Labouesse, Michel et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "A tension-dependent GIT/PIX/PAK pathway in C. elegans hemidesmosomes promotes epidermal morphogenesis"

Zahreddine, Hala, Landmann, Frederic, Zhang, Huimin, Labouesse, Michel

[C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany, 2010]

It has long been known that mutants with defective muscles arrest midway through elongation , but the underlying mechanism linking muscles to embryonic elongation remains unclear. Here we show that a mechanotransduction pathway operates between muscles and the epidermis, which transforms physical force into chemical signals and contributes to epidermal morphogenesis. Muscle contraction triggers extensive tension change in the epidermis. C. elegans hemidesmosomes (CeHDs), the intermediate filament-linked junctions attaching epidermis to the extracellular matrix at the muscle-epidermis interface, is critical for transmitting this tension . We identified the p21-activated kinase homologue PAK - 1 in a genome-wide RNAi screen for genes required to maintain CeHDs. Strong mutations in pak-1 combined with the viable allele of vab- 10 , a key protein of the hemidesmosome structure , destabilize CeHD s and result in severe epidermal defects. We found that PAK-1 is a novel CeHD component, whose kinase activity controls intermediate filament phosphorylation and their polymerization or anchoring to CeHDs. Interestingly, tension exerted on the epidermis by adjacent muscles is required for activating PAK-1 kinase . We further identified two PAK-1 co-factors, the GEF protein PIX-1 and the adaptor GIT-1, both localizing to the CeHDs, as novel players in this pathway. Biochemical and genetic approaches revealed that muscle tension maintains GIT - 1 in CeHDs to stimulate PAK-1 activity through PIX-1 and Rac GTPase CED-10. This novel mechanotransduction pathway promotes CeHD maturation into a junction able to resist m echanical stress. Our findings suggest that the C. elegans hemidesmosome is not only an attachment structure, but also a mechanosensor capable of transducing physical tension into chemical signals.

Michel Labouesse et al. (2008) Development & Evolution Meeting "Cytoskeleton and Junction Remodelling during C. elegans Embryonic Elongation"

Hala Zahreddine, Michel Labouesse, Christelle GALLY, Huimin Zhang, Frederic Wissler, Marie Diogon

[Development & Evolution Meeting, 2008]

Elongation of C. elegans embryos provides an invaluable system to investigate several cell biological processes occurring in all epithelia, such as establishment of cell polarity, junction assembly, cell shape changes, cytoskeleton dynamics. Embryonic elongation is a rather complex morphogenetic process since it involves distinct epidermal cells and muscles. While many important players controlling this process have been identified, their integration in time and space remain poorly defined. I will outline our recent progress in two areas that contribute to elongation, the spatiotemporal regulation of myosin II activity in the epidermis, and the mechanisms involved in remodelling fibrous organelles (FOs) in response to muscle tension. To define how myosin II activity is controlled we generated several GFP probes to visualize the actomyosin cytoskeleton in vivo. We dissected the mechanism of MLC-4/rMLC activation through phosphorylation at two conserved Ser and Thr residues. We showed that MLC-4 is primarily required in lateral seam cells, and that its activity is repressed in dorsal/ventral cells through a Rho-specific RhoGAP (RGA-2). In addition, we identified the two protein kinases that phosphorylate MLC-4 during elongation and completed the characterisation of myosin II structure by identifying its essential light chain. FOs are the adhesion complexes that attach dorsal/ventral epidermal cells to the extracellular matrix and thereby connect muscles to the cuticle. They are functionally and molecularly related to hemidesmosomes, since they contain a plakin (VAB-10) and intermediate filaments. To define how FOs are assembled and how they contribute to elongation, we performed a genome-wide RNAi enhancer screen looking for clones that decrease the viability of a weak vab-10 mutant. We identified 14 genes that have human homologues with predicted roles in transcription, RNA processing, protein stability or cytoskeleton remodelling. Most of these genes interact together and thus form a functional network. Among them, PAK-1/p21-activated kinase (a novel FO component) controls intermediate filament polymerization and/or anchoring, CRT-1/calreticulin the folding of the ECM component UNC-52/perlecan, and the HECT-domain E3-ligase EEL-1 the abundance of the UNC-52 receptor myotactin. We conclude that this functional network helps maintain FO structure when submitted to external tension.

