zipper

DEVELOPMENTAL BIOLOGY

Embryonic

Dramatic changes in the localization of conventional non-muscle myosin characterize early embryogenesis in Drosophila. During cellularization, myosin is concentrated around the furrow canals that form the leading margin of the plasma membrane as the membrane plunges inward to package each somatic nucleus into a columnar epithelial cell. During gastrulation, there is specific anti-myosin staining at the apical ends of those cells that change shape in regions of invagination. Both of these localizations appear to result from a redistribution of a cortical store of maternal myosin. In the preblastoderm embryo, myosin is localized to the egg cortex (the part of the egg immediately underneath the surface membrane), sub-cortical arrays of inclusions, and, diffusely, to the yolk-free periplasm. At the syncytial blastoderm stage, myosin is found within cytoskeletal caps associated with the somatic nuclei at the embryonic surface. Following the final syncytial division, these myosin caps give rise to the myosin rings observed during cellularization. These myosin localizations and the coincident changes in cell morphology are consistent with a key role for non-muscle myosin in powering cellularization and gastrulation during embryogenesis (Young, 1991).

Peanut and Sep1, a second Drosophila septin identified based on its homolog to yeast septins, colocalize to the leading edge in cellularizing embryos. During the interphase between nuclear divisions 13 and 14, actin first reorganizes from the set of caps over the nuclei to form a hexagonal actin-myosin network at the embryo cortex. During the process of cellularization, the cleavage furrows move into the embryo around each nucleus, with actin and myosin concentrate at their leading edges, and spectrin concentrates slightly behind the leading edges. At the beginning of cellularization, Sep1 also assumes a hexagonal pattern similar to (but apparently less uniform than) that of actin and myosin. The nonuniformity of the Sep1 staining is maintained until the end of cellularization. As cellularization proceeds, Sep1 is concentrated at the leading edges of the advancing cleavage furrows, although some diffuse staining is still observed at the embryo cortex and in the underlying cytoplasm. Both Sep1 and Pnut co-localize at least at the resolution of the light microscope. Examination of double-stained embryos reveal that the septins co-localize with actin and myosin at the very leading edge of the cleavage furrows, in a position distinctly ahead of spectrin (Fares, 1995).

Regulation of cytoskeletal dynamics is essential for cell shape change and morphogenesis. Drosophila embryos offer a well-defined system for observing alterations in the cytoskeleton during the process of cellularization, a specialized form of cytokinesis. During cellularization, the actomyosin cytoskeleton forms a hexagonal array and drives invagination of the plasma membrane between the nuclei located at the cortex of the syncytial blastoderm. Rho, Rac, and Cdc42 proteins are members of the Rho subfamily of Ras-related G proteins that are involved in the formation and maintenance of the actin cytoskeleton throughout phylogeny and in Drosophila. To investigate how Rho subfamily activity affects the cytoskeleton during cellularization stages, embryos were microinjected with C3 exoenzyme from Clostridium botulinum or with wild-type, constitutively active, or dominant negative versions of Rho, Rac, and Cdc42 proteins. C3 exoenzyme ADP-ribosylates and inactivates Rho with high specificity, whereas constitutively active dominant mutations remain in the activated GTP-bound state to activate downstream effectors. Dominant negative mutations likely inhibit endogenous small G protein activity by sequestering exchange factors. Of the 10 agents microinjected, C3 exoenzyme, constitutively active Cdc42, and dominant negative Rho have a specific and indistinguishable effect: the actomyosin cytoskeleton is disrupted, cellularization halts, and embryogenesis arrests. Time-lapse video records of embryos show that nuclei in injected regions move away from the cortex of the embryo, thereby phenocopying injections of cytochalasin or antimyosin. Rhodamine phalloidin staining reveals that the actin-based hexagonal array normally seen during cellularization is disrupted in a dose-dependent fashion. Additionally, DNA stain reveals that nuclei in the microinjected embryos aggregate in regions that correspond to actin disruption. These embryos halt in cellularization and do not proceed to gastrulation. It is concluded that Rho activity and Cdc42 regulation are required for cytoskeletal function in actomyosin-driven furrow canal formation and nuclear positioning (Crawford, 1998).

In mammalian cells, Rho and Cdc42 effectors function antagonistically. In competition are two distinct small GTPase protein-driven processes: the formation of stress fibers drive by Rho and the formation of filopodia driven by Cdc42. In Drosophila, Rho and Cdc42 effectors function antagonistically, but in contrast to the mammalian case the two phenotypes are indistinguishable. By this hypothesis, Rho and Cdc42 effectors function in independent pathways: Rho effector function is required for cellularization and maintenance of the actinomysin hexagonal array, whereas Cdc42 effector function antagonizes this process. Cdc42 effectors may inhibit Rho effector function directly, thereby phenocopying the disruption of Rho function generated by the microinjection of C3 exoenzyme or N19Rho. An alternative hypothesis that explains these data requires a mechanism of Rho subfamily regulation of the actomyosin cytoskeleton in a single pathway that depends only on Rho effector function during Drosophila cellularization. By this hypothesis, Rho and Cdc42 share a common factor that is required for Rho function, but is sequestered by GTP-bound Cdc42 (Crawford, 1998).

What is the role of myosin in cellularization? Since myosin function is necessary for cytokinesis and cellularization, a mechanism whereby myosin is regulated through phosphorylation by a Rho effector provides an attractive model. Because it is likely that force production for cytokinesis continuously requires activated myosin, it would be predicted that inhibition of Rho activity would block further progression of the cellularization front. Thus, agents that interfere with Rho activity would be expected to prevent the activation of myosin function and the formation of myosin bipolar filaments. It is possible that Rho effects on myosin can also explain the observed changes in the organization of actin. Indeed, Rho activity in the bundling of preexisting actin filaments may be directly or indirectly dependent on myosin function. Bipolar myosin filament formation can stimulate the formation and organization of filamentous actin; therefore, inhibition of myosin function may also explain the observed disruption of actin distribution (Crawford, 1998).

Cyclic reorganizations of filamentous actin, myosin II and microtubules in syncytial Drosophila blastoderms have been studied using drug treatments, time-lapse movies and laser scanning confocal microscopy of fixed stained embryos (including multiprobe three-dimensional reconstructions). These observations imply interactions between microtubules and the actomyosin cytoskeleton. They provide evidence that filamentous actin and cytoplasmic myosin II are transported along microtubules towards microtubule plus ends, with actin and myosin exhibiting different affinities for the cell's cortex. These studies further reveal that cell cycle phase modulates the amounts of both polymerized actin and myosin II associated with the cortex. Pseudocleavage furrow formation in the Drosophila blastoderm is analogous to how the mitotic apparatus positions the cleavage furrow for standard cytokinesis, and these findings are related to polar relaxation/global contraction mechanisms for furrow formation (Foe, 2000).

Laser scanning confocal microscope (hereafter LSCM) sectional views are provided of anterior ends of Drosophila embryos showing anaphase and interphase of cycle 9 (before centrosomes, microtubules and nuclei reach the cortex) and cycle 10 (the first round of bud formation and breakdown). In cycle 9, myosin II staining concentrates in a cortical rim during interphase but leaves the cortex during anaphase. Likewise, F-actin concentrates during interphase 9 in a cortical rim; the concentration attenuates greatly during anaphase 9. Throughout interphase 9, with no nuclei/asters near the cortex, cortical myosin II and F-actin co-localize. Similar waxing and waning of cortical F-actin and myosin II, co-localized and synchronized with globally synchronous mitotic cycles, is observed in cycle 8. Migrating nuclei with microtubule arrays reach the cortex 1 minute after interphase 10 begins. As telophase 9 ends and interphase 10 begins, F-actin and myosin II re-accumulate co-localized to high levels in a spatially uniform cortical rim. Two minutes after nuclei reach the cortex, cortical F-actin and myosin II are no longer co-localized but occur in the complementary patterns. Myosin II occurs at high levels between buds, but vacates the cortex where buds now protrude, while F-actin attains high levels precisely on the domes of the buds that myosin II has vacated. During anaphase 10, cortical levels of F-actin and myosin II are globally low. Cortical F-actin re-accumulation begins first near centrosomes at anaphase/telophase. Regardless of cortical fluctuations, high levels of myosin II staining occur throughout the embryo interior. When myosin dissociates from the cortex, it transiently boosts the concentration of myosin immediately beneath the cortex, but does not significantly boost the concentration of internal myosin globally, presumably because it is dispersing into an ocean of cytoplasmic myosin filling these large cells. Throughout the interior cytoplasm, F-actin occurs diffusely, and additionally in particles, but at lower levels than cortically (Foe, 2000).

When migrating nuclei with associated microtubule arrays first reach the periphery early in interphase 10, myosin II staining disappears from a small region immediately above the nuclei and, during the next 3 minutes of interphase and prophase, the holes vacated by myosin II staining enlarge into oblong holes. During metaphase and anaphase, the cortical myosin staining dims, then, during telophase, myosin staining returns brightly to the cortex except near centrosomes where it remains dim. Holes in the cortical myosin pattern, appearing in the cortex precisely when/where microtubule arrays first contact the cortex, are interpreted as implying an interaction between myosin II and microtubules (Foe, 2000).

Four hypotheses have been proposed about consecutive and simultaneous mitotic-cycle-modulated interactions between the cell cortex (the approximately 3 mm deep zone immediately underlying the plasma membrane), F-actin, myosin II, centrosomes and microtubules. When working together, they can explain these experimental results (Foe, 2000).

• Hypothesis H1: mitotic phase modulates the amount of myosin II associated with the cortex and the amount of actin polymerized into filaments at the cortex. Independent of the proximity of centrosomes and/or microtubules, both cortical myosin II and cortical F-actin increase from telophase through interphase then fall through metaphase and anaphase.
• Hypothesis H2: centrosomes, via the polymerization of the microtubule asters that they nucleate, boost the rate of actin polymerization nearby.
• Hypothesis H3: a kinesin-like activity attaches actin filaments to adjacent microtubules and transports actin filaments toward microtubule plus ends; this activity is weaker than the attachment of F-actin to the cortex.
• Hypothesis H4: a kinesin-like activity moves myosin II filaments towards microtubule plus ends; this is stronger than the attachment of myosin to the cell's cortex so that appropriately oriented microtubule arrays contacting the cortex strip myosin II off the plasma membrane.

The same four mechanical hypotheses that are proposed to explain pseudocleavage furrow and bud formation in the fly syncytial blastoderm, if operative in dividing mononucleate cells, could time and initiate localization of the actomyosin components of the contractile ring for cytokinesis. H1, by melting down prior interphase F-actin structures during metaphase, and H2 by causing F-actin to re-polymerize near centrosomes beginning in anaphase, in effect force a redeployment of the cell's actin just prior to beginning the specialized task of cytokinesis. By H1, the cortical concentrations of F-actin and myosin II, having fallen to low concentrations in metaphase-anaphase, rebuild in telophase. But the expanding telophase microtubule asters approaching the cortex at opposite poles of the cell would trigger, via H4, depletion of cortical myosin II filaments near the spindle poles where microtubules impinge, while simultaneously concentrating myosin II filaments by moving them through the cytoplasm towards the cell mid-zone. H4 will thus eventually concentrate myosin II in a three-dimensional disk whose perimeter will become the contractile ring. By H2, actin polymer will form coincidentally with microtubule outgrowth initially most concentrated near centrosomes, followed later in telophase by a migration of F-actin along astral microtubules away from centrosomes (by H3) and toward concentration in a three-dimensional disc-shaped volume centered at the equator where it will co-localize with myosin II. The 'rings' of cortical F-actin seen in late interphase, correspond to where these equatorial disc volumes intersect the cortex in the dividing mononucleate cell. The regions between buds where both F-actin and myosin II are present together at the cortex during interphase (though concentrations of F-actin may be higher elsewhere) constitute the so-called pseudocleavage furrows in the fly syncytial blastoderm. This region would be homologous to the cortex of the cleavage furrow, which constricts during cytokinesis (Foe, 2000 and references therein).

H1-H4 imply a bipolar 'global contraction-polar relaxation' mechanism for positioning the contractile apparatus for cytokinesis. Transient 'polar relaxation' would convert a global cortical contraction into a self-amplifying equatorial contraction. Computer simulations have shown how contracting an initially isotropic actomyosin meshwork into an equatorial belt aligns the filaments parallel to the equator, as in a contractile ring, positioning them to cleave a cell in two by a purse-string contraction. Plus-end transport along astral microtubules of cortical F-actin (H3) and of myosin II filaments (H4) could provide a mechanistic explanation for this oft-hypothesized polar relaxation, while simultaneously causing an increase in equatorial tension (assuming that cortical contractile strength is proportional to co-localized F-actin and myosin II concentrations). Either polar relaxation, or equatorial strengthening, or both together, can set the stage for the kind of actomyosin contraction-based cytokinesis that has been proposed (Foe, 2000 and references therein).

