BIOLOGICAL OVERVIEW

Vertebrates and invertebrates share common mechanisms for regulation of cell-cell adhesion, including a family of proteins known as cadherins. Cadherins are important in setting up boundaries between epithelial compartments, preventing cell populations with different fates from mixing with one another. They control orderly release of cells from epithelium during epithelial-mesenchymal transitions. shotgun, or DE-cadherin is a homolog of classic vertebrate cadherins. The level of shotgun, expressed from maternal mRNA throughout the entire blastoderm [Image], is high enough to sustain embryonic development except for severe disruption of Malpighian tubules (Uemura, 1996). Zygotic shg is expressed in all epithelial tissue except for presumptive mesoderm cells prior to gastrulation [Image].

The presence of shg expression in neuroectoderm, and its absence in delaminated neuroblasts, suggests an involvement of SHG in the critical process of delamination: the migration of neuroblasts from the ectoderm into the ventral nervous system (Tepass, 1996). Another cadherin, Cadherin-N shows a complementary expression pattern. It is expressed primarily in the nervous system but not in epithelial precursors of neural cells (Iwai, 1997).

E-cadherin is a cell adhesion molecule that provides one constituent of the adherens junction, a cellular junction localized at the apicolateral boundary of an epithelial cell. E-cadherin recruits other proteins to the region of cell-cell contact, such as Crumbs, a transmembrane protein involved in organizing the adherens junction, and Armadillo, a junction associated protein coded for by a pair rule gene involved in wingless signaling.

shotgun is involved in Malpighian tubule morphogenesis. Malpighian tubules are initiated by a local invagination from the hindgut primordia. In shg mutants, the Malpighian tubules are rounded and eventually disintegrate. SHG is concentrated at the apical poles of epithelial cell-cell junctions (Oda, 1994). Interestingly, the apical polarity of the cells is preserved in shg mutants as reflected in Crumbs protein distribution (Uemura, 1996).

Tracheal branch outgrowth and fusion are similarly defective in shotgun mutants. A "stall" phenotype is apparent, exhibiting delayed branch extention. armadillo mutants have similar effects on tracheal development. The role of Armadillo in tracheal morphogenesis is partly separate from that of Wingless, because the formation of dorsal tracheal trunks is only partially impaired in wg mutants. This means that the connection of wingless signaling to occurances at the apical junction is not sufficient to completely disrupt morphogenesis in wingless mutants (Uemura, 1996).

The primary role of shotgun is preservation of the integrity of Drosophila epithelium and it is downregulated during critical periods of epithelial-mesenchymal transition such as gastrulation and neurogenesis. As an epithelial cadherin, it is the central element of the adherens junction, a multiprotein complex that holds cells together and transduces adhesion signals across membranes. It is involved as the nucleation site of cytoskeleton to cell junctions. In zygotic shotgun mutants the level of alpha-catenin and Armadillo at the adherens junction is dramatically reduced, pointing to the importance of Drosophila E-cadherin in assembling the adherens junction (Tepass, 1996).

The molecular structures of classic cadherins show clear differences between chordate and nonchordate metazoans. Although nonchordate classic cadherins have cadherin superfamily-specific extracellular repeats (CRs) and a highly conserved cytoplasmic domain (CP), these cadherins have a unique extracellular domain that is absent from vertebrate and ascidian (phylum Urochordata) classic cadherins. This domain is termed here the primitive classic cadherin domain (PCCD). The PCCD consist of three motifs. The first is unique to nonchordate classic cadherins, and has been named the NC (nonchordate motif). The second is rich in cysteine residues and rather resembles EGF repeats, and has been named the CE (cysteine-rich EGF repeat-like motif). The third is similar to the laminin A globular domain, and is named the LAG (laminin A globular domain-like motif). These motifs constitute a domain that constitutes the primitive classic cadherin domain (PCCD). To understand the roles of the PCCD, a series of mutant forms of the Drosophila classic cadherin Shotgun were constructed. Biochemical analyses indicate that the last two CRs and PCCD form a special structure with proteolytic cleavage. Mutations in the PCCD do not eliminate the cell-cell-binding function of Shotgun in cultured cells, but prevent the cadherin from efficiently translocating to the plasma membrane in epithelial cells of the developing embryo. In addition, genetic rescue assays suggest that although CP-mediated control plays a central role in tracheal fusion, the role of the PCCD in efficient recruitment of Shotgun to apical areas of the plasma membranes is also important for dynamic epithelial morphogenesis. It is proposed that there is a fundamental difference in the mode of classic cadherin-mediated cell-cell adhesion between chordate and nonchordate metazoans (Oda, 1999).

When the amino acid sequences of the PCCDs of nonchordate classic cadherins are aligned, the PCCD of Shotgun shows 39.8%, 31.1%, and 27.5% amino acid sequence identity to those of Shotgun, LvG-cadherin (sea urchin), and HMR-1 (C. elegans), respectively. The PCCDs of Shotgun and Drosophila N-cadherin are the most similar to each other in all combinations. A phylogenetic tree was constructed based on the PCCDs, consistent with the tree based on the CPs. The PCCD consists of the NC, CE, and LAG. No significant similarity within the NC is found to sequences of other known proteins. The CEs have 10 or 12 conserved cysteine residues, and the LAGs have 4 conserved cysteine residues. Although sequences similar to the LAGs are found in a variety of proteins, including Drosophila Slit, Fat, Crumbs and rat neurexin I, the first two of the four cysteine residues are not conserved in proteins other than the nonchordate classic cadherins. These amino acid sequence data suggest that the PCCD forms a structure unique to nonchordate classic cadherins (Oda, 1999).

When full-length Shotgun tagged with green fluorescent protein (GFP) at the carboxyl terminus is expressed in Drosophila S2 cells, 150- and 110-kDa polypeptides named distal and proximal polypeptides (DP and PP) are detected in the cell lysates by anti-Shotgun monoclonal antibody DCAD2 and anti-GFP antibody, respectively, in addition to a small amount of the precursor (about 230 kDa). DP corresponds to the polypeptide detected in embryonic lysates by DCAD2, which recognizes the region between CR0 and CR1. A CP-deleted mutant molecule named dCPc3 is separated into a DP of the same size (150-kDa) and a PP of reduced size (80-kDa). Sequence analysis shows that protolytic removal of the signal peptide and proteolytic cleavage between G (aa 1010) and S (aa 1011) residues in the NC result in DP (aa 70-1010) and PP (aa 1011-1507). Deletion analysis indicates that the region covering CR4, CR5, NC, and CE (aa 738-1108) is necessary for NC cleavage, but this is not the case for most other regions, including the transmembrane segment (TM) (Oda, 1999).

The CR domain, PCCD, and CP are essential for the morphogenetic function of Shotgun. The morphogenetic abilities of Shotgun mutant molecules are not necessarily correlated with their adhesive abilities in S2 cells. Various extracellular-domain deletion mutants, dCR3h, dCR4h, and dEx, which have no cell-cell-binding ability, can produce positive morphogenetic effects on tracheal fusion, but dCPc3, dNc, dCL3, and inNcHA show no positive morphogenetic effects, despite having cell-cell-binding ability in S2 cells. These results indicate that CP-mediated control plays a central role in tracheal fusion. The CP binds to the catenins, which may mediate interaction with the actin cytoskeleton. This coupling is expected to translate the force of actin bundle contraction into cell shape changes. Even if cell-cell binding function is removed from Shotgun, such mechanical forces may induce incomplete tracheal fusion. Despite the intact CP, some mutations in the PCCD lead to complete loss of the morphogenetic ability of Shotgun. These mutant molecules fail in prompt translocation to the plasma membrane when expressed in epithelial cells of the developing embryo. Efficient recruitment of Shotgun to appropriate sites of the plasma membrane may be important for dynamic epithelial morphogenesis during development (Oda, 1999).

Shotgun protein must be translocated from the ER to the contact sites of the plasma membrane to participate in cell-cell binding. Exogenously expressed normal Shotgun immediately accumulates at apicolateral borders of the plasma membrane in tracheal cells and many other epithelial cells. Mutational analyses reveal important aspects of Shotgun translocation involving the last two CRs, PCCD, and CP. Mutations in the PCCD lead to strong accumulation of the molecules in subdomains of the cytoplasm probably corresponding to the ER. Deletion of the large extracellular region encompassing CRs 1-5 and the PCCD also reduce the efficiency of translocation. These observations favor the idea that the extracellular region, including the PCCD, has an active role in translocation out of the ER to the plasma membrane. Deficiencies in deletion mutants dCR3h or dCR4h have little effect on translocation, but dCR4 exhibits a defect similar to that of the PCCD mutant molecules. These observations suggested that the last two CRs are functionally linked with the PCCD, and that this set is required for normal translocation. It is intriguing that NC cleavage, which presumably takes place in the ER, also requires the same region. However, the results from mNcGSP and dCL3 indicate that NC cleavage is not a prerequisite for translocation and occurs even under conditions in which translocation is blocked. It is speculated that CRs 4 and 5 and the PCCD are an overlapping region required for NC cleavage and for translocation, but these two events occur independently. There are three possible explanations for the mechanism of the involvement of the PCCD in translocation. (1) The unique structure of the last two CRs and PCCD may interact in the ER lumen with some machinery that carries out protein sorting and directional transport. (2) The PCCD structure may be involved in a lateral interaction (between the synthesized cadherin proteins) that facilitates translocation. (3) The PCCD may have a role in folding the extracellular part of the synthesized cadherin in the ER lumen to form the compact structure necessary for efficient translocation. It is hypothesized that the common ancestor of chordate classic cadherins adopted an alternative translocation mechanism to the PCCD-mediated mechanism coincident with the structural change to small size (Oda, 1999).

The CP is another part essential for normal translocation of Shotgun. Although the CP-deleted molecule tends to translocate to the apical plasma membrane, it is not so actively concentrated into the sites of cell-cell contact. Despite the reduced efficiency of translocation, the mutant molecule lacking the large extracellular region appears to be recruited to the apical zones of cell-cell contact once it leaves the ER. It seems that the CP-mediated control plays an important role in efficient and directional translocation of Shotgun from the ER to the contact sites of the plasma membrane. Shotgun mutant molecules accumulating in the cytoplasm due to the lack of translocation appear to bind to Armadillo and alpha-catenin. This raises the possibility that normal Shotgun joins Armadillo and alpha-catenin before arriving at the plasma membrane (Oda, 1999).

In addition to the subcellular distributions of the classic cadherins, there are marked differences between chordates and nonchordates in junctional complexes. Adherens junctions (AJs) and septate junctions (SJs) are observed in a wide range of nonchordate metazoans by electron microscopy. The AJ is located at the apical-most portion of the cell, followed by the SJ. In chordates (including vertebrates and ascidians) tight junctions (TJs), in addition to AJs, are frequently found, but SJs have not been found. Exceptionally, in Amphioxus cephalochordata, neither TJs nor SJs have been observed. The TJ is positioned at the apical end of the lateral plasma membrane in chordate epithelial cells, followed by the AJ. These observations indicate that there are fundamental differences between chordate and nonchordate metazoans not only in cadherin-based cell-cell adhesion but also in other mechanisms of cell-cell connection (Oda, 1999).

