Developmental Cell

Previews with induced pulmonary hypertension, Cx43 is mislocalized (baso)laterally, as is EB1 (Chkourko et al., 2012). Now that Green and colleagues have found that desmoplakin is involved in normal Cx43 delivery to cardiac cell-cell borders (Patel et al., 2014), it will be important to explore the role of desmoplakin on Cx43 trafficking in non-AC-related heart failure as well. The study from Patel et al. (2014) provides an additional surprising and important detail. It is known that the transmembrane anchor of the desmosome, desmoglein, is a member of the cadherin family of proteins, which includes N-cadherin. However the authors found that desmoglein does not participate in EB1 stabilization of cortical microtubules or Cx43 delivery. In fact, their data suggest that desmoglein could even be a competitive inhibitor of this process. Because Ncadherin promotes Cx43 delivery (Shaw et al., 2007), it may be that desmoplakin clusters with adherens junction proteins to facilitate targeted delivery. Simply put, desmoplakin clusters with desmoglein for structural reasons, and desmoplakin could separately cluster with N-cadherin to promote microtubule capture and protein delivery. It has already been found that adherens junction-associated b-cat-

enin and p150(Glued) facilitate EB1based cortical capture ((Bellett et al., 2009; Ligon et al., 2001; Shaw et al., 2007) and Cx43 forward delivery (Shaw et al., 2007; Smyth et al., 2010). Together, all of these data indicate that desmoplakin is a new and important member of the microtubule capture complex. An objective in the field of protein forward trafficking is to elucidate the full path undertaken by membrane proteins, from their sorting within the Golgi apparatus to delivery to specific membrane subdomains. Understanding the key regulatory checkpoints along this path will provide targets to reverse inappropriate trafficking during development and disease. The intracellular forward trafficking path is likely not linear and involves other cytoskeletal components and facilitating proteins. The findings of Patel et al. (2014) introduce a new structural protein linkage to the cortical delivery machinery and bring us one step closer to understanding the directional cues that determine the complicated forward journeys of transmembrane proteins. The findings of Patel et al. (2014) also have importance to understanding arrhythmias of failing hearts. The cell biology of heart failure is entering an exciting era.

REFERENCES Asimaki, A., Tandri, H., Huang, H., Halushka, M.K., Gautam, S., Basso, C., Thiene, G., Tsatsopoulou, A., Protonotarios, N., McKenna, W.J., et al. (2009). N. Engl. J. Med. 360, 1075–1084. Asimaki, A., Kapoor, S., Plovie, E., Karin Arndt, A., Adams, E., Liu, Z., James, C.A., Judge, D.P., Calkins, H., Churko, J., et al. (2014). Sci. Transl. Med. 6, 40ra74. Bellett, G., Carter, J.M., Keynton, J., Goldspink, D., James, C., Moss, D.K., and Mogensen, M.M. (2009). Cell Motil. Cytoskeleton 66, 893–908. Chkourko, H.S., Guerrero-Serna, G., Lin, X., Darwish, N., Pohlmann, J.R., Cook, K.E., Martens, J.R., Rothenberg, E., Musa, H., and Delmar, M. (2012). Heart Rhythm 9, 1133–1140, e6. Ligon, L.A., Karki, S., Tokito, M., and Holzbaur, E.L. (2001). Nat. Cell Biol. 3, 913–917. Patel, D.M., Dubash, A.D., Kreitzer, G., and Green, K.J. (2014). J. Cell Biol. 206, 779–797. Shaw, R.M., Fay, A.J., Puthenveedu, M.A., von Zastrow, M., Jan, Y.N., and Jan, L.Y. (2007). Cell 128, 547–560. Smyth, J.W., Hong, T.T., Gao, D., Vogan, J.M., Jensen, B.C., Fong, T.S., Simpson, P.C., Stainier, D.Y., Chi, N.C., and Shaw, R.M. (2010). J. Clin. Invest. 120, 266–279. Smyth, J.W., Zhang, S.S., Sanchez, J.M., Lamouille, S., Vogan, J.M., Hesketh, G.G., Hong, T., Tomaselli, G.F., and Shaw, R.M. (2014). Traffic 15, 684–699.