Michel Labouesse et al. (2002) European Worm Meeting "One gene for two distinct plakins that provide outer and inner resistance to mechanical stress"

R Legouis, Ann M Rose, J M Bosher, B-S Hahn, A Gansmuller, Michel Labouesse, A Chisholm

[European Worm Meeting, 2002]

To identify genes required for epidermal integrity and embryonic morphogenesis, we previously performed a chromosomal deficiency screen looking for embryos that failed to elongate (Dev Dyn 210:19-32). We next focused on a promising deficiency, hDf17, and identified an embryonic lethal mutation, h1356, mapping under hDf17 that recapitulates part of its phenotype. We showed that h1356 corresponds to a putative null allele of the gene vab-10, previously defined by the viable allele e698, and identified another strong allele, ju281 which failed to complement h1356 and e698. We showed that vab-10 encodes two distinct plakins via a complex pattern of alternative splicing. Plakins are large proteins thought to bridge different cytoskeletal elements. Specifically, all VAB-10 isoforms share a common N-terminal domain with a predicted actin- binding domain. Their C-terminal domains are either similar to plectin, a protein known to bind actin and intermediate filaments, or MACF, a protein known to bind actin and microtubules (we refer to these isoforms as CePlectin and CeMACF, respectively). Sequencing of the three vab-10 alleles shows that h1356 affects the common N-terminal region, while ju281 and e698 are missense mutations affecting the unique CePlectin region. Using polyclonal antibodies against CePlectin, we found that it is a likely component of fibrous organelles, a structure found in the epidermis that provides a mechanical link between muscles and the cuticle across the epidermis. Similarly using two antisera against CeMACF we found that CeMACF forms a pattern of parallel bands in the epidermis that is interspersed with fibrous organelles.

Vuong, Thanh TK et al. (2013) International Worm Meeting "Mechanical forces in C. elegans embryo elongation."

Labouesse, Michel, Vuong, Thanh TK

[International Worm Meeting, 2013]

Embryo development requires precise coordination of mechanical forces and their failure can lead to diseases. During the morphogenesis of C. elegans embryo, the cooperation of epidermal acto-myosin network and muscle contractions is essential. The acto-myosin activity in the epidermis, which has been shown to be more important in lateral than in dorsal-ventral cells, squeeze the embryo and make it elongate [1]. Muscle contractions become active around 1.7-1.8 fold stage. They have been showed to induce a mechano-transduction pathway [2], which is important for elongation. However, it is unclear how the contractions along the anterior-posterior axis help to increase the length of the embryo. Our project aims to elucidate the mechanical role of muscle contractions and its coordination with acto-myosin forces. The experiments are designed following a working model where muscle contractions induce a change in the elasticity of the embryo. We are using a laser nano-dissection technique to investigate cortical tension and elasticity of epidermal cells before and after the onset of muscle contractions. In parallel, we are evaluating the relative changes of acto-myosin forces with a FRET sensor [3] inserted in HMP-1 - a component of the adherens junctions. I will present our observations and preliminary results of the epidermal cortex nano-dissection experiments and measures of acto-myosin forces exerted on adherens junctions. References 1.Gally C, Wissler F, Zahreddine H, Quintin S, Landmann F, Labouesse M. Myosin II regulation during C. elegans embryonic elongation: LET-502/ROCK, MRCK-1 and PAK-1, three kinases with different roles. Development. 2009 Sep;136(18):3109-19. Epub 2009 Aug 12. 2.Zhang H, Landmann F, Zahreddine H, Rodriguez D, Koch M, Labouesse M. A tension-induced mechanotransduction pathway promotes epithelial morphogenesis. Nature. 2011 Mar 3;471(7336):99-103. 3.Grashoff C, Hoffman B, Brenner M, Zhou R, Parsons M, Yang M, McLean M, Sligar S, Chen C, Ha T, Schwartz M. Measuring mechanical tension across vinculin reveals regulation of focal adhesion dynamics. Nature. 2010 July 8; 466(7303): 263-266.