Adhesion disengagement uncouples intrinsic and extrinsic forces to drive cytokinesis in epithelial tissues

Cytokinesis entails cell invagination by a contractile actomyosin ring. In epithelia, E-cadherin-mediated adhesion connects the cortices of contacting cells; thus, it is unclear how invagination occurs, how the new junction forms, and how tissue integrity is preserved. Investigations in Drosophila embryos first show that apicobasal cleavage is polarized: invagination is faster from the basal than from the apical side. Ring contraction but not its polarized constriction is controlled by septin filaments and Anillin. Polarized cleavage is due instead to mechanical anchorage of the ring to E-cadherin complexes. Formation of the new junction requires local adhesion disengagement in the cleavage furrow, followed by new E-cadherin complex formation at the new interface. E-cadherin disengagement depends on the tension exerted by the cytokinetic ring and by neighboring cells. This study uncovers intrinsic and extrinsic forces necessary for cytokinesis and presents a framework for understanding how tissue cohesion is preserved during epithelial division (Guillot, 2013).

Epithelial cells divide in the plane of the tissue, allowing the equal partitioning of polarity proteins. This study delineated two major events during epithelial cytokinesis that shed light on how this is controlled. Cleavage progresses along the apicobasal axis and is polarized, as it is faster from basal to apical. This is not due to polarized contraction of the ring but to apical anchoring of the ring to E-cad complexes. Second, cleavage occurs in the plane of junction and involves local adhesion disengagement. In contrast to standard cytokinesis, this study delineated intrinsic and extrinsic mechanical processes operating during epithelial cytokinesis. Contractility of the ring itself is dependent on septins and Anillin. Ring contraction is resisted by intercellular adhesion mediated by E-cadherin complexes and by tension from neighboring cells transmitted by adhesion. Thus, E-cad-based adhesion plays a pivotal role in epithelial cytokinesis by anchoring the contractile ring, while its disengagement uncouples intrinsic and extrinsic tensile activity (Guillot, 2013).

In Drosophila embryos, epithelial cells exhibit polarized cleavage furrow ingression. This is likely to be general in epithelial cells, albeit at different magnitudes. MDCK cells too divide from the basal side toward the apex, and neuroepithelial cells in vertebrates partition the basal body first before the more apical part of the cell. Polarized cleavage is not a property unique to epithelial cells, however. Embryonic cleavage in several species exhibit a range of patterns, from completely unilateral cleavage, as reported in jellyfish (Clytia and Beroe) and Ctenophores (Pleurobrachia), to partly asymmetric cleavage in the one-cell-stage C. elegans embryos). In the latter case, polarized ingression of the cleavage furrow is stochastic and correlates with heterogeneities in the recruitment of the actin crosslinker Anillin and of septins. In anillin and septin knockdowns, cleavage becomes symmetric. This contrasts with activators of MyoII, such as Rho kinase, which affects the speed of contraction but not its polarity. Thus, in nonepithelial cells, polarized cleavage is a purely autonomous process governed by heterogeneities in regulators of contractility. This study found, however, that in Drosophila embryos, polarized cleavage is not determined by polarized distribution of Anillin and septins or by differential biomechanical properties of the ring. Septins display a marginal yet significant enrichment basally, and Anillin is slightly enriched apically. However, invagination was still normally polarized along the apicobasal axis in both peanut mutants and anillin RNAi embryos, despite strong reduction in constriction rat. Moreover, no significant difference between apical and basal relaxation kinetics was detected following ablation in wild-types. The ablation kinetics reflects the relative effect of stiffness in the ring and friction internal to the ring and with the cytoplasm. With the caveat that the latter cannot be directly measured and and is assumed to be uniform, these ablation experiments indicate the relative stiffness in the ring. The fact that relaxation is faster (<5 s) than turnover of the internal components of the ring, such as MyoII, substantiates the idea that mostly the elastic relaxation of the ring was measured and not a quasi-static relaxation associated with turnover/movements of ring components (Guillot, 2013).

The rate of constriction was monotonic such that big rings and small rings contracted at a constant rate in wild-types but also in anillin or septin mutants, although it was strongly reduced in the latter cases. This contrasts with reports in C. elegans, where constriction was scaling with ring size, suggesting a mechanism based on disassembly of contractile units whose number scales with ring size. This difference may stem from the fact that cytokinesis is especially rapid in Drosophila embryos (about 150 s). Alternatively, it could reflect the epithelial nature of the divisions reported in this study (Guillot, 2013).

The evidence argues instead that polarized ingression depends largely on apical anchoring of the ring to E-cad complexes. First, E-cad complexes colocalize with the contractile ring for the most part of invagination. Second, ingression is symmetric in either e-cad or α-cat RNAi embryos. Although E-cad complexes, in particular α-cat, can recruit regulators of MyoII, this cannot explain polarized invagination of the ring, since apical and basal relaxations are not significantly different in wild-types and in α-cat RNAi embryos. E-cad complexes transmit actomyosin tension in epithelia. Two sets of observation support the idea that junctions exert pulling forces on the ring due to anchoring. The ring is stretched laterally as it constricts, and this requires apical junctions via e-cad and α-cat. The relative deformation of the ring following ablation is larger apically than basally, and this also requires cell junctions. It is striking that extrinsic and intrinsic regulators of the ring contraction have very different effects on ring dynamics. In the absence of Pnut or Anillin, the ring constriction is reduced but it is still polarized. However, following e-cad or α-cat depletion, ring constriction is normal but symmetric. It is concluded that the mechanical connection of E-cad complexes to the contractile ring causes polarized invagination. It is possible that, in other systems, both intrinsic and extrinsic regulation will operate in parallel to increase the cleavage asymmetry. This may be important in highly columnar epithelial cells or when adhesion is lower and unable to resist the ring tension (Guillot, 2013).

Polarized cleavage effectively separates apical and basal cleavage, adhesion complexes being a barrier separating the apical and lateral domains. The central problem becomes: How does cleavage occur at adherens junctions? This study delineated two critical phases in junctional cleavage. First, the adherens junctions invaginate with the actomyosin ring, consistent with the fact that the ring is anchored to the junctions. During this phase, E-cad intercellular adhesion is stable in the face of the tension exerted by the ring, and E-cad colocalizes with the ring at the point of coupling. Invagination of junctions then stops as E-cad levels decrease in this area. However, ring constriction continues and appears to detach from junctions. This is interpreted as a point of adhesion disengagement. Adhesion disengagement marks the formation of the new vertices and of the new junction between daughter cells. Electron microscopy images show this membrane disengagement. Consistent with this, the membrane still invaginates with the actomyosin ring), although E-cad is still not detected. Closer examination shows that E-cad monomers are present at this late stage of cytokinesis but that adhesion complexes form gradually from this stage onward. It is striking that adhesion is very locally (<1 μm out of ∼40 μm of junction perimeter) and transiently (∼200 s) perturbed during division. In the first 150 s, E-cad clusters immediately adjacent to the cleavage furrow remain in position as the junction invaginates. This suggests that the cortex can be extensively remodeled locally. It likely reflects the fact that tension induces membrane flows with respect to the actin-rich cortex and argues that E-cad-mediated adhesion does not prevent membrane flow during disengagement. Interestingly, local disengagement allows local cell deformation without affecting the overall shape of cell contacts. Consistent with the idea that adhesion is locally disengaged, the amount of E-cad has a strong impact on the timing and depth of junctional cleavage. Increasing E-cad delays disengagement (i.e., the formation of the new junction, inducing strong cell deformations. More generally, this implies that increasing adhesion may provide an efficient mechanism to prevent local cell-cell disengagement when internal tension is used to remodel junctions during morphogenesis. In apical constriction in the Drosophila mesoderm, actomyosin cables pull on the junctional cortex and reduce junction lengths. If adhesion was not strong enough, local disengagement would occur and junctions could not remodel. The fact that adhesion disengagement is local and transient during cytokinesis is also probably key to the overall maintenance of cell polarity and adhesion during epithelial division (Guillot, 2013).

It is proposed that adhesion disengagement is mechanically induced by tension in the cytokinetic ring and by tension from neighboring cells. When the cumulated tension is higher that the adhesive force, disengagement occurs. Consistent with this, disengagement and formation of the new junction is strongly delayed in mutants that reduce the constriction of the cytokinetic ring, namely, in septin mutants and in Anillin knockdown embryos. Likewise, ablation of neighboring cells delays disengagement. It is, however, possible that adhesion is also locally disrupted by either E-cad endocytosis or phosphorylation of β-cat/Arm (Guillot, 2013).

Adhesion complexes transmit cell tension exerted by neighboring cells. Surrounding junctions and, more specifically, MyoII cables oriented toward or near the cleavage furrow strongly affect furrow invagination when E-cad is present at high levels. The invagination in this case is very shallow, suggesting a tug of war between intrinsic (ring contraction) and extrinsic tension (MyoII cables in neighbors). This results in asymmetric furrows in the plane of junctions due to the asymmetric distribution of MyoII cables around the cell. When E-cad is expressed at lower levels, even if surrounding junctions are oriented toward the cleavage furrow, invagination is unaffected and symmetric. It is proposed that E-cad complexes sensitize cells to their mechanical environment. This may provide a mechanism for cells to integrate stress coming from the environment. It will be important to explore how E-cad levels may affect cells responsiveness to extrinsic stress during division by affecting the timing of the formation of the new junction by local disengagement and the resulting cell shape and topology (Guillot, 2013).

α-Catenin stabilises Cadherin-Catenin complexes and modulates actomyosin dynamics to allow pulsatile apical contraction

This study investigated how cell contractility and adhesion are functionally integrated during epithelial morphogenesis. To this end, the role of α-Catenin, a key molecule linking E-Cadherin-based adhesion and the actomyosin cytoskeleton, was analyzed during Drosophila embryonic dorsal closure, by studying a newly developed allelic series. α-Catenin was found to regulate pulsatile apical contraction in the amnioserosa, the main force-generating tissue driving closure of the embryonic epidermis. α-Catenin controls actomyosin dynamics by stabilising and promoting the formation of actomyosin foci, and also stabilises DE-Cadherin (Drosophila E-Cadherin, also known as Shotgun) at the cell membrane, suggesting that medioapical actomyosin contractility regulates junction stability. Furthermore, a genetic interaction was uncovered between α-Catenin and Vinculin, and a tension-dependent recruitment of Vinculin to amniosersoa apical cell membranes, suggesting the existence of a mechano-sensitive module operating in this tissue (Jurado, 2016).

How adhesion and actomyosin contractility are integrated at junctions is a fundamental question in morphogenesis. To tackle this, the role of α-Catenin, a key protein linking adherens junctions and the actin cytoskeleton, was analyzed in the context of Drosophila embryogenesis and in particular during dorsal closure. α-Catenin was found to regulates pulsatile actomyosin dynamics in apically contracting cells by stabilising and promoting actomyosin contractions. α-Catenin also stabilises DE-Cadherin at the cell membrane, suggesting that medioapical actomyosin contractility regulates junction stability. Furthermore, the results reveal an interaction between α-Catenin and Vinculin that could be important for DE-Cadherin stabilisation (Jurado, 2016).