The establishment of the contractile furrow is a mechanism requiring two independent steps: the release in tension at the poles, preceded by (or coincident with) a global increase in cortical tensioning. The global loss of actin and myosin from the cortex during metaphase/anaphase and their return beginning in telophase, which is observed in Drosophila embryos (H1), if phylogenetically general, could underlie the cyclic changes in cortical contractility. Operating together, H1-H4 are potentially capable of implementing a 'global contraction-polar relaxation' mechanism of furrow initiation for cytokinesis. If cortical contractile strength is proportional to co-localized F-actin and myosin II concentrations, then H3 and H4 would bring about equatorial strengthening of the cortical actomyosin meshwork, in principal also implementing 'equatorial stimulation'. Note that, in flattened cells with small asters, the microtubules in the mid-zone between spindle poles are positioned to execute the same actomyosin rearrangements as astral microtubules in spherical cells. Note also that H3 plus H4 can cause equatorial concentration of whatever cytoplasmic actomyosin network a cell contains, focusing internal forces on the furrow cortex with the potential to aid furrow invagination in non-spherical cells. Homologies between cleavage and pseudocleavage, while attractive, come with the caveat that cortical tensioning and microtubule outgrowth both occur earlier in the mitotic cycle during cytokinesis in echinoderms than during pseudocleavage furrow formation in Drosophila syncytia (Foe, 2000 and references therein).

In summary, this study has aimed to deduce, from descriptions of wild-type and drug-perturbed cytoskeletal kinematics of microtubules, F-actin and myosin II in syncytial Drosophila embryos, the specific ways that these filament systems must be interacting. The speculative mechanistic hypotheses that were deduced, H1-H4, are consistent with a large body of circumstantial and partial evidence reviewed above. This machinery, if phylogenetically general, could unify old ideas about cytokinesis with new molecular findings, reconcile polar relaxation with equatorial stimulation models of furrow formation, and homologize cytoskeletal pseudocleavage furrow formation in syncytia with cleavage furrow formation in mononucleate cells. Future revelations can be expected of the molecular details by which the products of an ensemble of key genes (e.g. anillin, centrosomin, diaphanous, KLP-3A, pavarotti, polo kinase, Rac 1, septins, etc.) collaborate to bring about and regulate the interactions between F-actin, myosin II, centrosomes and microtubules (Foe, 2000).

The effects of the absence of puckered (pucE69) and its overexpression were compared on the levels and organization of the actin cytoskeleton and nonmuscle myosin. In pucE69 mutants, the expression of myosin and actin does not change dramatically in the periphery of the cells in lateral regions of the embryo, but these proteins fail to accumulate along the leading edge of the epidermis. Cell shape changes proceed almost normally. In contrast, epidermal cells of embryos overexpressing puc fail to change their shapes and accumulate low levels of spatially disorganized myosin at the leading edge. In these embryos, actin fails to be expressed in the amnioserosa and its levels are reduced in the epidermis. Actin and myosin tend to form clumps in these epidermal cells. Thus puc is an essential component in the control of the different steps of dorsal closure progression and acts by modulating the apical accumulation of actin and myosin at the leading edge. These results correlate with those of the effects of overexpression of DN-Drac1 and Djun mutants, and further suggest a role for Puc in the control of JNK activity over the cytoskeleton (Martin-Blanco, 1998).

To explore the nature of the defects seen in the absence of diaphanous function, wild-type and dia mutant embryos were stained at nuclear cycles 11-13 with the DNA dye DAPI and with an antibody directed against F-actin. In the wild type, nuclei are positioned at the embryo cortex at interphase of nuclear cycles 11-13; a structure referred to as the actin cap is situated between each nucleus and the plasma membrane. During the transition to prophase, filament reorganization results in a concentration of actin at the edge of the caps. At metaphase, the resulting rings of cortical actin, together with associated plasma membrane, invaginate to form metaphase furrows. As viewed from above, actin staining at these furrows appears as a hexagonal array over the embryonic surface. In the sagittal view, actin staining at the metaphase furrow appears as a line between the metaphase nuclei. In dia-deficient embryos, severe structural changes in the actin cytoskeleton are manifested after nuclear cycle 11. Formation of the hexagonal actin arrays is disrupted during prophase and metaphase and there is an absence of actin staining between the metaphase nuclei. Similar patterns of staining are obtained when dia embryos are stained with antibodies directed against anillin (Drosophila gene: Scraps) and Peanut, other components of the metaphase furrow. There is thus a failure in the formation of the metaphase furrow. Consistent with the known role of metaphase furrows in maintaining nuclear organization, the nuclei in dia mutant embryos frequently exhibit abnormal spacing and, in some cases, fuse in subsequent nuclear cycles. These irregularities are readily apparent in contrast to the uniform pattern observed in the wild type. In regions in which cortical actin staining is weak or absent, nuclei are frequently found displaced into the interior of the embryo, although the centrosomes remain at the surface (Afshar, 2000).

To investigate whether the absence of metaphase furrows results from a failure in membrane invagination, dia embryos were stained with antibodies directed against myosin. In wild-type embryos, myosin localizes to the embryonic cortex between the actin caps at each interphase, appears at the tip of the invaginating membrane at prophase and disappears at metaphase. In dia embryos, myosin staining, albeit very weak and irregular, is detected between the actin caps at the cortex during interphase. At prophase, myosin, where detectable, remains at the cortex, with no detectable membrane pinching or invagination. Therefore, despite the presence of myosin at the cortex between actin caps, the membrane invagination that precedes metaphase furrowing is absent in dia embryos (Afshar, 2000).

Immunolocalization was used to determine whether Diaphanous plays a role in the recruitment of anillin and Peanut, a Drosophila septin. In wild-type embryos both anillin and Peanut localize to the embryonic cortex, between the actin caps at interphase. During prophase and metaphase, they localize to the metaphase furrow and their pattern of staining is similar to that of actin. In dia embryos, the staining patterns of both anillin and Peanut are very weak during interphase. Similarly, in dia embryos the localization of both of these proteins is disrupted during prophase and metaphase, when the metaphase furrow is being formed in wild-type embryos. Diaphanous is thus required for recruitment and proper localization of anillin and Peanut as well as myosin to the regions of membrane invagination (Afshar, 2000).

Lecuit, T., Samanta, R. and Wieschaus, E. (2002). Slam encodes a developmental regulator of polarized membrane growth during cleavage of the Drosophila embryo. Dev. Cell 2: 425-436. Medline abstract: 11970893

Zallen, J. A. and Wieschaus, E. (2004). Patterned gene expression directs bipolar planar polarity in Drosophila. Dev. Cell 6: 343-355. Medline abstract: 15030758

Slam servers as a molecular marker for polarized cell behavior revealing functions of Eve, Runt, Myosin II and Bazooka in germband extension

During convergent extension in Drosophila, polarized cell movements cause the germband to narrow along the dorsal-ventral (D-V) axis and more than double in length along the anterior-posterior (A-P) axis. This tissue remodeling requires the correct patterning of gene expression along the A-P axis, perpendicular to the direction of cell movement. A-P patterning information results in the polarized localization of cortical proteins in intercalating cells. In particular, cell fate differences conferred by striped expression of the even-skipped and runt pair-rule genes are both necessary and sufficient to orient planar polarity. This polarity consists of an enrichment of nonmuscle myosin II at A-P cell borders and Bazooka/PAR-3 protein at the reciprocal D-V cell borders. Moreover, bazooka mutants are defective for germband extension. These results indicate that spatial patterns of gene expression coordinate planar polarity across a multicellular population through the localized distribution of proteins required for cell movement (Zallen, 2004).

Polarized cell movement during convergent extension ultimately derives from the asymmetric localization of proteins that direct cell motility. Interestingly, intercalating cells in the Drosophila germband display a polarized localization of the ectopically expressed Slam protein (Lecuit, 2002). Slam is present in a bipolar distribution that correlates spatially and temporally with intercalary behavior. These observations indicate that Slam can serve as a molecular marker for polarized cell behavior. Pair-rule patterning genes expressed in stripes along the A-P axis are necessary for Slam localization and, conversely, altering the geometry of their expression is sufficient to reorient Slam polarity. An endogenous planar polarity in intercalating cells has been shown to be manifested by the accumulation of nonmuscle myosin II at A-P cell borders and Bazooka/PAR-3 at D-V cell borders. Moreover, germband extension is defective in bazooka mutant embryos, supporting a model where molecular polarization of the cell surface is a prerequisite for polarized cell movement. Therefore, differences in gene expression along the A-P axis may direct planar polarity in intercalating cells through the creation of molecularly distinct cell-cell interfaces that differ in migratory potential (Zallen, 2004).

Cell movement during germband extension is oriented along the D-V axis, suggesting a mechanism that restricts the productive generation of motility to dorsal and ventral cell surfaces. Molecules that are asymmetrically localized during convergent extension may therefore contribute to the spatial regulation of cell motility. Interestingly, intercalating cells in the Drosophila germband display a polarized localization of the ectopically expressed Slam protein, a novel cytoplasmic factor required for cellularization in the early embryo (Lecuit, 2002). While proteins such as Armadillo/β-catenin are uniformly distributed at the cell surface, ectopic Slam is enriched in borders between neighboring cells along the A-P axis. This polarized Slam population is present in a punctate apical distribution, coincident with the adherens junction component Armadillo/β-catenin. Therefore, intercalating cells have distinct apical junctional domains that differ in their capacity for Slam association (Zallen, 2004).

Interestingly, the polarized distribution of ectopic Slam protein is spatially and temporally correlated with intercalary behavior. Slam polarity is not observed in Stage 6 embryos prior to the onset of intercalation. Slam accumulation at A-P cell borders first appears in late Stage 7, when cells of the germband initiate intercalation, and reaches its full extent during the period of sustained intercalation in Stage 8. In contrast, Slam is uniformly distributed in cells of the head region and the dorsal ectoderm, tissues which do not undergo intercalary movements. These results indicate that the polarized distribution of ectopic Slam protein is specific to intercalating cells and that Slam can therefore serve as a molecular marker for the visualization of polarized cell behavior (Zallen, 2004).

The enrichment of Slam at borders between neighboring cells along the A-P axis is consistent with two modes of localization: Slam could mark one side of each cell in a unipolar distribution, or Slam could localize to both anterior and posterior surfaces in a bipolar pattern. To distinguish between these possibilities, mosaic embryos were generated where Slam-expressing cells were juxtaposed with unlabeled cells, using the Horka mutation to induce sporadic chromosome loss in early embryos. Slam protein accumulates at anterior and posterior boundaries of mosaic clone, indicating that ectopic Slam protein is targeted to both anterior and posterior surfaces of intercalating cells in a symmetric, bipolar distribution. The bipolar localization of ectopic Slam corresponds well with the bidirectionality of cell movement during germband extension, where cells are equally likely to migrate dorsally or ventrally during intercalation. Bipolar motility is also observed during convergent extension in the presumptive Xenopus and Ciona notochords and in Xenopus neural plate cells in the absence of midline structures (Zallen, 2004).

To extend the spatial and temporal correlation between Slam polarity and cell movement, it was asked if this polarized Slam localization is achieved in mutants that are defective for intercalation. Cell intercalation is dependent on the transcriptional cascade that generates cell fates along the A-P axis, in the direction of tissue elongation and perpendicular to the migrations of individual cells. A-P patterning reflects the hierarchical action of maternal, gap, and pair-rule genes. Cell fate differences along the A-P axis are abolished in embryos maternally deficient for the bicoid, nanos, and torso-like genes (referred to as bicoid nanos torso-like mutants), and these mutant embryos do not exhibit intercalary behavior. Ectopic Slam is correctly targeted to the apical cell surface in bicoid nanos torso-like mutants, but fails to adopt a polarized distribution in the plane of the epithelium (Zallen, 2004).

Downstream of the maternal patterning genes, gap genes establish overlapping subdomains along the A-P axis. A quadruple mutant for the gap genes knirps, hunchback, forkhead, and tailless lacks A-P pattern within the germband while retaining terminal structures. This quadruple mutant exhibits severely reduced cell intercalation, and mutant embryos also display a loss of Slam polarity. The absence of planar polarity in A-P patterning mutants correlates with a more hexagonal appearance of germband cells, in contrast to the irregular morphology of wild-type intercalating cells (Zallen, 2004).

