Rho GTPases are key signaling proteins controlling cell adhesion, cell morphology, cell migration, gene expression and membrane trafficking in response to extracellular signals. Theses proteins play a critical role throughout embryonic development of multicellular organisms and have been involved in the development of a large number of human pathologies including all kind of cancers and several genetic diseases. To fulfill their multiple tasks, Rho GTPases function in a coordinated manner. However, to date, the molecular basis of Rho GTPase regulation and coordination is far to being fully understood.
Our research aims to better understand Rho GTPase regulation during complex morphogenic events in multicellular organisms. To do so, we use developmental genetics and cell biology approaches, to dissect Rho GTPase regulatory machineries controlling epidermal morphogenesis in the embryos of the nematode Caenorhabditis elegans. We also use integrative genomic approaches to characterize and predict the functional interactions linking signaling proteins controling morphogenesis in C. elegans and human.
Rho GTPase regulation during epidermal morphogenesis in C. elegans embryos
Rho GTPases are molecular switches that cycle between an active GTP-bound form and an inactive GDP-bound form. When bound to GTP, they interact with specific effectors that mediate downstream signaling events. Three families of proteins control Rho GTPase activation level through catalysis of their nucleotide-exchange and GTP-hydrolysis activities: the Guanine nucleotide-Exchange Factors (GEFs); the GTPase-Activating Proteins (GAPs); and the GTP-Dissociation Inhibitors (GDIs). Together with effectors, regulators are thought to play a critical role in the control of Rho GTPase coordination. However, to date, these mechanisms are still poorly understood.
C. elegans constitutes an ideal model to study Rho GTPase function and regulation during morphogenesis. This is due to its relative simplicity in terms of cell number (a thousand cells) and organization, and its ease of genetic manipulation. In addition, Rho GTPases and their regulators are extremely well conserved between C. elegans and higher organisms. Six Rho GTPases (RHO-1, CED-10, RAC-2, MIG-2, CDC-42 and CRP-1) and forty regulators have been identified in the C. elegans genome.
Ventral enclosure of C. elegans embryos
The late phase of C. elegans embryonic development includes epidermal morphogenic events that enable the embryo to acquire its final tubular shape. One of these events, termed ventral enclosure, involves the migration of ventral hypodermal cells towards the ventral midline to cover the embryo in an epidermal layer. This event occurs in two phases. In the first phase, the anterior ventral hypodermal cells - referred to as the leading cells, migrate towards the ventral midline using large actin-rich protrusions (Figure 1 A), where they form junctions with their contralateral neighbours (Figure 1 B). Afterwards, the posterior ventral hypodermal cells, called the pocket cells, migrate towards the ventral midline using a contraction-dependent, purse-string mechanism, which is still poorly described. These migratory mechanisms are supported by signals from underlying neuroblasts.
During ventral enclosure, Rho GTPases control hypodermal cell migration in a cell-autonomous manner. They control the formation of actin-rich protrusions at the leading edge of migrating hypodermal cells and maintaining cell-cell junctions between them. This kind of migration of epithelial cells is called collective migration. It is evolutionary conserved and important for epithelial organ formation and repair as well as for carcinoma cell invasion. Our research characterizing singaling pathways controling collective migration of hypodermal cells during ventral enclosure is consequently of high interest for the advance of both basic sciences in cell biology and in oncology.
We recently characterized the zygotic function of RGA-7, a CDC-42/Cdc42 and RHO-1/RhoA-specific Rho GTPase-Activating Protein, which controls the formation of actin-rich protrusions at the leading edge of leading hypodermal cells (Figure 1 A) and the formation of new junctions between contralateral cells (Figure 1 B) (Ouellette et al., 2015). We showed that RGA-7 controls these processes in an antagonistic manner with the CDC-42 effector WSP-1/N-WASP and the CDC-42-binding proteins TOCA-1/2/TOCA1. Our study suggests that RGA-7 controls collective migration and junction formation between epithelial cells by spatially restricting active CDC-42 within cell-cell junctions (Ouellette et al., 2015). RGA-7 has three close homolgs in human: HMHA1, GMIP and PARG1, we are currently studying the function of these genes during collective migration of human carcinoma cells.
