HHS Public Access Author manuscript Author Manuscript
Curr Biol. Author manuscript; available in PMC 2016 March 08. Published in final edited form as: Curr Biol. 2016 March 7; 26(5): 647–653. doi:10.1016/j.cub.2015.12.070.
ER-PM contacts define actomyosin kinetics for proper contractile ring assembly Dan Zhang1,*, Tamara Bidone2, and Dimitrios Vavylonis2,* 1Temasek
Life Sciences Laboratory, 1 Research Link, 117604 Singapore
2Department
of Physics, Lehigh University, Bethlehem, PA 18015
Author Manuscript
Summary
Author Manuscript
The cortical endoplasmic reticulum (ER), an elaborate network of tubules and cisternae [1], establishes contact sites with the plasma membrane (PM) through tethering machinery involving a set of conserved integral ER proteins [2]. The physiological consequences of forming ER-PM contacts are not fully understood. Here, we reveal a kinetic restriction role of ER-PM contacts over ring compaction process for proper actomyosin ring assembly in Schizosaccharomyces pombe (S. pombe). We show that fission yeast cells deficient in ER-PM contacts exhibit aberrant equatorial clustering of actin cables during ring assembly and are particularly susceptible to compromised actin filament crosslinking activity. Using quantitative image analyses and computer simulation, we demonstrate that ER-PM contacts function to modulate the distribution of ring components and to constrain their compaction kinetics. We propose that ER-PM contacts have evolved as important physical modulators to ensure robust ring assembly.
Results and Discussion
Author Manuscript
In S. pombe, the reticulated ER network is anchored to the intracellular cortex mainly through the VAMP-associated proteins (VAPs) Scs2 and Scs22 [3]. The anillin-like protein Mid1 binds to the medial cortex through the ER meshwork gaps, recruiting actomyosin components into a broad band of cytokinetic “nodes” after it exits the nucleus at mitotic entry [4–8]. Driven by actomyosin interactions, Mid1 nodes subsequently condense into a medially positioned ring [9–11]. A previous study has shown that the lack of ER-PM contacts in scs2Δscs22Δ cells results in a faster compaction of Mid1 nodes and artificially restoring the ER-PM association in these cells decelerates this process, implying that ERPM contacts could physically impede node condensation [3] (see also Figures 1A, S1A, and S1B). To examine if the cortical ER indeed obstructs lateral movement of Mid1 nodes during compaction, we utilized scs2Δscs22Δ cells where residual ER-PM contacts can be visualized by the artificial luminal ER marker mCherry-ADEL [4]. Expression of mCherry-ADEL did
*
Address correspondence to Dan Zhang ([email protected]) or Dimitrios Vavylonis ([email protected]). Author Contributions DZ designed and performed experiments. TB and DV designed and performed simulations. DZ, TB and DV analyzed the data. DZ wrote the first draft. All authors contributed to the manuscript.
Zhang et al.
Page 2
Author Manuscript
not alter the ring compaction kinetics (Figures 1A and S1C). We traced the trajectories of individual nodes in early mitotic scs2Δscs22Δ cells and found that nodes at the ER-free PM surface moved at higher speeds than the ones surrounded by the ER meshwork (Figures 1B, 1C, and S1D). Of note, the cortical ER in between nodes appeared to be remodeled and removed when two nodes coalesced (arrowed in Figure 1B), suggesting that efficient node compaction requires the removal of physical barriers provided by ER-PM contacts. The average node velocity in scs2Δscs22Δ cells increased to almost twice as high as the wildtype during condensation (Figures 1D and S1E). Reinforcement of the ER-PM association by the artificial tether TM-mCherry-PHOsh3 [3] in either wild-type or scs2Δscs22Δ cells reduced average node velocity during mitosis (Figures 1D and S1E). Taken together, these data showed that the cortical ER restricts the lateral mobility of mitotic Mid1 nodes through its PM contacts.
Author Manuscript Author Manuscript
Unlike in wild-type cells where newly-formed mitotic Mid1 nodes distributed almost evenly along the cortex overlaying the nucleus, nodes in scs2Δscs22Δ cells frequently accumulated towards one half of the central perimeter, albeit without an apparent change in the initial width of the Mid1 band (Figures 2A, 2B, and S2A; FMAX measures this unevenness as the maximum percentage of integrated pixel intensity along half of the cell circumference in a radial projection). In addition, these Mid1 nodes often emerged as irregular large-sized clumps (Figure S2B). The cortical ER has been shown to physically insulate the PM, thus its reticulated morphology could provide a spatial cue for allocating large peripheral complexes and confining them into regularly spaced domains [3, 4]. Consistent with this proposed function, reestablishment of ER-PM contacts resolved Mid1 aggregates and corrected its nonuniform distribution in scs2Δscs22Δ cells (Figures 2A, 2B, and S2B). Expression of the control construct TM-mCherry in these cells did not alter Mid1 distribution (Figures 2A and 2B). These data indicated that the reestablished PM-associated ER network in scs2Δscs22Δ cells maintains morphological characteristics of the wild-type and hence promotes wild-type distribution of Mid1 nodes at early mitosis.
Author Manuscript
We observed a low incidence of abnormal cytokinetic events in scs2Δscs22Δ cells including multiple-septa formation (asterisked) and one-side septum deposition (arrowed in Figure 2C). Consistently, in a minor fraction of mitotic scs2Δscs22Δ cells (~10%), Mid1 rapidly compacted into a single prominent clump (Figure 2D). Concomitant abnormal clustering of equatorial actin cables was visualized by Lifeact-mCherry (Figure 2E). Similar clustering was also seen with myosin II marked by Rlc1-GFP (Figure S2C). Notably, Lifeact-mCherry expression increased the occurrence of aberrant F-actin clusters in late mitotic scs2Δscs22Δ cells, although it did not overtly affect ring assembly in the wild-type (Figure 2F). These results suggested that scs2Δscs22Δ cells are compromised in some aspect of ring assembly and are susceptible to subtle changes in actin properties introduced by the actin marker. We wondered if nonuniform node distribution and the faster compaction could account for abnormal actin cable clustering during ring formation. We tested this hypothesis using a computational 3D model of contractile ring assembly [12] based on the “search, capture, pull and release” (SCPR) mechanism [13]. To simulate the weakened physical barrier at the cortex resulting from the loss of ER-PM contacts, we decreased the node drag coefficient parameter ζnode determining node speed v in response to myosin pulling force Fmyo (v = Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 3
Author Manuscript
Fmyo/ζnode). As the average node velocity nearly doubled in scs2Δscs22Δ cells, the value of ζnode in simulated cells was set as half of the wild-type (400 pN s/µm [13]). Initial node distribution was scored by FMAX as in experiments (Figure S3A). Larger (smaller) values of FMAX correspond to more uneven (even) node distribution. Local node clumping (CL) was simulated by placing two nodes at adjacent positions, coarsely representing the anomalous initial Mid1 aggregates in scs2Δscs22Δ cells (the average node size almost doubled; see Figure S2B). We examined the individual and collective contributions of these parameters (i.e. ζnode, FMAX and CL) to ring assembly.