In response to maternal and gap genes, pair-rule patterning genes expressed in narrow stripes act in combination to assign each cell a distinct fate along the A-P axis. In particular, the even-skipped (eve) and runt pair-rule genes are essential for germband extension. This strong requirement for eve and runt during germband extension contrasts with the more subtle effects in mutants for other pair-rule genes such as hairy and ftz. Consistent with these defects in intercalation, eve and runt mutants also display aberrant Slam localization. These results establish a correlation between intercalary behavior and the polarized localization of the ectopic Slam marker (Zallen, 2004).

The Eve and Runt transcription factors ultimately direct Slam polarity and cell intercalation through the transcriptional regulation of target genes. To identify downstream effectors involved in this process, components of the noncanonical planar cell polarity (PCP) pathway, which is required for convergent extension in vertebrates, were examined. Germband extension occurs normally in the majority of embryos lacking the Frizzled and Frizzled2 receptors. Similarly, germband extension is unaffected in the absence of Dishevelled. Moreover, dishevelled mutants exhibit a normal polarization of the Slam marker. These results demonstrate that molecular and behavioral properties of planar polarity in the Drosophila germband do not require Frizzled or Dishevelled function (Zallen, 2004).

The polarized distribution of ectopic Slam in intercalating cells provides the first clue to a molecular distinction between D-V cell interfaces that generate productive cell motility and A-P interfaces that do not. However, endogenous Slam mRNA and protein are not detected during germband extension, indicating that Slam may not play a functional role in cell intercalation. Slam colocalizes with the Zipper nonmuscle myosin II heavy chain subunit during cellularization and when Slam is ectopically expressed at germband extension (Lecuit, 2002). Therefore, the endogenous distribution of myosin II was examined during germband extension in wild-type embryos. During cell intercalation, myosin II is present in a punctate distribution at the apical cell surface, colocalizing with the adherens junction component Armadillo/β-catenin. In Stage 8 embryos, apical myosin II protein accumulates at interfaces between cells along the A-P axis. Slam can enhance this polarized localization when ectopically expressed (Lecuit, 2002), suggesting that Slam and myosin II may associate with a common localization machinery. Myosin II polarity is not apparent in Stage 6 or early Stage 7 embryos that have not begun intercalation, indicating that the enrichment of myosin II at A-P interfaces is specific to intercalating cells (Zallen, 2004).

The localized distribution of myosin II is not as pronounced as that of ectopic Slam, suggesting that additional asymmetries contribute to the polarization of intercalating cells. To identify such proteins, the localization was examined of components implicated in cell polarity in other cell types. In particular, the PDZ domain protein Bazooka/PAR-3 participates in both apical-basal and planar polarity. Bazooka/PAR-3 also exhibits a polarized distribution in intercalating cells. Bazooka, like myosin II, is present in a punctate apical distribution, coincident with the adherens junction component Armadillo/β-catenin. However, in contrast to the accumulation of myosin II at A-P cell interfaces, Bazooka is enriched in the reciprocal D-V interfaces. Bazooka polarity is specific to intercalating cells, where it first appears at the onset of intercalary movements in late Stage 7. Bazooka polarity is not observed in cells of the head region, which do not undergo intercalation, nor is it observed in germband cells following the completion of germband extension at Stage 9 (Zallen, 2004).

To characterize the relationship between cell shape and the polarized localization of cortical proteins, the orientation of cell borders was measured as an angle relative to the A-P axis (with A-P interfaces closer to 90° and D-V interfaces closer to 0° and 180°). Interfaces from embryos stained for Bazooka and myosin II were ranked according to mean fluorescence intensity as a relative measure of protein distribution. These results illustrate that Bazooka and myosin II are enriched in distinct sets of cell-cell interfaces that adopt largely nonoverlapping orientations relative to the A-P axis. This quantitation confirms the visual impression from confocal images and demonstrates that the molecular composition of a cell surface domain is a reliable predictor of its orientation within the epithelial cell sheet (Zallen, 2004).

The polarized localization of Bazooka is abolished in the absence of A-P patterning information in bicoid nanos torso-like mutant embryos. A similar disruption of myosin II polarity is observed in A-P patterning mutants. The A-P patterning system may therefore mediate cell intercalation through the polarized accumulation of cell surface-associated proteins. Bazooka participates in a conserved protein complex containing the atypical PKC (DaPKC), and DaPKC is also enriched in D-V cell interfaces during germband extension (Zallen, 2004).

Zipper and dorsal closure

The small GTPase Rho is a molecular switch that is best known for its role in regulating the actomyosin cytoskeleton. Its role in the developing Drosophila embryonic epidermis during the process of dorsal closure has been investigated. By expressing the dominant negative DRhoA^N19 construct in stripes of epidermal cells, it has been confirmed that Rho function is required for dorsal closure and it is necessary to maintain the integrity of the ventral epidermis. Defects in actin organization, nonmuscle myosin II localization, the regulation of gene transcription, DE-cadherin-based cell-cell adhesion and cell polarity underlie the effects of DRhoA^N19 expression. Furthermore, these changes in cell physiology have a differential effect on the epidermis that is dependent upon position in the dorsoventral axis. In the ventral epidermis, cells either lose their adhesiveness and fall out of the epidermis or undergo apoptosis. At the leading edge, cells show altered adhesive properties such that they form ectopic contacts with other DRhoA^N19-expressing cells (Bloor, 2002).

Tension generated in the amnioserosa and the leading edge of the lateral epidermis independently contributes to the forces that drive dorsal closure. It has been proposed that nonmuscle myosin II activation generates tension in the leading edge and that this causes a leading edge intracellular actomyosin purse-string to shorten. Signaling downstream of RhoGTPase activates nonmuscle myosin II by modulating the level of myosin regulatory light chain phosphorylation. As such, expression of RhoA^N19 in epidermal stripes might disrupt contraction of the leading edge purse-string. Defects in actin and nonmuscle myosin II organization caused by RhoA^N19 expression are first observed at germband extension, up to 2 hours before purse-string formation. Thus, while actin and nonmuscle myosin II are localized at the leading edge in wild-type tissue, a purse-string structure is never formed in leading edge cells that express RhoA^N19. RhoA^N19 expression therefore effectively cuts the leading edge purse-string at multiple sites. This does not necessarily prevent progression of dorsal closure, confirming previous experiments which demonstrate that the integrity of the leading edge is not required for dorsal closure to continue to completion. It is concluded that small independent regions of leading edge in wild-type epidermal stripes can, in conjunction with contraction of the amnioserosa, migrate dorsally with relative normalcy (Bloor, 2002).

The question that arises is how do epidermal cells expressing RhoA^N19 move dorsally in the absence of a leading edge purse-string? These cells could hitchhike, i.e. they are pulled dorsally by the amnioserosa or dragged along with neighboring wild-type cells. Although spread and disorganized, dorsal RhoA^N19-expressing cells do maintain adhesion with wild-type neighbors and this might then allow passive RhoA^N19-expressing cells to move dorsally with wild-type tissue. This is consistent both with the inverse correlation between integrity of the ventral epidermis and the extent to which dorsal closure proceeds, as well as with observations on the distribution of tension at the embryo surface during dorsal closure. Thus, during the time that the epidermis lateral to the leading edge opposes dorsal closure, ventral failure of epidermal integrity (and hole formation) would release the tensional restraints on the remaining lateral epidermis, allowing it to move dorsally with more success. Similarly, in the absence of this release (i.e. the ventral epidermis retains its integrity and opposes dorsal movement of the epidermis), the leading edge is presumably no longer capable of generating sufficient force to drive dorsal closure to completion (Bloor, 2002).

Throughout development, a series of epithelial movements and fusions occur that collectively shape the embryo. They are dependent on coordinated reorganizations and contractions of the actin cytoskeleton within defined populations of epithelial cells. One paradigm morphogenetic movement, dorsal closure in the Drosophila embryo, involves closure of a dorsal epithelial hole by sweeping of epithelium from the two sides of the embryo over the exposed extraembryonic amnioserosa to form a seam where the two epithelial edges fuse together. The front row cells exhibit a thick actin cable at their leading edge. The function of this cable has been tested by live analysis of GFP-actin-expressing embryos in which the cable is disrupted by modulating Rho1 signaling or by loss of non-muscle myosin (Zipper) function. The cable serves a dual role during dorsal closure. It is contractile and thus can operate as a 'purse string,' but it also restricts forward movement of the leading edge and excess activity of filopodia/lamellipodia. Stripes of epithelium in which cable assembly is disrupted gain a migrational advantage over their wild-type neighbors, suggesting that the cable acts to restrain front row cells, thus maintaining a taut, free edge for efficient zippering together of the epithelial sheets (Jacinto, 2002).

To investigate the function of the actin cable during dorsal closure, advantage was taken of rho1 and also zip alleles that produce phenotypes in which the actomyosin cable disassembles part way through the dorsal closure process but in which the overall tissue movement typically does not fail. These mutants offered the opportunity to analyze the effects of cable loss at the cellular level without the complete disruption of tissue architecture. Also, they allowed the use of live GFP-actin embryos to analyze the dynamic cell behavioral response and determine how these behavioral changes influence the capacity of epithelial cells to participate in a coherent forward-sweeping movement (Jacinto, 2002).

Scanning electron micrographs of rho1 and zip embryos that complete dorsal closure reveal significant similarities. In both cases, combinations of amorphic or strong mutations, zip¹/zip^IIX62 or zip¹/zip¹ in the case of zip, and rho1^72O/rho1^72R in the case of rho1, were used. Both zip and rho1 mutant embryos exhibit the same dramatic defects in head involution, a tissue movement distinct from dorsal closure, but a clear, posterior dorsal hole is observed in none of the rho1 mutants, and in only 59% of the zip/zip^IIX62 or 8% of the zip¹/zip¹ embryos. The remaining zip mutants appear to complete dorsal closure successfully; although, frequently, like their rho1 counterparts, these embryos show puckering or segment misalignments along the closed midline seam. Interestingly, puckering has previously been observed in mutants in which epithelial movement is not properly downregulated at the midline (e.g., puckered), suggesting that the actin cable may be regulating cell spreading during some periods of dorsal closure. It is presumed that the phenotypic variation in zip embryos is due to differences in the maternal contribution to the myosin II RNA and protein pool, which must run out around the stage of dorsal closure (Jacinto, 2002).

Phenotypic similarities between zip embryos that complete closure of the dorsal hole and rho1 mutants are not surprising since rho1 and zip have previously been shown to interact genetically in Drosophila. Moreover, a signaling link has also been revealed in mammalian cell culture studies, with Rho-associated kinase (ROCK), a Rho effector, shown to regulate Myosin function, and thus the assembly and maintenance of cable-like stress fibers, by repression of Myosin Light Chain phosphatase and direct activation of the Myosin Light Chain. It is also likely that Rho1 regulates the cytoskeleton via alternative effectors. The kinase Protein kinase related to protein kinase N (see Rho1 Protein Interactions) has been shown to function downstream of Rho1 during Drosophila dorsal closure, and mDia, the murine homolog of Drosophila Diaphanous, has been shown to bind active Rho1 and contribute to the formation of stress fibers in mammalian cells by promoting actin polymerization. However, whether or not these signaling pathways regulate actomyosin contractility during dorsal closure has yet to be demonstrated. Interestingly, Rho1 also interacts with p120^ctn and regulates cadherin-based adherens junctions in the Drosophila embryo, suggesting that the leading edge disorganization seen in rho1 mutants may be not only a consequence of actin cable misregulation but also a result of defective adherens junctions (Jacinto, 2002).

Further similarities between the rho1 and zip embryos were revealed when higher-resolution SEM was used to look at the leading edge cells during the final stages of dorsal closure. In both mutants, the normally straight, and apparently taut, epithelial leading edge now appears to be relatively disorganized and to have lost tension, with front row cells extending increased levels of filopodial and lamellipodial protrusions compared to their wild-type counterparts (Jacinto, 2002).