Outside In: Inversion of Cell Polarity Controls Epithelial Lumen Formation George E. Davis1,* and Ondine B. Cleaver2 1Department

of Medical Pharmacology and Physiology, University of Missouri School of Medicine, Columbia, MO 65212, USA of Molecular Biology, University of Texas Southwestern Medical Center, Dallas, TX 75390, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.devcel.2014.10.011 2Department

Establishment of cell polarity is important for epithelial lumen formation, and the molecular mechanisms directing this process are only partially understood. In this issue of Developmental Cell, Bryant et al. (2014) show that disassembly, membrane translocation, and reassembly of podocalyxin complexes controls epithelial cell polarization and lumen formation in 3D matrices. Tubes are cellular structures, present within many tissues, whose lumens are lined by either epithelial or endothelial cells (ECs). Considerable work has identi-

fied key molecules and signaling events within these two cell types that are required for lumen and tube formation by controlling cell and cytoskeletal polariza-

140 Developmental Cell 31, October 27, 2014 ª2014 Elsevier Inc.

tion, as well as membrane trafficking events involved in these processes (Bryant et al., 2010; Bryant and Mostov, 2008; Datta et al., 2011; Davis et al.,

Developmental Cell

Previews 2011; Sacharidou et al., with single-lumen formation 2012). When generating between groups of epithelial lumen structures in a 3D envicells due to retention of these ronment, cells endocytose complexes at the basal membrane. Consistently, small membrane vesicles from pehairpin RNA suppression of ripheral membranes (i.e., protein phosphatase basal surface), which then 2A (PP2A), which dephostraffic via membrane transcyphorylates NHERF1, intertosis to create and expand an rupts lumen formation by apical membrane surface. interfering with the ability of Despite our advancing knowlpodocalyxin/NHERF1/ezrin edge of lumen formation, Figure 1. Podocalyxin-Dependent Polarity Inversion Controls complexes to reassemble considerably more informaEpithelial Lumen Formation during transcytosis. tion is necessary to underPodocalyxin (Podxl) and its interacting partners, NHERF1, NHERF2, and Ezrin When these events are disstand this fundamental (Ezr), control lumen formation via polarity inversion in 3D matrices. Integrin rupted by blockade of integrin cellular process, which plays signaling through a2b1 and a3b1 leads to FAK and p190A RhoGap activation and RhoA inactivation. This signal, in conjunction with PKC activation, leads to signaling, PKC activity, or critical functions in tissue disassembly of Podxl/NHERF1/Ezr complexes at the basal membrane (BM). inducing RhoA activation, development, differentiation, Membrane translocation events lead to nascent assembly of pre-apical memepithelial cell clusters were homeostasis, regeneration, brane surfaces called apical membrane initiating sites (AMIS), which then mature into apical membranes (AP) during lumen formation between multiple found to enter a state of and repair. epithelial cells. Podxl associates with NHERF2 during vesicular trafficking, front-rear polarity, wherein Now in Developmental Cell, while Podxl/NHERF1/Ezr complexes reform during apical membrane assemthey fail to remove podocaMostov and colleagues (Brybly in a manner dependent on the protein phosphatase, PP2A. lyxin from the basal memant et al., 2014) provide new brane while simultaneously insights into the molecular control of apical membrane biogenesis brane, and this is necessary for them to expressing it on the cell surface in an during epithelial morphogenesis. They reassemble at the developing apical asymmetrical manner. These clusters demonstrate that podocalyxin, an apically membrane. The authors further show lack distinct central lumens and actively expressed sialoprotein in epithelial and that disruption of these pathways leads migrate together in the direction of polarECs (Dekan et al., 1990), is an important to retention of the complexes at the basal ized podocalyxin expression, which reregulator of epithelial cell polarization surface, thereby blocking lumen forma- sembles the process of collective cell and lumen formation (Bryant et al., tion. As vesicles traffic toward the apical motility. Similar findings have been 2014). In response to a specific set of sig- surface, they reacquire both NHERF1 observed during EC tubulogenesis, in nals and membrane trafficking events, and Ezrin, which colocalize with podoca- which signaling