InteractiveFly: GeneBrief

bark beetle: Biological Overview | References

BIOLOGICAL OVERVIEW

In epithelia, specialized tricellular junctions (TCJs) mediate cell contacts at three-cell vertices. TCJs are fundamental to epithelial biology and disease, but only a few TCJ components are known, and how they assemble at tricellular vertices is not understood. This study describes a transmembrane protein, Anakonda (Aka: FlyBase name - Bark beatle), which localizes to TCJs and is essential for the formation of tricellular, but not bicellular, junctions in Drosophila. Loss of Aka caused epithelial barrier defects associated with irregular TCJ structure and geometry, suggesting that Aka organized cell corners. Aka is necessary and sufficient for accumulation of Gliotactin at TCJs, suggesting that Aka initiated TCJ assembly by recruiting other proteins to tricellular vertices. Aka's extracellular domain had an unusual tripartite repeat structure that might mediate self-assembly, directed by the geometry of tricellular vertices. Conversely, Aka's cytoplasmic tail is dispensable for TCJ localization. Thus, extracellular interactions, rather than TCJ-directed intracellular transport, appear to mediate TCJ assembly (Byri, 2015).

Epithelial cells are linked via intercellular junctions that provide paracellular diffusion barriers, maintenance of polarity, and cell-to-cell communication. Bicellular junctions (BCJs) make up the most abundant intercellular contacts in epithelia. They connect two neighboring plasma membranes and are organized into distinct complexes along the apical-basal axis, collectively forming the apical junctional belt. However, at certain positions along the cell perimeter, three cell corners meet and the bicellular junctional complex is disjointed. Here, specialized tricellular junctions (TCJs) connect epithelial cells. TCJs play important roles in epithelial barrier functions and cytoskeletal organization, and are preferential sites for trans-endothelial migration of neutrophils and metastatic cancer cells, as well as for the spreading of intracellular pathogens (Byri, 2015).

Due to the geometry of three-cell vertices, the sealing of the epithelium at these sites requires a dedicated junctional organization. Ultrastructural analyses using freeze-fracture electron microscopy (EM) showed that the zonula occludens, tight junctions (TJs) in vertebrates and septate junctions (SJs) in arthropods, changes characteristics when approaching a three-cell vertex. Instead of continuing parallel to the epithelial plane around the cell perimeter, the junctional strands extend basally, forming a bicellular seal along the apical-basal axis. In vertebrates, these structures are termed central sealing elements, and three such parallel TJ extensions enclose a narrow (approximately 10 nm) central canal at each TCJ. Similarly, in invertebrates the SJ strands turn by 90 degrees when approaching a tricellular corner, forming three parallel limiting strands that surround the tricellular juncture space, resembling the central sealing element in vertebrates. Within and perpendicular to the vertical juncture space are a series of diaphragms, which appear linked not only to three limiting septa, but also to the three cell corner membranes, thereby forming true tricellular contacts (Byri, 2015).

Despite the fundamental biological importance of TCJs, only few of their components are known (Furuse, 2014), and the mechanism of their localized assembly at three-cell vertices is not understood. In vertebrates, the Occludin family protein Tricellulin localizes to TCJs (Ikenouchi, 2005) and is recruited there by lipolysis-stimulated lipoprotein receptor (LSR) and related proteins (ILDR1 and ILDR2; Masuda, 2011; Higashi, 2013). Additionally, the cytoplasmic PDZ-domain-containing protein Tjp2iso3 associates with Tricellulin in Sertoli cells (Chakraborty, 2014). The Neuroligin-like transmembrane protein Gliotactin in Drosophila is the only TCJ protein characterized so far in invertebrates (Schulte, 2003). It is not clear what kind of cues direct the accumulation of these proteins at tricellular vertices, and the features of known TCJ proteins do not explain the distinct structure of three-cell contacts observed by EM. This study describes a transmembrane protein, Anakonda (Aka), which accumulates at TCJs in Drosophila epithelia. Aka can initiate TCJ assembly and may do so through its large extracellular domain, which exhibits an unusual triple-repeat structure (Byri, 2015).

Ultrastructural analyses revealed a unique junctional architecture at points in epithelia where three cells meet. Yet, only few TCJ-specific proteins are known, and how they assemble into a tightly localized complex exclusively at tricellular vertices is not understood. This study describes transmembrane protein Anakonda, which is shown to play a critical role in TCJ assembly and epithelial barrier formation. First, Aka was shown to localize to TCJs in ectodermal and endodermal epithelia, suggesting that Aka is a core TCJ component. Second, Aka was demonstrated to be specifically required for the assembly and correct geometry of tricellular, but not of bicellular SJs, and Aka was shown to localize to TCJs independently of bicellular SJs. Third, Aka is required for recruiting or maintaining Gli at tricellular vertices. Conversely, Gli is not required for TCJ localization of Aka. Fourth, Aka mis-expression causes premature accumulation of Gli at TCJs, indicating that Aka acts upstream of Gli in initiating TCJ formation (Byri, 2015).