Live imaging of mutant embryos shows a strong requirement for α-Catenin in the migration of the dorsal ridge primordia towards the dorsal midline, preventing the formation of the dorsal ridge and thus affecting both dorsal closure and head involution. These results reveal that the dorsal ridge is particularly sensitive to the levels of α-Catenin and suggest it is a key region that could mechanically coordinate both processes. Although it is clear that some of the defects observed during dorsal closure are a consequence of the defective dorsal ridge morphogenesis, this analysis shows that other cellular processes more specific to dorsal closure are affected. In particular, it was observed that the actin cable is disorganised and that the pulsatile apical contraction of the amnioserosa is abnormal (Jurado, 2016).

The defects observed at the level of amnioserosa apical cell oscillations could be a consequence of a defective actin cable, which would be acting as a ratchet and thus progressively restricting the expansion of apical cell area. However, several lines of evidence suggest that a ratchet mechanism stabilising the contracted state of amnioserosa cells is acting at the level of individual cells. In particular, the analysis performed in this study of actin oscillatory dynamics in α-Cat mutants suggests that the increase in the expansion half-cycle of amnioserosa apical cell oscillations could be due to an increase in the time interval between the appearance of consecutive foci. Thus, the results favour the idea that the Cadherin-Catenin complex has a role in promoting actomyosin oscillatory dynamics. How α-Catenin promotes actomyosin contractility remains to be elucidated, but it is likely to involve both direct and indirect (through other actin-binding proteins) interactions with the actin cytoskeleton. For example, an antagonistic interaction between α-Catenin and the Arp2/3 complex has been observed both in cell systems and in Drosophila embryos, raising the possibility that the actin-bundling activity of α-Catenin at adherens junctions, rather than the formation of Arp2/3-dependent networks, could be important for apical contraction (Jurado, 2016).

Interestingly, it was found that with the α-Cat²⁰⁴⁹ allele, adhesion dynamics are also defective, suggesting that medioapical actomyosin dynamics promote adherens junction stabilisation. In contrast, with the α-Cat⁴²¹ allele, which would bind constitutively to Vinculin in a context of defective medioapical actomyosin dynamics, DE-Cadherin stabilisation is recovered. This result suggests that the stabilisation of DE-Cadherin could be mediated by the binding of Vinculin to α-Catenin. This is in agreement with what has been observed in cell systems, where forms of α-Catenin that constitutively bind to Vinculin have decreased mobility. It was further shown that, although DE-Cadherin is stabilised in α-Cat⁴²¹ mutants, possibly due to the Vinculin-α-Catenin interaction, this stabilisation is not able to rescue normal medioapical actin dynamics. Thus, it is suggestd that direct binding of α-Catenin to actin through its actin-binding domain promotes the formation of medioapical actomyosin foci, whereas indirect binding to actin through Vinculin would promote junction stabilisation. Taken together, these data suggest that α-Catenin domains, through their interactions with other actin-binding proteins and actin, might differentially regulate actin dynamics (Jurado, 2016).

Finally, the results show that there is a tension-dependent recruitment of Vinculin at the membranes of amnioserosa cells, which could be mediated by α-Catenin. Interestingly, it has recently been found, in experiments using a heat-shock inducible Vinculin reporter, that the rate of change of Vinculin levels correlates with junctional tension. The results also suggest that Vinculin is able to perform an adhesive function when α-Catenin function is compromised. This could result from an α-Catenin-independent binding of Vinculin to E-Cadherin or from an interaction between Vinculin and other junctional proteins such as ZO-1 (also known as TJP1), which has been shown to recruit Vinculin to VE-cadherin junctions and increase cell-cell tension. However, given that ZO-1 can also interact with α-Catenin, it remains to be investigated whether the mechano-sensitivity of Vinculin is completely dependent on α-Catenin. Thus, it is likely that Vinculin is able to perform different functions depending on its developmental context. Interestingly, different mechanisms for Vinculin binding to Talin in integrin-mediated adhesion have recently been uncovered in different morphogenetic processes, meaning that Talin can sense different force vectors. Given that a role for Talin and integrin-mediated adhesion during dorsal closure has been uncovered, it would be interesting to investigate whether Vinculin is also involved in integrin-mediated adhesion at this stage. The results suggest that a tension-dependent module involving Vinculin is present in amnioserosa cells. An exciting avenue will be to identify the mechanisms and function of such module in the context of morphogenesis (Jurado, 2016).

Adherens junction length during tissue contraction is controlled by the mechanosensitive activity of actomyosin and junctional recycling

During epithelial contraction, cells generate forces to constrict their surface and, concurrently, fine-tune the length of their adherens junctions to ensure force transmission. While many studies have focused on understanding force generation, little is known on how junctional length is controlled. This study shows that, during amnioserosa contraction in Drosophila dorsal closure, adherens junctions reduce their length in coordination with the shrinkage of apical cell area, maintaining a nearly constant junctional straightness. This study reveals that junctional straightness and integrity depend on the endocytic machinery and on the mechanosensitive activity of the actomyosin cytoskeleton. On one hand, upon junctional stretch and decrease in E-cadherin density, actomyosin relocalizes from the medial area to the junctions, thus maintaining junctional integrity. On the other hand, when junctions have excess material and ruffles, junction removal is enhanced, and high junctional straightness and tension are restored. These two mechanisms control junctional length and integrity during morphogenesis (Sumi, 2018).

This study has shown that adherens junction length is actively controlled during epithelial contraction, leading to a constant junctional straightness upon cell contraction. It is proposed that two different cellular mechanisms are at the origin of this length control. First, a modulation of actomyosin localization from the medial areas toward junctions allows for integrity maintenance by increasing actomyosin contraction along the junctions upon junctional stretch. Possibly, E-cadherin dilution triggered by junctional stretching is responsible for the myosin recruitment, a possibility supported by the anti-correlation between junctional myosin and E-cadherin. Second, the junction removal rate depends on junctional straightness, maintaining a preferred junctional straightness. Presumably, the endocytosis machinery is at the origin of this maintenance. As junctional tension is dependent on junctional straightness, maintaining junctional straightness could also allow it to maintain junctional tension, which remains approximately constant during DC (Sumi, 2018).

These mechanisms rely on a complex interplay between junctional E-cadherin, actomyosin, and tension. Indeed, this study finds that (1) higher junctional E-cadherin levels are associated with lower myosin levels and (2) decreased junctional straightness is associated with lower junctional tension. These observations suggest that E-cadherin is an essential component for the control of adherens junction length and for actomyosin localization in the cell. In cell culture experiments, E-cadherin has already been determined to be a key player in cell-cell mechanical interplay and shown to interact closely with the actomyosin cortex. In the case of the AS tissue, an enrichment of actomyosin was observed at junctions when E-cadherin levels are reduced. A similar interaction has been previously identified in the context of cytokinesis. Ghd results indicate a general interplay between E-cadherin and myosin that could play a role in several morphogenetic rearrangements. Such an interplay is likely to involve actin-associated proteins that could respond to changes in mechanical states of the adherens junction. Consistent with this idea, previous studies identified dynamic enrichment of the formin Diaphanous or of vinculin specifically at adherens junctions (Sumi, 2018).

Interestingly, this study observed that when junctions are stretched and E-cadherin density decreased, the contractile pulses of the AS tissue are arrested, possibly due to the observed myosin recruitment to junctions. Such pulses of contraction have been identified in several tissues during Drosophila, chicken, and mouse embryo development. They have been associated with global tissue contraction and epithelial remodeling. This study has shown that the pulsatile activity can be tuned by junctional stretch and E-cadherin levels. The results therefore uncover a direct coupling between cell-cell adhesion and the contractile activity of the cell (Sumi, 2018).

Finally, the rate of junction removal appears to depend on junctional straightness. How is junction straightness sensed to dictate junction removal? Laser ablation experiments indicate that junctions with lower straightness also have lower mechanical tension. In single cells, it is known that the endocytic rate is modulated by the tension of the cell membrane. In a developmental context, endocytosis could also be tension sensitive, resulting in the control of junctional straightness through junctional tension. Alternatively, E-cadherin density could play a role in the control of the junction turnover rate. Indeed, an increase of E-cadherin density associated to a decrease in junction straightness could trigger an enhanced endocytic rate, therefore controlling junction straightness and tension. The mechanisms regulating junction length and integrity identified here involve the activity of two essential cellular machineries. They are therefore likely to act during many morphogenetic events that involve tissue contraction, such as neural tube closure, optic cup morphogenesis, or wound healing (Sumi, 2018).

Extracellular matrix stiffness cues junctional remodeling for 3D tissue elongation

Organs are sculpted by extracellular as well as cell-intrinsic forces, but how collective cell dynamics are orchestrated in response to environmental cues is poorly understood. This study applied advanced image analysis to reveal extracellular matrix-responsive cell behaviors that drive elongation of the Drosophila follicle, a model system in which basement membrane stiffness instructs three-dimensional tissue morphogenesis. Through in toto morphometric analyses of wild type and round egg mutants, this study found that neither changes in average cell shape nor oriented cell division are required for appropriate organ shape. Instead, a major element is the reorientation of elongated cells at the follicle anterior. Polarized reorientation is regulated by mechanical cues from the basement membrane, which are transduced by the Src tyrosine kinase to alter junctional E-cadherin trafficking. This mechanosensitive cellular behavior represents a conserved mechanism that can elongate edgeless tubular epithelia in a process distinct from those that elongate bounded, planar epithelia (Chen, 2019).

Follicles have an architecture that is typical of a number of animal organs, with several components that associate to form a 3D acinar epithelium surrounding a lumen. At the same time, the simplicity and highly regular development of the follicle lend themselves to comprehensive analyses. The follicle exhibits straightforward and symmetric geometry for much of its development, while its cells originate from only two stem cell populations and show limited differential fates. Follicles can be genetically manipulated using the powerful Drosophila toolkit, and are well-suited for imaging either in fixed preparations or when cultured live ex vivo (Chen, 2019).

Development of the follicle involves several conserved morphogenetic behaviors including initial primordial assembly, epithelial diversification, and collective cell migration. A major focus for mechanistic studies has been follicle elongation, during which the initially spherical organ transforms into a more tube-like ellipsoid shape. ~2-fold elongation is seen in ~40 h between follicle budding at stage 3 to the end of stage 8; eventually there is ~2.5-fold overall elongation when the egg is laid ~25 h later. This degree of elongation is similar to that in paradigmatic morphogenetic systems such as the amphibian neural plate and mesoderm, or the Drosophila germband. In the latter tissues, the main cellular behavior that drives elongation is convergent extension, as cells intercalate mediolaterally toward a specific landmark that is defined anatomically and/or molecularly. However, these tissues have defined borders, which create boundary conditions to instruct and orient cell behaviors. No such boundary is evident along the edgeless epithelium of the Drosophila follicle, and the cellular changes that drive elongation of this acinar organ are not known (Chen, 2019).

