Hutter, Harald et al. (2009) International Worm Meeting "GExplore: Multi-gene data mining for worm geneticists."

Chen, Nansheng, Hutter, Harald, Ng, Man-Ping

[International Worm Meeting, 2009]

The C. elegans genome contains some 20,000 genes, the majority of which has not been functionally characterized. Numerous genome wide data sets providing hints for gene function exist, yet mining those data to retrieve potential candidate genes for a particular function remains a challenge. Data most relevant for gene/protein function are - phenotype and expression (if known) - homology to other proteins (which might have been characterized already) - protein domains, which are frequently diagnostic for a biochemical function GExplore provides a simple database interface for worm biologists/geneticists for data mining at the gene/protein function level to help in experimental planning. It combines data downloaded from Wormbase with selected data sets related to gene expression or function, like SAGE data sets containing stage and tissue specific expression profiles. Data are preprocessed and organized for simple searches. The interface allows individual or combinatorial searches for proteins with certain domains, expression in certain tissues and/or certain phenotypes. Genes in a given genetic or physical interval can be listed to provide a convenient way to screen for candidate genes in mapping/cloning experiments. Information about available alleles can be displayed to simplify experimental planning. Literature related to entire lists of genes can be quickly retrieved and surveyed. GExplore operates on a local database and has fast response times, which is essential for exploratory searches. The interface is simple yet flexible enough to accommodate the addition of more data sets in the future. Since it is mainly based on data extracted from Wormbase it can be easily updated as new stable releases become available. GExplore is hosted under: http://genome.sfu.ca/gexplore/.

Harald Hutter (2002) European Worm Meeting "Axon guidance in the ventral cord of C. elegans"

Harald Hutter

[European Worm Meeting, 2002]

The ventral cord is the major longitudinal axon tract in C. elegans containing essential components of the motor circuit. It consists of two bundles running on either side of the ventral midline. Most axons enter the ventral cord after exiting the nerve ring in the head of the animal. Essentially all axons cross over to the right side near the retrovesicular ganglion leading to a highly asymmetrical ventral cord. Motorneuron axons originating from cell bodies at the ventral midline also exclusively grow in the right axon tract. The arrangement of processes within the cord is highly invariant implying that there is a precise navigational system guiding axons not just to the cord but also within the cord. The order of outgrowth of axons in the ventral cord during embryonic development has been studied by Richard Durbin1, who also initiated a first analysis on the pioneering function of certain early outgrowing axons. A number of genes are now known to affect guidance of axons in the ventral cord. The importance of particular guidance cues and pioneers for the navigation of individual classes of axons in the ventral cord can now be studied more systematically, since it is possible to visualize all major groups of ventral cord axons at the light microscopic level with GFP markers. To analyse pioneer-follower relationships several strains were made which express color variants of GFP in different groups of axons. The importance of pioneers like AVG was tested in laser ablation experiments. It was found that interneurons (glr-1 expressing) and D-type motorneurons are highly dependant on the pioneer AVG, whereas DA/DB motorneurons are almost not affected in the absence of AVG. Navigational mistakes are made individually. Errors of early outgrowing axons (DD) typically do not lead to mistakes of later outgrowing axons. Order of outgrowth apparently does not always reflect a pioneer-follower relationship. Similar observations were made in the absence of guidance cues like UNC-6. AVG and UNC-6 appear to be the main sources of navigational information for D-type motorneuron axons, but not so much for DA/DB motorneuron or interneuron axons indicating that different classes of neurons with axons in the ventral cord use different navigational cues. This analysis is currently extended to include other guidance cues and to test for redundancy among early outgrowing axons for the guidance of later outgrowing ones.

Harald Hutter et al. (2005) International Worm Meeting "The astacin family in C. elegans: evolution, expression and function"

Robert Johnsen, Marcus Schupp, Frank Mohrlen, Harald Hutter, Allan Mah, Donald Moerman

[International Worm Meeting, 2005]