Author Manuscript
Simulations reproduced formation of a huge node clump and clustered actin filaments during node condensation in scs2Δscs22Δ scenario (ζnode = 200 pN s/µm, FMAX = 0.75 and with/without CL; Figure 3A; Movies S1 and S2). In fact, further decrease in ζnode and increase in FMAX resulted in a more pronounced actin cable clustering, highly resembling experimental observations (Figure S3B; Movies S3 and S4). These parameters likely reflected physiological conditions of the most defective scs2Δscs22Δ cells, since the strength and extent of remnant ER-PM contacts vary among cells.
Author Manuscript
Simulations showed that uneven node distribution and CL contributed separately to emergence of unstable rings and high node porosity in runs with reduced drag coefficient (Figures 3B and S3C). Similar results were obtained when these two effects were combined (Figure 3B). Trends of compaction kinetics and node speed distributions in different simulated scenarios largely matched experimental statistics (Figures 3C and 3D). We also noticed that increase in drag coefficient decelerated node condensation and engendered more lagging nodes (Figures S3D and S3E), which was in agreement with our in vivo observations with cells expressing the artificial tether (Figure S3F). It further implies that suitable amounts and plasticity of ER-PM contacts have been evolved in wild-type S. pombe to provide an optimized physical impedance that keeps node condensation in check. Indeed, restoring ER-PM contacts in scs2Δscs22Δ cells, which simultaneously corrected Mid1 distribution and node compaction rate to wild-type levels, mostly rescued aberrant F-actin clustering (Figure 3E). To conclude, our simulations recapitulated the formation of node and actin clusters in cells lacking ER-PM contacts. They highlighted the importance of initial node distribution and restriction of node compaction kinetics for proper ring assembly.
Author Manuscript
Importantly, our numerical model predicted that faster node compaction kinetics (low ζnode) could challenge ring formation by narrowing the permissive range of actin crosslinking properties described by parameters spring constant kc and crosslinking threshold distance rc, which represent the concentration and persistency of bound crosslinking proteins respectively [14] (Figures 3F and S3G). For instance, a minor decrease in rc no longer enabled ring formation when ζnode was reduced (Figure 3G). Moreover, nodes organized into clumps when actin crosslinking capacity was further weakened (Figure S3H). To test this prediction, we examined ring assembly in scs2Δscs22Δ cells that lacked major actin cross-linkers. In fission yeast, α-actinin Ain1 plays a more prominent role in stabilizing actin bundles and aligning nodes during ring compaction as compared to fimbrin
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 4
Author Manuscript
Fim1 [14, 15]. Interestingly, while ain1Δ cells exhibited largely normal cell growth and division [15], ain1Δscs2Δscs22Δ cells displayed severe cytokinetic defects (Figure 4A). We did not observe similar abnormality in fim1Δscs2Δscs22Δ cells (data not shown). In line with our model-based predictions, actomyosin ring assembly was indeed impaired in ain1Δscs2Δscs22Δ cells where abnormal actin clusters were frequently observed (Figure 4B). Similar to scs2Δscs22Δ cells, Mid1 nodes condensed rapidly into a single large clump or occasionally a few disconnected clumps in ain1Δscs2Δscs22Δ cells (Figures 4C and S4A–C). Restoring ER-PM junctions in these cells largely prevented node clumping and hence corrected ring assembly defects (Figures 4D, S4B, and 4E). To avoid synthetic sickness of Lifeact-mCherry expression in scs2Δscs22Δ backgrounds, we utilized LifeactGFP to quantify ring morphologies in live cells (Figure S4D). Collectively, our data suggested that actin crosslinking activity became crucial for ring assembly in cells lacking ER-PM contacts.
Author Manuscript
Unlike scs2Δscs22Δ cells, initial node distribution and node mobility are not altered in ain1Δ cells [14]. Formation of a single Mid1 clump was observed in some ain1Δ cells but at much later stage (Figures S4A-C). The majority of Rlc1-GFP containing clumps in ain1Δ cells rearranged evenly into rings in late anaphase (9 out of 10 cells; Figure S4C). This implied the presence of a late pathway that could further promote ring assembly as proposed in other studies [14, 16]. Such ring correction mechanisms did not work efficiently in ain1Δscs2Δscs22Δ cells (6 out of 12 cells formed rings; Figure S4C), manifesting that ERPM contacts-associated modulation on proper node distribution and motility is important for robust ring formation.
Author Manuscript
The tight ER-PM junctions are physically unfavorable to the PM binding of large complexes and space-demanding cortical events, such as vesicle docking, fusion and membrane invagination [17, 18]. Extensive blockage of the PM by the ER cisternae in fission yeast mutants depleted of membrane shaping proteins results in aberrant localization of both interphase cell cycle regulator Cdr2 complex [19] and mitotic Mid1 nodes [3]. Interestingly, deficiency in ER-PM contacts barely affects the localization of Cdr2 [3] which recruits interphase Mid1 and leaves the cortex during mitosis [20, 21]. Likewise, ER-PM contacts seemed to have little effect on the slow diffusion of prior mitotic nodes (Figure S1F). Uneven distribution of mitotic Mid1 in scs2Δscs22Δ cells is likely related to its distinct cortical anchoring mechanism from interphase [22] and further enhanced by the lack of motility restriction. How ER-PM contacts limit mitotic Mid1 dynamics is still ambiguous. Besides the steric constraints, it is tempting to consider possible effects of close ER-PM contacts on membrane fluidity and lipid raft organization of the PM [2].