To observe the dynamic behavior of these cells, time-lapse confocal microscopy was used to analyze the final stages of dorsal closure in living, wild-type embryos versus rho1 and zip embryos, expressing GFP-actin fusion proteins using the GAL4-UAS system. The fusion protein was driven in the epidermis by the epithelial driver e22cGAL4, and only embryos that subsequently closed the dorsal hole are described. As expected from SEM observations, the cytoskeletal architecture of the leading edge that normally characterizes wild-type dorsal closure is lost in both rho1 and zip mutants. The actin cable, which is clearly apparent in the wild-type leading edge from stage 13 onward, fails to assemble or disassembles during the dorsal closure process in rho1 and zip mutants. The disassembly of the actin cable is temporally coincident with a transition from an organized to disorganized leading edge. In both mutants, loss of cable is also coincident with more exuberant filopodial extensions than in wild-type leading edge cells, and these filopodia frequently coalesce to form lamellipodia. In wild-type embryos, lamellipodia are generally more a feature of the final stages of dorsal closure, as opposing epithelial fronts make contact with one another (Jacinto, 2002).

The coalescing of filopodia to form lamellipodia in mutant embryos leads to a broader extent of protrusion per unit length of leading edge cells and an increase of up to 300% in the total protrusion area extending from these cells. It is suggested that the increased level of actin-based protrusions seen in both rho1 and zip embryos at a time when the leading edge cable is disassembling may reflect that the cable serves some function in repressing excessive protrusive activity in these front row cells. The link could simply be due to a greater availability of free actin monomers, but it might also be a consequence of changes in membrane and cortical actin stiffness at the free edge in these cells. In this regard, myosin II has been shown to play a role in maintaining the integrity and stiffness of the cortical cytoskeleton during Dictyostelium morphogenesis. It is not clear from this data whether downregulation of Rho in any way feeds back on Cdc42 activity, but no increase in thickness or apparent contractility of the actin cable is observed in cells expressing dominant-negative versions of Cdc42 (Jacinto, 2002).

These observations indicate that maintenance of a fully formed and functioning actin cable is not an absolute requisite for closure of the dorsal hole. In order to test further how a failure to activate Rho1 and assemble an actin cable might influence a cell's capacity to participate in dorsal closure, the enGAL4 driver was used to express both GFP-actin and either dominant-negative Rho1 (Rho^N19) or constitutively active Rho1 (Rho^V14) in 4- to 5-cell-wide epithelial stripes. These embryos were then imaged live or, alternatively, fixed and costained with Alexa594-phalloidin to reveal the actin machinery of all the cells, including the intervening wild-type stripes that do not express GAL4. Initial live analysis of these embryos, from the earliest stages of dorsal closure, revealed differences between leading edge cells that are blocked in Rho1 signaling and their wild-type neighbors. Rho1^N19-expressing cells fail to assemble an actin cable but do express broad filopodia and lamellipodia. Without a cable, they do not constrict at their leading edge in the way that their wild-type neighbors clearly do. Subsequently, these mutant Rho1^N19 stripes of cells sweep forward, apparently released from their usual constraints, overspilling and displacing their wild-type neighbors. Frequently, the dorsal edges of intervening wild-type stripes are enveloped by adjacent Rho1^N19-expressing stripes, and this wild-type tissue is consequently trapped back from the leading edge. By the time that the dorsal hole is closed, most of the midline seam epithelium is taken up by Rho1^N19-expressing cells that clearly have a migration advantage over their wild-type, cable-assembling neighbors. In the converse experiment in which GFP-actin was expressed and a constitutively active Rho1 construct (Rho1^V14) was expressed using the enGAL4 driver, precisely the opposite effect is seen. Now, Rho1^V14-expressing cells are more constricted than their wild-type neighbors at early stages, and they are subsequently outcompeted during dorsal closure, such that wild-type stripes tend to dominate the leading edge during dorsal closure. Thus, when dorsal closure is complete, the midline seam epithelium is largely wild-type (Jacinto, 2002).

These results suggest that the actomyosin cable has a dual role to play during dorsal closure. It is a driver of leading edge cell contractility from the earliest stages of dorsal closure, but it is also required for restraining the leading edge epithelial cells and maintaining a coherent and taut epithelial margin during the later phases of dorsal closure, particularly as the epithelial faces are being zippered together. It is proposed that, in addition to the previously described 'purse string' function, actomyosin cables may have a more subtle, but equally important, support role during morphogenetic episodes such as dorsal closure. These cytoskeletal structures, regulated by Rho activity, function to maintain epithelial coherence during coordinated epithelial movements, particularly at free epithelial edges where a taut front enables efficient zippering together and alignment of cells at a zipper seam, in a process that is perhaps analogous to zippering closed a very full luggage bag (Jacinto, 2002).

In a sense, the restraining function of the cable during dorsal closure is analogous to that which operates for the band of actin at the periphery of a tissue culture clone of epithelial cells. In this case, the growth factor, scatter-factor, can trigger disruption of the restraining actin band as well as dissolution of cell:cell junctions, and cells at the periphery are consequently now able to break free of their neighbors (Jacinto, 2002).

The dual role for the actomyosin cable during dorsal closure, operating both as a 'purse string' and as a cell restrainer, is likely to be a finely balanced operation. Modulation of normal Rho1 activity, as in the experiments reported above, results in either overcontractility of cells or their release from leading edge constraints. These data support previous evidence suggesting that multiple cytoskeletal events drive dorsal closure but demonstrate that these events must be finely balanced to ultimately produce the precisely organized zippering required for perfect midline fusion of the two epithelial faces (Jacinto, 2002).

Zipper promotes the asymmetric segregation of cell fate determinants by cortical exclusion rather than active transport

Cell fate diversity can be achieved through the asymmetric segregation of cell fate determinants. In the Drosophila embryo, neuroblasts divide asymmetrically and in a stem cell fashion. The determinants Prospero and Numb localize in a basal crescent and are partitioned from neuroblasts to their daughters (GMCs). Nonmuscle myosin II regulates asymmetric cell division by an unexpected mechanism, excluding determinants from the apical cortex. Myosin II is activated by Rho kinase and restricted to the apical cortex by the tumor suppressor Lethal (2) giant larvae. During prophase and metaphase, myosin II prevents determinants from localizing apically. At anaphase and telophase, myosin II moves to the cleavage furrow and appears to “push” rather than carry determinants into the GMC. Therefore, the movement of myosin II to the contractile ring not only initiates cytokinesis but also completes the partitioning of cell fate determinants from the neuroblast to its daughter (Barros, 2003).

Class II myosins are barbed end-directed motors that form bipolar filaments. The filaments bind actin and initiate contraction when the two ends of the bipolar filament pull in opposite directions. Myosin's mode of action makes it unlikely that myosin II could transport cargo from one side of the cell to the other, except perhaps by progressive contraction along the cortex. The lack of colocalization of myosin II and Miranda in neuroblasts further implies that myosin II does not transport Miranda directly. The data suggest, first, that myosin II is required to maintain an intact cortical actin cytoskeleton and, second, that active myosin modifies the actin cytoskeleton at the apical cortex to exclude Miranda binding. The C. elegans myosin II may act in a similar fashion, as it appears to limit PAR-3 to the anterior of the zygote (Barros, 2003).

Several alternative approaches were taken to inactivate myosin II in neuroblasts. First, germline clones of Sqh were analyzed. In severe sqh^1GLC embryos, levels of the regulatory light chain are greatly reduced from early development, and the heavy chain is found only in inactive aggregates. The actin cytoskeleton is disrupted, and neither Lgl nor Miranda localize to the cortex. Miranda concentrates instead at the spindle microtubules. Therefore, active myosin is necessary from early development to organize the actin cytoskeleton, which is in turn required for Lgl and Miranda to localize to the cortex (Barros, 2003).

Myosin II is activated by phosphorylation of its regulatory light chain by Rho kinase. Myosin was inactivated at the time of neuroblast cell division by inhibition of Rho kinase. Myosin II no longer localizes at the apical neuroblast cortex but instead spreads into the cytoplasm, and basal protein localization is disrupted. Although F-actin and Lgl remain uniformly at the cortex, cell fate determinants are now found around the entire cell cortex, demonstrating that apical cortical myosin is required to confine determinants to the basal half of the cell. Inhibition of Rho kinase also blocks cytokinesis, although the defect in basal protein localization is unlikely to be the consequence of mitotic arrest or a block in cytokinesis. First, basal protein localization is not disrupted in neuroblasts arrested in mitosis by colcemid treatment. Second, mitosis occurs without cytokinesis in pebble mutants, but the resultant polyploid neuroblasts still localize Numb and Prospero asymmetrically. Finally, the loss of asymmetry resulting from Rho kinase inhibition can be rescued by expression of a constitutively active form of the myosin II regulatory light chain (SqhE20E21). It is concluded that myosin II is required to restrict cell fate determinants to the basal cortex (Barros, 2003).

Myosin II localizes to the apical cortex of metaphase neuroblasts. Why is myosin localization/activity asymmetric? Lgl binds myosin II heavy chain directly and inhibits myosin filament formation. This binding is regulated by phosphorylation of Lgl; this phosphorylation inhibits its interaction with myosin II in vitro. If Lgl negatively regulates myosin activity and localization, then myosin should be uniformly distributed in an lgl mutant. Indeed, in lgl^1GLC mutants myosin II no longer concentrates apically but is found uniformly around the cortex. Most Miranda protein is released from the cortex and binds microtubules, again suggesting that myosin excludes Miranda from the cortex (Barros, 2003).

Myosin II localizes to the entire cortex in lgl mutants and thereby prevents Miranda binding basally. In Drosophila neuroblasts in which Lgl levels are reduced (zygotic lgl¹ mutants), Miranda is released from the cortex. Miranda localization can be rescued by simultaneously reducing the level of myosin II (zip¹ zygotic mutants). Reducing the level of active myosin may restore the balance between the levels of Lgl and myosin, enabling the remaining myosin to concentrate apically (Barros, 2003).

How does myosin II restrict neuroblast proteins to the basal side of the cell cortex? Myosin II and Miranda occupy primarily opposite sides at the neuroblast cortex: myosin II is concentrated at the apical cortex while Miranda localizes as a basal crescent. As myosin II shifts to the cleavage furrow, Miranda is segregated into the forming GMC. The apical F-actin compartment may be modified by myosin II to exclude binding of basal proteins like Miranda. Active myosin II requires Rho kinase activity and depends on inactivation of Lgl at the apical cortex by aPKC (Betschinger, 2003). Ectopic expression of a nonphosphorylatable form of Lgl, in which the conserved aPKC-dependent phosphorylation sites are mutated from Serines to Alanines (Lgl-3A), results in mislocalization of Miranda around the neuroblast cortex (Betschinger, 2003). The data support a spatially regulated interaction between myosin II and Lgl. Myosin is apically localized in wild-type neuroblasts, corresponding to the domain in which Lgl is inactivated by aPKC. In lgl mutants, myosin is no longer restricted apically but localizes around the entire cell cortex. Conversely, when nonphosphorylatable Lgl is expressed in neuroblasts, myosin is inhibited throughout the cell and drops off the cortex. It is proposed that myosin II is activated and can form filaments at the apical cortex, where phosphorylated Lgl is inactive and unable to bind myosin II. Myosin may then modify the actin cytoskeleton to prevent the binding of Miranda. At the basal cortex, in the absence of aPKC, Lgl is active and can bind and inhibit myosin. Myosin cannot form filaments, which are required for it to bind to the actin cortex. As a result, Miranda can bind to the basal cortex (Barros, 2003).

At anaphase, myosin II moves to the equator and appears to “push” cell fate determinants into the daughter cell. This movement is regulated in an Lgl-independent fashion and occurs whether myosin is restricted to the apical cortex or is uniformly cortical (as in lgl mutants). Cortical myosin is essential, however, to efficiently segregate determinants into the GMC at telophase (telophase rescue). In neuroblasts expressing Lgl-3A, myosin II is cytoplasmic, and determinants are not partitioned to the daughter cell. Nonetheless, at telophase, myosin seems to be recruited from the cytoplasm, since it still accumulates to the cleavage furrow. Thus three separate steps of myosin regulation in neuroblasts can be defined. First, myosin forms an apical crescent. This is positively regulated by Rho kinase and negatively regulated by Lgl. Second, cortical myosin moves to the equator. This movement occurs independently of Lgl. Third, cortical and cytoplasmic myosin accumulates at the cleavage furrow, a step that is also Lgl independent. Rho Kinase activation seems to be important for all three steps of myosin II regulation. When Rho kinase is inhibited, myosin falls into the cytoplasm, and there is no cleavage furrow formation (Barros, 2003).

In conclusion, these results demonstrate that myosin II acts downstream of Lgl and the apical protein complex to regulate the segregation of cell fate determinants. Myosin II does not negatively regulate basal protein targeting, as has previously been suggested nor does it transport determinants directly. Instead, it is proposed that myosin II acts in a novel fashion, excluding determinants from the apical cortex and 'pushing' them into the GMC at anaphase and telophase. Myosin II might modify the actin cytoskeleton to prevent determinants binding, although the actual structure formed and the physical change in the actin cytoskeleton remains to be determined (Barros, 2003).