cascades and proteins podocalyxin and associated proteins lyxin again at the epithelial apical mem- such as Rasip1, Arhgap29, and cerebral switch from a basal to an apical mem- brane (Figure 1), and the authors further cavernous malformation (CCM)-1/2 supbrane position, thereby controlling lumen show that podocalyxin is required for press RhoA during lumen formation (Davis formation (Figure 1). The authors first NHERF1 and ezrin to target apically. et al., 2011; Sacharidou et al., 2012; Xu demonstrate that small, two- to three- Thus, podocalyxin-containing complexes et al., 2011). Overall, these detailed studies provide cell clusters of MDCK cells in 3D Matrigel are necessary for development of polarlocalize podocalyxin to a basal region at ized single-lumen structures in MDCK important new insights into how apicalbasal polarization is established during the cell-extracellular matrix (ECM) inter- cysts. face. At this location, podocalyxin forms Bryant et al. next identified the molecu- morphogenic events in 3D matrices. One a complex with the PDZ scaffold protein, lar events that govern podocalyxin of the many strengths of this study is its NHERF1, and the actin-binding protein, complex disassembly and subsequent temporal analysis of signaling and protein ezrin. Podocalyxin is then removed from membrane translocation events, demon- trafficking (Bryant et al., 2014), an essenbasal membrane sites and transported strating that RhoA activity must be tial approach in the molecular analysis of within Rab11a-containing vesicles (con- suppressed during the lumen formation a complex morphogenic process that octaining a different NHERF, NHERF2) to process. They found that integrin-depen- curs over days. Many of the basic princian apical membrane position, initiating dent (i.e., a2b1 and a3b1) activation of ples and signaling interactions identified formation of a single lumen compartment focal adhesion kinase (FAK) and FAK- in this work will be relevant to other at the center of a group of polarized dependent phosphorylation of p190A morphogenic processes, including lumen epithelial cells (i.e., polarity inversion) RhoGAP led to inactivation of RhoA/Rho formation of different epithelial cell types, (Figure 1). kinase and that this is a necessary step as well as the developing endothelium Bryant et al. demonstrate that these for podocalyxin translocation. In addition, (Bryant et al., 2014; Davis et al., 2011). processes require integrin-ECM signaling PKCbII and PKCa (to a lesser extent) Interestingly, in mouse knockout studies and protein kinase C (PKC)-dependent were found to cause phosphorylation- with podocalyxin, the principle defect phosphorylation. These events result in dependent dissociation of podocalyxin/ observed was in the kidney, where a faildisassembly of podocalyxin/NHERF1/ NHERF1/ezrin complexes. Blockade of ure to produce urine (anuria) suggested Ezrin complexes within the basal mem- PKCa/b with chemical inhibitors interferes disruption of epithelial tube networks Developmental Cell 31, October 27, 2014 ª2014 Elsevier Inc. 141

Developmental Cell

Previews and/or glomerular ultrafiltration due to loss of specialized podocyte foot processes (Doyonnas et al., 2001). By contrast, endothelial tubulogenesis and lumen formation were normal in these animals (Doyonnas et al., 2001), despite the fact that podocalyxin is apically expressed in ECs (Dekan et al., 1990). Although it is clear that epithelial and endothelial tubulogenic mechanisms have similar features, these cell types are functionally different. The mechanisms controlling apical-basal polarization in these cells, while similar, also have distinct features (Datta et al., 2011; Davis et al., 2011; Sacharidou et al., 2012). In contrast to epithelial polarization, EC polarization depends on unique signals derived from blood flow and shear stress on the EC apical surface, the vascular basement membrane com-

posed of different laminin isoforms (compared to epithelial basement membranes), and the close association of mural cells along the basal surface (Davis et al., 2011; Sacharidou et al., 2012). Together, these studies demonstrate common themes in both epithelial and endothelial systems as to how molecular switches control cell polarization. Bryant et al. (2014) provide a key lesson for future studies using different cell types and model systems by demonstrating that detailed molecular studies are necessary to elucidate the underlying basis for complex morphogenic processes, including lumen and tube formation.

