PerspectiveCell Biology

Embryonic Clutch Control

See allHide authors and affiliations

Science  09 Mar 2012:
Vol. 335, Issue 6073, pp. 1181-1182
DOI: 10.1126/science.1220388

All embryos, from worms to humans, are shaped during development by morphogenetic steps that tug, bend, fold, and sculpt epithelial sheets into forms that resemble, or are the precursors of, the final adult structure (1). Most of these changes are the consequence of constrictions of the apical surfaces of epithelial cells that are powered by pulsatile contracting cytoskeletal (actomyosin) networks. On page 1232 of this issue, Roh-Johnson et al. (2) show that, just as in a car where the power of the engine is linked to forward movement by means of a clutch, clutch control is also the rate-limiting step for contracting cells in tissues.

One of the best-studied examples of apical constriction driving morphogenetic episodes is gastrulation in the fly Drosophila melanogaster, in which a strip of approximately 1200 epithelial cells buckles inward and invaginates to internalize the presumptive mesoderm of the embryo (3). Variations of this process drive gastrulation in all organisms, as well as other morphogenetic events such as neural tube formation in vertebrates (4), which gives rise to the brain and spinal cord, and to formation of the optic and otic vesicles, which develop into the human eye and inner ear, respectively (see the figure). Concerted apical constrictions of cells are generated by the assembly and contractility of actomyosin networks composed of myosin II molecules that tug on actin filaments (5). These networks are linked to the plasma membrane by adherens junctions, protein complexes that also weld one cell to its neighbor (6). Live imaging of the fly embryo has revealed the pulsatile and ratchet-like nature of cell contractions during gastrulation and other morphogenetic processes, with periods of shrinkage and resting before another round of contraction (5, 7). This contractile machinery is exquisitely responsive to mechanical cues from neighboring cells and tissue; gentle probing with a needle can trigger actomyosin assembly and drive epithelial invagination (8); similarly, disrupting adherens junctions, which changes tissue tensions, can alter the polarity of apical constrictions (9). The presumption has been that the regulatory step for apical constrictions is determined by when and where cells assemble and contract their actomyosin networks.

Cell apical constrictions during embryonic morphogenesis.

(A) Concerted apical constrictions in epithelial sheets drive invaginations to form structures such as the otic cup from which the human inner ear derives. (B) Constricting cells become wedge-like to drive many morphological movements. (C) As in a car, the engine (actomyosin flow) during gastrulation can run without output. Only when the clutch is engaged and the motor is directly linked to the effector (apical plasma membrane) can there be movement of the apical membrane.


Roh-Johnson et al. challenge and extend this idea through a series of elegant studies in the worm Caenorhabitis elegans. In this model organism, two neighboring endodermal precursor cells (Ea and Ep) constrict their apical surfaces and ingress beneath the surface during gastrulation. The authors followed actomyosin network dynamics in these cells by tracking fluorescently tagged myosin II and, concomitantly, measured changes in the shape of the apical membrane. They observed that myosin moves centripetally toward the center of the cell at a constant pace, whereas the apical cell surface initially shrinks slowly, or not at all. At early stages, actomyosin flow (the engine) runs for several minutes before any cell constriction actually happens. Only later does myosin movement occur in unison with the movement of “contact zones” (adherens junctions). After a transitional period of “slippage,” the flow of actomyosin is only later efficiently coupled to constriction of the apical cell surface. The authors observed very similar mechanisms operating during Drosophila gastrulation, and it might now be feasible to live image and observe whether the same is true for vertebrate embryos, too (10).

What explains the initial lack of linkage between actomyosin flow and cell shrinkage? Roh-Johnson et al. suggest that a “molecular clutch” engages the myosin II engine with the apical membrane so that actomyosin contractility can drive shrinkage of the apical cell surface. Because cortical tension generated within the apical actomyosin network increased only a little after clutch engagement, constriction of the apical plasma membrane must reflect a change in efficiency of the link between actomyosin and contact zones.

Clutch control of myosin flow is not an entirely novel concept. In the migrating neural growth cone, the actomyosin engine is continually running, but only when positive guidance cues are received does a region of lamellae engage the clutch and extend forward (11). What are the benefits to having a clutch during morphogenesis? It may be easier to synchronize the start of a morphogenic event if the engines of all the participating cells are already running and clutch engagement is the rate-limiting step. Indeed, Roh-Johnson et al. observed that cells “rev” their actomyosin engines at least once or twice before the first sign of cell contraction is seen. It might also be easier to subtly respond to local cues (for example, the faltering contraction by a neighboring cell) by slightly altering the degree of engagement of the clutch—as one might when starting a car on hills of different inclines—than by crudely switching the engine on or off.

What is the molecular basis of this clutch and how is it regulated? Roh-Johnson et al. show that disruption of the small GTP-binding protein Rac, or of various components of adherens junctions (cadherin and the catenins), disengages actomyosin flow from cell contractions in the worm embryo. Further clues may come from observing the altered linkage between actomyosin contractions and the generation of tissue tension in fly embryos that harbor mutations in either of two key transcription factors, snail and twist (9). Or, possibly clutch machinery is regulated by a factor called folded gastrulation, a diffusible signal that synchronizes cell constrictions during fly gastrulation (12). Perhaps the ezrin-radixin-moesin family of proteins, which both transduces signals in the cell as well as links actomyosin with the plasma membrane (13), is a potential candidate for clutch proteins. We can now shift attention slightly away from the engine and unravel how the clutch works although, as those who drive manual transmission cars can attest, clutches are not always the easiest of gadgets to get to grips with.


View Abstract

Stay Connected to Science

Navigate This Article