Larval and Pupal development

Zygotically expressed myosin II is required for leg disc eversion. Depletion of Spaghetti squash, the myosin regulatory light chain, results in curved, thickened upper legs and defective wings. Depletion of SQH often results in a complete disorder in the usually perfect hexagonal packing of the ommatidia of the eye, as well as the development of an anterior notch on the eye. The time at which SQH is withheld is critical in full penetrance of the eye phenotype (Edwards, 1996).

Adult

Stable intercellular bridges called ring canals form following incomplete cytokinesis, and interconnect mitotically or meiotically related germ cells during spermatogenesis. Ring canals in Drosophila melanogaster males are surprisingly different from those previously described in females. Mature ring canal walls in males lack actin and appear to derive directly from structural proteins associated with the contractile ring. Ring canal assembly in males, as in females, initiates during cytokinesis with the appearance of a ring of phosphotyrosine epitopes at the site of the contractile ring. Following constriction, actin and myosin II disappear. However, at least four proteins present at the contractile ring remain: the three septins (Pnut, Sep1 and Sep2) and anillin. In sharp contrast, in ovarian ring canals, septins have not been detected, anillin is lost from mature ring canals and filamentous actin is a major component. In both males and females, a highly branched vesicular structure, termed the fusome, interconnects developing germ cells via the ring canals and is thought to coordinate mitotic germ cell divisions. In males, unlike females, the fusome persists and enlarges following cessation of the mitotic divisions, developing additional branches during meiosis. During differentiation, the fusome and its associated ring canals localize to the distal tip of the elongating spermatids (Hime, 1996).

Rho-associated kinase indirectly affects the activity of myosin II in the development of planar cell polarity

Drosophila Rho-associated kinase (Rok) works downstream of Fz/Dsh to mediate a branch of the planar polarity pathway involved in ommatidial rotation in the eye and in restricting actin bundle formation to a single site in developing wing cells. The primary output of Rok signaling is regulating the phosphorylation of nonmuscle myosin regulatory light chain (Winter, 2001), and hence the activity of myosin II. Drosophila myosin VIIA, the homolog of the human Usher Syndrome 1B gene, also functions in conjunction with this newly defined portion of the Fz/Dsh signaling pathway to regulate the actin cytoskeleton (Winter, 2001).

Rok signaling regulates the phosphorylation of nonmuscle myosin regulatory light chain (MRLC), and hence the activity of myosin II. Does the phosphorylation state of MRLC modify the multiple hair phenotype of dishevelled mutants? Use was made of a series of mutant spaghetti squash (sqh) transgenes (sqh codes for the Drosophila MRLC) with point mutations in the primary (Ser-21) and secondary (Thr-20) phosphorylation sites, changing them either to glutamic acid (phosphomimetic), or to nonphosphorylatable alanine. Can the phosphorylation state of MRLC also modulate Fz/Dsh signaling? An examination was made to determine whether the phosphomimetic and nonphosphorylatable forms of MRLC could directly modify the dsh¹ multiple hair phenotype. Introducing one copy of sqhE20E21 reduces the number of multiple hair cells in dsh¹ mutants by 5-fold. sqhE21, or sqhA20E21, also suppresses the dsh¹ phenotype by more than 2-fold. In contrast, introduction of sqhA21 into the dsh¹ background enhances the multiple hair phenotype. The involvement of MRLC in the Fz/Dsh pathway was also examined using the Fz-overexpression assay. Reducing the wild-type sqh gene dosage from two to one, by introducing a single copy of the sqh^AX3 null allele, results in a 2-fold suppression of the multiple hair phenotype caused by Fz overexpression. These results support the notion that MRLC functions in the PCP pathway to restrict F-actin bundle assembly to a single site (Winter, 2001).

MRLC phosphorylation in response to Rok activation would be predicted to modify the conformation and elevate the catalytic activity of its associated heavy chain, Zipper (Zip). Does Zip also participate in regulating actin distribution/wing hair number in response to Fz/Dsh? Loss of one copy of the zip gene enhances the dsh¹ phenotype by 4.5-fold, consistent with the genetic interaction data between fz/dsh and sqh. These results suggest that myosin II functions positively downstream of Fz/Dsh in regulating actin prehair development (Winter, 2001).

The localization of Zip protein in wing cells further supports its role downstream of Fz/Dsh. At the apical surface of the pupal wing cell, Zip is asymmetrically localized to the distal portion of the cell, where prehair growth initiates. This distal localization appears to coincide, temporally, with prehair initiation. To test whether Zip localization could be modified by Fz/Dsh signaling, Zip distribution at the apical surface was examined in dsh¹ mutants. Instead of being concentrated in the distal region of the cell, Zip is concentrated in the center of the cell, where prehairs form in dsh¹ mutants (Winter, 2001).

Does reduction in myosin II/Zip activity also result in the multihair phenotype? Use was made of the hypomorphic zip⁰²⁹⁵⁷, since zip and sqh null mutations appear to be cell lethal in the wing. As is the case with rok, some homozygous zip⁰²⁹⁵⁷ wing cells possess multiple F-actin prehairs (Winter, 2001).

Tests were performed to see if the gene crinkled (ck) is involved in the Fz/Dsh signaling pathway regulating wing hair number because (1) ck mutant cells in the wing lead to multiple hair and split hair phenotypes, and (2) ck encodes the Drosophila myosin VIIA protein. Mutations in mouse myosinVIIA lead to stereocilia disorganization and the formation of multiple bundles of stereocilia (Winter, 2001 and references therein).

Reduction of ck activity potently suppresses the dsh¹ multiple hair phenotype. This result contrasts with the result that zip¹ enhances the dsh¹ multiple hair phenotype, and suggests that the two myosin heavy chains have opposing effects in regulating prehair assembly (Winter, 2001).

Both myosin heavy chain genes were tested for their ability to interact with the hs-fz induced multiple hair phenotype, and again it was found that they have opposing effects. Surprisingly, loss of one copy of zip slightly but significantly enhances the late hs-fz multiple hair phenotype, while loss of one copy of ck markedly suppresses this phenotype. These results are the reverse of what one would expect based on their interactions with dsh¹, and suggests the possibility that there is a signal from Fz to Ck that is independent of Dsh, or that the multiple hair phenotypes resulting from hypo- or hyper-activity of the Fz/Dsh pathway arise via distinct biochemical mechanisms (Winter, 2001).

To further assess the nature of the relationship between the two myosins, the effect of raising or lowering the activity of MRLC on the ck phenotype was tested. The multiple hair phenotype in animals homozygous for a weak ck mutation is enhanced by one copy of the sqhE20E21 transgene (and hence, a probable increase in myosin II activity), but not by a sqhA20A21 transgene. Taken together, these experiments suggest that a balance between the activities of myosin II and myosin VIIA is important in regulating wing hair number (Winter, 2001).

Unlike other characterized PCP mutants that affect both orientation and number of wing hairs, the primary defect in Drok² clones appears to be the presence of multiple hairs per cell, with little or no wing hair orientation defect. This suggested that Rok and what lies downstream are involved in transmitting a subset of the Fz/Dsh signal. Supporting this idea, it was found that tubP-Drok and sqhE20E21 suppress the multiple hair phenotype of dsh¹, but not the hair misorientation phenotype. Additional data supporting this conclusion comes from observing the site of prehair initiation. Prehairs emerge aberrantly from the center of dsh¹ mutant cells, rather than from the distal vertex as seen in wild type cells. Such mispositioning of prehair initiation correlates with the failure to acquire the proper distal orientation. While tubP-Drok expression suppresses multiple prehair formation, it does not affect the site of F-actin initiation in dsh¹. Finally, the hair orientation defect resulting from Fz overexpression (via hs-fz) at 24 hours is suppressed by reducing dsh gene dosage but not that of RhoA, rok, sqh or ck. Taken together, these observations suggest that separate mechanisms allow Fz/Dsh to independently regulate the number and the orientation of prehairs, and that only the former involves Rok signaling (Winter, 2001).

The data presented in this study suggest that the Rok/myosin II pathway is involved in regulating the number -- but not orientation -- of the wing hair. What then are the components that regulate wing hair orientation? One possibility is that a bifurcation of the pathway occurs at the level of RhoA, with a separate effector pathway regulating wing hair orientation. In the eye, the JNK pathway has been implicated in functioning downstream of RhoA in regulating ommatidial polarity. However, the function of the JNK pathway in the wing has not been described, and a signaling pathway that regulates transcription is unlikely to encode the requisite spatial information necessary for selection of the site of prehair initiation. Therefore, it is likely that a separate signal from or upstream of RhoA may control the selection of the F-actin assembly site, and therefore the orientation of the wing hair (Winter, 2001 and references therein).

By what mechanism do myosins restrict F-actin bundle formation? In light of the finding that myosin II is concentrated at the site of prehair formation, it seems plausible that myosin II is actively involved in either the recruitment of F-actin to the prehair site, or that it directly participates in the assembly of actin bundles, or both. Studies of mammalian myosin II provide a precedent for a role in the formation of F-actin bundles. Phosphorylation of MRLC promotes a conformational change in myosin II from a folded to an extended state that readily forms multivalent bipolar filaments capable of binding multiple actin filaments. This is thought to result in F-actin bundling and stress fiber formation (Winter, 2001 and references therein).

It appears that in the developing wing, the level of MRLC phosphorylation/myosin II activity must be within an optimal range to establish the formation of a single hair. It is possible that the efficiency of F-actin bundle formation is regulated by MRLC phosphorylation in a manner similar to the control of stress fiber formation. If one further assumes that there are certain bundling substrates present only at limiting concentrations (e.g., F-actin itself), then one would predict that the assembly of one F-actin bundle would reduce the probability of forming a second bundle. When MRLC phosphorylation falls below some threshold level (e.g., in rok mutant cells), the efficiency of primary bundle formation is reduced, and thus the concentration of the limiting substrate remains at sufficient levels to support the assembly of secondary bundles/prehairs. Conversely, if MRLC is hyperphosphorylated (e.g., in Fz-overexpressing cells), the bundling efficiency may increase such that the threshold concentration for bundle formation would be reduced, thereby increasing the probability of assembling multiple bundles/prehairs. Future studies will be required to determine the detailed mechanisms involved (Winter, 2001).

In addition to nonmuscle myosin II, which resembles the myosin II from skeletal muscle, there exists a large class of unconventional myosins that have different properties and potential functions in nonmuscle cells. For instance, several different classes of unconventional myosins are expressed in inner ear epithelium with different subcellular localization. Mutations in three of the unconventional myosins, myosin VI, VIIA, and XV, cause hearing/balancing defects in mice, two of which when mutated in humans result in deafness. Of particular interest in the context of this study is myosin VIIA, mutations of which are responsible for mouse shaker-1 and human Usher's syndrome 1B. Loss-of-function ck (Drosophila Myosin VIIA) mutants exhibit a multiple hair and split wing hair phenotype. ck exhibits strong genetic interactions with components of the signal transduction pathway defined in this study, and has the opposite effects as that of myosin II. The seemingly antagonistic relationship between myosin II and myosinVIIA may suggest a mechanism in which the balance of the activities or stoichiometry of these two myosins is critical for the common process they regulate. For example, myosin II and myosin VIIA may share some common, limiting component(s) required for their activity. Thus, by reducing the myosin VIIA dose, myosin II has a larger share of the common component(s) and thus its activity is upregulated (Winter, 2001 and references therein).

Effects of Mutation or Deletion

Myosin II is required during oogenesis for follicle cell migration and cytoplasmic "dumping" of nurse cell contents (Wheatley, 1995 and Edwards, 1996), it is required during early development for the posterior migration of nuclei (Wheatley, 1995), it is required for dorsal closure late in embryonic development (Young, 1993) and it is required for development of the eye in larvae and for the eversion of the leg disc and proper formation of the wing during pupation (Edwards, 1996).

Embryos that lack functional myosin display defects in dorsal closure, head involution, and axon patterning. Analysis of cell morphology and myosin localization during dorsal closure in wild-type and homozygous mutant embryos demonstrates a key role for myosin in the maintenance of cell shape and suggests a model for the involvement of myosin in cell sheet movement during development. These experiments, in conjunction with the observation that cytokinesis also requires myosin, suggest that the processes of cell shape change in morphogenesis and cell division are intimately and mechanistically related (Young, 1993).