Astacins are an evolutionary conserved family of secreted metalloproteases found in all animals. Members of this family are implicated in such different processes as digestion of food or pattern formation during embryonic development. C. elegans has the largest number of astacin genes among animals (39 genes). A comparison with C. briggsae indicates that most members of this family originated before these two species were separated in evolution, since only four astacins are unique to C. elegans and only one is unique to C. briggsae. Only three members of this gene family have been functionally characterized and these are involved in either hatching (hch-1), molting (nas-37) or cuticular collagen processing (dpy-31/toh-2). In order to get an idea of the functional diversity of this gene family in C. elegans we analysed the expression pattern of all astacins with promoter-GFP constructs and SAGE analysis (see http://elegans.bcgsc.bc.ca/ for link to this data). One third of the astacins are not represented in any of the developmental or tissue specific SAGE libraries posted to date, and for one quarter we did not observe GFP expression in the reporter strain. However, combining SAGE and GFP data we obtained tissue specfic expression information for all but three astacins, indicating that the two approaches can be complementary. With only a few exceptions, astacins tags in SAGE libraries were not abundant. GFP expression levels reflected this low message abundance for many family members. Several astacins not represented in SAGE libraries show clear GFP expression limited to a few cells or to certain stages in development indicating that there is a detection limit for SAGE when the abundance of a message is either very low or limited to a few cells. The digestive tract (pharynx, gut) and the hypodermis are tissues expressing a variety of different astacins, suggesting that a large fraction of the astacins might play a role in food digestion and molting/cuticular collagen processing. Only a few members are expressed in body wall muscle or the nervous system. To initiate a functional analysis we selected certain members for targeted knock-out. So far we have isolated mutants in nas-4, nas-5 and nas-39 (tolloid homolog), which are viable and show no obvious developmental defects.

Harald Hutter (2003) International Worm Meeting "Axon guidance in the ventral cord of C. elegans: The role of pioneers and extracellular guidance cues"

Harald Hutter

[International Worm Meeting, 2003]

The ventral cord is the major longitudinal axon tract in C. elegans consisting of two bundles running on either side of the ventral midline. The majority of the axons runs in the right axon tract, despite an overall left-right symmetry of the nervous system. The arrangement of neuronal processes within the ventral cord is highly invariant implying that there is a precise navigational system guiding axons not just to the cord but also within the cord. Ventral cord axons grow out sequentially, suggesting that there might be dependancies between early and late outgrowing axons (pioneer-follower relationship). To systematically analyse such dependancies, transgenic strains were made expressing two or three color variants of GFP in different groups of axons. The importance of pioneers like the AVG neuron, which is the first neuron to send an axon into the right ventral cord tract, was tested in laser ablation experiments. It was found that certain classes of interneurons and motorneurons are highly dependant on this pioneer, whereas other classes of motorneurons are almost not affected by its absence. Surprisingly, neither AVG nor pioneers supposedly leading nerve ring axons to the right side of the ventral cord (RIF, RIG, SABV) are essential for the generation of left-right asymmetry of the ventral cord. Navigational errors like crossing of the ventral midline by individual axons do not seem to influence other nearby axons, which continue on their normal path, but might commit a similar mistake independantly. Frequently misrouting of early outgrowing axons does not lead to errors of later outgrowing axons, suggesting that the order of outgrowth does not necessarily reflect a strict pioneer-follower relationship. On the other hand, a pairwise correlation of axon outgrowth defects of different classes of neurons revealed subtle dependancies for certain combinations, e.g. between interneurons and PVQ axons, indicating that early outgrowing axons have some influence on the behaviour of late outgrowing axons. A detailed analysis of defects in known axon guidance mutants revealed that the different classes of neurons use different combinations of guidance cues. Even early outgrowing axons like AVG and DD-type motorneuron axons, which should depend almost exclusively on extracellular cues, are affected rather differently by the absence of known extracellular guidance signals. Taken together this analysis shows: not all the early outgrowing axons act as pioneer; later outgrowing axons can interpret the same navigational cues as those used by the pioneers; different axons in the same axon bundle are fairly independant in their outgrowth using different combinations of guidance cues for their navigation.

Steimel, Andreas et al. (2009) International Worm Meeting "The non-classical cadherin fmi-1 mediates pioneer-follower axon guidance in the ventral nerve cord."