Recent work has shown that mechanical heterogeneity patterned not within the cells of the follicle, but instead within its underlying basement membrane (BM), instructs organ shape (Crest, 2017). Specifically, a gradient of matrix stiffness that is low at the poles and peaks in the organ center provides differential resistance to luminal expansion, leading to tissue elongation. Construction of this pattern relies in part on a collective migration of cells around the follicle equatorial axis, leading to global tissue rotation. But how the cells of the epithelium respond to stiffness cues and engage in the dynamics that actually elongate the organ along the anterior-posterior (A-P) axis remains unexplored (Chen, 2019).

This study has identified an unexpected cell behavior that drives follicle elongation and demonstrates its control by a regulatory axis that responds to BM stiffness cues, thus connecting extracellular mechanical properties to intracellular signaling that drives intercellular morphogenesis in vivo (Chen, 2019).

All three morphogenetic behaviors that are engines for tissue elongation in other systems (cell shape changes, cell division, and cell rearrangement) are seen during follicle elongation. Average cellular elongation changes little during follicle elongation, and an initial cell shape anisotropy-lengthening around the organ circumferential axis-is not perturbed in a round egg mutant. Additionally, tissue rotation alone is insufficient to direct elongation-driving cell dynamics. The manipulations carried out in this study show that oriented cell division is not essential for elongation, and its absence can be compensated by other polarized cell behaviors, as also seen in e.g., the pupal thorax. Moreover, addition of cells to the elongating axis of the follicle continues at stage 7, after all mitoses have ceased. Instead, the data suggest that the major tissue-elongating behavior involves changes in cell orientation. In particular, cells in the follicle anterior shift the direction of their long axis towards the A-P axis. Without changing average cellular elongation, this reorientation of anisotropically-shaped cells changes neighbor relationships; it is suggested that this results in net cellular intercalation along the A-P axis (Chen, 2019).

The number of frank intercalations that occur during follicle elongation appears relatively limited. For instance, of the ~850 cells in a post-mitotic stage 7 follicle, only ~3 are added to the meridional arc during ~8 h prior to stage 8, inducing a ~10% change in aspect ratio. This relative paucity of intercalation events contrasts not only with the rapidly developing Drosophila germband, but also with other elongating tissues such as the Drosophila pupal wing and thorax and the vertebrate neural tube and mesoderm. Three major differences bear consideration here. First, significant growth (~9-fold) of both epithelial cells and the follicle overall occur during the 24-34 h that span stages 4 to 8, while the other systems largely rearrange a constant tissue volume. Second, the other systems have defined boundaries to impose vectorial orientation of cell behaviors, while the topologically continuous follicle epithelium lacks a boundary in the equatorial axis. Third, follicle elongation does not require conventional PCP morphogenetic signaling, nor is there evidence of PCP Myosin localization that remodels cell junctions. Instead, the instructive force seems to be patterned anisotropic resistance to growth, wherein softer BM at the poles triggers regional changes in cell behavior. Indeed, the largest changes in follicle cell orientation are seen at the anterior pole, and genetic manipulation in this region alone can prevent elongation. It is important to note that the data do not identify BM stiffness differences between the anterior and the posterior prior to cell orientation changes in the former, nor differences in pSrc levels, but do identify differences in Ecad dynamics in the latter. It is speculated that A-P patterned fates in the follicle epithelium prevent elongation behaviors in the posterior, consistent with the observation that follicles lacking posterior fate specification elongate at both poles. Overall, these differences with bounded epithelia undergoing elongation via convergent extension emphasize the new perspectives required for analysis of elongating tubular organs (Chen, 2019).

How do cells sense BM stiffness to change their orientations? The elongation-defective phenotype of Rack1-depleted follicles points to one mechanism involving Src. Rack1 is a Src-inhibiting protein, and its loss, with associated increases in cellular Src activity, perturbs follicle elongation. In Rack1-depleted follicles, initial cell orientation is not greatly altered, but anterior follicle cells remain static and are unable to shift their orientation during stages 6-8. This suggests that proper regulation of Src is required for the dynamic reapportionment of cell shape that is reflected in this switch. One familiar regulatory target of Src is integrins and their associated proteins, but no defects in integrin-dependent cell migration are seen when Rack1 is depleted from follicle cells. However, Src has also been implicated in directly regulating AJ remodeling through effects on Ecad trafficking, and FRAP analysis reveals compromised Ecad dynamics when follicles cells lack Rack1. Src is considered a mechanotransducer, and pSrc levels anticorrelate with BM stiffness in wild type, mutant and manipulated follicles. It is therefore proposed that Src mediates the instructive cue provided by BM stiffness, inducing AJ remodeling to drive morphogenesis. Whether the AJ-regulating apical Src seen in follicles is directly phosphorylated by basal integrin activation, or results from an indirect intracellular signaling cascade, remains to be investigated (Chen, 2019).

The defective tissue topology and reduced Ecad mobility seen in Rack1-depleted follicles suggests that Src is required for both elongation-driving and tissue disorder-minimizing cell rearrangements. This role of Src in follicle elongation raises interesting parallels with a second edgeless epithelium: the Drosophila trachea. In this established tubulogenesis model, gain as well as loss of Src42A activity results in shortened but broader tubules, which result from inappropriately oriented cells. Interestingly, depletion of Src42A from the follicle, like its activation, also induced elongation defects, while Src controls tracheal Ecad dynamics, perhaps in response to an ECM, albeit apically localized. Thus, in both organs Src42A could mediate stiffness-cued AJ dynamics that change the orientation of cell eccentricity. Since mammalian kidney tubule development also shows Src-dependence, these results suggest that ECM-mediated control of cell junctions via Src may be a general mechanism for morphogenesis of edgeless epithelia (Chen, 2019).

The data described in this work stem from a comprehensive analysis of follicle morphogenesis, which was enabled by ImSAnE software. ImSAnE allowed identification of critical cell dynamics near the follicle poles, which are most subject to distortion from conventional analyses, and distinguished the cellular basis underlying overtly similar elongation phenotypes. For instance, it revealed that under fat2 depletion, cells throughout the follicle are defective in planar polarized cell orientation from early stages, whereas in Rack1 follicles defective orientation is limited to the anterior and occurs only later, when elongation behaviors initiate. Furthermore, some of the ImSAnE-based morphometric findings are concordant with those recently reported based on conventional imaging. This work focuses on stages 4 through 8, and does not address cell dynamics that elongate the follicle prior to and following these stages. However, it does provide a framework for true in toto analysis of this simple organ, at multiple developmental stages and genotypes. As genetic screens and biophysical studies in the follicle are extended, the quantitative imaging platform reported here will provide a bridge towards mathematical and mechanical modeling of this flourishing system for in toto organ morphogenesis (Chen, 2019).

Adherens junction remodelling during mitotic rounding of pseudostratified epithelial cells

Epithelial cells undergo cortical rounding at the onset of mitosis to enable spindle orientation in the plane of the epithelium. In cuboidal epithelia in culture, the adherens junction protein E-cadherin recruits Pins/LGN/GPSM2 and Mud/NuMA to orient the mitotic spindle. In the pseudostratified columnar epithelial cells of Drosophila, septate junctions recruit Mud/NuMA to orient the spindle, while Pins/LGN/GPSM2 is surprisingly dispensable. This study shows that these pseudostratified epithelial cells downregulate E-cadherin as they round up for mitosis. Preventing cortical rounding by inhibiting Rho-kinase-mediated actomyosin contractility blocks downregulation of E-cadherin during mitosis. Mitotic activation of Rho-kinase depends on the RhoGEF ECT2/Pebble and its binding partners RacGAP1/MgcRacGAP/CYK4/Tum and MKLP1/KIF23/ZEN4/Pav. Cell cycle control of these Rho activators is mediated by the Aurora A and B kinases, which act redundantly during mitotic rounding. Thus, in Drosophila pseudostratified epithelia, disruption of adherens junctions during mitosis necessitates planar spindle orientation by septate junctions to maintain epithelial integrity (Aguilar-Aragon, 2020).

Adherens junctions have long been thought to be continuously essential for maintaining epithelial form and function. The current findings demonstrate transient loss of adherens junctions during division of pseudostratified epithelial cells, an event that involves adherens junction remodelling during the extensive rounding up of cell shape in mitosis. This study has furthermore shown that loss of adherens junctions is a direct consequence of the increased Rho activity and actomyosin contractility that drives mitotic rounding, which is both necessary and sufficient to regulate the level of junctional E-cadherin in the pseudostratified wing imaginal disc epithelium of Drosophila. These findings are consistent with previous observations that adherens junctions can be removed via E-cadherin endocytosis upon planar polarised Rho activation and actomyosin-driven junctional shrinkage generated during cell-cell rearrangements in the Drosophila embryo and recent optogenetic experiments in human cells. However, the global loss of E-cadherin observed during mitosis of pseudostratified epithelial cells is unprecedented and may be uniquely required by the rapid transformation of these cells from their highly columnar shape to a rounded sphere at mitosis, which involves a rapid increase in apical area and junctional length. This change in cell chape, driven by global actomyosin contractility during mitotic rounding, may both spread out junctions and disrupt cadherin-cadherin contacts between neighbouring cells to favour endocytosis (Aguilar-Aragon, 2020).

Notably, loss of E-cadherin does not seem to occur during mitosis in cuboidal epithelial cells, which undergo a much milder cell shape change during mitotic rounding, such as the Drosophila follicle cell epithelium or in many human cultured epithelial cell lines. Indeed, E-cadherin was reported to play essential roles in planar spindle orientation in cultured human cells. Thus, the pseudostratified epithelia of Drosophila face a unique challenge of orienting the mitotic spindle in the plane of the epithelium without the use of adherens junctions as a cue, which may explain why these cells instead rely on septate junctions, while planar spindle orientation can occur normally in the cuboidal follicle cell epithelium before septate junctions form. In the absence of septate junctions, spindle-orienting factors such as Pins, Mud, Dlg and Scrib localise to lateral membranes, overlapping with E-cadherin, which can directly interact with Scrib in both Drosophila and human cells (Aguilar-Aragon, 2020).

The findings of this study also shed light on the molecular mechanisms linking the cell cycle with control of Rho activation during mitotic rounding. Downstream of the master mitotic kinase Cdk1, a key role for both Aurora A and B kinases (acting redundantly) was identified in initiating mitosis and maintaining cortical rounding, with Aurora B then acting alone to drive furrow formation during cytokinesis. Aurora kinases are known to activate the key cell cycle kinase Plk1/Polo, which can then activate RacGAP1/MgcRacGAP/CYK4/Tum and ECT2/Pbl, possibly via the kinesin-like protein MKLP1/KIF23/ZEN4/Pav. One report suggested that Aurora B also acts to directly phosphorylate RacGAP1/MgcRacGAP/CYK4/Tum on S387, but this study found that a CRISPR-knockin mutation of this site (tum_S387A) is homozygous viable and fertile in Drosophila. In addition, Aurora B can also directly phosphorylate MKLP1/KIF23/ZEN4/Pav to oligomerise and activate RacGAP1/MgcRacGAP/CYK4/Tum and ECT2/Pbl during cytokinesis, via a mechanism involving plasma membrane association of clustered C1 domains from the RacGAP1/MgcRacGAP/CYK4/Tum protein. Accordingly, it was found that a CRISPR-knockin mutation of this site (pavS734A/S735A) is homozygous lethal in Drosophila, suggesting that Aurora A/B could act via this complex during mitotic rounding. Indeed, MKLP1/KIF23/ZEN4/Pav undergoes cell cycle-dependent re-localisation from the nucleus (in interphase) to the cytoplasm (in mitosis), including a clear localisation to the entire plasma membrane during mitosis and then to the cleavage furrow during cytokinesis. The results show that mitotic activation of the MKLP1/KIF23/ZEN4/Pav binding partner RacGAP1/MgcRacGAP/CYK4/Tum also involves translocation from the nucleus to the cytoplasm, similar to ECT2/Pbl, where all three proteins are then available to bind the entire plasma membrane and generate global Rho activity to drive cortical contractility and loss of adherens junctions during mitotic rounding (Aguilar-Aragon, 2020).