Author Manuscript
For ring assembly, a highly dynamic process, S. pombe appears to have evolved specific ranges of actomyosin kinetics adaptive to its physiology. Since the ER is attached to the cortex, the physical barrier effect of ER-PM contacts on the dynamics of ring components may have been incorporated as a default module for wild-type cells. Our work shows that the removal of such a physical module alters distribution and kinetics of cytokinetic nodes, resulting in defective ring assembly particularly vulnerable to fluctuations in actin crosslinking activity.
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 5
Author Manuscript
We propose that the cortical ER acts as an important physical modulator that sets actomyosin dynamics to ensure robust ring formation. The ER-PM association is prominent in plants, yeast and excitable cells in metazoans [23–26]. A similar kinetics-modulating role of the cortical ER could influence proper function of cortical dynamic networks in other cell types.
Experimental Procedures S. pombe strains and reagents S. pombe cells were grown at 24 °C for all experiments. The design of artificial ER-PM tethers TM-GFP-PHOsh3 and TM-mCherry-PHOsh3 were previously described [3]. The cell wall dye Calcofluor White was from Sigma-Aldrich (St. Louis, MO). Alexa Fluor 568Phalloidin for actin staining was from Life Technologies (Carlsbad, CA).
Author Manuscript
We modeled cytokinetic ring formation by extending prior 3D Brownian Dynamics simulations [12]. Other detailed information on strain list, microscopy, image analysis and computational modeling can be found in Supplemental Experimental Procedures.
Supplementary Material Refer to Web version on PubMed Central for supplementary material.
Acknowledgments We are grateful to S. Oliferenko, N. Naqvi and G. Jedd for suggestions on the manuscript. Many thanks are due to S. Oliferenko, M. Balasubramanian and JQ. Wu for sharing reagents used in this work. DZ was funded by the Temasek Trust and TB, DV by NIH grant R01GM098430.
Author Manuscript
References
Author Manuscript
1. De Craene JO, Coleman J, Estrada de Martin P, Pypaert M, Anderson S, Yates JR 3rd, Ferro-Novick S, Novick P. Rtn1p is involved in structuring the cortical endoplasmic reticulum. Mol Biol Cell. 2006; 17:3009–3020. [PubMed: 16624861] 2. Henne WM, Liou J, Emr SD. Molecular mechanisms of inter-organelle ER-PM contact sites. Curr Opin Cell Biol. 2015; 35:123–130. [PubMed: 26025028] 3. Zhang D, Vjestica A, Oliferenko S. Plasma membrane tethering of the cortical ER necessitates its finely reticulated architecture. Curr Biol. 2012; 22:2048–2052. [PubMed: 23041194] 4. Zhang D, Vjestica A, Oliferenko S. The cortical ER network limits the permissive zone for actomyosin ring assembly. Curr Biol. 2010; 20:1029–1034. [PubMed: 20434336] 5. Sohrmann M, Fankhauser C, Brodbeck C, Simanis V. The dmf1/mid1 gene is essential for correct positioning of the division septum in fission yeast. Genes Dev. 1996; 10:2707–2719. [PubMed: 8946912] 6. Bahler J, Steever AB, Wheatley S, Wang Y, Pringle JR, Gould KL, McCollum D. Role of polo kinase and Mid1p in determining the site of cell division in fission yeast. J Cell Biol. 1998; 143:1603–1616. [PubMed: 9852154] 7. Paoletti A, Chang F. Analysis of mid1p, a protein required for placement of the cell division site, reveals a link between the nucleus and the cell surface in fission yeast. Mol Biol Cell. 2000; 11:2757–2773. [PubMed: 10930468] 8. Wu JQ, Kuhn JR, Kovar DR, Pollard TD. Spatial and temporal pathway for assembly and constriction of the contractile ring in fission yeast cytokinesis. Dev Cell. 2003; 5:723–734. [PubMed: 14602073]
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 6
Author Manuscript Author Manuscript Author Manuscript
9. Motegi F, Nakano K, Mabuchi I. Molecular mechanism of myosin-II assembly at the division site in Schizosaccharomyces pombe. J Cell Sci. 2000; 113(Pt 10):1813–1825. [PubMed: 10769212] 10. Arai R, Mabuchi I. F-actin ring formation and the role of F-actin cables in the fission yeast Schizosaccharomyces pombe. J Cell Sci. 2002; 115:887–898. [PubMed: 11870208] 11. Wu JQ, Sirotkin V, Kovar DR, Lord M, Beltzner CC, Kuhn JR, Pollard TD. Assembly of the cytokinetic contractile ring from a broad band of nodes in fission yeast. J Cell Biol. 2006; 174:391–402. [PubMed: 16864655] 12. Bidone TC, Tang H, Vavylonis D. Dynamic network morphology and tension buildup in a 3D model of cytokinetic ring assembly. Biophys J. 2014; 107:2618–2628. [PubMed: 25468341] 13. Vavylonis D, Wu JQ, Hao S, O'Shaughnessy B, Pollard TD. Assembly mechanism of the contractile ring for cytokinesis by fission yeast. Science. 2008; 319:97–100. [PubMed: 18079366] 14. Laporte D, Ojkic N, Vavylonis D, Wu JQ. alpha-Actinin and fimbrin cooperate with myosin II to organize actomyosin bundles during contractile-ring assembly. Mol Biol Cell. 2012; 23:3094– 3110. [PubMed: 22740629] 15. Wu JQ, Bahler J, Pringle JR. Roles of a fimbrin and an alpha-actinin-like protein in fission yeast cell polarization and cytokinesis. Mol Biol Cell. 2001; 12:1061–1077. [PubMed: 11294907] 16. Huang Y, Yan H, Balasubramanian MK. Assembly of normal actomyosin rings in the absence of Mid1p and cortical nodes in fission yeast. J Cell Biol. 2008; 183:979–988. [PubMed: 19075108] 17. West M, Zurek N, Hoenger A, Voeltz GK. A 3D analysis of yeast ER structure reveals how ER domains are organized by membrane curvature. J Cell Biol. 2011; 193:333–346. [PubMed: 21502358] 18. Stradalova V, Blazikova M, Grossmann G, Opekarova M, Tanner W, Malinsky J. Distribution of cortical endoplasmic reticulum determines positioning of endocytic events in yeast plasma membrane. PLoS One. 2012; 7:e35132. [PubMed: 22496901] 19. Martin SG, Berthelot-Grosjean M. Polar gradients of the DYRK-family kinase Pom1 couple cell length with the cell cycle. Nature. 