Hutter, Harald, Steimel, Andreas

[International Worm Meeting, 2009]

The ventral cord axon tracks are established through sequential outgrowth of pioneer and follower axons. The PVP axon pioneers the left axon track closely followed by the PVQ axon. In an EMS screen for animals with defects in ventral cord axon guidance we isolated the fmi-1 allele rh308 (1). fmi-1(rh308) animals display strong PVP and PVQ axon guidance defects. Interestingly the pioneer-follower relationship between PVP and PVQ axons is disrupted in fmi-1(rh308) animals. PVQ axons cross the ventral midline independently of the PVP axons or even leave the ventral cord. They stop prematurely in virtually all fmi-1(rh308) animals. Further fmi-1 alleles were isolated in genetic screens for defects in synapse formation and HSN axon guidance in the Jin and Garriga Lab respectively. In 68% of fmi-1(rh308) animals HSN axons fail to join the ventral cord axon tracks and circle around the vulva. Moreover HSN axons stop before reaching the nerve ring in nearly all fmi-1(rh308) animals. Interneuron axons that extend along pioneers in the right axon track are affected in 31% of fmi-1(rh308) animals. Fmi-1 is the homologue of Drosophila Flamingo and vertebrate CELSR1, 2 and 3, which are known to function in neuronal development (2). FMI-1 is characterized by a unique domain composition with eight cadherin repeats, laminin G and EGF modules and a G-protein coupled receptor domain. A 2.6 kb fmi-1 promoter GFP construct is mainly expressed in neurons including PVP, PVQ and HSN. Expression starts during gastrulation before axons grow out, persists throughout all larval stages and decreases noticeably in adults. To determine the subcellular localization of FMI-1 we fused the fmi-1 transcript to GFP. FMI-1::GFP is predominantly localized to axons and rescues PVQ defects in fmi-1(rh308) animals. Mosaic analysis revealed that fmi-1 acts cell-autonomously in PVP and PVQ axons. Interestingly fmi-1 is necessary in both PVP and PVQ axons for correct PVQ axon guidance. An fmi-1 construct containing only the cadherin repeats in the extra-cellular domain can nearly completely rescue PVP-independent PVQ axon outgrowth. These data suggest a model, where FMI-1 acts as homophilic adhesion molecule between pioneer and follower neurons to ensure that follower axons extend along the pioneer axons. 1. Hutter et al., Dev Biol 284 (2005) 260-272; 2. Takeichi, Nat Rev Neurosci. 2007 Jan;8(1):11-20.

Suh, Jinkyo et al. (2013) International Worm Meeting "GExplore updated: more genome-scale data mining for worm researchers."

Suh, Jinkyo, Hutter, Harald

[International Worm Meeting, 2013]

Previously we implemented a simple web-based database interface for genome-scale data mining at the protein (rather than the DNA) level for experimental planning - hosted at: http://genome.sfu.ca/gexplore/. The interface allows individual or combinatorial searches for proteins with certain domains, expression in particular tissues or specific phenotypes. Genes in a given interval can be retrieved for vetting candidate genes in positional mapping experiments. The database, which contains also GO terms and homology information, has been updated regularly using Wormbase data. Recently new datasets including stage- and sex-specific RNAseq expression data (see Gerstein et al., Science 330:1775-87) have been processed and added. We are currently preparing a major update of GExplore to include several novel features and datasets. We included protein domain data using Pfam and SMART annotations for all core species in Wormbase (C. elegans, C. briggsae, C. brenneri, C. japonica, C. remanei and P. pacificus). We extended the manually curated list of protein domains that can be searched by simple abbreviations, e.g. 'IG' to retrieve all proteins containing Pfam or SMART domains referring to ImmunoGlobulin domains, to over 450 domains. Proteins containing a particular Pfam or SMART domain can be identified using Pfam or SMART identifiers. In addition to a graphic display of the domain organization of proteins, the protein sequences are available for display and download. A new interface allows users to observe where mutations are located within a protein of interest. The output can be limited to mutations affecting certain protein domains. Our database contains mutations accessible through Wormmart and will be updated as more recent Wormbase data will become available for general data mining. The database also includes mutations isolated by the Million Mutation Project, which have been released recently by the Moerman and Waterston labs (Thompson et al. submitted). Independently of GExplore we maintain a web-based interface to search for mutations from the Million Mutation Project (hosted at http://genome.sfu.ca/mmp/). We will present the current progress in updating the GExplore website and underlying database.

Bhat, Jaffar M. et al. (2015) International Worm Meeting "A novel role for AEX-3, a guanine nucleotide exchange factor (GEF) for the Rab3 GTPase, in axon guidance."