Drosophila is an ideal metazoan model system for analyzing the role of nonmuscle myosin-II during development. In Drosophila, myosin function is required for cytokinesis and morphogenesis driven by cell migration and/or cell shape changes during oogenesis, embryogenesis, larval development and pupal metamorphosis. The mechanisms that regulate myosin function and the supramolecular structures into which myosin incorporates have not been systematically characterized. The genetic screens described here identify genomic regions that uncover loci that facilitate myosin function. The nonmuscle myosin heavy chain is encoded by a single locus, zipper. Contiguous chromosomal deficiencies that represent approximately 70% of the euchromatic genome were screened for genetic interactions with two recessive lethal alleles of zipper in a second-site noncomplementation (SSNC) assay for the malformed phenotype. Malformation in the adult leg reflects aberrations in cell shape changes driven by myosin-based contraction during leg morphogenesis. Of the 158 deficiencies tested, 47 behave as second-site noncomplementors of zipper. Two of the deficiencies are strong interactors, 17 are intermediate and 28 are weak. Finer genetic mapping reveals that mutations in cytoplasmic tropomyosin and viking (collagen IV) behave as second-site noncomplementors of zipper during leg morphogenesis and that zipper function requires a previously uncharacterized locus, E3.10/J3.8, for leg morphogenesis and viability (Halsell, 1998).

Collagen IV is a basement membrane collagen; its function has been demonstrated during morphogenesis in C. elegans and Drosophila. Mutations in the emb-9 locus [Collagen 1(IV)] and the let-2 locus [Collagen 2(IV)] cause defects during the late morphogenetic stage that result in embryonic lethality. In Drosophila, two Collagen IV genes, localize to the SSNC interval uncovered by Df(2L)sc19-5. Mutations in viking act as SSNCs of zip during leg morphogenesis. Levels of Collagen IV are detectable throughout the life cycle of the fly, with high levels detected until the onset of pupation. During leg morphogenesis, the basal lamina detaches from the disc epithelium, and proteolysis of Collagen IV is thought to make the basement membrane more extensible, thus facilitating leg elongation. This site-directed cleavage of Collagen IV occurs in response to ecdysone. Identification of Stubble as a necessary gene product during leg imaginal disc morphogenesis also demonstrates a role for proteolysis during this process. The Stubble locus encodes a type II transmembrane serine protease; Stubble interacts genetically with road1 and zipEbr. However, since the existing viking mutations are P-element insertions, it is more likely that these mutations affect the expression of Collagen IV, rather than its proteolysis, during metamorphosis. Basal cytoplasmic extensions have been observed in the imaginal discs of several insects and are thought to be important for morphogenetic movements. Perhaps mutations in viking interfere with such processes in the Drosophila leg imaginal disc; if so, this effect would occur prior to the observed proteolysis (Halsell, 1998 and references).

rib and raw mutations prevent cells in a number of tissues from assuming specialized shapes, resulting in abnormal tubular epithelia and failure of morphogenetic movements such as dorsal closure. Mutations of zipper, which encodes the nonmuscle myosin heavy chain, suppress the phenotypes of rib and raw, suggesting that rib and raw are not directly required for myosin function. Abnormal formation of the actin cytoskeletal structures underlying embryonic cuticular hairs suggests possible roles for rib and raw in organizing the actin cytoskeleton. The actin prehair structures are absent in rib mutants and abnormally shaped in raw mutants, indicating that the two genes have different functions required for organizing the actin cytoskeleton (Blake, 1999).

The fact that zip mutations suppress many of the mutant phenotypes of rib and raw is inconsistent with the hypothesis that either rib or raw is directly required for contraction of the actin cytoskeleton by myosin. Nevertheless, the suppression of the rib and raw phenotypes by zip mutations might be observed if the rib and raw products regulate myosin contraction either by repression or by controlling the direction of contraction. Alternatively, both the effect of the mutations on cell shape and the suppression by zip mutations could be observed if rib and raw contribute to remodeling of the actin cytoskeleton by involvement in the organization of the actin filaments. The counteraction of rib and raw mutant phenotypes by zip mutations could then occur if the normal activities of the rib and raw products on the cytoskeleton oppose to some extent the activity of myosin (Blake, 1999).

The effect of the rib and raw mutations on hairs and denticles of the embryonic cuticle offers support for the hypothesis that the gene products are active in organizing actin. In late embryogenesis, bundles of filamentous actin form epidermal extensions around which cuticular structures are secreted. Some of these cuticular structures are the external apparatus of sensory organs and others are nonsensory projections, the dorsal and lateral cuticular hairs and ventral denticles. The denticles and hairs, both sensory and nonsensory, have various shapes and orientations and are organized in stereotypical, segmentally repeated patterns. The actin cytoskeletal supports of the cell extensions can be observed in stage 16 and 17 embryos by staining with rhodamine labeled phalloidin. Both rib and raw mutations alter the morphology of the F-actin structures, but mutations of each gene have different effects on the structures (Blake, 1999).

In normal embryos actin bundles form a prehair in the cells that secrete sensory hairs, but no prehairs form in rib embryos. rib mutants lack hairs and denticles almost completely, leaving only a few isolated cuticular hairs and denticles. The remaining hairs are either much longer than normal or are abnormally curved. At junctions of three or more cells, rib embryos display intense actin spots, some of which could be sockets of sense organs. In addition, F-actin, which in wild-type epidermal cells accumulates in the cytoplasm and subsequently dissipates, remains in the cytoplasm of rib epidermal cells. The observation of cytoplasmic F-actin accumulation that disappears prior to formation of the actin prehair is consistent with the possibility that actin filaments begin to form in the cytoplasm and are recruited into the prehair structures. The rib product is apparently required for the formation of the larger actin structures from smaller actin filaments that form in the cytoplasm (Blake, 1999).

The cells of raw mutant embryos do form projections, albeit abnormally shaped ones. In raw mutants, hairs are generally disorganized in appearance, may be inappropriately clustered, and are often forked or branched. These are the same types of abnormalities described for embryos mutant for forked (f) and singed (sn), two genes that encode actin bundling proteins. Thus, the raw product could have a role in bundling or otherwise organizing actin filaments. The formation of the F-actin prehair structures might be independent of the activity of myosin. Zygotic expression of zip is not obviously required for formation of actin prehairs and predenticles since both form normally in zip mutants. If myosin does not affect the organization of actin into prehair structures, zip mutations would not be expected to alter the phenotype of rib mutants with respect to the failure of formation of prehairs. However, zip mutation suppresses the rib phenotype, causing a substantial increase in the number of denticles and hairs present on the embryonic cuticle of rib;zip mutants, as compared to rib mutants. Thus, zip counteracts the effect of rib mutations for each of the rib phenotypes. The suppression of the cuticular hair phenotype of raw mutations by zip is less obvious, but the severity of the branching and forking characteristic of the raw prehair phenotype is reduced in raw;zip double mutants. The fact that a zip mutation causes actin prehairs and predenticles to form more normally in rib and raw mutants indicates that myosin antagonizes the formation of the actin structures (Blake, 1999).

Although rib and raw have similar effects on the ability of cells to elongate, the differences in the effects of mutations of the two genes on the actin structures that underly cuticular hairs suggest that the two gene products have different functions with respect to the actin cytoskeleton. Distinct functions for rib and raw products are consistent with the observation that raw;rib double mutants are far more defective than embryos mutant for either of the genes individually. In the double mutants many of the affected tissues are greatly reduced in size and the embryos are generally very delicate. The extreme phenotype of the double mutant could be the result of separate defects in the actin cytoskeleton. The evidence presented provides further support for the hypothesis that rib and raw products have functions necessary for cytoskeletal activity, either in a structural or regulatory capacity. The data also indicate that the gene products are not required for myosin to apply force to the actin cytoskeleton. Because the products are essential for formation of the actin models of cuticular hairs and denticles, they could function directly in organizing actin filaments. Defects in reorganization of actin filaments of the cortical cytoskeleton could also explain the abnormal cell shapes associated with rib and raw mutants. However, as is the case with other rib and raw phenotypes, lack of zygotic zip activity suppresses the effect of mutations of the genes on hair and denticle formation in the embryonic cuticle. Therefore the rib and raw products could also act by repressing myosin or controlling its activity in some other way. Analysis of the rib and raw products will likely be necessary to resolve the issue (Blake, 1999).

A dynamic actomyosin cytoskeleton drives many morphogenetic events. Conventional nonmuscle myosin-II (myosin) is a key chemomechanical motor that drives contraction of the actin cytoskeleton. The regulation of myosin activity has been explored by performing genetic screens to identify gene products that collaborate with myosin during Drosophila morphogenesis. Specifically, a screen was performed for second-site noncomplementors of a mutation in the zipper gene that encodes the nonmuscle myosin-II heavy chain. A single missense mutation in the zipper^Ebr allele gives rise to its sensitivity to second-site noncomplementation. The Rho signal transduction pathway has been identified as necessary for proper myosin function. A lethal P-element insertion interacts genetically with zipper. Subsequently this second-site noncomplementing mutation has been shown to disrupt the RhoGEF2 locus. Two EMS-induced mutations, previously shown to interact genetically with zipper^Ebr, disrupt the RhoA locus. Further, their molecular lesions have been identified and it has been determined that disruption of the carboxyl-terminal CaaX box gives rise to their mutant phenotype. Finally, it has been shown that RhoA mutations themselves can be utilized in genetic screens. Biochemical and cell culture analyses suggest that Rho signal transduction regulates the activity of myosin. These studies provide direct genetic proof of the biological relevance of regulation of myosin by Rho signal transduction in an intact metazoan (Halsell, 2000).

To identify loci encoding gene products that collaborate with nonmuscle myosin during morphogenesis, second-site noncomplementation screens were performed for the malformed adult leg phenotype (mlf). Depletion of myosin during leg imaginal disc morphogenesis results in mlf. A collection of 268 single, lethal P-element insertional mutations on the second chromosome were screened for genetic interactions with the zip^Ebr allele. Fourteen insertions failed to complement zip^Ebr. The strength of the genetic interaction is arbitrarily defined on the basis of the percentage of flies of the appropriate genotype that exhibit the malformed phenotype: weak interactions show penetrance of 10%-25% while intermediate interactions are 25%-75% penetrant. Eleven of the lethal P-element insertions identified are weak interactors. Three of the insertions are intermediate interactors. Two of these intermediate interactors are not second-site noncomplementing loci but are new zipper alleles, exhibiting intraallelic complementation. The third intermediate interacting mutation, l(2)04291, causes mlf flies in trans to zip^Ebr with a penetrance of 38% (Halsell, 2000).

The P-element insertion of l(2)04291 disrupts the RhoGEF2 locus. Genomic DNA flanking the P-element insertion was recovered by plasmid rescue, and by sequencing flanking DNA it was discovered that the P element lies within an intron that interrupts the 5' UTR of the RhoGEF2 gene (Barrett, 1997; Hacker, 1998). To further confirm that the genetic interaction observed with zip^Ebr results from a mutation in RhoGEF2, two EMS-induced mutant RhoGEF2 alleles, 1.1 and 4.1, were tested in the malformed leg assay. Both alleles interact with zip^Ebr; the penetrance of the malformed phenotype in double heterozygous flies is 33% with the RhoGEF2^1.1 allele and 27% with the RhoGEF2^4.1 allele, comparable to that seen with the original P-insertional allele (Halsell, 2000).

In addition to the malformed legs observed in flies double heterozygous for mutant RhoGEF2 and zip^Ebr, malformed wings were observed at comparable frequencies. Between 80% and 97% of the flies exhibiting a malformed leg phenotype also exhibit malformed wings. In contrast, most other loci that interact with zipper do not exhibit significant wing defects. Malformed wings are rarely observed when the legs are wild type. Taken together, these data indicate a requirement for RhoGEF2 during myosin-driven leg and wing imaginal disc morphogenesis (Halsell, 2000).