2009; 459:852–856. [PubMed: 19474792] 20. Almonacid M, Moseley JB, Janvore J, Mayeux A, Fraisier V, Nurse P, Paoletti A. Spatial control of cytokinesis by Cdr2 kinase and Mid1/anillin nuclear export. Curr Biol. 2009; 19:961–966. [PubMed: 19427212] 21. Akamatsu M, Berro J, Pu KM, Tebbs IR, Pollard TD. Cytokinetic nodes in fission yeast arise from two distinct types of nodes that merge during interphase. J Cell Biol. 2014; 204:977–988. [PubMed: 24637325] 22. Sun L, Guan R, Lee IJ, Liu Y, Chen M, Wang J, Wu JQ, Chen Z. Mechanistic insights into the anchorage of the contractile ring by anillin and Mid1. Dev Cell. 2015; 33:413–426. [PubMed: 25959226] 23. Sparkes IA, Frigerio L, Tolley N, Hawes C. The plant endoplasmic reticulum: a cell-wide web. Biochem J. 2009; 423:145–155. [PubMed: 19772494] 24. Prinz WA, Grzyb L, Veenhuis M, Kahana JA, Silver PA, Rapoport TA. Mutants affecting the structure of the cortical endoplasmic reticulum in Saccharomyces cerevisiae. J Cell Biol. 2000; 150:461–474. [PubMed: 10931860] 25. Schneider MF. Control of calcium release in functioning skeletal muscle fibers. Annu Rev Physiol. 1994; 56:463–484. [PubMed: 8010748] 26. Wu MM, Luik RM, Lewis RS. Some assembly required: constructing the elementary units of store-operated Ca2+ entry. Cell Calcium. 2007; 42:163–172. [PubMed: 17499354]
Author Manuscript Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 7
Author Manuscript Author Manuscript Author Manuscript
Figure 1.
Author Manuscript
The cortical ER restricts Mid1 node motility during ring compaction. (A) Quantification of node compaction time. Error bars, 2 × standard deviation (SD); P-values, two-tailed t-test. (B) Time-lapse spinning disk confocal images of scs2Δscs22Δ cells co-expressing Mid1GFP and mCherry-ADEL. Shown are top focal plane images from two framed regions. Right panels show 2 × magnification of the dashed frame at indicated time points. White dotted lines, the initiate position of the traced nodes. Image contrast is reported using heat map wedges. (C–D) Distributions of node velocities (mean±SD; n indicates the number of measurements). Scale bars, 1 µm. See also Figure S1.
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 8
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Figure 2.
ER-PM contacts are required for regular distribution of mitotic Mid1 nodes and proper actomyosin ring assembly. (A) Spinning disk confocal micrographs of early mitotic cells expressing indicated proteins. (B) Quantification of uneven node distribution at early mitotic cells. P-values, two-tailed t-test. (C) Epifluorescence and DIC images of Calcofluor stained cells. SI, Septation index. (D–E) Time-lapse maximum z-projection spinning disk confocal images of cells expressing indicated proteins. (F) Spinning disk confocal images of Lifeact-
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 9
Author Manuscript
mCherry and GFP-ADEL in indicated cells. Percentages of late mitotic cells with abnormal actin clusters are included. Scale bars, 5 µm. See also Figure S2.
Author Manuscript Author Manuscript Author Manuscript Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 10
Author Manuscript Author Manuscript Author Manuscript Figure 3.
Author Manuscript
3D model of cytokinetic ring assembly reproduces abnormal clustering of actin cables in cells lacking ER-PM contacts and predicts vulnerability of these cells to change of actin crosslinking properties. (A) Representative snapshots at 540–600 s in simulations with indicated parameters (actin in purple; nodes in green; ζnode = 400 pN s/µm for WT; ζnode = 200 pN s/µm for cells lacking ER-PM contacts; FMAX = 0.75 for initial uneven node distribution; CL for initial node clustering). (B) Fraction of cells with stable actin rings, defined as rings that persist through 600–700 s independent of node localization (6–22 simulations per indicated scenario). Error bars, 90% Clopper-Pearson confidence intervals. (C) Half width of broad band (2 × SD of node longitudinal positions) vs time in indicated scenarios (average of six runs). (D) Instantaneous node velocity distributions between 50
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 11
Author Manuscript
and 80 s, evaluated at 6 s intervals (n = 6 simulations). (E) Spinning disk confocal images of Lifeact-mCherry and artificial constructs in scs2Δscs22Δ cells. Percentages of late mitotic cells with abnormal actin clusters are included. Scale bars, 5 µm. (F) Node circumferential porosity vs actin crosslinking parameters rc and kc, at 500 s, with indicated ζnode (average of four simulations). Permissive ranges of rc and kc for ring formation are outlined. Dashed lines indicate that rings also form outside of the boundary with low probability. (G) Snapshots of simulations with indicated parameters. See also Figure S3.
Author Manuscript Author Manuscript Author Manuscript Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 12
Author Manuscript Author Manuscript Author Manuscript Figure 4.
Author Manuscript
Actin crosslinking activity is crucial for actomyosin ring assembly in cells lacking ER-PM contacts. (A) Epifluorescence and DIC images of Calcofluor stained cells. White asterisks, multi-septa; white arrows, one-side septum deposition; SI, Septation index. (B) Maximum zprojection epifluorescence images of Alexa Fluor 568-Phalloidin stained cells. 2 × magnification of the framed regions and the percentages of cells with abnormal actin clusters are included. (C–D) Time-lapse maximum z-projection spinning disk confocal images of cells expressing indicated proteins. (E) Spinning disk confocal images of LifeactGFP and artificial constructs in indicated cells. The percentages of late mitotic cells with abnormal actin clusters are included. Black asterisks, aberrant actin clusters. Scale bars, 5 µm. See also Figure S4.