Bryant, D.M., Datta, A., Rodrı´guez-Fraticelli, A.E., Pera¨nen, J., Martı´n-Belmonte, F., and Mostov, K.E. (2010). Nat. Cell Biol. 12, 1035–1045. Bryant, D.M., Roignot, J., Datta, A., Overeem, A.W., Kim, M., Yu, W., Peng, X., Eastburn, D.J., Ewald, A.J., Werb, Z., and Mostov, K.E. (2014). Dev. Cell, in press. Published online October 8, 2014. Datta, A., Bryant, D.M., and Mostov, K.E. (2011). Curr. Biol. 21, R126–R136. Davis, G.E., Stratman, A.N., Sacharidou, A., and Koh, W. (2011). Int Rev Cell Mol Biol 288, 101–165. Dekan, G., Miettinen, A., Schnabel, E., and Farquhar, M.G. (1990). Am. J. Pathol. 137, 913–927. Doyonnas, R., Kershaw, D.B., Duhme, C., Merkens, H., Chelliah, S., Graf, T., and McNagny, K.M. (2001). J. Exp. Med. 194, 13–27.

REFERENCES

Sacharidou, A., Stratman, A.N., and Davis, G.E. (2012). Cells Tissues Organs (Print) 195, 122–143.

Bryant, D.M., and Mostov, K.E. (2008). Nat. Rev. Mol. Cell Biol. 9, 887–901.

Xu, K., Sacharidou, A., Fu, S., Chong, D.C., Skaug, B., Chen, Z.J., Davis, G.E., and Cleaver, O. (2011). Dev. Cell 20, 526–539.

Multigenerational Chromatin Marks: No Enzymes Need Apply William G. Kelly1,* 1Biology Department, Emory University, Atlanta, GA 30322, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.devcel.2014.10.008

Epigenetic memory stably maintains and transmits information during genome replication. Recently in Science, Gaydos et al. (2014) show that repressive chromatin marks exhibit transgenerational stability in the absence of chromatin-modifying enzymes in Caenorhabditis elegans, in contrast to work in flies suggesting that such proteins mediate stable inheritance of epigenetic modifications. Epigenetic processes drive heritable alterations in gene expression independently of changes in DNA sequences. While changes in gene expression always occur in cells in response to alterations in their environment, epigenetic heritability requires that these changes (which can be activating or repressive) are stable during and after cell division and that they no longer requires the factor(s) that initiated the change. A classic example is the maintenance of Hox gene expression patterns in fly embryos after the initiating transcription factors are no longer present. The conserved PcG enzyme E(z) or EZH2 adds methyl

groups to histone H3 on lysine 27 (H3K27me), and trimethylation of this residue (H3K27me3) is strongly correlated with gene repression in many organisms. E(z) acts in a conserved complex, called PRC2, along with Esc/EED; EED can bind H3K27me3 and recruit the complex to its own mark in a feedforward loop: PRC2 is recruited to nucleosomes with preexisting H3K27me3 to maintain or increase its regional enrichment (Margueron and Reinberg, 2011). Recent work in C. elegans from the group of Susan Strome now provides further insight into the mechanisms underlying PRC2/H3K27me3-mediated trans-

142 Developmental Cell 31, October 27, 2014 ª2014 Elsevier Inc.

mission of epigenetic memory (Gaydos et al., 2014). The contribution of modified histones to epigenetic memory has been debated because DNA replication involves massive dilution of chromatin-associated proteins and, possibly, extensive replacement of parental histones (reviewed in Campos et al., 2014). How might any histone-based information be faithfully transferred between cell divisions, let alone between generations, in the face of such histone dynamics? One possibility is that histonemodifying complexes, rather than the histone marks themselves, are anchored to target loci. This mechanism is supported