Aka protein contains a large extracellular domain with a conserved tripartite repeat structure. This unique structure distinguishes Aka from other known TCJ proteins. Tricellulin, a four-pass transmembrane protein in vertebrates, localizes to the central sealing elements, suggesting that it participates in the specialized bicellular contacts that surround each three-cell vertex (Ikenouchi, 2005). Tricellulin is recruited to TCJs by the immunoglobulin domain transmembrane protein LSR, but how LSR localizes to TCJs is still unclear (Masuda, 2011). In Drosophila, Gli, a transmembrane protein with a cholinesterase domain lacking catalytic activity, accumulates at TCJs, but is also involved in bicellular SJ organization. Phosphorylation-dependent endocytic turnover and indirect association with Dlg are required for accumulation of Gli at TCJs. However, the mechanisms underlying TCJ-specific localization are not understood for Gli or any other TCJ protein (Byri, 2015).

The results suggest that Aka acts at an early step during TCJ assembly by recruiting other TCJ components, including Gli, to tricellular vertices. TCJ localization of Aka and Gli does not require Aka's C terminus, including the PDZ-binding motif. Similarly, Gli was found to localize to TCJs independently of its PDZ-binding motif (Schulte, 2006). These findings suggest that Aka and Gli might be targeted to TCJs through their extracellular domains, rather than through a cytoplasmic localization machinery. Because mislocalized Aka protein is not sufficient to recruit Gli to ectopic locations, Aka might not interact directly with Gli, or only do so in the context of TCJs. Consistent with this notion, we were not able to detect co-immunoprecipitation of Aka and Gli. Interestingly, overexpression of Gli (Schulte, 2006) or absence of Aka causes Gli to spread from TCJs to BCJs. Thus, Gli does not have an intrinsic propensity to localize to TCJs, but by default localizes to the apicolateral membrane domain occupied by SJs along the cell perimeter. TCJ accumulation of Gli therefore requires (1) association with the apicolateral membrane and (2) Aka-dependent recruitment of Gli specifically to TCJs. The findings show that unlike Gli, Aka does not depend on bicellular SJs for its localization to TCJs, and suggesting that Aka protein might have intrinsic properties that lead to its accumulation at TCJs (Byri, 2015).

Tricellular vertices display a unique geometry where three plasma membranes are in close proximity and at fixed angles. The exceptional curvature of the plasma membrane at these sites implies distinct physical properties compared to bicellular contacts. Intriguingly, it was found that TCJ assembly depends on Aka expression in all three cells adjoining a vertex, suggesting that extracellular interactions between Aka molecules from different cells are essential for TCJ formation. Considering the size of Aka's extracellular domain (303 kDa), it is likely that a single Aka protein spans the entire TCJ canal (25-30 nm), although shorter Aka isoforms may constitute different structures. Shorter Aka species might carry out functions that may or may not be related to the function of full-length Aka at TCJs. Alternatively, Aka fragments may reflect turnover of TCJ complexes, e.g., during junctional remodeling (Byri, 2015).

It is tempting to speculate that the three repeat regions in full-length Aka protein, with apparent similar domain organization, could make equal contacts with the three cell corners in the plane perpendicular to the central TCJ canal. Possibly, interaction with membrane components in the tricellular region occurs via the three SR domains, since such domains are found in receptors that recognize a wide range of molecular patterns, including surface proteins, carbohydrates, lipids, lipopolysaccharides, and peptidoglycans associated with pathogens or apoptotic cells. Because such three-way contacts may, for steric reasons, only be possible at vertices, they might selectively stabilize Aka complexes at these sites. Computer simulations suggest that the enrichment of Aka at three-cell vertices could theoretically be enhanced by the reduction of dimensionality in the tricellular region, which may promote stacking interactions between Aka molecules within the central TCJ canal. This idea is consistent with the regular structure of the TCJ diaphragms and the equal dihedral angles near TCJs as observed with electron microscopy. Stacking might occur with Aka molecules rotated by 120 or 240 degrees within the stack, perhaps in a helical array, depending on the contributing cell. Notably, such a scenario would explain the finding that TCJ formation requires Aka protein production by all three cells adjoining a vertex. A high priority of future studies will be to investigate the arrangement of Aka proteins at TCJs. Together, the current results on Aka localization, the geometry of three-cell vertices, and the triple-repeat structure of Aka protein suggest a mechanism of TCJ formation, which is promoted by self-assembly of Aka at tricellular contacts. Such self-assembly might additionally involve interactions with other membrane-associated or extracellular components, and could cooperate with bicellular adhesion molecules that zip up bicellular contacts. It will be interesting to test whether the geometry of tricellular vertices and the specific properties of Aka protein are sufficient to direct its accumulation to TCJs. Conversely, perturbed TCJ geometry upon loss of Aka may have long-range effects, such as the loss of cell-cell adhesion that this study observed in late-stage aka embryos. Interestingly, depletion of Tricellulin from mammalian cells affects cell shape and the F-actin network, suggesting that TCJs may in fact have organizing activity on the entire cell (Byri, 2015).

