the corresponding non-tagged versions (−), probed with either peroxidase-conjugated anti-HA antibody (α−HA) or peroxidase anti-peroxidase complex (PAP), for detection of PA. Ten micrograms of total protein was loaded as input. See also Figures S4A and S4B. (G) Confocal fluorescence micrographs of yeast expressing either GFP-Myo5 and Osh2mCherry (upper panels) or Myo5-mCherry and Osh3-GFP (lower panels). Images of individual channels and the merge are shown. In the merge, Osh2-mCherry and Myo5-GFP are red and green, respectively. Magnified insets of adjacent patches are shown (right). Scale bar, 1μm. See also Figure S4 and Movie S4. Fig. 5. Osh2 and Osh3 Link Myo5 to Scs2 (A) Proposed link between the endocytic sites and the cER. (B and C) Immunoblots of anti-HA agarose precipitates from yeast expressing PA-Myo5 or Scs2-HA (+) or the corresponding nontagged versions (−), in cells either expressing (+) or not expressing (−) Osh2 and/or Osh3 (B) or the indicated WT or mutant Osh2 expressed in a osh2Δ strain, from centromeric plasmids under their own promoter (C), probed with either peroxidase-conjugated anti-HA antibody (α−HA) or PAP. The graphs represent the average ±SEM percentage of PA-Myo5 pulled down with Scs2−HA, normalized to the corresponding WT, from at least 3 independent experiments. Student's t test p values are indicated. Significant differences are in red. (D) Confocal fluorescence micrographs of yeast expressing Sec61-mCherry, Osh2-YFP, and Abp1-CFP. Images of individual channels and the merge are shown. In the merge, Sec61mCherry, Osh2-YFP, and Abp1-CFP are red, yellow, and blue, respectively. Magnified insets and the corresponding 3D intensity plots of adjacent patches are shown (bottom). Arrowheads indicate partially overlapping Sec61-mCherry, Osh2-YFP, and Abp1-CFP signals. Scale bar, 1μm. (E) Electron micrographs showing PM invaginations associated with cER rims, labeled with Immunogold particles decorating Osh2−HA and Osh3−HA. Scale bar, 100 nm. (F) Graph representing the Osh2−HA and Osh3−HA gold distance to the PM (GDPM) versus the length of the associated invagination (IL) for 36 invaginations associated with ER rims. (G) Left: graph representing the average ±SEM of the ER versus gold relative positions (ERRP versus GRP) for the indicated endocytic proteins. Data for Sla1 and Myo5 were described in Idrissi et al. (2012). Student's t test p values are indicated. Significant differences are in red. Right: scheme representing the centroids of the positions for the ER rims