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Curr Biol. Author manuscript; available in PMC 2016 March 08. Published in final edited form as: Curr Biol. 2016 March 7; 26(5): 647–653. doi:10.1016/j.cub.2015.12.070.
ER-PM contacts define actomyosin kinetics for proper contractile ring assembly Dan Zhang1,*, Tamara Bidone2, and Dimitrios Vavylonis2,* 1Temasek
Life Sciences Laboratory, 1 Research Link, 117604 Singapore
2Department
of Physics, Lehigh University, Bethlehem, PA 18015
Author Manuscript
Summary
Author Manuscript
The cortical endoplasmic reticulum (ER), an elaborate network of tubules and cisternae [1], establishes contact sites with the plasma membrane (PM) through tethering machinery involving a set of conserved integral ER proteins [2]. The physiological consequences of forming ER-PM contacts are not fully understood. Here, we reveal a kinetic restriction role of ER-PM contacts over ring compaction process for proper actomyosin ring assembly in Schizosaccharomyces pombe (S. pombe). We show that fission yeast cells deficient in ER-PM contacts exhibit aberrant equatorial clustering of actin cables during ring assembly and are particularly susceptible to compromised actin filament crosslinking activity. Using quantitative image analyses and computer simulation, we demonstrate that ER-PM contacts function to modulate the distribution of ring components and to constrain their compaction kinetics. We propose that ER-PM contacts have evolved as important physical modulators to ensure robust ring assembly.
Results and Discussion
Author Manuscript
In S. pombe, the reticulated ER network is anchored to the intracellular cortex mainly through the VAMP-associated proteins (VAPs) Scs2 and Scs22 [3]. The anillin-like protein Mid1 binds to the medial cortex through the ER meshwork gaps, recruiting actomyosin components into a broad band of cytokinetic “nodes” after it exits the nucleus at mitotic entry [4–8]. Driven by actomyosin interactions, Mid1 nodes subsequently condense into a medially positioned ring [9–11]. A previous study has shown that the lack of ER-PM contacts in scs2Δscs22Δ cells results in a faster compaction of Mid1 nodes and artificially restoring the ER-PM association in these cells decelerates this process, implying that ERPM contacts could physically impede node condensation [3] (see also Figures 1A, S1A, and S1B). To examine if the cortical ER indeed obstructs lateral movement of Mid1 nodes during compaction, we utilized scs2Δscs22Δ cells where residual ER-PM contacts can be visualized by the artificial luminal ER marker mCherry-ADEL [4]. Expression of mCherry-ADEL did
*
Address correspondence to Dan Zhang ([email protected]) or Dimitrios Vavylonis ([email protected]). Author Contributions DZ designed and performed experiments. TB and DV designed and performed simulations. DZ, TB and DV analyzed the data. DZ wrote the first draft. All authors contributed to the manuscript.
Zhang et al.
Page 2
Author Manuscript
not alter the ring compaction kinetics (Figures 1A and S1C). We traced the trajectories of individual nodes in early mitotic scs2Δscs22Δ cells and found that nodes at the ER-free PM surface moved at higher speeds than the ones surrounded by the ER meshwork (Figures 1B, 1C, and S1D). Of note, the cortical ER in between nodes appeared to be remodeled and removed when two nodes coalesced (arrowed in Figure 1B), suggesting that efficient node compaction requires the removal of physical barriers provided by ER-PM contacts. The average node velocity in scs2Δscs22Δ cells increased to almost twice as high as the wildtype during condensation (Figures 1D and S1E). Reinforcement of the ER-PM association by the artificial tether TM-mCherry-PHOsh3 [3] in either wild-type or scs2Δscs22Δ cells reduced average node velocity during mitosis (Figures 1D and S1E). Taken together, these data showed that the cortical ER restricts the lateral mobility of mitotic Mid1 nodes through its PM contacts.
Author Manuscript Author Manuscript
Unlike in wild-type cells where newly-formed mitotic Mid1 nodes distributed almost evenly along the cortex overlaying the nucleus, nodes in scs2Δscs22Δ cells frequently accumulated towards one half of the central perimeter, albeit without an apparent change in the initial width of the Mid1 band (Figures 2A, 2B, and S2A; FMAX measures this unevenness as the maximum percentage of integrated pixel intensity along half of the cell circumference in a radial projection). In addition, these Mid1 nodes often emerged as irregular large-sized clumps (Figure S2B). The cortical ER has been shown to physically insulate the PM, thus its reticulated morphology could provide a spatial cue for allocating large peripheral complexes and confining them into regularly spaced domains [3, 4]. Consistent with this proposed function, reestablishment of ER-PM contacts resolved Mid1 aggregates and corrected its nonuniform distribution in scs2Δscs22Δ cells (Figures 2A, 2B, and S2B). Expression of the control construct TM-mCherry in these cells did not alter Mid1 distribution (Figures 2A and 2B). These data indicated that the reestablished PM-associated ER network in scs2Δscs22Δ cells maintains morphological characteristics of the wild-type and hence promotes wild-type distribution of Mid1 nodes at early mitosis.
Author Manuscript
We observed a low incidence of abnormal cytokinetic events in scs2Δscs22Δ cells including multiple-septa formation (asterisked) and one-side septum deposition (arrowed in Figure 2C). Consistently, in a minor fraction of mitotic scs2Δscs22Δ cells (~10%), Mid1 rapidly compacted into a single prominent clump (Figure 2D). Concomitant abnormal clustering of equatorial actin cables was visualized by Lifeact-mCherry (Figure 2E). Similar clustering was also seen with myosin II marked by Rlc1-GFP (Figure S2C). Notably, Lifeact-mCherry expression increased the occurrence of aberrant F-actin clusters in late mitotic scs2Δscs22Δ cells, although it did not overtly affect ring assembly in the wild-type (Figure 2F). These results suggested that scs2Δscs22Δ cells are compromised in some aspect of ring assembly and are susceptible to subtle changes in actin properties introduced by the actin marker. We wondered if nonuniform node distribution and the faster compaction could account for abnormal actin cable clustering during ring formation. We tested this hypothesis using a computational 3D model of contractile ring assembly [12] based on the “search, capture, pull and release” (SCPR) mechanism [13]. To simulate the weakened physical barrier at the cortex resulting from the loss of ER-PM contacts, we decreased the node drag coefficient parameter ζnode determining node speed v in response to myosin pulling force Fmyo (v = Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 3
Author Manuscript
Fmyo/ζnode). As the average node velocity nearly doubled in scs2Δscs22Δ cells, the value of ζnode in simulated cells was set as half of the wild-type (400 pN s/µm [13]). Initial node distribution was scored by FMAX as in experiments (Figure S3A). Larger (smaller) values of FMAX correspond to more uneven (even) node distribution. Local node clumping (CL) was simulated by placing two nodes at adjacent positions, coarsely representing the anomalous initial Mid1 aggregates in scs2Δscs22Δ cells (the average node size almost doubled; see Figure S2B). We examined the individual and collective contributions of these parameters (i.e. ζnode, FMAX and CL) to ring assembly.