Bhat, Jaffar M., Hutter, Harald

[International Worm Meeting, 2015]

During embryonic development neurons differentiate and extend axons and dendrites that have to reach their appropriate targets. This task is highly regulated and is achieved by using a set of conserved guidance cues and receptors. The first neurons which extend their processes are called 'pioneer' neurons. Pioneers establish the major axon tracts which are used by the latter outgrowing axons (followers) to extend along. In C. elegans the AVG neuron pioneers the right axon tract of the ventral nerve cord, the major longitudinal axon tract in the animal. The cues used by AVG for axonal outgrowth and navigation are not known yet as the available mutants for known guidance cues and receptors do not show significant defects in AVG navigation or outgrowth. Moreover, simple genetic screens have not yielded new mutants with penetrant AVG axon guidance defects. Nidogen (nid-1), a basement membrane component which plays a role in the positioning of longitudinal axons, has weakly penetrant AVG axon navigation defects. We performed an enhancer screen in a nid-1 mutant background and isolated a novel allele (hd148) of aex-3. aex-3(hd148) mutants have highly penetrant AVG axon navigation defects, which are nid-1 dependent. In addition aex-3(hd148); nid-1 mutant animals also show axon navigation defects in interneurons, DD/VD and DA/DB motor neurons as well as in AVK and HSN neurons. AEX-3 is a guanine nucleotide exchange factor (GEF) for the Rab3 and Rab27 GTPases thereby activating Rab3 and Rab27. These GTPases are important for trafficking of synaptic vesicle precursors as well as synaptic vesicle release. We found that rab-3 has nid-1 dependent AVG axon navigation defects whereas aex-6/Rab27 does not show these defects. Our genetic interaction data also suggests that aex-3 and rab-3 are in the same genetic pathway. This suggests that AEX-3 activates Rab3 but not Rab27 GTPase in the context of AVG axon navigation. Recently it has been shown that syntaxin 1 (unc-64) and DCC (unc-40) coassociate with each other and this interaction is required for Netrin (unc-6) dependent guidance of migrating axons. We found that both unc-6 and unc-64 have AVG axon navigation defects which are nid-1 dependent and that both aex-3 and unc-64 are in the same genetic pathway. This suggests that UNC-64 likely plays a role in the fusion of AEX-3 activated vesicles which go through exocytosis. It is known that localized exocytosis and endocytosis is important for the steering and turning of growth cone. It is possible that AEX-3 is involved in localized exocytosis of vesicles at the growth cone thereby contributing to the correct navigation of the growth cone. .

Taylor, Jesse et al. (2015) International Worm Meeting "Genetic Screens Reveal Genes Regulating Ventral Nerve Cord Asymmetry in C. elegans."

Taylor, Jesse, Unsoeld, Thomas, Hutter, Harald

[International Worm Meeting, 2015]

The ventral nerve cord (VNC) is the major longitudinal axon tract of the nematode C.elegans. The two fascicles of the VNC extend to the left and right of the ventral midline over a basement membrane. The majority of axons are found in the right tract resulting in a noticeable asymmetry between the two tracts. Interestingly the anterior most portion of the VNC is actually symmetrical in appearance, with all but four axons decussating into the right VNC tract just posterior to the nerve ring. Known guidance cues have little or no effect on the navigation at this point of decussation suggesting the existence of additional cues. Forward genetic screens in our lab identified four mutants displaying mildly penetrant symmetrical VNCs and midline crossing defects of axons extending in the VNC. Two of the four mutants, hd131 and hd133, have undergone preliminary mapping and whole genome sequencing and likely define previously uncharacterized axon guidance genes. A third allele, hd132, was identified as a loss of function mutation in unc-34 the enabled/VASP homolog. The fourth mutant from the screens was identified as an allele of a previously uncharacterized transmembrane collagen, col-99. col-99 encodes a type II transmembrane collagen consisting of a small intracellular domain and a large extracellular domain containing several collagen domain repeats. Sequencing revealed our allele (hd130) as an early nonsense mutation likely resulting in complete loss-of-function. Further characterization of axonal trajectories in col-99 mutant animals identified significant axonal defects in most VNC neurons. Axonal navigation defects were also observed in other longitudinal axon tracts including the dorsal nerve cord and the two dorsal sublateral tracts. During axonal outgrowth in the embryo, col-99::GFP reporter expression is prominent in the hypodermis but not the nervous system suggesting a non cell-autonomous role for COL-99. Sequence comparison with vertebrate homologs revealed a conserved cleavage site indicating that COL-99 may be shed from the cell surface for function. Presently we explore the idea that the ectodomain of COL-99 is shed by proprotein convertases and localizes to the basement membrane. Current evidence suggests COL-99 is likely acting as a guidance cue for outgrowing axons potentially through interactions with basement membrane components and discoidin domain receptors on extending axons.