and Immunogold particles labeling Osh2/3, Myo5, and Sla1 for an invagination of 150 nm in length. The centroid was calculated using the average ±SEM ERRP and ERDI for the cER rim, the GRP, and the GDI for the ORPs and the analogous parameters described in Idrissi et al. (2012) for Myo5 and Sla1. (H) Kymographs of time-lapse confocal fluorescence movies from the cortex of rtn1Δ, osh2Δrtn1Δ, or osh2Δosh3Δrtn1Δ cells expressing Sec61-mCherry and Sla1-GFP or Sla1-YFP, showing the assembly and dynamics of several Sla1 patches (green), relative to the ER (red). Arrowheads indicate immobile Sla1-GFP cortical patches. See also Figure S5E. (I) Average ±SEM frequency of Sla1 patches associated with cER rims, cER-sheltered, or cER-free cortical regions in the indicated cells. (J) Average ±SEM frequency of WT, v-shape, or flat Sla1 patches at cER rims in the indicated cells for 3 experiments, where at least 200 patches in at least 60 cells were analyzed per strain. Student's t test p values are indicated. Significant differences are in red. See also Figure S5. Figure S5, related to Figure 5. (A and B) Immunoblots of anti-HA agarose precipitates (IP (HA)) from membrane fractions of yeast expressing the indicated PA, GFP or HA-tagged proteins (+) or the corresponding non-tagged versions (-), either probed with a peroxidase-conjugated anti-HA antibody (α-HA), with PAP for detection of PA, or with a rabbit serum against Gas1 or a mouse antiGFP antibody and the adequate peroxidase-conjugated anti-IgG. 10 µg of total protein were loaded to control protein expression in the inputs. (C) Time-lapse confocal fluorescence micrographs of cropped endocytic patches from WT cells expressing Sec61mCherry, Abp1-CFP and Osh2-YFP. Images of individual channels and the merge are shown. In the merge, Sec61-mCherry, Abp1-CFP and Osh2-YFP are shown in red, cyan and yellow, respectively. (D) Plots of the average ± SEM relative fluorescent intensities of Osh2-YFP and Abp1-CFP against time of double color wide-field time lapse fluorescence movies of cortical patches of WT yeast expressing the indicated markers. N=3. (E) Fluorescence micrographs (lower panels) and kymographs of the indicated regions (upper panels) of life-cell confocal fluorescence movies of rtn1∆, rtn1∆ osh2∆ or rtn1∆ osh2∆ osh3∆ cells expressing Sla1-GFP or YFP and Sec61mCherry. The merge of individual chanels are shown. Sec61-mCherry is depicted in red and Sla1-GFP or YFP is depicted in green. Images were taken every 2 seconds. Bar=1µm Figure S6, related