Author Manuscript
Simulations reproduced formation of a huge node clump and clustered actin filaments during node condensation in scs2Δscs22Δ scenario (ζnode = 200 pN s/µm, FMAX = 0.75 and with/without CL; Figure 3A; Movies S1 and S2). In fact, further decrease in ζnode and increase in FMAX resulted in a more pronounced actin cable clustering, highly resembling experimental observations (Figure S3B; Movies S3 and S4). These parameters likely reflected physiological conditions of the most defective scs2Δscs22Δ cells, since the strength and extent of remnant ER-PM contacts vary among cells.
Author Manuscript
Simulations showed that uneven node distribution and CL contributed separately to emergence of unstable rings and high node porosity in runs with reduced drag coefficient (Figures 3B and S3C). Similar results were obtained when these two effects were combined (Figure 3B). Trends of compaction kinetics and node speed distributions in different simulated scenarios largely matched experimental statistics (Figures 3C and 3D). We also noticed that increase in drag coefficient decelerated node condensation and engendered more lagging nodes (Figures S3D and S3E), which was in agreement with our in vivo observations with cells expressing the artificial tether (Figure S3F). It further implies that suitable amounts and plasticity of ER-PM contacts have been evolved in wild-type S. pombe to provide an optimized physical impedance that keeps node condensation in check. Indeed, restoring ER-PM contacts in scs2Δscs22Δ cells, which simultaneously corrected Mid1 distribution and node compaction rate to wild-type levels, mostly rescued aberrant F-actin clustering (Figure 3E). To conclude, our simulations recapitulated the formation of node and actin clusters in cells lacking ER-PM contacts. They highlighted the importance of initial node distribution and restriction of node compaction kinetics for proper ring assembly.
Author Manuscript
Importantly, our numerical model predicted that faster node compaction kinetics (low ζnode) could challenge ring formation by narrowing the permissive range of actin crosslinking properties described by parameters spring constant kc and crosslinking threshold distance rc, which represent the concentration and persistency of bound crosslinking proteins respectively [14] (Figures 3F and S3G). For instance, a minor decrease in rc no longer enabled ring formation when ζnode was reduced (Figure 3G). Moreover, nodes organized into clumps when actin crosslinking capacity was further weakened (Figure S3H). To test this prediction, we examined ring assembly in scs2Δscs22Δ cells that lacked major actin cross-linkers. In fission yeast, α-actinin Ain1 plays a more prominent role in stabilizing actin bundles and aligning nodes during ring compaction as compared to fimbrin
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 4
Author Manuscript
Fim1 [14, 15]. Interestingly, while ain1Δ cells exhibited largely normal cell growth and division [15], ain1Δscs2Δscs22Δ cells displayed severe cytokinetic defects (Figure 4A). We did not observe similar abnormality in fim1Δscs2Δscs22Δ cells (data not shown). In line with our model-based predictions, actomyosin ring assembly was indeed impaired in ain1Δscs2Δscs22Δ cells where abnormal actin clusters were frequently observed (Figure 4B). Similar to scs2Δscs22Δ cells, Mid1 nodes condensed rapidly into a single large clump or occasionally a few disconnected clumps in ain1Δscs2Δscs22Δ cells (Figures 4C and S4A–C). Restoring ER-PM junctions in these cells largely prevented node clumping and hence corrected ring assembly defects (Figures 4D, S4B, and 4E). To avoid synthetic sickness of Lifeact-mCherry expression in scs2Δscs22Δ backgrounds, we utilized LifeactGFP to quantify ring morphologies in live cells (Figure S4D). Collectively, our data suggested that actin crosslinking activity became crucial for ring assembly in cells lacking ER-PM contacts.
Author Manuscript
Unlike scs2Δscs22Δ cells, initial node distribution and node mobility are not altered in ain1Δ cells [14]. Formation of a single Mid1 clump was observed in some ain1Δ cells but at much later stage (Figures S4A-C). The majority of Rlc1-GFP containing clumps in ain1Δ cells rearranged evenly into rings in late anaphase (9 out of 10 cells; Figure S4C). This implied the presence of a late pathway that could further promote ring assembly as proposed in other studies [14, 16]. Such ring correction mechanisms did not work efficiently in ain1Δscs2Δscs22Δ cells (6 out of 12 cells formed rings; Figure S4C), manifesting that ERPM contacts-associated modulation on proper node distribution and motility is important for robust ring formation.
Author Manuscript
The tight ER-PM junctions are physically unfavorable to the PM binding of large complexes and space-demanding cortical events, such as vesicle docking, fusion and membrane invagination [17, 18]. Extensive blockage of the PM by the ER cisternae in fission yeast mutants depleted of membrane shaping proteins results in aberrant localization of both interphase cell cycle regulator Cdr2 complex [19] and mitotic Mid1 nodes [3]. Interestingly, deficiency in ER-PM contacts barely affects the localization of Cdr2 [3] which recruits interphase Mid1 and leaves the cortex during mitosis [20, 21]. Likewise, ER-PM contacts seemed to have little effect on the slow diffusion of prior mitotic nodes (Figure S1F). Uneven distribution of mitotic Mid1 in scs2Δscs22Δ cells is likely related to its distinct cortical anchoring mechanism from interphase [22] and further enhanced by the lack of motility restriction. How ER-PM contacts limit mitotic Mid1 dynamics is still ambiguous. Besides the steric constraints, it is tempting to consider possible effects of close ER-PM contacts on membrane fluidity and lipid raft organization of the PM [2].