RhoA^E3.10 genetically behaves as a severe allele, yet molecularly results from a single amino acid change that converts a cysteine at position 189 to a tyrosine residue. This missense mutation causes severe effects because it alters the first residue, cysteine, in the CaaX box. The CaaX box is a common feature of members of the Ras-superfamily of small GTPases. Functionally, the cysteine residue is the site of a post-translational prenylation modification. Subsequent to this modification further lipid modifications may occur, and in most cases, the final three amino acids are removed. These modifications are required for proper association of the small GTPase and the membrane; without this association, the GTPase is nonfunctional. These functional relationships have been demonstrated for numerous Ras superfamily members, including Rho. Site-directed mutagenesis that changes the CaaX box cysteine to serine of the S. cerevisiae RhoA homolog, Rho1, results in the failure of the mutated Rho1 protein to repartition from the cytosolic compartment to the membrane. Further, these Rho1 mutant cells fail to grow. In mammalian tissue culture, CaaX box-mutated RhoB cannot be lipid modified, and these cells lose their ability to become transformed in sensitized backgrounds. Therefore, it is likely that the RhoA^E3.10-encoded protein cannot be post-translationally modified, resulting in a complete loss of RhoA function. Similarly, the nonsense mutation at residue 180 in the J3.8 allele would remove the CaaX box and an additional nine amino acids and, therefore, would also behave as a severe RhoA allele (Halsell, 2000).

However, on the basis of the differences observed in their genetic interactions with Df(2R)Jp1 and their levels of reduced viability in trans to zip^Ebr, RhoA^E3.10 appears to be a more severe allele than RhoA^J3.8. It is hypothesized that the protein encoded by RhoA^E3.10 may have a partial dominant-negative effect because it does not repartition properly. On the other hand, the premature stop codon in RhoA^J3.8 may give rise to an unstable gene product. Since appropriate antibodies directed against Rho are not yet available, this alternative cannot be adequately evaluated (Halsell, 2000).

Studies reveal that multiple processes require myosin function throughout Drosophila development, including oogenic cell migrations, larval cytokinesis, and imaginal disc morphogenesis. Strong or null alleles of zipper are embryonic lethal, fail during dorsal closure, and give rise to embryos with dorsal cuticular holes. Additionally, myosin immunolocalization studies suggest that myosin is required during stages not yet tested functionally, including embryonic cellularization and gastrulation. RhoGEF2 and RhoA also function at least during a subset of the morphogenetic processes that require myosin (Halsell, 2000 and references therein).

Mutations in the Drosophila RhoGEF2 gene have been identified by three distinct means: phenotypic suppression of ectopically expressed RhoA (Barrett, 1997); genetic screens for maternally encoded molecules required during early Drosophila embryogenesis (Hacker, 1998), and genetic screening for molecules required for myosin function (this study). Maternal depletion of RhoGEF2 results in defects during gastrulation (Barrett, 1997; Hacker, 1998). Specifically, embryos lacking maternal RhoGEF2 fail during apical constriction of ventral furrow cells. Interestingly, myosin localizes to the apical ends of these ventral furrow cells. This observation coupled with the genetic interaction between RhoGEF2 and myosin during leg morphogenesis suggests that RhoGEF2 may exert some of its effect during gastrulation via the activity of myosin in these cells (Halsell, 2000 and references therein).

RhoA mutations are recessive embryonic lethals. Zygotic depletion of RhoA results in an anterior dorsal hole in the cuticle. This defect has been characterized as a dorsal closure phenotype. Dorsal closure is an embryonic morphogenetic event in which the lateral epidermis moves over the dorsal side of the embryo, ultimately fusing along the midline. If dorsal closure fails, then cuticular holes result. Typically, these holes are more posteriorly localized than those observed in RhoA mutants. However, certain zipper alleles give rise to cuticular holes that extend from the posterior one-third of the embryo to the anterior end. These extensive cuticular holes are consistent with the head involution defects observed in zipper mutants and may reflect combined defects in head morphogenesis and dorsal closure. Therefore RhoA loss-of-function mutations may more accurately represent a particular sensitivity in head morphogenesis to perturbation rather than being dorsal closure mutants per se (Halsell, 2000 and references therein).

Nonetheless, RhoA function during dorsal closure has been implicated by analysis of embryos expressing dominant negative RhoA transgenes. In wild-type embryos, the leading-edge cells and the adjacent lateral cells elongate during dorsal closure. When dominant-negative RhoA is driven in the leading edge by utilizing the GAL-4 UAS system, stretching of the leading cells initiates but is ultimately lost, and the lateral cells never elongate. The Jun-kinase signal transduction cascade acts during dorsal closure and induces expression of the TGFß gene, decapentaplegic (dpp), in the leading-edge cells. Leading-edge dpp expression is a prerequisite for elongation of the flanking lateral cells. In the dominant-negative RhoA embryos, dpp expression is wild type, therefore the authors suggest that RhoA acts upstream of a separate transcriptional pathway. Three observations suggest that RhoA may function directly upstream of myosin in the leading edge. (1) It has been shown that RhoA signaling is necessary for myosin-driven cell shape changes during leg imaginal disc morphogenesis. (2) zipper mutants lose myosin in the leading-edge cells, and, subsequently, the leading-edge cells fail to elongate. (3) Myosin is delocalized in leading-edge cells expressing dominant negative RhoA. Taken together, these results suggest that RhoA signaling may have a direct cellular output at the level of myosin activity in the leading-edge cells and may not exert its effect via a transcriptional pathway (Halsell, 2000 and references therein).

Numerous pharmacological, cell culture, and biochemical studies implicate the Rho subfamily of GTPases as signal transducers upstream of actin cytoskeleton rearrangements and myosin regulation. In Drosophila, injection of mutant forms of Rho or Cdc42 proteins induces gross malformations in the actomyosin cytoskeleton, disrupting a specialized embryonic cytokinesis known as cellularization. When dominant-negative Rac1 is expressed at later stages of embryogenesis, the actomyosin cytoskeleton is disrupted in the leading-edge cells during dorsal closure. In Swiss 3T3 cells, the Rho GTPase induces the formation of actin stress fibers. Further, it has been demonstrated that contractility of the actin cytoskeleton, presumably mediated by myosin, is required for stress fiber formation and that this contractility is downstream of Rho signal transduction (Halsell, 2000 and references therein).

In metazoans, nonmuscle myosin and smooth muscle-based contractility depend on the phosphorylation state of the noncovalently bound regulatory light chain. Molecularly, activated Rho may modulate the phosphorylation state of the regulatory light chain. Biochemical analysis reveals that activated Rho binds and activates a variety of effectors, including a group of serine/threonine kinases known as Rho kinase/ROK and p160ROCK/ROKß. In vitro biochemical assays reveal that Rho kinase can phosphorylate the regulatory light chain at its activating sites and induce myosin activity. Further, Rho kinases phosphorylate the myosin binding subunit of myosin phosphatase and thus repress its activity; the net result is a further increase in the phosphorylation state of the regulatory light chain (Halsell, 2000 and references therein).

Genetic screens for morphogenesis defects in C. elegans have also identified mutations in loci encoding Rho signal transduction components. Mutations in the C. elegans Rho kinase locus, let-502, disrupt embryonic elongation, while mutations in the regulatory subunit of the myosin phosphatase gene, mel-11, suppress the let-502 morphogenetic defect (Wissmann, 1997). These results suggest that Rho signal transduction is upstream of myosin-driven morphogenesis in C. elegans. This hypothesis cannot be tested directly because myosin mutations that affect cell sheet morphogenesis have not been identified in C. elegans. Nonmuscle myosin is encoded by more than one locus and functional redundancy of these loci may preclude the isolation of morphogenetic myosin mutations (Halsell, 2000 and references therein).

ribbon is thought to be required for generating specialized cell shapes. For instance, during dorsal closure, leading edge cells of the lateral epidermis fail to elongate in rib mutants. rib mutants also show abnormal dilation of salivary gland lumina in late embryogenesis, suggesting that either rib is also required at late stages to maintain organ shape or loss of early rib function indirectly causes the late lumenal dilation. rib appears to control cell shapes by regulating the cytoskeleton. During dorsal closure, a band of actin and myosin forms at the dorsal margin of leading edge cells. In rib embryos, the actin band is narrower and myosin heavy chain (MHC) is absent from leading edge cells. Thus, rib may be required for the localization or organization of cytoskeletal components. zip encodes a nonmuscle MHC and is required in many of the same tissues as rib; however, strong loss-of-function mutations in zip suppress the distended lumenal phenotype of rib salivary glands, suggesting that rib does not positively regulate myosin activities. Instead, rib may repress myosin contraction or regulate the direction of contraction, perhaps by providing a balancing force to the direction of basal myosin contractions. These studies reveal a role for rib in coordinating directed cell migration, a process that clearly involves actin/myosin dynamics. Thus, rib may modulate actin/myosin behavior for cell movement and cell shape during both tissue formation and tissue homeostasis. If rib is responding to signaling pathways, rib could be a critical factor linking signaling events to changes in the cytoskeleton (Bradley, 2001).

Nonmuscle myosin-II is a key motor protein that drives cell shape change and cell movement. The function of nonmuscle myosin-II has been analyzed during Drosophila embryonic myogenesis. Nonmuscle myosin-II and the adhesion molecule, PS2 integrin (Myospheroid), colocalize at the developing muscle termini. In the paradigm emerging from cultured fibroblasts, nonmuscle actomyosin-II contractility, mediated by the small GTPase Rho, is required to cluster integrins at focal adhesions. In direct opposition to this model, it has been found that neither nonmuscle myosin-II nor RhoA appear to function in PS2 clustering. Instead, PS2 integrin is required for the maintenance of nonmuscle myosin-II localization and the cytoplasmic tail of the ß_PS integrin subunit is capable of mediating this PS2 integrin function. Embryos that lack zygotic expression of nonmuscle myosin-II fail to form striated myofibrils. In keeping with this, a PS2 mutant that specifically disrupts myofibril formation is unable to mediate proper localization of nonmuscle myosin-II at the muscle termini. In contrast, embryos that lack RhoA function do generate striated muscles. Finally, nonmuscle myosin-II localizes to the Z-line in mature larval muscle. It is suggested that nonmuscle myosin-II functions at the muscle termini and the Z-line as an actin crosslinker and acts to maintain the structural integrity of the sarcomere (Bloor, 2001).

The myogenic function of nonmuscle myosin-II has been analyzed by using Drosophila genetics to manipulate the levels of nonmuscle myosin-II heavy chain, PS2 integrin, and RhoA GTPase in vivo in the developing larval muscles. Both nonmuscle myosin-II and PS2 colocalize at muscle termini. However, in contrast to models based on cultured fibroblasts, there is no evidence for either nonmuscle myosin-II or RhoA function in PS2 clustering. Instead, the maintenance of nonmuscle myosin-II localization at muscle termini is dependent on the presence of PS2 integrin and the cytoplasmic tail of the ßPS integrin subunit is sufficient for this. Further, nonmuscle myosin-II maintenance at the muscle termini is compromised in if^SEF, a ß_PS2 integrin subunit mutant that specifically disrupts myofibril formation. Through the analysis of actin distribution in the musculature of living wild-type and mutant embryos, it has been demonstrated that RhoA-independent nonmuscle myosin-II function is required for the proper sarcomeric organization of the muscle cytoskeleton. Finally, since nonmuscle myosin-II localizes to the Z-line in late larval muscle, it has been suggested that nonmuscle myosin-II functions at both the muscle termini and the Z-line to maintain the structural integrity of the sarcomere (Bloor, 2001).

When fibroblasts in culture attach to ECM substrates through their cell surface integrin receptors, they dramatically redistribute these receptors such that they become clustered at focal adhesions. Integrin ligand binding also induces the actin cytoskeleton to rearrange into stress fibers, which terminate at focal adhesions and connect to the cytoplasmic domains of the clustered integrins. Pharmacological inhibitors of contractility block this complex cellular response, providing evidence that nonmuscle myosin-II driven contractility is essential for integrin clustering. At least a subset of nonmuscle myosin-II based contractility is dependent on RhoGTPase-mediated phosphorylation events. This suggests a model for focal adhesion and stress fiber formation in which, when diffuse cell surface integrins bind ECM ligand, they associate with actin filaments and activate Rho. In turn, Rho activates nonmuscle myosin-II, driving the formation of nonmuscle myosin-II filaments and increasing contractility. This increases the tension exerted on the actin cytoskeleton causing actin filaments to bundle and align into stress fibers (Bloor, 2001).

Bundling drives actin-associated cell surface integrins into clusters and focal adhesions are formed. The localization of PS2 integrin at Drosophila muscle termini is an ideal system in which to test this model in vivo. Indeed, consistent with the possibility that RhoA-mediated nonmuscle myosin-II contractility drives PS2 integrin clustering in the larval musculature, genetic evidence implicates RhoA-dependent activation of nonmuscle myosin-II in multiple morphogenetic pathways during Drosophila development. Furthermore, previous studies have shown that an uncharacterized intracellular mechanism is capable of driving integrin localization to the muscle termini. Despite this, this study shows that genetic depletion of either nonmuscle myosin-II or RhoA fails to disrupt PS2 localization (Bloor, 2001).