to Figure 6. (A) Kymographs of life-cell confocal fluorescence movies of Sla1-GFP cortical patches in the indicated strains. Images were taken every 2 seconds. (B) Average life span ± SEM of cortical Sla1-GFP patches for strains in A. At least 50 patches in at least 10 cells were analyzed per strain. (C) Frequency of normal, v-shape and flat Sla1-GFP cortical patches in the indicated strains. The average frequency ± SEM of 3 independent experiments where a minimum of 100 patches in at least 12 cells were recorded, is shown. Figure S7, related to Figure 7. (A) Immunoblot (upper panel) and Ponceau read staining (lower panel) of immunoblots of protein extracts from osh2∆ strains bearing centromeric plasmids with the indicated WT or mutant Osh2. 20 µg of total protein were separated by SDS-PAGE, transferred to nitrocellulose and stained with Ponceau red and subsequently decorated with peroxidase-conjugated anti-HA antibody. (B) Fluorescence micrographs of myo5∆ osh2∆ strains expressing Myo5-mCherry and the indicated Osh2-ORD∆ C-terminally tagged with GFP, expressed from single copy plasmids under their own promotors. Images using the individual channels and the merge are shown. In the merge, Osh2 proteins are in green and Myo5 is in red. Bar = 1 µm. (C) Widefield fluorescence micrographs of the indicated cells overexpressing Myo5 or Myo5-W1123S and expressing GFP-Myo5, stained with 5 µg/ml of filipin to visualize PM sterols. In the merge, filipin is shown in red and GFP-Myo5 in green. Arrows indicate the position of Myo5 accumulations. (D) Kymographs of time-lapse wide field fluorescence movies of Sla1-YFP cortical patches in the a myo3∆ myo5∆ strain expressing the indicated WT or mutant Myo5 or Myo5-Las17 chimeras. Images were taken every 2 seconds. E) Average ± SEM lifespan of Sla1-YFP cortical patches for the strains in C. A minimum of 50 patches in at least 15 cells were analyzed per protein and strain. (F) Immunoblot (upper panel) and Ponceau (lower panel) of immunoblots of protein extracts from bearing the indicated proteins. (G) Fluorescence micrographs of strains expressing the indicated proteins tagged with GFP. (H) Kymographs of life-cell confocal fluorescence movies of Sla1-YFP cortical patches in the indicated strains. Images were taken every 2 seconds. (I) Average life span ± SEM of cortical Sla1-YFP patches for strains in C. At least 50 patches in at least 10 cells were analyzed per strain. (J) Frequency of normal, v-shape and flat Sla1-YFP cortical patches in the indicated strains. The average ± SEM frequency of 3 independent experiments where a minimum of 100 patches in at least 12 cells were recorded, is shown. (K) Immunoblot (upper panel) and Ponceau read staining (lower panel) of immunoblots of protein extracts from osh3∆ strains bearing centromeric plasmids with the indicated WT or mutant Osh3. 