Embryonic elongation of C. elegans
Once the embryo is fully enclosed it start elongating. This elongation involves communication between hypodermal and body-wall muscle cells and the control of the contraction of the circumferential actin filaments (CAFs) located at the apical pole of the hypodermal cells. Elongation is divided into an early phase between comma and 1.75-fold stages, and a late phase between 1.75- and 4-fold stages. The early phase of elongation is controlled by contraction state of the ventral, lateral and dorsal hypodermal cells, while the late phase involves the synergistic action of hypodermal and body-wall muscle cells contraction. Within hypodermis, contraction and relaxation of CAFs involve the regulation of myosin-light chains (MLC) phosphorylation state by specific kinases (LET-502, MRCK-1, PAK-1) and phosphatase MEL-11.
We recently showed that that PIX-1/bPIX/COOL, PAK-1/PAK1 and LET-502/ROCK control the rate of elongation, and the antero-posterior morphology of the embryos (Martin et al., 2014; Martin et al., 2016 JoVE). We recently developed quantitative confocal microscopy methods enabling the caracterisation of epidermal morphogenesis at a single cell level. This emthod allowed us to show that a Rho/ROCK program control morphogenesis in lateral cells while a Rac-like PIX-1/PAK-1 program controls morphogenesis of dorsal hypodermal cells during early elongation (Martin et al., 2016 JCB; commented by Short B., 2016). This study also revealed that cell autonomous antagonisms between Rho and Rac programs allows hypodermal cells to switch from one program to another when one gets genetically compromised. We are currently studying the molecular events enabling hypodermal cells to adopt a specific morphogenic program and the implication of cell-to-cell heterogeneity observed in the hypodermis for morphogenesis during early elongation.
Characterization and prediction of genetic interactions
A genetic interaction (GI) between two genes generally indicates that the phenotype of a double mutant differs from what is expected from each individual mutant. In the last decade, genome scale studies of quantitative genetic interactions were completed using mainly synthetic genetic array (SGA) technology and RNA interference in yeast and C. elegans. These studies raised questions regarding the functional interpretation of GIs, the relationship of genetic and molecular interaction networks, the usefulness of GI networks to infer gene function and co-functionality, the evolutionary conservation of GI, etc.
While GIs have been used for decades to dissect signaling pathways in genetic models, their functional interpretations are still not trivial. The existence of a GI between two genes does not necessarily imply that these two genes code for interacting proteins or that the two genes are even expressed in the same cell. In fact, a GI only implies that the two genes share a functional relationship. These two genes may be involved in the same biological process or pathway; they may also be involved in compensatory pathways with unrelated apparent function. Considering the powerful opportunity to better understand gene function, genetic relationship, robustness and evolution, provided by a genome-wide mapping of GIs, several in silico approaches have been employed to predict GIs in unicellular and multicellular organisms. Most of these methods used weighted data integration.
We developed integrative genomics approaches to characterize and predict genetic interactions in C. elegans (Lee et al., 2010). We recently published a literature review on that topic (Boucher and Jenna, 2013) (Figure 3) and characterized the modular structure of the C. elegans genetic interaction within pathways (Boucher et al., 2016). We are currently studying the structure of similar interactomes in human and the plasticity of these interactomes in human carcinoma cells.
Figure 1: Ventral enclosure
RGA-7 spatially controls active CDC-42 at cell-cell junctions during collective migration and expansion of newly formed junctions. (From Ouellette et al., in Press)
Figure 2: Early elongation
Model for signaling pathways controlling embryonic elongation. A) Schematic representation of an embryo during early elongation. Anterior is at the left and dorsal side on the top. The blue plan indicates the location of the transversal sectioning of the hypodermal cells represented in panel B. B) Signaling pathways in the dorsal, lateral and ventral hypodermis in the anterior part of the embryo. (From Martin et al., 2014)
Fig. 3:Abstraction levels of systems
Representation of the six abstraction levels of biological systems. Note that, while each gene/protein can be followed from one abstraction level to another, the relationship linking it with its neighbours is different at each level. (From Boucher and Jenna, 2013)
PIX-1 expression in embryos
Staining of embryos expressing PIX-1::GFP fusion protein with an anti-GFP antibody; antibody staining apical junctions of epithelial cells MH27; and DAPI for nucleus (From Martin et al., 2014)