Maternally contributed RNAs and proteins support the early stages of Drosophila embryogenesis. Both zip and RhoA are maternally expressed and, in the absence of zygotic expression, this maternal contribution is sufficient for development to proceed normally until stage 14. Subsequent to this, depletion of maternal gene product results in defects in epidermal morphogenesis. Since the localization of PS2 integrin to the muscle termini occurs during stages 15 and 16, it seems likely that maternally contributed gene product is depleted from zip and RhoA mutant embryos prior to PS2 clustering. Thus, the presence of PS2 at muscle termini in zip and RhoA mutants suggests that neither nonmuscle myosin-II nor RhoA are required for PS2 clustering. However, the absolute amount of these gene products required to localize PS2 may be lower than that required for continued epidermal morphogenesis. Alternatively, these gene products may perdure longer in the mesoderm than in the developing epidermis (Bloor, 2001).

It is possible to eliminate the maternal contribution of a gene by generating mutant clones in the female germline. However, germline clones of zip and RhoA null mutations fail to make eggs. A novel technique was used to address the role of maternal nonmuscle myosin-II in PS2 localization. p127-l(2)gl, a nonmuscle myosin-II heavy chain binding protein, was overexpressed in a zip² mutant background. The p127-l(2)gl protein binds and sequesters maternal non-muscle myosin-II heavy chain, effectively titrating the available levels of nonmuscle myosin-II and antagonizing its function. In these embryos, the epidermal morphogenesis defects associated with zip² mutations are enhanced; however, muscle attachment and PS2 localization are unaffected. Muscle abnormalities are, however, observed in these and in zip zygotic null embryos: a variable subset of ventral muscles is deleted. Interestingly the affected muscles, VA1, VA2, and VA3, are derived from two muscle progenitors that arise from the same cluster of mesodermal cells. One progenitor divides to produce the muscle founder cells for VA1 and VA2. Subsequently, the other progenitor divides to produce the VA3 muscle founder and an adult muscle founder cell. The lineage and temporal relationships between these cells are reflected in the frequency at which these muscles are deleted: VA3 is more commonly deleted than VA1 and VA2, which are always either both present or both deleted. Although it is unclear whether these are the last progenitors to divide, it seems likely that defects in nonmuscle myosin-II-dependent cytokinesis are the basis of this phenotype. The fact that depletion of nonmuscle myosin-II can affect myogenic events occurring during stages 11 and 12 without affecting the localization of PS2 that occurs during stages 15 and 16 further supports the contention that nonmuscle myosin-II is not required for PS2 localization (Bloor, 2001).

An alternative approach to the analysis of gene function is the ectopic expression of dominant negative constructs. Although the interpretations of such experiments are not always unambiguous, ectopic expression of the dominant negative RhoA^N19 construct has implicated RhoA function in biological processes not revealed by maternal and zygotic mutational analyses. For example, driving RhoA^N19 expression in the early mesoderm disrupts invagination of this tissue, phenocopying embryos that lack both maternal and zygotic DRhoGEF2 expression. In this study, the 24B GAL4 driver was used to express UAS-RhoA^N19 in the mesoderm during later stages of development (Bloor, 2001).

24B-driven expression can be detected by stage 10 and strong expression occurs from stage 13 onward. Thus, this approach will compromise endogenous RhoA function during late stages of mesoderm development. Indeed 24B-driven UAS-RhoA^N19 expression affects the development of the visceral mesoderm and causes defects in somatic muscle patterning similar to those seen in zip mutant embryos. However, no defects were detected in somatic muscle attachment or PS2 localization in these embryos. Thus, while it is not absolutely certain that maternally supplied nonmuscle myosin-II or RhoA do not contribute to PS2 localization, these experiments provide strong evidence against this possibility (Bloor, 2001).

To exert tension on an underlying substrate, a cell must form a strong transmembrane connection between the substrate and its contractile cytoskeleton. At Drosophila muscle attachments this connection is mediated by localized PS2 integrin. As such, these structures superficially resemble focal adhesions, a point further emphasized by the localization of the focal adhesion protein integrin-linked kinase (ILK) to these sites. In contrast to the recruitment of intracellular proteins to focal adhesions, it has been shown that the initial localization of nonmuscle myosin-II to Drosophila muscle termini is not dependent on integrin clustering. Similarly, the localization of PAK and ILK are both independent of PS2 integrin. Furthermore, PS2 is also not required for the formation of an electron-dense hemiadherens junction at the muscle termini. However, among the proteins known to localize to the muscle termini, nonmuscle myosin-II is so far unique in that only it is dependent on PS2 integrin for its continued localization at these sites (Bloor, 2001).

Interestingly, the localization of nonmuscle myosin-II becomes PS2-dependent at the same developmental stage at which the PS2-dependence of muscle adhesion becomes apparent. When integrins bind ECM ligand, they are thought to undergo a conformational change that displaces the cytoplasmic tail of the integrin alpha subunit, exposing protein binding sites on the cytoplasmic tail of the ß subunit, a process known as outside-in signaling. It is possible that the accessibility of the ßPS cytoplasmic tail for protein-protein binding regulates PS2-dependent nonmuscle myosin-II localization. This is supported by the observation that, in the absence of endogenous PS2, the ßPS cytoplasmic tail is sufficient to keep nonmuscle myosin-II at the muscle termini. Excitingly, biochemical studies show that peptides derived from the cytoplasmic tail of the vertebrate integrin ß3 subunit are able to interact with the tail of nonmuscle myosin-II. In addition, this interaction has recently been demonstrated in cultured cells, suggesting that the maintenance of nonmuscle myosin-II localization might occur through a direct binding reaction with the ßPS cytoplasmic tail (Bloor, 2001).

PS2 integrin is required in muscle both for attachment to the epidermis and for the generation of sarcomeric ultrastructure. These data suggest that the sarcomeric function of PS2 is due, at least in part, to its role in maintaining nonmuscle myosin-II at the muscle termini. A possible function for nonmuscle myosin-II at the muscle termini is to physically link PS2 integrin to the muscle cytoskeleton by directly binding both PS2 and actin. One prediction of this model is that, in the absence of nonmuscle myosin-II, the muscle cytoskeleton will detach from the muscle termini, but that PS2 will continue to mediate adhesion of the muscle sarcolemma to the tendon matrix of the muscle-attachment site. Such a phenotype is observed in embryos mutant for ilk, the gene that encodes ILK. However, this disconnection of the muscle cytoskeleton from the muscle termini is clearly distinct from that observed in the muscles of zip mutant embryos, where muscle actin remains connected to the muscle termini, but fails to organize into sarcomeres. While this does not rule out a role for nonmuscle myosin-II in connecting PS2 and actin, it certainly argues against nonmuscle myosin-II being the primary component of this link (Bloor, 2001).

An alternative function for nonmuscle myosin-II is suggested by the demonstration that nonmuscle myosin-II localizes to the Z-line in the somatic muscles of third instar larvae. Both the Z-line and the muscle termini are ultrastructural elements that transmit tensile stress during muscle contraction. It is possible that nonmuscle myosin-II might function as an actin-crosslinking protein at these sites to help maintain their structural integrity. This role is generally assumed to be a function of alpha-actinin, an actin crosslinking protein that is the major component of muscle termini and Z-lines in both vertebrates and invertebrates. Mutations in the single Drosophila alpha-actinin gene do cause terminal defects and sarcomeric abnormalities. Surprisingly though, embryos that lack both maternal and zygotic alpha-actinin expression hatch and lethality does not occur until the end of the first larval instar. Furthermore, the organization of actin into a striated pattern of I-bands is unaffected in alpha-actinin mutants. This implicates other actin-binding proteins in the maintenance of actin organization at the muscle termini and Z-line and the data suggest that nonmuscle myosin-II maybe one such protein (Bloor, 2001).

There is no clear model of how nonmuscle myosin-II might fulfill this function. Intriguingly, myosin heads from a single filament have been shown to be able to bind parallel actin thin filaments of opposite orientation. One speculation is that nonmuscle myosin-II might be capable of such behavior, but how such cross-links would occur at cellular levels of ATP is not clear. One possibility is that nonmuscle myosin-II in the muscle termini and Z-line is in a 'catch' muscle state in which tension is maintained without turnover of ATP. By this scenario, nonmuscle myosin-II would not function in contraction, but would serve as an effective actin crosslinker. A noncontractile function for nonmuscle myosin-II in myofibrillogenesis would explain why RhoA appears to have no role in this process. It is interesting to note that, in Dictyostelium, a contraction-independent function of non-muscle myosin-II has been shown to be important for the generation of cortical tension (Bloor, 2001).

Finally, PS2 has been shown to be present at the sarcolemma above the Z-line in cultured Drosophila myotubes. It is therefore possible that PS2 somehow functions to maintain nonmuscle myosin-II within the Z-line. Indeed, adhesion between Z-line-associated hemiadherens junctions and the muscle basement membrane fails in the absence of PS2. This suggests that the integrity of the sarcomeric muscle cytoskeleton requires it to be connected to the ECM at both the muscle termini and the Z-line and that this connection is mediated by nonmuscle myosin-II and PS2 integrin (Bloor, 2001).

Drosophila Myosin-binding substrate (MBS), the homolog of mammalian MBS, was identified to study the roles of myosin phosphatase in morphogenesis. Myosin phosphatase negatively regulates nonmuscle myosin II through dephosphorylation of the myosin regulatory light chain (MRLC: Spaghetti squash). Myosin phosphatase's regulatory myosin-binding subunit, MBS, is responsible for regulating the myosin phosphatase catalytic subunit in response to upstream signals and for determining myosin phosphatase's substrate specificity (Mizuno, 2002).

Embryos defective for both maternal and zygotic MBS demonstrate a failure in dorsal closure. In the mutant embryos, the defects are mainly confined to the leading edge cells which fail to fully elongate. Ectopic accumulation of phosphorylated MRLC is detected in the lateral region of the leading edge cells, suggesting that the role of MBS is to repress the activation of nonmuscle myosin II at the subcellular location for coordinated cell shape change. Aberrant accumulation of F-actin within the leading edge cells may correspond to the morphological aberrations of such cells. Similar defects were seen in embryos overexpressing Rho-associated kinase, suggesting that myosin phosphatase and Rho-kinase function antagonistically. The genetic interaction of MBS with mutations in the components of the Rho signaling cascade also indicates that MBS functions antagonistically to the Rho signal transduction pathway. The results indicate an important role for myosin phosphatase in morphogenesis (Mizuno, 2002).

To examine whether defects in the dorsal closure in the embryos lacking MBS or overexpressing wild-type Rho kinase are due to an aberrant activation of nonmuscle myosin II, the genetic interactions with zipper (zip), which encodes the heavy chain of nonmuscle myosin II, were analyzed. About 25% of the progeny from crossing the females transheterozygous with DMBS^P2 and Df(3L)th117 to the males heterozygous for DMBS^E1 are embryonically lethal. It was expected that a reduction in the gene dosage of zip⁺ would suppress the defects in the MBS mutant or Rho-kinase-expressing embryos. When DMBS^P2/Df(3L)th117 females are mated with males heterozygous for both DMBS^E1 and zip^Ebr, half of the embryos defective for both maternal and zygotic MBS should be heterozygous for zip^Ebr. As expected, the embryonic lethality was reduced to nearly half that of the corresponding cross. Similarly, the heterozygosity for zip^Ebr considerably suppresses lethality due to ectopic wild-type Rho kinase expression. These results strongly suggest that either loss of MBS⁺ or overexpression of wild-type Rho-kinase causes hyperactivation of nonmuscle myosin II through increasing the levels of phosphorylation of MRLC (Mizuno, 2002).

zip^Ebr is a point mutation reported to be highly sensitive to genetic backgrounds. About 70% of the flies transheterozygous between zip^Ebr and zip⁰²⁹⁵⁷ have malformed wings with varying degrees of severity. Although zip^Ebr is recessive, a considerable percentage of the flies heterozygous for both zip^Ebr and the mutations in the components of the Rho signaling pathway such as DRho1 and DRhoGEF2 produced similar defects. A half reduction of Drok, which encodes Rho-kinase, also dominantly enhances zip^Ebr. This indicates the involvement of the Rho signaling pathway and its effector, Rho-kinase, in the myosin function of adult wing morphogenesis. When the flies are also heterozygous for DMBS^E1, wing malformation is significantly suppressed, suggesting that MBS functions antagonistically to the Rho signaling pathway (Mizuno, 2002).

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

Please e-mail comments/corrections to brodyt@codon.nih.gov