Author Manuscript
For ring assembly, a highly dynamic process, S. pombe appears to have evolved specific ranges of actomyosin kinetics adaptive to its physiology. Since the ER is attached to the cortex, the physical barrier effect of ER-PM contacts on the dynamics of ring components may have been incorporated as a default module for wild-type cells. Our work shows that the removal of such a physical module alters distribution and kinetics of cytokinetic nodes, resulting in defective ring assembly particularly vulnerable to fluctuations in actin crosslinking activity.
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 5
Author Manuscript
We propose that the cortical ER acts as an important physical modulator that sets actomyosin dynamics to ensure robust ring formation. The ER-PM association is prominent in plants, yeast and excitable cells in metazoans [23–26]. A similar kinetics-modulating role of the cortical ER could influence proper function of cortical dynamic networks in other cell types.
Experimental Procedures S. pombe strains and reagents S. pombe cells were grown at 24 °C for all experiments. The design of artificial ER-PM tethers TM-GFP-PHOsh3 and TM-mCherry-PHOsh3 were previously described [3]. The cell wall dye Calcofluor White was from Sigma-Aldrich (St. Louis, MO). Alexa Fluor 568Phalloidin for actin staining was from Life Technologies (Carlsbad, CA).
Author Manuscript
We modeled cytokinetic ring formation by extending prior 3D Brownian Dynamics simulations [12]. Other detailed information on strain list, microscopy, image analysis and computational modeling can be found in Supplemental Experimental Procedures.
Supplementary Material Refer to Web version on PubMed Central for supplementary material.
Acknowledgments We are grateful to S. Oliferenko, N. Naqvi and G. Jedd for suggestions on the manuscript. Many thanks are due to S. Oliferenko, M. Balasubramanian and JQ. Wu for sharing reagents used in this work. DZ was funded by the Temasek Trust and TB, DV by NIH grant R01GM098430.
Author Manuscript
References
Author Manuscript
1. De Craene JO, Coleman J, Estrada de Martin P, Pypaert M, Anderson S, Yates JR 3rd, Ferro-Novick S, Novick P. Rtn1p is involved in structuring the cortical endoplasmic reticulum. Mol Biol Cell. 2006; 17:3009–3020. [PubMed: 16624861] 2. Henne WM, Liou J, Emr SD. Molecular mechanisms of inter-organelle ER-PM contact sites. Curr Opin Cell Biol. 2015; 35:123–130. [PubMed: 26025028] 3. Zhang D, Vjestica A, Oliferenko S. Plasma membrane tethering of the cortical ER necessitates its finely reticulated architecture. Curr Biol. 2012; 22:2048–2052. [PubMed: 23041194] 4. Zhang D, Vjestica A, Oliferenko S. The cortical ER network limits the permissive zone for actomyosin ring assembly. Curr Biol. 2010; 20:1029–1034. [PubMed: 20434336] 5. Sohrmann M, Fankhauser C, Brodbeck C, Simanis V. The dmf1/mid1 gene is essential for correct positioning of the division septum in fission yeast. Genes Dev. 1996; 10:2707–2719. [PubMed: 8946912] 6. Bahler J, Steever AB, Wheatley S, Wang Y, Pringle JR, Gould KL, McCollum D. Role of polo kinase and Mid1p in determining the site of cell division in fission yeast. J Cell Biol. 1998; 143:1603–1616. [PubMed: 9852154] 7. Paoletti A, Chang F. Analysis of mid1p, a protein required for placement of the cell division site, reveals a link between the nucleus and the cell surface in fission yeast. Mol Biol Cell. 2000; 11:2757–2773. [PubMed: 10930468] 8. Wu JQ, Kuhn JR, Kovar DR, Pollard TD. Spatial and temporal pathway for assembly and constriction of the contractile ring in fission yeast cytokinesis. Dev Cell. 2003; 5:723–734. [PubMed: 14602073]
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 6
Author Manuscript Author Manuscript Author Manuscript
9. Motegi F, Nakano K, Mabuchi I. Molecular mechanism of myosin-II assembly at the division site in Schizosaccharomyces pombe. J Cell Sci. 2000; 113(Pt 10):1813–1825. [PubMed: 10769212] 10. Arai R, Mabuchi I. F-actin ring formation and the role of F-actin cables in the fission yeast Schizosaccharomyces pombe. J Cell Sci. 2002; 115:887–898. [PubMed: 11870208] 11. Wu JQ, Sirotkin V, Kovar DR, Lord M, Beltzner CC, Kuhn JR, Pollard TD. Assembly of the cytokinetic contractile ring from a broad band of nodes in fission yeast. J Cell Biol. 2006; 174:391–402. [PubMed: 16864655] 12. Bidone TC, Tang H, Vavylonis D. Dynamic network morphology and tension buildup in a 3D model of cytokinetic ring assembly. Biophys J. 2014; 107:2618–2628. [PubMed: 25468341] 13. Vavylonis D, Wu JQ, Hao S, O'Shaughnessy B, Pollard TD. Assembly mechanism of the contractile ring for cytokinesis by fission yeast. Science. 2008; 319:97–100. [PubMed: 18079366] 14. Laporte D, Ojkic N, Vavylonis D, Wu JQ. alpha-Actinin and fimbrin cooperate with myosin II to organize actomyosin bundles during contractile-ring assembly. Mol Biol Cell. 2012; 23:3094– 3110. [PubMed: 22740629] 15. Wu JQ, Bahler J, Pringle JR. Roles of a fimbrin and an alpha-actinin-like protein in fission yeast cell polarization and cytokinesis. Mol Biol Cell. 2001; 12:1061–1077. [PubMed: 11294907] 16. Huang Y, Yan H, Balasubramanian MK. Assembly of normal actomyosin rings in the absence of Mid1p and cortical nodes in fission yeast. J Cell Biol. 2008; 183:979–988. [PubMed: 19075108] 17. West M, Zurek N, Hoenger A, Voeltz GK. A 3D analysis of yeast ER structure reveals how ER domains are organized by membrane curvature. J Cell Biol. 2011; 193:333–346. [PubMed: 21502358] 18. Stradalova V, Blazikova M, Grossmann G, Opekarova M, Tanner W, Malinsky J. Distribution of cortical endoplasmic reticulum determines positioning of endocytic events in yeast plasma membrane. PLoS One. 2012; 7:e35132. [PubMed: 22496901] 19. Martin SG, Berthelot-Grosjean M. Polar gradients of the DYRK-family kinase Pom1 couple cell length with the cell cycle. Nature. 