The existence of aka homologs in invertebrates and cephalochordates correlates with the presence of SJs in these groups. However, the proposed self-assembly model for TCJ formation in Drosophila may apply also to vertebrates, although the corresponding proteins remain to be discovered. A better understanding of TCJ assembly will be a key step toward elucidating how these poorly characterized cellular structures provide epithelial barrier function, while at the same time allowing the passage of migrating lymphocytes, metastatic cells, and intracellular pathogens (Byri, 2015).

Bark beetle controls epithelial morphogenesis by septate junction maturation in Drosophila

Epithelial tissues separate body compartments with different compositions. Tight junctions (TJs) in vertebrates and septate junctions (SJs) in invertebrates control the paracellular flow of molecules between these compartments. This epithelial barrier function of TJs and SJs must be stably maintained in tissue morphogenesis during cell proliferation and cell movement. This study shows that Bark beetle (Bark), a putative transmembrane scavenger receptor-like protein, is essential for the maturation but not the establishment of SJs in Drosophila. Embryos that lack bark establish functional SJs, but due to rudimentary septae formation during subsequent embryonic development, these become non-functional. Furthermore, cell adhesion is impaired at the lateral cell membrane and the core protein complexes of SJs are mis-localised, but appear to form otherwise normally in such embryos. A model is proposed in which Bark acts as a scaffold protein that mediates cell adhesion and mounting of SJ core complexes during cell rearrangement in tissue morphogenesis (Hildebrandt, 2015).

This study describes the characterisation and functional analysis of the putative scavenger receptor protein Bark during epithelial morphogenesis. Bark was found to be required for the maturation but not the establishment of SJs. Furthermore, Bark is involved in epithelial cell adhesion during SJ maturation (Hildebrandt, 2015).

The initiation and establishment of SJ formation appear to be independent of bark activity. This conclusion is based on the observation that the ultrastructural analysis of developing wild-type and bark mutant SJs is similar during stage 15 and 16. At stage 15, single septae have already formed in tracheal and epithelial cells of wild-type and bark mutant embryos, even though the typical ladder-like septae structure of later stages is not yet established. At stage 16 electron-dense material appears in SJs of bark mutant embryos as found in wild-type embryos. Thus, the ultrastructural analysis suggests a normal formation of SJs in bark mutants until stage 16. In addition, integral SJ components are also normally localised in bark mutants at stage 15 suggesting correct assembly of the SJs. Furthermore, the exocytosis of Serpentine and Vermiform into the tracheal lumen mediated by SJs during stage 16 is not affected in bark mutants, suggesting normal SJ function independent of bark. Remarkable is the delay of tracheal dorsal trunk elongation in bark mutants. Embryos in which the initial SJ formation is affected as found in mega display dorsal trunk elongation about two hours earlier than observed in bark mutants. Thus, establishment of SJs seems to be normal in bark mutants, since dorsal trunk elongation is indicative for disrupted SJs. Taken together, these results suggest that Bark does not critically participate in the initial morphogenesis and functional properties of SJs (Hildebrandt, 2015).

In contrast, subsequent SJ maturation (stage 17) strongly depends on Bark. Rhodamine-dextran injection experiments revealed that the transepithelial barrier function of SJs is compromised in bark mutants. In such embryos the wild-type ladder-like SJ structure is disrupted, only rudimentary septae are formed and cell adhesion is impaired. This phenotype is distinct from the archetypal SJ phenotype observed in mutants of the SJ core complexes, e.g., in mega or Nrg mutant embryos. In these mutants the septae are either reduced in number or are absent, while cell adhesion seems not to be affected, i.e., the uniform spacing between the plasma membranes of adjacent epithelial cells is maintained. Individual bark mutant embryos develop a great diversity of septae ranging from no detectable septae in places where septae would normally form to rudimentary septae and up to wild-type like septae. Furthermore, such mutant embryos show an erratic spacing of the epithelial plasma membranes; in extreme cases the plasma membranes detach from each other resulting in gaps between the cells. Such cell adhesion defects of bark epithelial cells have also been observed in mutants of Gliotactin, a marker for tricellular junctions, but so far in no other mutant that affects SJs (Hildebrandt, 2015).