Apaydin, Ahmet et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "Analysis of Exosome Secretion in C. elegans"

Labouesse, Michel, Apaydin, Ahmet

[C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany, 2010]

Exosomes are 40-100nm membrane vesicles that are secreted when multivesicular bodie s (MVBs) fuse with the apical plasma membrane instead of committing proteins to lysosomal degradation. Transmembrane proteins, lipid raft components, mRNAs and miRNAs are amongst exosome components, echoing their origins. A broad range of cells secrete exosomes; neurons, tumors, intestinal epithelial cells and cells of hematopoietic origin. Thus, exosomes are increasingly being viewed as an important channel for intercellular communication and potential therapeutic targets. Key questions in the field are how proteins are sorted into exosomes, the distinction between secretory and degradative vesicles, whether or not they co-exist in the same MVBs, and how exosome secretion is achieved and controlled. These questions remain by and large unanswered. Previously our lab has implicated the V0 subcomple x of the V-ATPase proton pump in exosome secretion in the C. elegans epidermis. Building upon this body of work I performed a targeted RNAi screen, looking for cuticular defects under DIC microscopy as a criterion for abnormal secretion, similar to what had been observed after V0 disruption. Out of more than 1000 candidates involved in vesicular trafficking and various other processes, nearly 50 genes were selected for further characterization, based on cuticle defects and distribution of VHA-5 (a V0 component). In particular, VHA-5::mRFP was imaged by spinning disk confocal microscopy and the different classes of VHA-5 puncta were automatically categorized after image segmentation. We are focusing on 10 promising genes that cause strong alae defects and have various VHA-5 localization patterns. I am further characterizing them by TEM and in vivo imaging of fluorescent fusion reporters. In addition, using a novel technique of Correlative Light and Electron Microscopy (CLEM) I am defining the identity VHA-5 positive structures.

Kolotuev, Irina et al. (2009) International Worm Meeting "Correlative microscopy in C. elegans."

Labouesse, Michel, Kolotuev, Irina, Schwab, Yannick

[International Worm Meeting, 2009]

Correlative light and electron microscopy (CLEM) approach is aimed to bridge the data acquired with different imaging media. It is successfully applied in cell culture systems, however, the use of this approach in multicellular organisms raises technical problems due to the addition of a third dimension and of their motility, which makes difficult the retrospective location of the region of interest (ROI). We are interested in cellular processes associated with cytoskeletal rearrangements, tubulogenesis and secretion, and have realized the need to use CLEM to address a wide range of questions impossible without a precise correlation. We developed a CLEM technique that enables easy processing of all stages and is adequate both for morphology and immuno EM specimen preparations. Using it enables observation of single animals in real time, HPF fixation and flat embedding. Our major improvement over previous protocols has been the development of a precise mapping system that considerably simplifies and speeds up the retrospective location of the ROI within 1mm. Our method allows us to acquire thin sections of the ROI in about one hour starting from the in-block specimen. We chose to illustrate the technique and its potential through two examples. a)exc-4 is a chloride intracellular channel mutant affecting the excretory system. The terminal defect has been previously shown to result in a huge vacuole in the excretory canal. We were interested to see the nature of damage in embryonic and early larval stages and could follow the same specimen from the location of the vacuolized region in DIC to exactly the same region in TEM. We found that the mutation results in hyperfusion of the inner lumen with the canal composing canaliculi and further deposition of agglomerates in the oversized inner lumen. b)che-14, a homolog of the transmembrane Dispatched protein, was previously proposed to be involved in secretion and tube formation. Its subcellular localization had only been inferred based on GFP reporter fluorescence. Using animals with CHE-14::GFP construct and GFP antiserum, we found by immuno-EM gold particles at the apical surface of epithelial and sensory support cells, in accord with GFP reporter studies. Moreover, we could locate precisely the seam cells that showed mosaic expression of GFP construct, side by side with non-expressing cells, confirming the powerful ability of the technique to locate the exact ROI. We believe that this technique has a potential to contribute to correlative studies in any small animal (we could successfully apply it to fly embryos) and will facilitate the use of the time demanding procedure of specimen preparation for TEM in different developmental stages.

Kolotuev, Irina et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "C. elegans prox-1 promotes excretory canal formation and amphid sheath cell tubule development"

Schwab, Yannick, Labouesse, Michel, Kolotuev, Irina

[C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany, 2010]