Viveiros, Ryan et al. (2009) International Worm Meeting "Investigation of myoblast migration during embryogenesis in Caenorhabditis elegans."

Viveiros, Ryan, Moerman, Donald, Hutter, Harald

[International Worm Meeting, 2009]

C. elegans body wall muscle is formed after a series of well-orchestrated steps. Initially, embryonic muscle cells accumulate under the hypodermal seam cells, at the left and right lateral midline, and shortly thereafter, begin to migrate dorsally and ventrally, to form the final four muscle quadrants present upon hatching. As the myoblasts migrate they are still dividing, as are many other cells in their immediate environment. This means the cell-cell contact of cells during migration is dynamic and can vary from animal to animal. This situation creates an environment where the extracellular matrix (ECM) and cell surface contacts are in constant flux, which begs the questions as to how these cells navigate unerringly to their final destination. Though a number of ECM and cell surface molecules have previously been shown the be involved in cell migration, positioning and attachment, including perlecan, type IV collagens, the NCAM, cadherin and integrin families, and the UNC-6/netrins, previous studies have not shown any of these molecules to affect the early muscle migration. Using RNAi to target ECM components, we found that loss of laminin results in a number of muscle migration defects. Analysis of the laminin knockdown phenotype reveals defects in the aforementioned dorsal and ventral muscle cell migrations, as well as a posterior displacement of the anterior-most ventral muscle cells. Investigation of this posterior displacement has led to the identification of a previously un-described anterior migration event of the embryonic muscle cells, MSapappp, MSapaaap, MSapapap and MSsappapp. This anterior migration is dependent upon the extension of processes from the anterior-most muscle cells, as these processes appear to be absent in the ventral anterior-most cells of the laminin knockdown embryos. Using spinning disk confocal microscopy we observe that the extension of the muscle processes to the anterior coincides with the onset of elongation and is concurrent with the dorsal and ventral migrations. Wildtype analysis has also revealed the persistence of cell-cell contacts between the dorsally and ventrally migrating muscle cells with the degree of contact increasing towards the posterior. These contacts are eventually lost as the embryo elongates. While laminin is required for a subset of these processes, their extension does not require pat-3/beta-integrin, suggesting either this is an integrin independent migratory event, or that there is a yet to be identified beta-integrin. Investigations into what other genes are involved in mediating these anterior migrations are currently underway.

Viveiros, Ryan et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "Investigation of muscle migration during embryogenesis in Caenorhabditis elegans"

Moerman, Donald, Viveiros, Ryan, Hutter, Harald

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

C. elegans body wall muscle is formed after a series of ordered steps. Initially, embryonic muscle cells accumulate under the hypodermal seam cells and shortly thereafter, begin to migrate dorsally and ventrally, to form the final four muscle quadrants present upon hatching. As the myoblasts migrate they are still dividing, as are many other cells in their immediate environment. This means the cell-cell contact of cells during migration is dynamic and can vary from animal to animal. This situation creates an environment where the extracellular matrix (ECM) and cell surface contacts are in constant flux, which begs the questions as to how these cells navigate unerringly to their final destination. To better characterize these muscle cells we used a membrane-bound GFP to visualize their membrane dynamics. This allowed us to identify filapodia and lamelapodia in the embryonic muscle cells and led to the identification of processes extending from the anterior-most muscle cells. Using spinning disk confocal microscopy, we observe that these extensions to the anterior coincide with the onset of elongation and are concurrent with the dorsal and ventral migrations. Using RNAi and mutant analysis to identify genes involved in mediating these migrations, we found that laminin, the a-integrin INA-1, the ephrin VAB-2, and its receptor VAB-1 and the ROBO receptor SAX-3 are required for proper muscle migration. Disrupting any of these genes results in defects in the extension of the ventral anterior processes, which leads to the posterior displacement of the ventral muscle quadrants. As most of these genes have a known role in hypodermal morphogenesis, we are currently investigating whether the improper muscle migration is due to hypodermal defects.