The phenotypic differences between bark and archetypal SJ lack-of-function mutant embryos are also observed in gain-of-function experiments. Overexpression of Bark does not interfere with normal development or the barrier function of SJs and rescues the bark mutant phenotype. In contrast, overexpression of other SJ components causes mis-localisation of the components and a disruption of the barrier function. Thus, it has been proposed that SJ components are functionally interdependent. The observation that the level of Bark is not critical for bark function supports the argument that bark mediates a distinct role during SJ maturation in late embryogenesis (Hildebrandt, 2015).

The detachment of lateral cell membranes, which occasionally deteriorate and form gaps between the epithelial cells of bark mutant embryos, indicates that Bark plays also a role in epithelial cell adhesion in addition to its function in SJ integrity. As Bark represents a large transmembrane protein, it may mediate the cell-cell adhesion through homophilic interactions. This possibility was tested, and putative homophilic Bark binding was analyzed in a cell aggregation assay, but no homophilic Bark binding was detected. However, homophilic binding in the embryo might not been detected, since posttranslational modifications may not occur in the cell culture system. Such modifications could include the attachment of sugar moieties via several potential glycosylation sites noted in the extracellular Bark domain and/or binding of sugar moieties to Bark via its lectin domain. Heterophilic Bark binding with an already identified SJ component is also not very likely, since the lack of such components has no effect on cell adhesion in the region of SJs. Thus, it is speculated that Bark mediates its cell adhesion function by homophilic binding, which depends on specific posttranslational modifications, or in conjunction with an unknown interaction partner (Hildebrandt, 2015).

The distinctive feature of SJs is that they must maintain their functional properties, in particular the control of the transepithelial barrier function, while cells rearrange during tissue morphogenesis. Thus, epithelial layers must be able to simultaneously alter cell-cell-contacts, shuffle SJ protein components and establish functional SJ structures, which create distinct fluid compartments. Given these features, the questions that come to mind are: What is the molecular basis of these distinct requirements and what is the role of Bark? (Hildebrandt, 2015).

The recent finding of a stable SJ multiprotein core complex is an important step in understanding SJ protein dynamics. The SJ core complex is preassembled intracellularly before its incorporation into the SJs at the plasma membrane. Interestingly, the core complex seems to be stable even in cells actively rearranging their contacts (Oshima, 2011). FRAP experiments indicate that the core complex is also stable in bark mutants, but the morphology and function of SJs are severely affected. The mis-localisation of SJ core components along the basolateral cell membrane in stage 17 bark mutants suggest that the SJ core complexes do not properly coalesce to assemble SJs in the apicolateral membrane region. This observation is consistent with previous binding studies suggesting a direct interaction of Bark with Mega or another SJ core protein (Jaspers, 2012). Thus, it is speculated that Bark provides a scaffold-like matrix serving as a platform for the SJ core complexes, which assemble into functional SJs. The potential Bark matrix is not necessary for establishment of SJs, but becomes essential during tissue morphogenesis, i.e., during SJ maturation to ensure that SJ core complexes remain well-ordered and able to sustain the epithelial barrier function. This role of Bark becomes particularly apparent during morphogenesis of the embryonic tracheal system, which is established by extensive cell shape changes and cell-cell rearrangements. Such a possible role would explain why a lack of Bark leads not only to cell adhesion defects and SJ failure, but also to a severely mis-shapen, convoluted tracheal system (Hildebrandt, 2015).

A scenario that would then allow for cell-cell rearrangement is that Bark is regulated in a way that it may detach from defined regions within the SJs and thereby reduce both cell adhesion and SJ integrity, which results in SJ core complex release. Such opened SJ sub-regions may in turn tolerate cell-cell rearrangements, while nearby sub-regions may still contain functional septae that establish the transepithelial barrier function. Consistent with this view is the observation that Bark localises to recycling endosomes within intracellular compartments. Thus, Bark may shuttle from sites of SJ breakdown to sites of SJ assembly via the recycling endosomal pathway. At the site of SJ assembly Bark might mediate dual functions in cell adhesion and providing anchor points for the SJ core complexes (Hildebrandt, 2015).