2009; 459:852–856. [PubMed: 19474792] 20. Almonacid M, Moseley JB, Janvore J, Mayeux A, Fraisier V, Nurse P, Paoletti A. Spatial control of cytokinesis by Cdr2 kinase and Mid1/anillin nuclear export. Curr Biol. 2009; 19:961–966. [PubMed: 19427212] 21. Akamatsu M, Berro J, Pu KM, Tebbs IR, Pollard TD. Cytokinetic nodes in fission yeast arise from two distinct types of nodes that merge during interphase. J Cell Biol. 2014; 204:977–988. [PubMed: 24637325] 22. Sun L, Guan R, Lee IJ, Liu Y, Chen M, Wang J, Wu JQ, Chen Z. Mechanistic insights into the anchorage of the contractile ring by anillin and Mid1. Dev Cell. 2015; 33:413–426. [PubMed: 25959226] 23. Sparkes IA, Frigerio L, Tolley N, Hawes C. The plant endoplasmic reticulum: a cell-wide web. Biochem J. 2009; 423:145–155. [PubMed: 19772494] 24. Prinz WA, Grzyb L, Veenhuis M, Kahana JA, Silver PA, Rapoport TA. Mutants affecting the structure of the cortical endoplasmic reticulum in Saccharomyces cerevisiae. J Cell Biol. 2000; 150:461–474. [PubMed: 10931860] 25. Schneider MF. Control of calcium release in functioning skeletal muscle fibers. Annu Rev Physiol. 1994; 56:463–484. [PubMed: 8010748] 26. Wu MM, Luik RM, Lewis RS. Some assembly required: constructing the elementary units of store-operated Ca2+ entry. Cell Calcium. 2007; 42:163–172. [PubMed: 17499354]
Author Manuscript Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 7
Author Manuscript Author Manuscript Author Manuscript
Figure 1.
Author Manuscript
The cortical ER restricts Mid1 node motility during ring compaction. (A) Quantification of node compaction time. Error bars, 2 × standard deviation (SD); P-values, two-tailed t-test. (B) Time-lapse spinning disk confocal images of scs2Δscs22Δ cells co-expressing Mid1GFP and mCherry-ADEL. Shown are top focal plane images from two framed regions. Right panels show 2 × magnification of the dashed frame at indicated time points. White dotted lines, the initiate position of the traced nodes. Image contrast is reported using heat map wedges. (C–D) Distributions of node velocities (mean±SD; n indicates the number of measurements). Scale bars, 1 µm. See also Figure S1.
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 8
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Figure 2.
ER-PM contacts are required for regular distribution of mitotic Mid1 nodes and proper actomyosin ring assembly. (A) Spinning disk confocal micrographs of early mitotic cells expressing indicated proteins. (B) Quantification of uneven node distribution at early mitotic cells. P-values, two-tailed t-test. (C) Epifluorescence and DIC images of Calcofluor stained cells. SI, Septation index. (D–E) Time-lapse maximum z-projection spinning disk confocal images of cells expressing indicated proteins. (F) Spinning disk confocal images of Lifeact-
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 9
Author Manuscript
mCherry and GFP-ADEL in indicated cells. Percentages of late mitotic cells with abnormal actin clusters are included. Scale bars, 5 µm. See also Figure S2.
Author Manuscript Author Manuscript Author Manuscript Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 10
Author Manuscript Author Manuscript Author Manuscript Figure 3.
Author Manuscript
3D model of cytokinetic ring assembly reproduces abnormal clustering of actin cables in cells lacking ER-PM contacts and predicts vulnerability of these cells to change of actin crosslinking properties. (A) Representative snapshots at 540–600 s in simulations with indicated parameters (actin in purple; nodes in green; ζnode = 400 pN s/µm for WT; ζnode = 200 pN s/µm for cells lacking ER-PM contacts; FMAX = 0.75 for initial uneven node distribution; CL for initial node clustering). (B) Fraction of cells with stable actin rings, defined as rings that persist through 600–700 s independent of node localization (6–22 simulations per indicated scenario). Error bars, 90% Clopper-Pearson confidence intervals. (C) Half width of broad band (2 × SD of node longitudinal positions) vs time in indicated scenarios (average of six runs). (D) Instantaneous node velocity distributions between 50
Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 11
Author Manuscript
and 80 s, evaluated at 6 s intervals (n = 6 simulations). (E) Spinning disk confocal images of Lifeact-mCherry and artificial constructs in scs2Δscs22Δ cells. Percentages of late mitotic cells with abnormal actin clusters are included. Scale bars, 5 µm. (F) Node circumferential porosity vs actin crosslinking parameters rc and kc, at 500 s, with indicated ζnode (average of four simulations). Permissive ranges of rc and kc for ring formation are outlined. Dashed lines indicate that rings also form outside of the boundary with low probability. (G) Snapshots of simulations with indicated parameters. See also Figure S3.
Author Manuscript Author Manuscript Author Manuscript Curr Biol. Author manuscript; available in PMC 2016 March 08.
Zhang et al.
Page 12
Author Manuscript Author Manuscript Author Manuscript Figure 4.
Author Manuscript
Actin crosslinking activity is crucial for actomyosin ring assembly in cells lacking ER-PM contacts. (A) Epifluorescence and DIC images of Calcofluor stained cells. White asterisks, multi-septa; white arrows, one-side septum deposition; SI, Septation index. (B) Maximum zprojection epifluorescence images of Alexa Fluor 568-Phalloidin stained cells. 2 × magnification of the framed regions and the percentages of cells with abnormal actin clusters are included. (C–D) Time-lapse maximum z-projection spinning disk confocal images of cells expressing indicated proteins. (E) Spinning disk confocal images of LifeactGFP and artificial constructs in indicated cells. The percentages of late mitotic cells with abnormal actin clusters are included. Black asterisks, aberrant actin clusters. Scale bars, 5 µm. See also Figure S4.
Curr Biol. Author manuscript; available in PMC 2016 March 08.