20 µg of total protein were separated by SDS-PAGE and processed as in A. Oligonucleotides GGATCTAGAAGCCGCCACTAACTTTTTGTTG GATAGTGCTCGGATCCCCGGGTTAATTAA Fisher Ede1F2 Ede1 C-terminal tagging GGAAGAAGTACAAAAAGAAGACGAAATGGT CCATTAGAATTCGAGCTCGTTTAAAC Fisher Ede1R1 Ede1 C-terminal tagging AATGCTACTGCATCAAATCCGTTTGGATTCC GGATCCCCGGGTTAATTAA Fisher Sla1.F2.D Sla1 C-terminal tagging TTATCCTATAAAATCTTAAAATACATTAATG AATTCGAGCTCGTTTAAAC Fisher Sla1R1.U Sla1 C-terminal tagging TTCCCCAGCAATTATGTGTCTTTGGGCAACC GGATCCCCGGGTTAATTAA Fisher Abp1.F2.D Abp1 C-terminal tagging AATATAATAGCATGACGCTGACGTGTGATTG AATTCGAGCTCGTTTAAAC Fisher Abp1.R1.U Abp1 C-terminal tagging GTTTACTAAGAACCTCGTTCCAGGATTTTCTG ATTTGATGCGGATCCCCGGGTTAATTAA Fisher Sec61.1400D.F 2 Sec61 C-terminal tagging GCGATTTTTTTTTTCTTTGGATATTATTTTCAT TTTATATGAATTCGAGCTCGTTTAAAC Fisher Sec61.1483U.R 1 Sec61 C-terminal tagging AGGCATGCCTACTATAACCAGGATGATGATC GGATCCCCGGGTTAATTAA Fisher Sla2.F2.D Sla2 C-terminal tagging ATCTTTATATATAAAAAGTACAATTCATGAG AATTCGAGCTCGTTTAAAC Fisher Sla2.R1.U Sla2 C-terminal tagging AAACAATCGCATAACCGCACGTATATACACG CACACACCTATCAATCACAATGAGTAAAGGA GAAGAACTTTTCACTGG Fisher ABP1.- 50D.CFP Abp1-CFP Cterminal tagging TCTGCGTCGATCTCTCTCGAGTGAGTAGTAT AATCAATAGGTTCCAAAGCTTTGTATAGTTC ATCCATGCCATGTG Fisher ABP1.53U.CFP Abp1-CFP Cterminal tagging AACCAAGAATTCATGTCTTCGTGCTCTCTGC A Fisher OSH1.1D.EcoR I pB42-OSH1-N AACCAACTCGAGTTAGAAAATATCAGCACA ATCTTTAAAG Fisher OSH1.2580U.X hoI pB42-OSH1-N AACCAACCGAATTCATGTCTAGGGAAGACTT GTCCAT Fisher OSH2.1D.EcoR I pB42-OSH2-N AACCAACTCGAGTCTTTTTTGCACTGTAGTG ACAG Fisher OSH2.2580U.X hoI pB42-OSH2-N AACCAAGAATTCGTGCAAAAAAGAAAAGAA GAATATTTAC Fisher OSH2.2518D.E coRI pJG4-5OSH2-C AACCAACTCGAGTTAAAAAATGTCACCACAA TCTTTC Fisher OSH2.3840U.X hoI pJG4-5OSH2-C AACCAAGAATTCATGGAAACAATTGATATAC AAAATCG Fisher OSH3.1D.EcoR I pB42-OSH3-N AACCAACTCGAGTTATTTGGTTTTTTGATGAT GAGGG Fisher OSH3.1811U.X hoI pB42-OSH3-N AACCAACAATTGAGCGCCCAATCCTCAAC Fisher OSH3.1813D.M feI pJG4-5OSH3-C AACCAACTCGAGTCACCAGAGTTGAGAAAT ATCAGAC Fisher OSH3.2991U.X hoI pJG4-5OSH3-C AACCAAGATATCAATGGGCTCCAAAAAACT GAC Fisher OSH6.1D.EcoR V.2 pJG4-5OSH6 AACCAACTCGAGCTATTGTTTTGCTGGGTTCT G Fisher OSH6.1344U.X hoI pJG4-5OSH6 