The Transmembrane Proteins M6 and Anakonda Cooperate to Initiate Tricellular Junction Assembly in Epithelia of Drosophila

Cell vertices in epithelia comprise specialized tricellular tricellular junctions (TCJs) that seal the paracellular space between three adjoining cells. Although TCJs play fundamental roles in tissue homeostasis, pathogen defense, and in sensing tension and cell shape, how they are assembled, maintained, and remodeled is poorly understood. In Drosophila, the transmembrane proteins Anakonda (Aka) and Gliotactin (Gli) are TCJ components essential for epithelial barrier formation. Additionally, the conserved four-transmembrane-domain protein M6, the only myelin proteolipid protein (PLP) family member in Drosophila, localizes to TCJs. PLPs associate with cholesterol-rich membrane domains and induce filopodia formation and membrane curvature, and Drosophila M6 acts as a tumor suppressor, but its role in TCJ formation remained unknown. This study shows that M6 is essential for the assembly of tricellular, but not bicellular, occluding junctions, and for barrier function in embryonic epithelia. M6 and Aka localize to TCJs in a mutually dependent manner and are jointly required for TCJ localization of Gli, whereas Aka and M6 localize to TCJs independently of Gli. Aka acts instructively and is sufficient to direct M6 to cell vertices in the absence of septate junctions, while M6 is required permissively to maintain Aka at TCJs. Furthermore, M6 and Aka are mutually dependent for their accumulation in a low-mobility pool at TCJs. These findings suggest a hierarchical model for TCJ assembly, where Aka and M6 promote TCJ formation through synergistic interactions that demarcate a distinct plasma membrane microdomain at cell vertices (Wittek, 2020).

<

Palmitoylation of proteolipid protein M6 promotes tricellular junction assembly in epithelia of Drosophila

Tricellular junctions (TCJs) seal epithelial cell vertices and are essential for tissue integrity and physiology, but how TCJs are assembled and maintained is poorly understood. In Drosophila, the transmembrane proteins Anakonda (Aka, also known as Bark), Gliotactin (Gli) and M6 organize occluding TCJs. Aka and M6 localize in an interdependent manner to vertices and act jointly to localize Gli, but how these proteins interact to assemble TCJs was not previously known. This study shows that the proteolipid protein M6 physically interacts with Aka and with itself, and that M6 is palmitoylated on conserved juxta-membrane cysteine residues. This modification promotes vertex localization of M6 and binding to Aka, but not to itself, and becomes essential when TCJ protein levels are reduced. Abolishing M6 palmitoylation leads to delayed localization of M6 and Aka but does not affect the rate of TCJ growth or mobility of M6 or Aka. These findings suggest that palmitoylation-dependent recruitment of Aka by M6 promotes initiation of TCJ assembly, whereas subsequent TCJ growth relies on different mechanisms that are independent of M6 palmitoylation (Schleutker, 2024).

How specialized junctional complexes with distinct adhesive and occluding properties are built at three-cell contacts remains a fundamental open question in epithelial biology. Three TCJ-specific proteins, Aka, Gli and M6, are known in Drosophila, but how they interact to assemble TCJs was not known. This study reports insights into the early steps of TCJ assembly, which depends on interactions between the transmembrane proteins Aka and M6. It was first shown that all M6 isoforms share the elements required for vertex localization. Second, it was demonstrated that M6 is palmitoylated on a conserved cysteine cluster and that this modification is required for efficient initial localization of M6 to vertices, but not for subsequent TCJ growth and maintenance, indicating that different mechanisms control the early and late phases, respectively, of TCJ formation. Third, it was shown that Aka interacts with M6 in a palmitoylation-dependent manner, whereas M6 interacts with itself independently of palmitoylation, possibly forming homotypic clusters like the mammalian M6 homolog GPM6a (Schleutker, 2024).

These findings provide a molecular basis for the interdependent localization of Aka and M6 at TCJs, although it is not yet clear which physical or chemical features of vertices mediate localization of proteins to these sites. Aka engages in homotypic trans-interactions that depend on its expression in three adjacent cells and on its large extracellular domain, but not on its cytoplasmic domain, for accumulating at vertices. This suggests that the specific geometry of vertices or the presence of three adjacent plasma membranes act to recruit or maintain Aka localization. M6 is required for vertex localization of Aka in a permissive fashion. These findings suggest possible mechanisms underlying this role. In one scenario, M6 might stabilize Aka in the plasma membrane by preventing its internalization. Supporting this idea is the finding that in the absence of M6 Aka was barely detectable at the plasma membrane but instead distributed throughout the cytoplasm. Alternatively, association with M6 might concentrate Aka in a vertex-specific plasma membrane domain and thereby facilitate trans-interactions between Aka molecules on adjacent cells. This is reminiscent of the role of tetraspanins, which interact laterally among themselves and with other proteins to form large assemblies called tetraspanin-enriched microdomains (TEMs) or a tetraspanin web that controls the clustering and activity of TM proteins, such as EGFR or integrins (Schleutker, 2024).