The formation of tubular structures during organogenesis is a general process that remains poorly characterized at the molecular level. In C. elegans the excretory canal and amphid sheath tubules serve as models of tubulogenesis. We previously identified the mutation rdy-3(mc41) in a genetic screen for secretion and osmoregulatory defects. rdy-3(mc41) larvae die at the L1/L2 stage with no excretory canal arms and no DiI uptake in the amphids. We cloned rdy-3 and found that it corresponds to the C. elegans Prox1 homologue (aka ceh-26); we suggest renaming it prox-1. Prospero-related homeobox transcription factor (Prox1) is indispensable for the development of various tissues (notably D. melanogaster neurons, and mice lymphatic vessels), however its downstream targets are unknown. prox-1(mc41) is a 102bp deletion in the homeobox domain. We fully rescued mc41 phenotypes with a cosmid containing prox-1, and partially with a prox-1 cDNA driven by a heat-shock promoter. prox-1 expression is visible in a subset of head and body neurons and in the excretory canal. Using fluorescent reporter analysis and correlative TEM techniques we found that prox-1(mc41) does not cause cell fate transformation, but influences the process of excretory inner lumen formation and canal extension. Similarly, we found that the amphid sheath cell terminal tube is affected, while amphid neurons appear normal. To search for prox-1 downstream targets we used evolutionary bioinformatics based on the data of genome-wide analysis of fly Prospero binding sites. The search identified exc-5, an important factor in excretory canal extension and lumen formation, plus several other targets from the same pathway. Real time PCR showed upregulation of exc-5 in the prox-1 background and bioinformatic analysis revealed a pu tative Prox1 binding site. We propose that C.elegans can serve as a model to further characterize PROX-1 function in lumen formation and tubulogenesis.

Gally, Christelle et al. (2013) International Worm Meeting "A genome-wide RNAi screen to identify new players of a muscle-to-epidermis mechanotransduction pathway essential for embryonic elongation."

Aubry, Agnes, Labouesse, Michel, Gally, Christelle

[International Worm Meeting, 2013]

The interplay between different layers or tissues adds a level of complexity in the understanding of how organs develop. As an example, loss of muscle activity in the C. elegans embryo results in embryonic elongation arrest at the 2-fold stage (Pat phenotype) for a reason that has long remained unclear. We recently unraveled a mechanotransduction pathway between muscles and epidermis that accounts to a large extent for the Pat phenotype (Nature, 471, 99-103, 2011). Specifically, during the second phase of embryonic elongation, muscle contractions activate the serine/threonine p21-activated kinase PAK-1 in the epidermis and trigger the remodeling of hemidesmosomes (HDs) connecting the apical and basal sides of the epidermis through intermediate filaments. PAK-1 controls intermediate filament phosphorylation and their anchoring to HDs. The most upstream known component of this pathway is the adaptor protein GIT-1, whose localization to HDs requires muscle activity. However, since git-1 mutants do not induce a two-fold arrest, we predict that muscle contractions also trigger a second parallel pathway, probably to activate actomyosin remodelling. In order to identify this putative parallel pathway, we have undertaken a genome-wide RNAi screen in the synthetic git-1(tm1962) mutant background, looking for enhancers that lead to an elongation arrest. As an approach, we used 96-well worm liquid cultures of L4 mothers (see Nat Protoc, 1, 1617-20, 2006), looking for late embryonic arrests in their progeny. In the primary screen, we have identified 78 candidates that induce body-morphology defects as well as an elongation arrest in the git-1 mutant in a more effective way than in the wild-type N2 background. We have confirmed this git-1 enhancement for 39 candidates in a secondary screen. Among them, four genes encode HD components and six genes are involved in muscle differentiation/function or attachment to the extracellular matrix, hence validating our approach. Now, we are focusing on genes that were not previously linked to embryonic elongation. We are currently testing their involvement in HD organization as well as in actomyosin organization and function, and will present our preliminary results.

Quintin, Sophie et al. (2011) International Worm Meeting "Microtubules in embryonic morphogenesis."

Quintin, Sophie, Labouesse, Michel, Bender, Ambre

[International Worm Meeting, 2011]