AACCAACCCGGGAATGGCTCTCAATAAACTA AAGAATATAC Fisher OSH7.1D.SmaI. 2 pJG4-5OSH7 AACCAACTCGAGCTAATTCTTTTGGATTCCA TGC Fisher OSH7.1314U.X hoI pJG4-5OSH7 CAACATATTCAATGCTACTGCATCAAATCCG TTTGGATTCCGGAT Fisher Sla1.3695D.F2 p111SLA1-YFP GTTTTAGTTATTATCCTATAAAATCTTAAAAT ACATTAAGAATTCGAGCTCGTTTAAAC Fisher Sla1.3775U.R1 p111SLA1-YFP TTCCCCAGCAATTATGTGTCTTTGGGCAACC GGATCCCCGGGTTAATTAA Fisher ABP1.F2.D p33ABP1-CFP AATATAATAGCATGACGCTGACGTGTGATTG AATTCGAGCTCGTTTAAAC Fisher ABP1.R1.U p33ABP1-CFP AAACAATCGCATAACCGCACGTATATACACG CACACACCTATCAATCACAATGAGTAAAGGA GAAGAACTTTTCACTGG Fisher ABP1.- 50D.CFP p33CFP-ABP1 TCTGCGTCGATCTCTCTCGAGTGAGTAGTAT AATCAATAGGTTCCAAAGCTTTGTATAGTTC ATCCATGCCATGTG Fisher ABP1.53U.CFP p33CFP-ABP1 CACAGAACAAAAACCTGCAGGAAACGAAGA TAAATCGAATTCCCGGGGATCCGGTGATG Fisher URA3D.D p33ABP1-mCherry GTTAGTGGCCTAGGCAGCTGGACGT GACATAATATTCATTTACGTACATATGATTT GAGTG Fisher URA3D.U p33ABP1-mCherry AACCAAGAATTCATGTCTTCGTGCTCTCTGC A Fisher OSH1.1D.EcoR I pB42-OSH2-NPPPVP* AACCAACTCGAGTTAGAAAATATCAGCACA ATCTTTAAAG Fisher OSH1.2580U.X hoI pB42-OSH2-NPPPVP* GCTCCCAAGCATGCTGCTGCCGCAGTTGCAA ATGAAACAGATAACGACAGCC Fisher OSH2.PPPVPAAAVA.D pB42-OSH2-NPPPVP* GGCTGTCGTTATCTGTTTCATTTGCAACTGCG GCAGCAGCATGCTTGGGAGC Fisher OSH2.PPPVPAAAVA.U pB42-OSH2-NPPPVP* AACCAAGAATTCGACCTATAGATAGTATGCC CAATTG Fisher OSH2.- 580D.EcoRI p111OSH2 AACCAACTCGAGCGAAGGGTTATGGTAACC G Fisher OSH2.4152U.X hoI p111OSH2 AACCAAGAATTCATGTCTTCGTGCTCTCTGC A Fisher OSH1.1D.EcoR I p111osh2-PPPVP* AACCAACTCGAGTTAGAAAATATCAGCACA ATCTTTAAAG Fisher OSH1.2580U.X hoI p111osh2-PPPVP* GCTCCCAAGCATGCTGCTGCCGCAGTTGCAA ATGAAACAGATAACGACAGCC Fisher OSH2.PPPVPAAAVA.D p111osh2-PPPVP* GGCTGTCGTTATCTGTTTCATTTGCAACTGCG GCAGCAGCATGCTTGGGAGC Fisher OSH2.PPPVPAAAVA.U p111osh2-PPPVP* GAATCTACTTGAAAAGAATCTGAGAG Fisher Osh2.820D p111osh2-FFAT* CTCTGTTAATTCAGTTACTTCATCCACCAATT CCGCGGCAGCAGCAGCCTCATCAGCATCAGA TGCTTC Fisher OSH2.FYDAAA.U p111osh2-FFAT* AAGAAGAGATCATGATTTGAAAGATTGTGGT GACATTTTTCGGATCCCCGGGTTAATTAA Fisher OSH2.3808.D.F 2 p111OSH2-HA ATACAAGTACCAGGAAAAAAGCTCGCATAA AAAAGGCGTGGAATTCGAGCTCGTTTAAAC Fisher OSH2.3891U.R 1 p111OSH2-HA AAGACTATCTATGGATAAGGACGATAGACC AAAGATTAGTCGGATCCCCGGGTTAATTAA Fisher OSH2.2651D.F 2 p111osh2-ORD-HA ATACAAGTACCAGGAAAAAAGCTCGCATAA AAAAGGCGTGGAATTCGAGCTCGTTTAAAC Fisher OSH2.3891U.R 1 p111osh2-ORD-HA AAGAAGAGATCATGATTTGAAAGATTGTGGT GACATTTTTCGGATCCCCGGGTTAATTAA Fisher OSH2.3808.D.F 2 p111osh2-PPPVP*- HA ATACAAGTACCAGGAAAAAAGCTCGCATAA AAAAGGCGTGGAATTCGAGCTCGTTTAAAC Fisher OSH2.3891U.R 1 p111osh2-PPPVP*- HA AAGAAGAGATCATGATTTGAAAGATTGTGGT