The complex of MKLP1/KIF23/ZEN4/Pav and RacGAP1/MgcRacGAP/CYK4/Tum is often referred to as the 'centralspindlin' complex, due to its association with the spindle midzone at anaphase. However, the use of this term was avoided because association with the spindle midzone is not required for the function of this complex in activating ECT2/Pbl and Rho at the plasma membrane during cytokinesis. Furthermore, the results show that the same complex also functions prior to anaphase or cytokinesis in activating ECT2/Pbl and Rho for cortical rounding and downregulation of adherens junctions from the very onset of mitosis (Aguilar-Aragon, 2020).

Importantly, the function of the three cell cycle-regulated Rho activators discussed in this study - MKLP1/KIF23/ZEN4/Pav, RacGAP1/MgcRacGAP/CYK4/Tum and ECT2/Pbl - on the regulation of adherens junctions appears to be dose-dependent and cell shape change-dependent. The sudden re-localisation of these proteins to the cytoplasm at mitosis drives a global increase in Rho activation, cell shape change and loss of junctions. In contrast, the relatively low level of these proteins in the cytoplasm during interphase appears to contribute to the normal maintenance of Rho activity at the junctional actomyosin ring to maintain adherens junctions-at least in mammalian cells. In Drosophila, an interphase role of ECT2/Pbl in contributing to Rho activation and maintenance of adherens junctions is also plausible and should be most easily distinguished from the mitotic role in non-dividing epithelial cells. However, in the ovarian follicle cell epithelium, silencing of ECT2/Pbl by RNAi affected cytokinesis (leading to larger cells) but did not affect overall epithelial architecture after cells arrest their proliferation, suggesting that any interphase function may be obscured by redundancy with other RhoGEFs in this tissue. Note also that, in this cuboidal epithelium, mitotic rounding itself is more subtle and ECT2/Pbl is dispensable for planar spindle orientation and epithelial integrity (Aguilar-Aragon, 2020).

Finally, it is noted that ECT2/Pbl was initially reported to be capable of acting as GEF for another GTPase, Cdc42, in addition to Rho 38. This activity suggested a possible role in regulating the apical Cdc42-Par6-aPKC complex in epithelial polarity and in mitosis. In Drosophila, ECT2/Pbl was found to drive transient apical spreading of the Cdc42-Par6-aPKC complex during mitotic rounding of pupal notum epithelial cells. This study found evidence for a similar function of ECT2/Pbl in activating Cdc42 during interphase in the post-mitotic ovarian follicle cell epithelium, although the loss-of-function phenotype of ECT2/Pbl is obscured due to redundancy with other Cdc42 GEFs such as beta-PIX. Loss of both ECT2/Pbl and beta-PIX reduced the level of the Cdc42-Par6-aPKC complex localising to the apical domain of post-mitotic follicle cells and also reduced apical ZO-1 localisation in human Caco2 intestinal epithelial cells in culture. In contrast, the gain-of-function phenotype caused by ECT2/Pbl overexpression is a neoplastic tumour-like phenotype and clearly involves ectopic spreading of the apical Cdc42-Par6-aPKC complex, in addition to persistent cell rounding in both cuboidal follicle cells and pseudostratified wing epithelial cells (Aguilar-Aragon, 2020).

In conclusion, the findings provide new insights into the cell biology of mitotic rounding, identifying remodelling of adherens junctions as a key event in pseudostratified epithelia, where rounding is extensive, but not cuboidal epithelia, where rounding is more subtle. These results are consistent with the hypothesis that Rho activation and actomyosin contractility can stabilise adherens junctions in the absence of mechanical strain, but that Rho activation can induce E-cadherin endocytosis above a critical strain threshold, be it either junctional shrinkage or expansion, both of which may alter the geometry of the junctional actomyosin ring and disrupt cadherin-cadherin contacts between neighbouring cells to favour endocytic internalisation of E-cadherin. The results also clarify the molecular mechanisms linking cell cycle control machinery, particularly the Aurora A and B kinases, with Rho activation and mitotic rounding. Lastly, this work may have direct relevance to certain human epithelial cancers, such as lung cancer or glioma, where overexpression of ECT2/Pbl has been reported to correlate for poor prognosis, and where the findings suggest it could drive not only disruption of epithelial polarity via activation of Cdc42 or Rac but also loss of adherens junctions via sustained Rho activation to promote tumour progression (Aguilar-Aragon, 2020).

Tissue mechanical properties modulate cell extrusion in the Drosophila abdominal epidermis

The replacement of cells is a common strategy during animal development. In the Drosophila pupal abdomen, larval epidermal cells (LECs) are replaced by adult progenitor cells (histoblasts). Previous work showed that interactions between histoblasts and LECs result in apoptotic extrusion of LECs during early pupal development. Extrusion of cells is closely preceded by caspase activation and is executed by contraction of a cortical actomyosin cable. This study identified a population of LECs that extrudes independently of the presence of histoblasts during late pupal development. Extrusion of these LECs is not closely preceded by caspase activation, involves a pulsatile medial actomyosin network, and correlates with a developmental time period when mechanical tension and E-cadherin turnover at adherens junctions is particularly high. This work reveals a developmental switch in the cell extrusion mechanism that correlates with changes in tissue mechanical properties (Michel, 2020).

This study has analyzed the replacement of LECs by adult histoblasts in the pupal abdominal epidermis in Drosophila. Previous work has shown that the interaction between the growing nest of histoblasts and LECs is required for LEC extrusion. LEC extrusion is closely preceded by caspase activation and involves contraction of a supracellular actomyosin cable. This study has identified a second hotspot of cell extrusion at the dorsal midline of the pupal abdomen. Cell extrusion at the dorsal midline occurs mainly late during development and is independent of histoblasts, but depends on the presence of neighboring LECs. LEC extrusion is preceded by a uniform rise in caspase activity throughout the larval tissue and involves pulsatile contractions of a medial actomyosin network. Mechanical tension on adherens junctions of LECs increases during pupal development. High mechanical tension during late pupal development depends in part on caspase activity and correlates with a high turnover of E-cadherin. E-cadherin turnover depends on Dynamin-dependent endocytosis and blocking Dynamin severely affects LEC extrusion. This work reveals a novel, histoblast-independent mechanism by which LECs extrude during late pupal development (Michel, 2020).

Extrusion of LECs located at the border of histoblast nests during early pupal development involves the remodeling and weakening of adherens junctions between the extruding LEC and its neighboring cells, and the accumulation of Myosin II at the cortices of both the extruding LEC and its neighboring cells. Cortical Myosin II accumulation in neighboring cells results in the formation of a supracellular actomyosin cable. Contraction of this cable facilitates the extrusion of the LEC (Teng, 2017). The data indicate that extrusion of LECs at the dorsal midline during late pupal development involves a distinct mechanism. In addition to a cortical pool of Myosin II and F-actin, Myosin II and F-actin was found also in an apical medial region of the extruding LECs. Similar accumulations of Myosin II and F-actin have been reported in several other epithelia, including in the Drosophila embryo in ventral furrow cells of the epidermis and in amnioserosa cells. The medial actomyosin network is pulsatile in these cells resulting in concomitant fluctuations in apical cell area. Similarly, it was found that in extruding LECs located in the vicinity of the dorsal midline, apical cell area fluctuates and apical constriction correlates with increased medial Myosin II intensity. Over longer time periods, apical cell area shrinks, indicating a ratchet-like mechanism. A similar ratchet-like contraction mechanism is involved in the Drosophila embryo during neuroblast delamination (Simoes, 2017) and the extrusion of amnioserosa cells. The reason for these different extrusion mechanisms of LECs during early and late pupal development remains unclear. It is noted, however, that mechanical tension on LECs increases approximately 5- to 10-fold between early (20 h APF) and late (27 h APF) pupal development, suggesting that overall tissue mechanical tension influences the mechanism by which cells extrude (Michel, 2020).

How is the extrusion of LECs at the dorsal midline initiated? Several mechanisms have been described that drive cell extrusion, including cell competition and tissue crowding. These two mechanisms, however, appear not to play a major role in the extrusion of LECs located at the dorsal midline of the pupal abdominal epidermis. Tissue crowding induces, for example, caspase-dependent cell extrusion at the dorsal midline in the pupal notum. Experimental induction of clones expressing an activated form of Ras, RasV12, leads to the crowding of neighboring wild-type cells and to their extrusion. In the abdominal epidermis, the growth or migration of histoblasts could result in crowding of the LECs. During the analyzed developmental time period, the average area of non-extruding LECs located at the dorsal midline, however, remains nearly constant, indicating that these LECs do not become crowded. Moreover, LECs located at the dorsal midline still extrude when histoblast division is halted or histoblasts are genetically ablated, providing further evidence that tissue crowding is not involved in the extrusion of these cells. Finally, cell competition refers to the out competition of slow-growing cells by fast-growing cell. Again, the finding that LEC extrusion at the dorsal midline is independent of histoblast division, strongly argues that cell competition does not have a major influence on the extrusion of these LECs (Michel, 2020).

During early pupal development, local interactions between histoblasts and LECs trigger caspase activation in LECs. Caspase activation then closely precedes extrusion of LECs, cell by cell, at the border of histoblast nests, indicating that caspase activation triggers extrusion (Teng, 2017). In LECs at the dorsal midline, by contrast, caspase activity, as measured by Apoliner cleavage, shows a population-wide fluctuation during pupal development. Caspase activity peaks during the observed developmental time period twice, around 20 h APF and around 26-29 h APF. Caspase activation in these cells does not appear to trigger their extrusion, because the first peak of caspase activity at around 20 h APF takes place several hours before LECs at the dorsal midline start to extrude. Interestingly, the second rise of caspase activity between 24 h and 26 h APF precedes the extrusion of LECs at the dorsal midline. Taken together with the observation that LEC extrusion at the dorsal midline requires caspase activity, these results indicate that the second rise of caspase activity is a prerequisite for LECs to extrude at the dorsal midline. How the population-wide fluctuations of caspase activity in LECs at the dorsal midline arise will be interesting to explore (Michel, 2020).