S-palmitoylation was identified as a post-translational modification, which, although not essential for vertex localization of M6 or for TCJ formation, significantly enhances the efficiency and robustness of this process. Supporting this notion, palmitoylation of M6 becomes essential for TCJ formation and viability under conditions of reduced TCJ protein concentration. How could palmitoylation promote targeting of M6 to vertices? In many TM proteins, including claudins, tetraspanins and mammalian GPM6a, palmitoylation of juxtamembrane cysteine residues mediates association with cholesterol-rich membrane domains. Negative plasma membrane curvature at vertices might favor accumulation of cholesterol and could thereby attract palmitoylated proteins. Consistent with this idea, methyl-%beta;-cyclodextrin-induced depletion of cholesterol from the plasma membrane impaired the vertex localization of palmitoylated angulin-1 in cultured mammalian cells, although direct evidence for elevated cholesterol levels at cell vertices is lacking thus far (Schleutker, 2024).

Lack of M6 palmitoylation reduces, although it does not completely abolish, its interaction with Aka, resulting in delayed accumulation of M6 and Aka at vertices. How could palmitoylation affect the interaction of M6 with Aka? Structural predictions using AlphaFold and PPM3 (Predicting Protein position in Membranes) suggest that M6 protein adopts a tilted orientation in the membrane. Palmitoylation of the juxtamembrane cysteine cluster could stabilize the tilt and help to resolve a potential hydrophobic mismatch resulting from the different lengths of the four TMDs, thereby promoting a confirmation of M6 that is favorable for interaction with Aka. Palmitoylation of LRP6 has been proposed to act in this way by resolving a hydrophobic mismatch that otherwise causes the protein to be retained in the ER. Similarly, tetraspanins (e.g. CD151), require palmitoylation for interaction with other TM proteins (Schleutker, 2024).

Intriguingly, eliminating palmitoylation of M6 revealed a distinct requirement of this modification for efficient initial accumulation of M6 at vertex tips, whereas the subsequent extension of these clusters along the apical-basal vertex axis apparently relies on a different mechanism. This might include homotypic M6-M6 interactions, which was found to be independent of M6 palmitoylation and which might aid in organizing an M6-enriched plasma membrane domain, reminiscent of TEMs. Interestingly, in mammalian cells, palmitoylation of juxtamembrane cysteine residues has been proposed to play an analogous role in targeting angulin-1 to cholesterol-enriched membrane domains at vertices, where angulin-1 recruits tricellulin. However, angulin-1 vertex localization appears to be dispensable for maintaining tricellulin localization after its initial recruitment to TCJs, suggesting that tricellulin is maintained at vertices through association with other factors, for example, with claudin-based TJ strands. Analogously, in Drosophila, M6 palmitoylation is required early during TCJ formation for efficient targeting of M6 and Aka to vertices, whereas Aka is subsequently stabilized at TCJs through interaction with components of SJ limiting strands, including Gli (Esmangart de Bournonville, 2020). Thus, although TCJs of vertebrates and invertebrates comprise different sets of proteins, the current findings suggest that in both cases juxtamembrane palmitoylation contributes to targeting TM proteins to cell vertices during TCJ assembly (Schleutker, 2024).

Coordination of septate junctions assembly and completion of cytokinesis in proliferative epithelial tissues

How permeability barrier function is maintained when epithelial cells divide is largely unknown. This study has investigated how the bicellular septate junctions (BSJs) and tricellular septate junctions (TSJs) are remodeled throughout completion of cytokinesis in Drosophila epithelia. Following cytokinetic ring constriction, the midbody assembles, matures within SJs, and is displaced basally in two phases. In a first slow phase, the neighboring cells remain connected to the dividing cells by means of SJ-containing membrane protrusions pointing to the maturing midbody. Fluorescence recovery after photobleaching (FRAP) experiments revealed that SJs within the membrane protrusions correspond to the old SJs that were present prior to cytokinesis. In contrast, new SJs are assembled below the adherens junctions and spread basally to build a new belt of SJs in a manner analogous to a conveyor belt. Loss of function of a core BSJ component, the Na+/K+-ATPase pump Nervana 2 subunit, revealed that the apical-to-basal spread of BSJs drives the basal displacement of the midbody. In contrast, loss of the TSJ protein Bark beetle indicated that remodeling of TSJs is rate limiting and slowed down midbody migration. In the second phase, once the belt of SJs is assembled, the basal displacement of the midbody is accelerated and ultimately leads to abscission. This last step is temporally uncoupled from the remodeling of SJs. It is proposed that cytokinesis in epithelia involves the coordinated polarized assembly and remodeling of SJs both in the dividing cell and its neighbors to ensure the maintenance of permeability barrier integrity in proliferative epithelia (Daniel, 2018).