During C. elegans morphogenesis, the embryo elongates from a bean to a worm shape within 3 hours, without any cell division. Elongation, which ultimately reduces embryo circumference and lenghthens its A-P axis by 4-fold, depends on 3 processes: 1- actomyosin tension 2- a mechanotransduction pathway between muscles and the epidermis 3- microtubules (MT). In 1986, J. Priess showed that treatments with MT-depolymerising drugs affect elongation, however how MT precisely act is still unknown. We address this question by inducing the expression of spastin -a MT-severing protein- in various cell types. We use an operon-like expression system in order to co-express the spastin protein (SPAS-1) with an mCherry reporter, and express it in all epidermal cells (lin-26 promoter), in the lateral seam cells (P_ceh-16), in the dorso-ventral cells (P_elt-3 and P_dpy-7), or in all cells using a heat-shock promoter (hsp). MTs are indeed disrupted in mCherry-expressing cells, as visualised with a tubulin::GFP reporter. MT degradation is proportional to the dose of spastin expressed in the cells, the best results being obtained with the strong hsp and the dpy-7 promoters. Besides, early SPAS-1 induction causes cell division defects in the early blastomeres, suggesting that MT function is efficiently disrupted. We found that embryonic lethality is higher with the hsp promoter vs any epidermal promoter (35% vs below 10%). Importantly, a deleted version of SPAS-1 did not cause lethality. Using videomicroscopy, we could define the time window in which SPAS-1 induction causes embryonic phenotypes, namely just before the onset of elongation. Interestingly, if SPAS-1 is induced after the lima-bean stage, it no longer affects morphogenesis. Similar results were observed when SPAS-1 was expressed under the dpy-7 promoter, which becomes active after the onset of elongation; the embryos elongate normally although they lack MTs. However, MTs might provide some stiffness, since a transient and local bulge is often observed where they are degraded. Finally we found that the Rho kinase let-502 mutants are hypersensitive to SPAS-1 expression, arguing that actomyosin tension could compensate for the lack of MTs. In parallel, to test if some genes required for elongation also affect MT dynamics, we have generated an integrated P_dpy-7::EB-1::GFP line. We observed that the MT growth rate is higher in seam cells than in dorso-ventral cells. Interestingly this difference is almost abolished in the let-502 mutants, which are blocked in elongation and lack actomyosin tension. In conclusion, there would be a close link between actin and MTs in morphogenesis: i) actomyosin tension could compensate the lack of MTs in the epidermis, ii) MT growth rate in the seam cells depends on a normal actomyosin tension. Further tests should confirm or infirm this hypothesis.

Shafaq-zadah, Massi et al. (2009) International Worm Meeting "EM analysis of the adaptor complex AP-1 loss of function phenotype reveals polarity and endocytosis defects."

Schwab, Yannick, Michaux, Gregoire, Shafaq-zadah, Massi, Labouesse, Michel

[International Worm Meeting, 2009]

We are interested in the function of the AP-1 complex, a clathrin adaptor which is implicated in membrane traffic. AP-1 loss of function induces a developmental arrest at the 3-fold stage (see also Shafaq-zadah & Michaux abstract). In order to study this phenotype, we decided to use electron microscopy (EM). WT adults were RNAi treated by feeding for three hours and the embryos were fixed 10 hours later by high pressure freezing, followed by freeze substitution. We examined hypodermal and intestinal cells for polarity and trafficking defects. In the hypodermis, all intracellular organelles looked normal, including ER, Golgi and various classes of endosomes and lysosomes. Although the cuticle was sometimes irregular in thickness alae could be identified even in 3-fold arrested embryos, indicating no major problem in cuticle synthesis and secretion, as also shown by a normal morphology and number of light multivesicular bodies (MVB) containing exosomes (Liegeois et al, JCB, 2006). Furthermore it demonstrates that the arrest in elongation does not affect the rest of the developmental process. However apical junctions were abnormal, too long and mislocalised along the lateral membrane instead of being subapical; additionally seam cells were sometimes bulging. We are now planning to perform immuno-EM to analyse the precise localisation of various junction markers. In the intestine, several phenotypes were observed. First, there was an accumulation of MVBs and lysosomes between the apical pole and the nucleus. We also observed that small and dense vesicular structures of about 100nm in diameter and restricted to the apical cortex were absent in AP-1(RNAi) embryos. Based on the disappearance of RAB-11::GFP positive apical endosomes in these embryos, as seen by confocal microscopy, we propose that these vesicular structures easily identified by EM could be apical recycling endosomes. Immuno-EM will allow us to test that hypothesis. We also observed defects at the apical pole with microvilli disorganised and/or abnormally long. Finally, apical junctions displayed abnormalities in length, although not misplaced as in the hypodermis. A surprising observation was that the intestinal lumen was full of cellular debris, as well as the space between the embryo and the eggshell. We interpret that as leaking of cells or of parts of cells which then float in the egg cavity and can be ingested; basically the embryo is feeding on itself. This is likely due to the loss of adherence between hypodermal cells, leading to cellular debris detaching from the hypodermal cells. Indeed these cells often have irregular shapes compared to control embryos; contrary to this intestinal cells do not look affected in a similar way.