This study found that mechanical tension acting on LEC adherens junctions increases during pupal development. Increased mechanical tension is independent of the proliferation of histoblasts, but requires caspase activity, which in turn is necessary for LEC extrusion. It is speculated that LEC extrusion during late pupal development may, in part, facilitate the increased mechanical tension on adherens junctions of the remaining LECs. Mechanical tension may influence cell extrusion in different ways and by different mechanisms. In the pupal notum, for example, a similar developmental increase in mechanical tension has been observed, but, unlike the situation in the pupal abdomen, increased mechanical tension correlates with fewer cell extrusions. This different behavior may result from the different characteristics of the two epithelia. In the notum, cell extrusion occurs within a diploid, proliferating tissue with columnar cells. On the other hand, LECs of the pupal abdomen are polyploid, non-proliferating and squamous in shape. In these respects, the abdominal LECs resemble more the amnioserosa cells of the embryo (Michel, 2020).

How could an increase in mechanical tension on LEC adherens junctions facilitate extrusion? LEC extrusion involves the remodeling of adherens junctions of the extruding cell (Teng, 2017). Adherens junction remodeling depends on E-cadherin. The turnover of E-cadherin was found to have increased at 27 h APF compared with 20 h APF, thus positively correlating with increased mechanical tension. The turnover of E-cadherin depends on lateral diffusion on the plasma membrane and on trafficking between a cytosolic, presumably vesicular, pool of E-cadherin and a junctional pool of E-cadherin. The exchange between these two pools requires Dynamin-dependent endocytosis. Consistent with this notion, E-cadherin turnover was reduced in LECs of flies mutant for shibire¹, which encodes Dynamin. Interestingly, extrusion of LECs in the lateral region of the abdominal segments is nearly blocked in shibire¹ mutants. Although the possibility cannot be excluded that Dynamin-dependent endocytosis of proteins other than E-cadherin are important for LEC extrusion, these data lend support for the idea that high E-cadherin turnover is required for extrusion of these LECs. Surprisingly, a subset of LECs, those located in medial regions of the segment, do extrude in shibire1 mutants. These findings reveal an unexpected complexity of LEC extrusion that will require future work to be resolved (Michel, 2020).

The following model is proposed of how LECs in the pupal abdominal epidermis extrude. During early pupal development (20 h APF) mechanical tension on LEC junctions is low and E-cadherin turnover is low. LECs extrude only if they receive a specific 'extrusion signal' from the histoblasts. Extrusion is mediated by cortical actomyosin, which forms a contractile supracellular cable driving apical cell constriction (Teng, 2017). Later during pupal development (27 h APF), mechanical tension on LEC junctions, E-cadherin turnover and caspase activity are high. Stochastic extrusions of LECs commence, which are mediated by a pulsatile medial actomyosin network that leads to ratchet-like apical cell contractions (Michel, 2020).

Collectively stabilizing and orienting posterior migratory forces disperses cell clusters in vivo

Individual cells detach from cohesive ensembles during development and can inappropriately separate in disease. Although much is known about how cells separate from epithelia, it remains unclear how cells disperse from clusters lacking apical-basal polarity, a hallmark of advanced epithelial cancers. Using live imaging of the developmental migration program of Drosophila primordial germ cells (PGCs), this study shows that cluster dispersal is accomplished by stabilizing and orienting migratory forces. PGCs utilize a G protein coupled receptor (GPCR), Tre1, to guide front-back migratory polarity radially from the cluster toward the endoderm. Posteriorly positioned myosin-dependent contractile forces pull on cell-cell contacts until cells release. Tre1 mutant cells migrate randomly with transient enrichment of the force machinery but fail to separate, indicating a temporal contractile force threshold for detachment. E-cadherin is retained on the cell surface during cell separation and augmenting cell-cell adhesion does not impede detachment. Notably, coordinated migration improves cluster dispersal efficiency by stabilizing cell-cell interfaces and facilitating symmetric pulling. This study demonstrates that guidance of inherent migratory forces is sufficient to disperse cell clusters under physiological settings and present a paradigm for how such events could occur across development and disease (Lin, 2020).

The developmental migration of Drosophila PGCs is an excellent model to study how actomyosin contractility is deployed to separate cells from non-epithelial clusters. PGCs are a group of 30-40 cells born at the posterior of the embryo. During gastrulation, PGCs are swept into the interior of the embryo, where they reside as a tight cluster in a rosette configuration enveloped by the endoderm. PGCs subsequently separate and individually transmigrate through the endoderm as it undergoes a developmentally programmed epithelial-to-mesenchymal transition (EMT). How cluster separation is achieved mechanistically remains elusive. Known autonomous proteins required for PGC cluster dispersal are the orphan G-protein-coupled receptor (GPCR), Tre1 and its associated Gβγ subunit, consisting of Gβ13F and Gγ1, as well as the small Rho GTPase, RhoA, suggesting the involvement of contractile forces. However, actomyosin dynamics during PGC cluster dispersal remain uncharacterized and how Tre1 influences the spatiotemporal dynamics of these networks is unknown (Lin, 2020).

This work harnesses two-photon live imaging to provide an in vivo description of how actomyosin contractility is deployed to disperse cell clusters lacking apical-basal polarity under physiological conditions. In contrast to current models of epithelial delamination, cluster dispersal does not involve a sustained downregulation of cell-cell adhesion or augmented force production and is surprisingly robust to increased levels of adhesion. Rather, inherent migratory forces are co-opted to liberate cells. This is accomplished through the sensing of a directed migration cue via the GPCR, Tre1. Tre1 signaling stabilizes and orients migratory polarity radially from the cell cluster, thereby positioning posterior myosin II dependent contractile forces towards cell-cell interfaces in the cluster interior. This collective radial polarity stabilizes cell-cell interfaces and enables symmetric tugging, increasing the efficiency of cluster dispersal (see PGC clusters disperse by directing individual migratory polarity outward to collectively remove cell-cell adhesions). Symmetric tugging, however, is not absolutely necessary for cell separation, as individual WT PGCs can still detach from tre1 PGC clusters, albeit less efficiently. Subsequent detachment requires sustained pulling on cell-cell adhesions provided by a stable migratory polarity. Thus, randomly migrating cells, equally capable of contractile force production, are unable to separate because they do not pull on cell-cell adhesions in a given orientation for a sufficient period of time. A caveat to this model is that this study has not directly shown that migrating PGCs exert posterior pulling forces, as this is technically challenging at the depth where PGC cluster dispersal occurs. However, posterior pulling forces have been clearly demonstrated in various cell types utilizing a rearward driven 3D migration mode which closely resembles PGC migration in Drosophila (Lin, 2020).

Mechanistically, the migration-based cluster dispersal mechanism described in this study harbors many commonalities with hepatocyte growth factor (HGF) mediated epithelial scattering. Myogenic precursors induced to delaminate by ectopic application of HGF maintain expression of N-cadherin, the cardinal adhesion molecule originally linking them to the dermomyotome. Similarly, HGF induced scattering of Madin-Darby canine kidney (MDCK) cells does not involve direct alterations in E-cadherin. Instead, HGF promotes motility and strengthens cell-ECM attachment through integrins, which in turn generate a local increase in tension on cell-cell adhesions until they are physically disrupted. PGCs, on the other hand, continue to utilize E-cadherin to adhere to another cellular substrate, the surrounding endoderm, to pull away from each other. For both PGCs and HGF stimulated MDCK cells, the absence of free space to migrate is sufficient to block dispersal (Lin, 2020).

Cell ensembles frequently exhibit collective migration during development and disease. How is group cohesion maintained if contractile migratory forces are sufficient to disrupt cell-cell adhesion? In collectively migrating squamous cell carcinoma (SSC) cells, primary colorectal cancer explants, and Xenopus neural crest, actomyosin contractility is enriched along the group perimeter and is actively suppressed from cell to cell interfaces in SSC cells to prevent cell detachment. This suppression relies on Discoidin domain receptor 1 (DDR1), which acts in a noncanonical manner at cell-cell interfaces by recruiting Par3 and Par6 to control the localization of RhoE to antagonize RhoA activity. Depletion of DDR1 and other members of the complex result in elevated levels of active Myosin II at cell-cell margins and individual cell migration away from the group, leading to group dispersal. This is strikingly similar to the mechanism uncovered in this study during PGC cluster dispersal and that of HGF mediated cell scattering. Recent work has also revealed an alternative strategy to reduce RhoA signaling at cell-cell junctions of collectively migrating SSC cells through Snail dependent expression of claudin-11, which activates Src to recruit p190RhoGAP to cell-cell interfaces (Li, 2019). In collectively migrating Drosophila border cells, E-cadherin tension is also reduced in the interior of the migrating cell cluster. Thus, while the suppression of contractile forces at cell-cell contacts appears to be a general principle to ensure cohesion in collectively migrating cell populations, PGCs actively direct actomyosin contractility towards cell-cell interfaces to separate. It will be interesting to assess whether pathological cell aggregates can be coaxed to disperse through a similar mechanism (Lin, 2020).

This work demonstrates that cluster dispersal can be driven by the concerted reorientation of migratory actomyosin forces towards cell-cell interfaces. For PGCs, this is accomplished by sensing a directed migration cue through a GPCR, Tre1. However, in principle, any migratory cue, such as ECM or substrate stiffness, could be sufficient to mediate this reorientation. Subsequent cell-cell separation does not require any alterations of the inherent actomyosin forces driving migration. Rather, these forces must be applied on cell-cell junctions for a sufficient period of time, highlighting a potential safeguard against erroneous cell detachment. An open question is the identity of the Tre1 ligand. Given that PGCs are directed to migrate toward the endoderm within a tightly enclosed pocket, the ligand is likely to be surface bound, which would concur with the known roles of Tre1 in orienting neuroblast division56 and immune cell extravasation. Overall, it is anticipated that directed motility-based separation will have general relevance for individual cell detachment events from clustered cell groups lacking apical-basal polarity in development and disease (Lin, 2020).

Regulation of apical constriction via microtubule- and Rab11-dependent apical transport during tissue invagination

The formation of an epithelial tube is a fundamental process for organogenesis. During Drosophila embryonic salivary gland (SG) invagination, Folded gastrulation (Fog)-dependent Rho-associated kinase (Rok) promotes contractile apical myosin formation to drive apical constriction. Microtubules (MTs) are also crucial for this process and are required for forming and maintaining apicomedial myosin. However, the underlying mechanism that coordinates actomyosin and MT networks still remains elusive. This study shows that MT-dependent intracellular trafficking regulates apical constriction during SG invagination. Key components involved in protein trafficking, such as Rab11 and Nuclear fallout (Nuf), are apically enriched near the SG invagination pit in a MT-dependent manner. Disruption of the MT networks or knockdown of Rab11 impairs apicomedial myosin formation and apical constriction. MTs and Rab11 are required for apical enrichment of the Fog ligand and the continuous distribution of the apical determinant protein Crumbs (Crb) and the key adherens junction protein E-Cadherin (E-Cad) along junctions. Targeted knockdown of crb or E-Cad in the SG disrupts apical myosin networks and results in apical constriction defects. These data suggest a role of MT- and Rab11-dependent intracellular trafficking in regulating actomyosin networks and cell junctions to coordinate cell behaviors during tubular organ formation (Le, 2021).