How permeability barrier function is maintained when epithelial cells divide is largely unknown. This study has investigated how the bicellular septate junctions (BSJs) and tricellular septate junctions (TSJs) are remodeled throughout completion of cytokinesis in Drosophila epithelia. Following cytokinetic ring constriction, the midbody assembles, matures within SJs, and is displaced basally in two phases. In a first slow phase, the neighboring cells remain connected to the dividing cells by means of SJ-containing membrane protrusions pointing to the maturing midbody. Fluorescence recovery after photobleaching (FRAP) experiments revealed that SJs within the membrane protrusions correspond to the old SJs that were present prior to cytokinesis. In contrast, new SJs are assembled below the adherens junctions and spread basally to build a new belt of SJs in a manner analogous to a conveyor belt. Loss of function of a core BSJ component, the Na+/K+-ATPase pump Nervana 2 subunit, revealed that the apical-to-basal spread of BSJs drives the basal displacement of the midbody. In contrast, loss of the TSJ protein Bark beetle indicated that remodeling of TSJs is rate limiting and slowed down midbody migration. In the second phase, once the belt of SJs is assembled, the basal displacement of the midbody is accelerated and ultimately leads to abscission. This last step is temporally uncoupled from the remodeling of SJs. It is proposed that cytokinesis in epithelia involves the coordinated polarized assembly and remodeling of SJs both in the dividing cell and its neighbors to ensure the maintenance of permeability barrier integrity in proliferative epithelia (Daniel, 2018).

Epithelial growth requires the formation of cell-cell junctions and physical separation of daughters upon cell division. This study has characterized a multicellular mechanism that coordinates completion of cytokinesis and de novo formation of SJ in both cuboidal and columnar epithelia: (1) following actomyosin ring constriction, the midbody forms just basal to the AJ within the SJ and matures into an intercellular bridge that is displaced basally. (2) Neighboring interphase cells maintain SJ contacts, the finger-like protrusions connecting to the maturing midbody of the dividing cells. At this position, the ménage à quatre formed contains pre-existing SJs and is stable. (3) Novel BSJs are assembled below the AJs and above the finger-like protrusions and spread toward the basal side, thereby driving the basal displacement of the finger-like protrusion plus midbody. Once the new BSJ is complete, the midbody is positioned outside the SJ belt and basal displacement accelerated. Overall, it is proposed that this multicellular process is to ensure the maintenance of the permeability barrier throughout cytokinesis (Daniel, 2018).

The data show that, as in embryos, newly synthesized SJ components assemble into stable protein complexes exhibiting a low rate of diffusion, endocytosis, and recycling. These stable complexes assemble below the AJs and then propagate toward the basal pole, in a manner analogous to a conveyor belt, both in interphase and during cytokinesis (Daniel, 2018).

Loss of the core SJ component Nrv2 prevents both assembly of stable BSJ complexes and basal displacement of the midbody. Thus, it is proposed that the apico-basal flux of newly assembled BSJs is the propelling force for the basal displacement of the finger-like protrusions connected to the midbody. The conveyor belt model first predicts that newly synthesized and/or slowly recycling SJ components are delivered apically in the vicinity of AJs. The model also predicts that SJ disassembly occurs at the basal rim of SJ belt. Future work will determine what regulates the polarized traffic of SJ components and the disassembly of SJs to control both the positioning and thickness of the SJ belt (Daniel, 2018).

This study sheds light on the remodeling of TSJs at cytokinesis. Bark and Gli are initially detected as puncta next to the midbody, prior to the pearl necklace distribution along the finger-like protrusions. These dotted structures could represent new TSJs, as, at this location, a new three-way contact is formed between the two daughter cells and a neighbor. However, FRAP analyses suggest that these punctae do not contain exchangeable components and are not moving laterally. Moreover, the appearing of Gli and Bark punctae in mitotic cells is faster (20 min post-anaphase) than the de novo assembly of TSJs in interphase cells (~80% recovery in 90 min). Analyses of the punctae by TEM will be necessary to test whether they are bona fide TSJs or components in the process of assembly. Alternatively, the old TSJs between dividing cells and neighbors prior to mitosis could undergo a change in distribution from a uniform distribution into clusters, giving rise to the pearl necklace. Redistribution could represent an intermediate step in TSJ disassembly and may explain the fast kinetics of assembly of pearl necklace structure. Future work using, for example, photoconvertible probes would help address this issue (Daniel, 2018).