GACATTTTTCGGATCCCCGGGTTAATTAA Fisher OSH2.3808.D.F 2 p111osh2-FFAT*- HA ATACAAGTACCAGGAAAAAAGCTCGCATAA AAAAGGCGTGGAATTCGAGCTCGTTTAAAC Fisher OSH2.3891U.R 1 p111osh2-FFAT*- HA AAGAAGAGATCATGATTTGAAAGATTGTGGT GACATTTTTCGGATCCCCGGGTTAATTAA Fisher OSH2.3808.D.F 2 p111osh2-GFP ATACAAGTACCAGGAAAAAAGCTCGCATAA AAAAGGCGTGGAATTCGAGCTCGTTTAAAC Fisher OSH2.3891U.R 1 p111osh2-GFP AAGACTATCTATGGATAAGGACGATAGACC AAAGATTAGTCGGATCCCCGGGTTAATTAA Fisher OSH2.2651D.F 2 p111osh2-ORD- GFP ATACAAGTACCAGGAAAAAAGCTCGCATAA AAAAGGCGTGGAATTCGAGCTCGTTTAAAC Fisher OSH2.3891U.R 1 p111osh2-ORD- GFP AACCAAGAATTCATGGAAACAATTGATATAC AAAATCG Fisher OSH3.1D.EcoR I pJG4-5.OSH3-GOLD AACCAACCCTCGAGTCATACTCTAAATAACA TATCCTGTTGAGT Fisher Osh3.654U.Xho I pJG4-5.OSH3-GOLD AACCAACCGAATTCCATGGGACAAGGTCGTT ACTTG Fisher Osh3.567D.Eco RI pJG4-5.OSH3-PH AACCAACCCTCGAGTCACGTGGGCGTGAGGT G Fisher Osh3.1062U.Xh oI pJG4-5.OSH3-PH AACCAACCGAATTCATGGCTACGACCAAAA GTGCA Fisher Osh3.1065D.Ec oRI pJG4-5.OSH3-L AACCAACTCGAGTTATTTGGTTTTTTGATGAT GAGGG Fisher OSH3.1811U.X hoI pJG4-5.OSH3-L AACCAACCGAATTCCATGGGACAAGGTCGTT ACTTG Fisher Osh3.567D.Eco RI pJG4-5.OSH3-PH-L AACCAACTCGAGTTATTTGGTTTTTTGATGAT GAGGG Fisher OSH3.1811U.X hoI pJG4-5.OSH3-PH-L AACCAACCGAATTCCATGGGACAAGGTCGTT ACTTG Fisher Osh3.567D.Eco RI pGST-OSH3-PH AACCAACCCTCGAGTCACGTGGGCGTGAGGT G Fisher Osh3.1062U.Xh oI pGST-OSH3-PH AACCAACCGAATTCCATGCACAAGAAATTGC TGGAATCACAT Fisher Osh2.2079D.Ec oRI.GST pGST-OSH2L(PPPVP) AACCAACTCGAGTCTTTTTTGCACTGTAGTG ACAG Fisher OSH2.2580U.X hoI pGST-OSH2L(PPPVP) AAGAAAAAATCATGACTTTAAAGATTGTGCT GATATTTTCCGGATCCCCGGGTTAATTAA Fisher OSH1.F2 Osh1 C-terminal tagging ATACAAATGAACGAGTGTTATTGTGACTACA TTGCACAGCGAATTCGAGCTCGTTTAAAC Fisher OSH1.R1 Osh1 C-terminal tagging GGTTGCACTCCTTATCTTGGTTTTAGGATGGT TCTACAGA CGGATCCCCGGGTTAATTAA Fisher SCS2.F2 Scs2 C-terminal tagging AATATATATTTAGAATACAGCTATATCCTCA ATCTCCCTA GAATTCGAGCTCGTTTAAAC Fisher SCS2.R1 Scs2 C-terminal tagging AAGTTAGCTATTCTTGTTTGAAATGAAAAAA AAAAAGCACGAATTCGAGCTCGTTTAAAC Fisher RTN1.928.U.R1 Rtn1 C-terminal tagging TACAAAAAACTTGCAAAATGAATTGGAAAA AAACAACGCT CGGATCCCCGGGTTAATTAA Fisher RTN1.845D.F2 Rtn1 C-terminal tagging AGTGGGAGCTCATGACGATATGGACAATGGT GATGATTGGCGGATCCCCGGGTTAATTAA Fisher Las17.d.F2 Las17 C-terminal tagging CCAATCATCACCATTGTCCATATCGTCATGA GCTCCCACTGAATTCGAGCTCGTTTAAAC Fisher Las17.u.R1 Las17 C-terminal tagging TGACGGATTGAAAGGTCTATTTCCTACAAGT TACTGTAAACGGAT Fisher BZZ1.F2 Bzz1 C-terminal tagging GCCAGGGAAAATATTTAATAGTTTCAGTTCA TTCCTTCGTGAATT Fisher BZZ1.R1 Bzz1 C-terminal tagging