MTs have a crucial role in stabilizing apical myosin during epithelial morphogenesis both in early Drosophila embryos and in the Drosophila SG. In the SG, MTs interact with apicomedial myosin via Short stop, the Drosophila spectraplakin, emphasizing a direct interplay between the MT and the apical myosin networks. The data of this study reveal another key role of MTs in regulating protein trafficking to control the apical myosin networks during tissue invagination. During SG invagination, a network of longitudinal MT bundles is observed near the invagination pit. These data show apical enrichment of Rab11 in the same area is MT dependent and that this enrichment is important for forming the apicomedial myosin networks, suggesting a link between localized intracellular trafficking along MTs to apical myosin regulation (Le, 2021).

The dorsal/posterior region of the SG, where Rab11 is apically enriched in a MT-dependent manner, correlates with localized Fog signaling activity that promotes clustered apical constriction. Disruption of MTs or Rab11 knockdown reduces Fog signals in the apical domain of SG cells and causes dispersed Rok accumulation and defective apicomedial myosin formation. It is consistent with a previous study that the absence of Fog signal results in dispersed apical Rok and defects in apicomedial myosin formation. It is therefore proposed that MT- and Rab11-dependent apical trafficking regulates Fog signaling activity to control apical constriction during epithelial tube formation through transporting the Fog ligand. As recycling of membrane receptors to the cell surface plays an important role in regulating overall signaling activity, it is possible that Rab11 is involved in recycling the as yet unidentified SG receptor(s) of Fog to regulate Fog activity in the SG. Indeed, several GPCRs are recycled via Rab11. During epithelial invagination in early Drosophila embryogenesis, the concentration of the Fog ligand and receptor endocytosis by 'β-arrestin-2 have been shown as coupled processes to set the amplitude of apical Rho1 and myosin activation. It is possible that the movement of Fog receptor(s) that have internalized as a stable complex with 'β-arrestin is recycled back to the cell surface by Rab11. The Fog signaling pathway represents one of the best understood signaling cascades controlling epithelial morphogenesis. Although best studied in Drosophila, the pathway components have also been identified in other insects, suggesting a more widely conserved role for Fog signaling in development. Further work needs to be done to fully understand the regulatory mechanisms underlying the trafficking of Fog and its receptor(s) during epithelial morphogenesis (Le, 2021).

Analysis of apicomedial myosin shows that reduced Rab11 function not only causes a decrease of the myosin intensity but also causes myosin to be dispersed rather than forming proper myosin web structures in the apicomedial domain of SG cells. These data support the idea that Rab11 function is required for both concentration and spatial organization of apicomedial myosin. This can be explained by the combined effect of multiple cargos that are transported by Rab11, including Fog, Crb, and E-Cad. Time-lapse imaging of myosin will help determine how the dynamic behavior of apicomedial myosin is compromised when Rab11 function is disrupted (Le, 2021).

During branching morphogenesis in Drosophila trachea, MTs and dynein motors have a critical role in the proper localization of junctional proteins such as E-Cad. This is consistent with the observations with MT-dependent uniform distribution of E-Cad at adherens junctions in the invaginating SG, suggesting a conserved role of MT-dependent intracellular trafficking in junctional remodeling and stabilization during epithelial tube formation. The data further suggest that the MT networks and Rab11 have key roles in apical distribution of Crb and E-Cad in the SG and that proper levels of apical and junctional proteins are important for apical constriction during SG invagination. Based on these data, it is proposed that MT- and Rab11-dependent apical trafficking of Crb and E-Cad is critical for apical constriction during SG invagination. Alternatively, MTs have an additional role in assembling/anchoring these apical components through the regulation of unidentified molecules. Recent studies in Drosophila mesoderm invagination showed that MTs help establish actomyosin networks linked to cell junction to facilitate efficient force transmission to promote apical constriction. In Ko (2019), however, MT-interfering drugs and RNAi of CAMSAP end-binding protein were used to prevent MT functions and the effect cannot be directly compared with the current data where spastin was used to sever existing MTs. Direct monitoring of MT-dependent transport of Crb and E-Cad during SG invagination will help clarify the mechanism (Le, 2021).

On knockdown of crb or E-Cad, less prominent apicomedial myosin web structures are observed in invaginating SGs, suggesting a requirement of Crb and E-Cad in proper organization of apical actomyosin networks during SG tube formation. Crb acts as a negative regulator of actomyosin dynamics during Drosophila dorsal closure and during SG invagination. It is possible that proper Crb levels are required for modulating myosin activity both in the apicomedial domain and at junctions during SG invagination. Both of which contribute to apical constriction and cell rearrangement, respectively. Anisotropic localization of Crb and myosin was observed at the SG placode boundary, where myosin accumulates at edges where Crb is lowest. Planar polarization of Rok at this boundary is modulated through phosphorylation by Pak1 downstream of Crb. A further test will help understand whether and how Crb might affect junctional myosin dynamics and SG invagination. As contractile actomyosin structures exert forces on adherens junction to drive apical constriction, it is speculated that apical constriction defects on E-Cad RNAi might be due to reduction of cell adhesion and/or of improper force transmission. It will be interesting to determine if the coordination of apical and junctional proteins and apical cytoskeletal networks through intracellular trafficking is conserved during tubular organ formation in general (Le, 2021).

Dhc is also apically enriched in the dorsal/posterior region of the invaginating SG. The data show that knockdown of Dhc64C not only affects Rok accumulation and apicomedial myosin formation but also disrupts MT organization in the SG. These data are consistent with previous findings that cytoplasmic dynein is associated with cellular structures and exerts tension on MTs. For example, dynein tethered at the cell cortex can apply a pulling force on the MT network by walking toward the minus end of a MT. Dynein also scaffolds the apical cell cortex to MTs to generate the forces that shape the tissue into a dome-like structure. In interphase cells, the force generated by dynein also regulates MT turnover and organization (Le, 2021).

In klar mutants, on the other hand, MT organization is not affected in the SG, suggesting that reduction of dynein-dependent trafficking by loss of klar does not cause changes in the MT networks. Notably, although the intensity of apicomedial myosin does not change on Dhc64C knockdown or in the klar mutant background, formation of apicomedial myosin web structures is affected. These data suggest a possible scenario that dynein function is not required for myosin concentration in the apical domain but is only needed for the spatial organization of apicomedial myosin. However, it cannot be ruled out that the zygotic knockdown of Dhc64C by RNAi is not strong enough to affect the intensity of apicomedial myosin. Dhc64C has strong maternal expression and is essential for oogenesis and early embryo development. Embryos with reduced maternal and zygotic pools of Dhc64C showed a range of morphological defects in the entire embryo, some of which were severely distorted. Precise roles for dynein and dynein-dependent trafficking in regulating apicomedial myosin formation remain to be elucidated (Le, 2021).

Golgi-dependent reactivation and regeneration of Drosophila quiescent neural stem cells

The ability of stem cells to switch between quiescent and proliferative states is crucial for maintaining tissue homeostasis and regeneration. In Drosophila, quiescent neural stem cells (qNSCs) extend a primary protrusion, a hallmark of qNSCs. This study found that qNSC protrusions can be regenerated upon injury. This regeneration process relies on the Golgi apparatus that acts as the major acentrosomal microtubule-organizing center in qNSCs. A Golgi-resident GTPase Arf1 and its guanine nucleotide exchange factor Sec71 promote NSC reactivation and regeneration via the regulation of microtubule growth. Arf1 physically associates with its new effector mini spindles (Msps)/XMAP215, a microtubule polymerase. Finally, Arf1 functions upstream of Msps to target the cell adhesion molecule E-cadherin to NSC-neuropil contact sites during NSC reactivation. These findings have established Drosophila qNSCs as a regeneration model and identified Arf1/Sec71-Msps pathway in the regulation of microtubule growth and NSC reactivation (Gujar, 2023).

This work has established Drosophila qNSC protrusion as a regeneration model. It shares two features with other regeneration models, including adult rat retinal ganglion cells (RGCs) and adult C. elegans motor neurons. First, Drosophila qNSCs exhibit age-dependent decline in regeneration capability. This is similar to age-dependent decline in axon regeneration capacity after axotomy in C. elegans mechanosensory neurons and in corticospinal and rubrospinal neurons in the adult mammalian CNS. Second, alterations in microtubule dynamics affect regeneration. The effects of microtubule dynamics on qNSC regeneration are in line with findings in the mammalian CNS neurons that microtubule manipulation promotes axon regeneration after injury (Gujar, 2023).

One interesting aspect of Drosophila qNSC regeneration is that the distal tip of the protrusion does not degenerate after severing but is capable of the regeneration similar to the proximal end. Likely, certain cues or signals from the neuropil to which the protrusion is attached play an important role in the regeneration. It would be worthwhile to understand how the damaged plasma membrane is repaired during qNSC regeneration. Interestingly, re-sealing of the plasma membrane during axon regeneration is dependent on actin and calcium (Gujar, 2023).

This study shows that Golgi functions as a MTOC in qNSCs. The centrosomes may take over Golgi as the MTOC when NSCs re-enter the cell cycle as the proximity between the immature centrosomes and the apical Golgi likely facilitates this transition. Golgi is required for the regeneration of qNSC protrusions upon injury, likely via microtubule growth. Similarly, moderately stable microtubules are required for efficient axon regeneration in central neurons following spinal cord injury. The anterograde transport of cargo, such as mitochondria and vesicles, is also important for their delivery to lesioned axonal tip for axon regeneration. Although it was previously thought that qNSCs retract their protrusions before cell-cycle re-entry, recent work has shown that qNSCs can retain their protrusions throughout the first post-reactivation division (Gujar, 2023).

We also observed that the primary protrusion of larval brain NSCs persists during reactivation and subsequently the first division, although it remains unknown if this applies to most of the reactivating NSCs. Observations also suggest that regeneration of the protrusion precedes NSC reactivation at the event of injury (Gujar, 2023).

This study has identified Arf1 and Sec71 as key regulators of acentrosomal microtubule nucleation, growth, and orientation in the primary protrusion of Drosophila qNSCs. The role of Arf1 in regulating microtubule growth is unexpected, as it is well established that Arf1-6 proteins are critical for membrane trafficking, while Arf-like (Arl) proteins such as Arl2 are important for microtubule functions. Interestingly, mammalian Arf1 recruits tubulin-binding cofactor E (TBCE), which is known to associate with the microtubule regulator Arl2 during microtubule polymerization, to the Golgi, and Arf1 overexpression increases tubulin abundance in motor neurons (Gujar, 2023).

It will be particularly interesting to investigate a similar microtubule regulatory role for Arf1 in other Drosophila cell types as well as in different organisms (Gujar, 2023).

This study reports that in Drosophila qNSC regeneration, the distal tip is capable of the regeneration similar to the proximal end, and focus was placed on the intrinsic role of Arf1 and Msps in regeneration. It remains to be tested whether they could also function in neuropil to facilitate the regeneration of the distal protrusion. EB1-GFP comets going into PIS Golgi and coming out from it in a few seconds were assigned as 'not from Golgi,' which may underestimate the number of Golgi-derived EB1-GFP comets. Furthermore, due to technical limitations, it was not possible to pinpoint the changes in microtubule dynamics or cargo transport in qNSCs after injury. Further studies are warranted to better understand how microtubule growth and dynamics enhance qNSC regeneration (Gujar, 2023).