Finally, loss of Bark resulted in acceleration of the midbody basal displacement. The de novo assembly of TSJs in the pearl necklace structure could be the rate-limiting step that imposes the slower speed of basal displacement. Alternatively, if pearl necklace represents the remodeling of prior TSJ connections with neighboring cells, TSJ disassembly may be the rate-limiting step. Regardless, the data argue that the TSJ components present in the pearl necklace slow down the midbody basal displacement driven by the BSJ-mediated conveyor belt (Daniel, 2018).

This study reveals differences in midbody maturation and organization of the intercellular bridge between isolated and epithelial cells. In epithelial cells, the densely packed microtubule array that normally sits in the center of the intercellular bridge and the Flemming body were not observed. Instead, an intercellular bridge was found with a uniform density of microtubules that progressively disappear. The electron-dense material detected along the plasma membrane of the intercellular bridge could act as a rigid scaffold preventing membrane constriction. The SJ core components plus MyoII and Pnut at the interface made between the intercellular bridge and the finger-like protrusions might be part of the electron-dense material and therefore contribute to membrane rigidity. Alternatively, this scaffold could prevent tension release within the intercellular bridge, therefore preventing premature abscission (Daniel, 2018).

It was also found that abscission is asymmetric, with the bridge remnant internalized and then degraded in one of the daughter cells in a mode similar to the internalization and autophagy described for isolated cells. However, it cannot be excluded that abscission could also occur on both sides of the bridge, leading to its release followed by its recapture by endocytosis in some epithelial cells (Daniel, 2018).

Finally, whereas epithelial cells seem to be set for abscission about 40 min following the onset of anaphase, cytoplasmic isolation had not occurred 5 hr later. Photoconversion data furthermore suggest that cytoplasmic isolation takes place about one hour prior to one of the sisters entering mitosis, possibly at the mid to late G2 phase as reported for germline stem cells (Daniel, 2018).

Despite the opposite apico-basal positioning of permeability barrier relative to the mechanical barrier, a number of similarities are encountered in vertebrates' and invertebrates' epithelial cells. First, mechanical barrier transmission during epithelial cytokinesis is a multicellular process. Second, the permeability barrier is also maintained throughout epithelial cytokinesis. Third, the recruitment of tricellular junction components in close vicinity to the midbody supports the idea that polarized delivery, membrane composition, and topology at midbody are engineered for tricellular junction remodeling. Fourth, epithelial cytokinesis is polarized along the apico-basal axis, leading to the embedding of midbody en route to abscission within the permeability barrier both in vertebrates and in Drosophila. Based on these findings, it is anticipated that the coordination between permeability barrier transmission and completion of cytokinesis in Drosophila is conserved in vertebrates for maintenance of tissue integrity in proliferative epithelia (Daniel, 2018).

The claudin Megatrachea protein complex

Claudins are integral transmembrane components of the tight junctions forming trans-epithelial barriers in many organs, such as the nervous system, lung, and epidermis. In Drosophila three claudins have been identified that are required for forming the tight junctions analogous structure, the septate junctions (SJs). The lack of claudins results in a disruption of SJ integrity leading to a breakdown of the trans-epithelial barrier and to disturbed epithelial morphogenesis. However, little is known about claudin partners for transport mechanisms and membrane organization. This study presents a comprehensive analysis of the claudin proteome in Drosophila by combining biochemical and physiological approaches. Using specific antibodies against the claudin Megatrachea for immunoprecipitation and mass spectrometry, 142 proteins associated with Megatrachea were identified in embryos. The Megatrachea interacting proteins were analyzed in vivo by tissue-specific knockdown of the corresponding genes using RNA interference. Known and novel putative SJ components were identified, such as the gene product of CG3921. Furthermore, the data suggest that the control of secretion processes specific to SJs and dependent on Sec61p may involve Megatrachea interaction with Sec61 subunits. Also, the findings suggest that clathrin-coated vesicles may regulate Megatrachea turnover at the plasma membrane similar to human claudins. As claudins are conserved both in structure and function, these findings offer novel candidate proteins involved in the claudin interactome of vertebrates and invertebrates (Jaspers, 2012).