AAGAAAGAAGCATGATTGGTCTGATATTTCT CAACTCTGGCGGATCCCCGGGTTAATTAA Fisher OSH3.F2 Osh3 C-terminal tagging CAAAGGGGGGAAAAAAGTTCACTTGATGTC ATCAAGGCATGAATTCGAGCTCGTTTAAAC Fisher OSH3.R1 Osh3 C-terminal tagging CCTATCCAAAGAGTGTTACACGCCGTAATGA CATACCTGAAGCTGCACGGATCCCCGGGTTA ATTAA Fisher OSH3.1866D.F 2 Osh3 C-terminal tagging GAATCTACTTGAAAAGAATCTGAGAG Fisher Osh2.820D p111OSH2-ORD4 CCATCTACTACAAAAAGAGTTGCGAAACCTT TGAACCCATTTCTAGGTGAG Fisher Osh2.Osh4.D p111OSH2-ORD4 CTCACCTAGAAATGGGTTCAAAGGTTTCGCA ACTCTTTTTGTAGTAGATGG Fisher Osh2.Osh4.U p111OSH2-ORD4 CCAAAGAGAGTTGTGGGACGAAGAAAAGGA AATTGTTTTGCGGATCCCCGGGTTAATTAA Fisher Osh4.F2 p111OSH2-ORD4HA ATACAAGTACCAGGAAAAAAGCTCGCATAA AAAAGGCGTGGAATTCGAGCTCGTTTAAAC Fisher OSH2.3891U.R 1 p111OSH2-ORD4HA GAATCTACTTGAAAAGAATCTGAGAG Fisher Osh2.820D p111OSH2-ORD6 GATATGACAAGAATGACGTTACCAACTTTTA TTTTAGAAAAAAAATCGATGCTT Fisher Osh2.Osh6.D p111OSH2-ORD6 AAGCATCGATTTTTTTTCTAAAATAAAAGTT GGTAACGTCATTCTTGTCATATC Fisher Osh2.Osh6.U p111OSH2-ORD6 ATACAAGTACCAGGAAAAAAGCTCGCATAA AAAAGGCGTGGAATTCGAGCTCGTTTAAAC Fisher OSH2.3891U.R 1 p111OSH2-ORD6HA GAAGATGATTGAAAACGAAAAGCAGAACCC AGCAAAACAACGGATCCCCGGGTTAATTAA Fisher Osh6.F2 p111OSH2-ORD6HA ATACAAGTACCAGGAAAAAAGCTCGCATAA AAAAGGCGTGGAATTCGAGCTCGTTTAAAC Fisher OSH2.3891U.R 1 p111OSH2-YFP ATTGGGTTAAAAGAAGAGATCATGATTTGAA AGATTGTGGTGACATTTTTCGTACGCTGCAG GTCGACAG Fisher Osh2.3800d.f2 p111OSH2-YFP CGTTCGAACCGCATTAGTACCAGAGCTCAAG GAAGGAA Fisher OMIG34 p33MYO5-(W1123S) TAGGCTCGGCGATAGAGTACC Fisher OMIG38 p33MYO5-(W1123S) AAG AAC CAA GCG GTT CGT CTC TCG GGA AAC Fisher OMIG212 p33MYO5-(W1123S) GTT TCC CGA GAG ACG AAC CGC TTG GTT CTT Fisher OMIG213 p33MYO5-(W1123S) CCTTTTGATAATTTCCTTATCTGAAGAGTTTA GGAGTCCCATCCAATCATCTTCCTCTTCATCT TC Fisher Las17.40U.Myo 5.AllU p33MYO5-LAS17 TACAGTTCGTTACTTTAAGTGTTGATAGGCG TGATTTAATATGGCTATCTTAAAAAGAGGAG CTAG Fisher Las17.- 40D.Myo51D p33MYO5-LAS17 CCTTTTGATAATTTCCTTATCTGAAGAGTTTA GGAGTCCCATCCAATCATCTTCCTCTTCATCT TC Fisher Las17.40U.Myo 5.AllU p33MYO5- (W1123S)-LAS17 TACAGTTCGTTACTTTAAGTGTTGATAGGCG TGATTTAATATGGCTATCTTAAAAAGAGGAG CTAG Fisher Las17.- 40D.Myo51D p33MYO5- (W1123S)-LAS17 AGTGGGAGCTCATGACGATATGGACAATGGT GATGATTGGCGGATCCCCGGGTTAATTAA Fisher Las17.d.F2 p33MYO5-LAS17GFP CCAATCATCACCATTGTCCATATCGTCATGA GCTCCCACTGAATTCGAGCTCGTTTAAAC Fisher Las17.u.R1 p33MYO5-LAS17GFP AGTGGGAGCTCATGACGATATGGACAATGGT GATGATTGGCGGATCCCCGGGTTAATTAA Fisher Las17.d.F2 p33MYO5- (W1123S)-LAS17GFP CCAATCATCACCATTGTCCATATCGTCATGA GCTCCCACTGAATTCGAGCTCGTTTAAAC Fisher Las17.u.R1 p33MYO5- (W1123S)-LAS17GFP Table S1. Oligonucleotides used.