Tethering Piezo channels to the actin cytoskeleton for mechanogating via the cadherin-beta-catenin mechanotransduction complex

The mechanically activated Piezo channel plays a versatile role in conferring mechanosensitivity to various cell types. However, how it incorporates its intrinsic mechanosensitivity and cellular components to effectively sense long-range mechanical perturbation across a cell remains elusive. This study shows that Piezo channels are biochemically and functionally tethered to the actin cytoskeleton via the cadherin-beta-catenin mechanotransduction complex, whose perturbation significantly impairs Piezo-mediated responses. Mechanistically, the adhesive extracellular domain of E-cadherin interacts with the cap domain of Piezo1, which controls the transmembrane gate, while its cytosolic tail might interact with the cytosolic domains of Piezo1, which are in close proximity to its intracellular gates, allowing a direct focus of adhesion-cytoskeleton-transmitted force for gating. Specific disruption of the intermolecular interactions prevents cytoskeleton-dependent gating of Piezo1. Thus, a force-from-filament model is proposed to complement the previously suggested force-from-lipids model for mechanogating of Piezo channels, enabling them to serve as versatile and tunable mechanotransducers. This study shows biochemical and functional findings to demonstrate that Piezo channels are physically linked to the actin cytoskeleton via the cadherin-β-catenin-vinculin mechanotransduction complex and the direct interaction of E-cadherin to key mechanogating domains of Piezo1, prompting a proposed force-from-filament or tether model for the Piezo channel. In line with this proposed gating model, previous studies have shown that pharmacological disruption of the actin cytoskeleton impairs Piezo1-mediated currents evoked by cell indention or elastomeric pillar deflection at cell-substrate contact points (Wang, 2022).

Several lines of evidence suggest that the regulatory effect of E-cadherin on Piezo1 might not depend on E-cadherin's homotypic interactions between neighboring cells. First, the interaction and functional regulation of Piezo is preserved in the E-cadherin-W2 mutant, which is defective in homotypic dimerization. Second, Piezo1-E-cadherin interactions and functional regulation do not depend on high concentration of Ca2+, which is required for E-cadherin's homotypic dimerization. Third, functional potentiation of Piezo1-mediated currents by E-cadherin could be detected in individual cells without contacting cells and in cell-attached membranes. Together, these data demonstrate that E-cadherin-mediated adhesions junctions might not be necessarily needed for the proposed tethered gating of Piezo channels. Instead, F-actin-derived force transduction might modulate the mechanosensitivity of Piezo channels. In line with this, endogenous traction forces have been reported to locally activate Piezo1 (Wang, 2022).

On the basis of the identification of the transmembrane gate located in the inner helix (IH) and the cytosolic lateral plug gates that physically block the three lateral ion-conducting portals, it is proposed that Piezo channels might use a dual-gating mechanism, in which the transmembrane gate is dominantly controlled by the top extracellular cap, while the lateral plug gates are controlled by the peripheral blade-beam apparatus through a plug-and-latch mechanism. The cap domain is embedded in the center of the nano-bowl shaped by the curved Piezo1-membrane system and sits right on the top of the transmembrane pore of Piezo1 (see A proposed mechanogating model for Piezo channel integrating both the cadherin-mediated tether model and the intrinsic force-from-lipids model). Structural analysis has revealed that motion of the cap domain is strictly coupled to the transmembrane gate residing in the pore-lining inner helix. Furthermore, either deleting or crosslinking the cap completely abolished mechanical activation of Piezo1. These studies have led to a proposal that the cap domain functions as a critical mechanogating domain to predominantly control the transmembrane gate. Following the transmembrane pore, Piezo channels use three lateral portals equipped with three spliceable lateral plug gates as their intracellular ion-conducting routes. Interestingly, the three lateral plug gates physically block the lateral portals and are strategically latched onto the central axis for coordinated gating. Remarkably, this study has identified that the extracellular ectodomain of E-cadherin directly interacts with the extracellular cap domain of Piezo1, while its cytosolic tail with the TM region might interact with the cytosolic linker and CTD of Piezo1, which are in close proximity to the lateral plug gates. Thus, a physical interaction of E-cadherin with Piezo1 might allow a direct focus of cytoskeleton-transmitted force on the extracellular cap domain and the intracellular linker and CTD to gate the transmembrane gate and cytosolic lateral plug gates, respectively, providing a structural basis for a tether model for mechanogating of both gates. Furthermore, the Cap domain plays a critical role in determining the inactivation kinetics of Piezo channels. The direct interaction of the extracellular domain of E-cadherin with the Cap domain might provide a structural basis underlying the effect of E-cadherin in slowing the inactivation kinetics. Thus, these findings are consistent with the existence of tether-mediated gating of Piezo channel (Wang, 2022).

In addition to the tether model proposed in the present study, the Piezo channel might use its signature bowl-shaped structural feature and intrinsic mechanosensitivity to adopt the force-from-lipids model for mechanogating. Structural analysis reveals large conformational changes of the highly curved blades, which might be converted to the cytosolic lateral plug gates by a lever-like motion of the featured beam structure. The highly curved blade has been shown to undergo reversible flattening at biologically relevant pressures. Furthermore, this study has identified a serial of key mechanotransduction sites along the blade-beam-lateral plug gate, which might constitute an intramolecular mechanotransduction pathway for converting long-range conformational changes from the distal blades to the cytosolic lateral plug gates. Collectively, these studies suggest that the blade-beam apparatus might serve as the molecular basis for a force-from-lipids model to gate the cytosolic lateral plug gates (Wang, 2022).

Given the existence of the structurally and functionally distinct molecular bases for the proposed force-from-filament and force-from-lipids models, it is proposed that the Piezo channel might incorporate these two distinct yet non-exclusive gating models to serve as a versatile mechanotransducer. Using the dual mechanogating models, Piezo channels are not only able to constantly monitor changes in local membrane curvature and tension but also overcome the obstacle of limited membrane tension propagation within an intact cellular membrane to effectively detect long-range mechanical perturbation across a cell or between cells. Indeed, Piezo1 can mediate both localized and whole-cell mechanical responses regardless of whether the mechanical stimuli are either exogenously applied or endogenously originated (Wang, 2022).

The dual mechanogating model might also allow Piezo channels to flexibly tune to variable cellular context and to distinct forms of mechanical stimuli. For instance, the two gating models might act synergistically to sensitize the mechanosensitivity of Piezo channels. In line with this, the mechanosensitivity of Piezo1 was measured higher in intact cell membranes (T50 ~ 1.4 mN/m) than in artificial lipid bilayers (T50 ~ 3.4 mN/m) and membrane blebs that are presumed to lack actin cytoskeleton (T50 ~ 4.5 mN/m). Importantly, the convergence of the tether model and force-from-lipids model at the cytosolic lateral plug gates might allow a synergistic gating. The two models might also be preferentially used in a cellular context-dependent manner. For instance, actin-directed force rather than membrane tension might act dominantly to activate Piezo1 at the cell-substrate interface. In contrast, localized perturbation of membrane tension might mainly activate Piezo1 via the force-from-lipids model (Wang, 2022).

The identification of a direct interaction between the Piezo channel-based and the cadherin-based mechanotransduction systems might open a novel avenue in understanding of cellular mechanotransduction in general. Although this study has focused on E-cadherin to illustrate the proof of principle for the proposed tethered gating model of the Piezo channel, cadherins constitute a large superfamily of adhesion receptors, which are widely expressed in epithelium, endothelium, and the nervous system. Indeed, this study has found that N-cadherin, P-cadherin, and VE-cadherin all biochemically and functionally interact with Piezo1. Furthermore, the demonstration of a physical interaction between Piezo channels and cadherin raises the possibility that Piezo channels might also influence cadherin-mediated mechanotransduction process through at least two foreseeing mechanisms: Piezo channel-mediated Ca2+ signaling (e.g., the observed bell-shaped dose dependence of Piezo1-E-cadherin interaction on Ca2+ and a reciprocal force transmission from Piezo channels to the E-cadherin-mediated mechanotransduction complex. For instance, membrane tension-induced conformational changes of the bowl-shaped Piezo-membrane system might disturb the embedded Cap domain and consequently the ectodomain of E-cadherin, which is involved in both trans and cis dimerization of E-cadherin and the formation of adhesion junctions. Future studies will investigate how the two prominent mechanotransduction systems are integrated to regulate cellular mechanotransduction processes (Wang, 2022).

A definitive proof of the proposed tether model might require structural determination and analysis of the complex composed of Piezo1, E-cadherin, β-catenin, vinculin, and actin cytoskeleton. Furthermore, the physiological importance of the Piezo-cadherin-&betql-catenin-F-actin complex requires further investigation (Wang, 2022).

PROTEIN STRUCTURE

Amino Acids - 1507

Structural Domains

The Drosophila homolog of classic vertebrate cadherins is DE-cadherin. The extracellular domain of DE-cadherin has six cadherin-specific repeats, although the first repeat seems to be cleaved off upon maturation, and the cytoplasmic domain shows significant identity to that of classic vertebrate cadherins. DE-cadherin is distinguishable from its vertebrate counterparts by a large insertion between the last cadherin repeat and the transmembrane domain, with local sequence similarity to Fat, Laminin A chain, Slit, and Neurexin I at the proximal region of the extracellular domain. This domain has two motifs, one a cysteine rich domain similar to Fat EGF repeat 3. The other is a laminin A globular-domain (G-domain) repeat (Oda, 1994).

The enormous size of Cadherin-N is primarily due to the presence of 15 cadherin repeats in the extracellular region, presenting a contrast to the 4 repeats in all vertebrate classic cadherins and the 6 repeats in Shotgun, the Drosophila E-cadherin. Both Drosophila cadherins have insertions of similar sequences between the last extracellular cadherin repeat and the membrane-spanning segment. The insert contains a series of subdomains: an Fcc box (fly classic cadherin box), a cysteine-rich segment (C-rich 1), a laminin A (See Drosophila Laminin A) globular segment (LmA-G), and another cysteine-rich segment (C-rich 2). The Fcc box, comprising 170 amino acids is defined as such because database searches with its sequences identify only the comparable regions of Shotgun as a relative. Similar sequences are not found in vertebrate cadherins. The whole Cadherin-N LmA-G displays 25% sequence identity to mouse laminin A and to the presynaptic transmembrane protein Neurexin. The Cadherin-N cytoplasmic domain is much more similar to those domains of Shotgun and vertebrate classic cadherins with respect to both size and sequence. The intracellular domains of the two Drosophila cadherins and mouse N-cadherin range between 157 and 160 amino acids in length, and have 37% and 46% sequence identity in any combination among them, with the higher figure representing the Cadherin-N identity to murine N-cadherin, and the lower representing Shotgun identity to murine N-cadherin. The degree of sequence conservation between the two Drosophila cadherins (41% identity) is lower than the 63% identity between N- and E-cadherin in the same vertebrate species, as for example, in mice (Iwai, 1997).

date revised:   25 April 2024

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.