Poulin, Gino et al. (2021) International Worm Meeting "RBBP-5 regulated methylation at Histone 3 Lysine 4 promotes longevity in C. elegans"

Vintilla, Adriana, Fisher, Kate, Whitmarsh, Alan, Samsudin, Nurulhafizah, Fei, Pang Yoke, Velichkova, Gabriel, Poulin, Gino

[International Worm Meeting, 2021]

Histone 3 lysine 4 methylation (H3K4me) is an epigenetic mark regulating transcription, metabolism, and longevity. H3K4me necessitates, in addition to KMT2 enzymes, the conserved core components WDR5, ASH2, and RBBP5. In C. elegans, deficiency in WDR-5 or ASH-2 reduces H3K4me3 levels and extends lifespan. Transgenerational experiments whereby only the parental generation is deficient in WDR-5 or ASH-2 leads to incomplete reprogramming of the next generation resulting in long-lived descendants. Whether RBBP-5 has similar functions remains unclear. Herein, using spike in ChIP-seq, we show that RBBP-5 is required for H3K4 mono- and multi-methylation and adults lacking RBBP-5 are short-lived. In contrast with WDR-5 or ASH-2 deficiencies, our transgenerational experiments show that reprogramming of the next wild type generation is normal. Instead, we revealed that RBBP-5-deprived descendants originating from mothers heterozygous for the rbbp-5 deficiency inherit a wild type lifespan. However, at the fifth generation the rbbp-5(-) short-lived phenotype fully manifest. Using RNA-seq and reporter assays, we found that RBBP-5 is important to maintain mitochondrial bioenergetic and core translational functions. We also show that restoring RBBP-5 expression in the soma is sufficient to recapitulate the wild type lifespan, hence ruling out the germlines as the primary site for H3K4me longevity promoting functions. Collectively, this work shows RBBP-5 regulated H3K4me promotes longevity and wild type lifespan can be epigenetically transmitted to H3K4me-depleted descendants most likely by a mechanism alleviating mitochondrial and translational dysfunctions in somatic cells.

Poulin, Gino et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "Methylation and demethylation activities of a C. elegans MLL-like complex attenuate RAS signalling"

Fisher, Kate, Wilson, Jon, Poulin, Gino, Southall, Stacey

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

The conserved Mixed Lineage Leukaemia (MLL) complex deposits activating methyl marks on histone tails through a methyltransferase (MT) activity. Here we provide in vivo evidence that in addition to methylation, the C. elegans MLL-like complex can remove specific methyl marks linked to repression of transcription. This supports the proposed model in which the MLL complex orchestrates both the deposition and the removal of methyl marks to activate transcription. We have uncovered the MLL-like complex in a large-scale RNAi screen designed to identify attenuators of RAS signalling during vulval development. We have also found that the histone acetyltransferase complex, NuA4/TIP60, cooperates with the C. elegans MLL-like complex in the attenuation of RAS signalling. Critically, we show that both complexes regulate a common novel target and attenuator of RAS signalling, AJM-1 (Apical Junction Molecule-1). Therefore, the C. elegans MLL-like complex cooperates with the NuA4/TIP60 complex to regulate the expression of a novel effector, AJM-1.

Gino Poulin et al. (2003) International Worm Meeting "Genome wide RNAi screen for new SynMuV genes"

Andrew Fraser, Yan Dong, Hui Ge, Julie Ahringer, Marc Vidal, Gino Poulin

[International Worm Meeting, 2003]

Proper vulval development requires coordination between the Ras, Wnt, Notch and the SynMuv (synthetic multiple vulvae) pathways. Vulval development is induced by Ras signalling in three of six precursor cells; vulval development is inhibited in the remaining three cells, in part, by the synthetic multivulval (SynMuv) genes. These latter genes, which antagonise RAS signalling, form two functionally redundant groups (class A and class B). Animals lacking a gene in either class have a normal vulva, but in SynMuVA;SynMuVB double mutants more than three of the six precursor cells form vulval tissue resulting in the multivulval, or Muv, phenotype. A subset of the SynMuv genes include the tumor suppressor Rb and components of the NuRD chromatin remodelling complex. The functional relationships between the SynMuv genes, how their different roles are integrated, and what targets they regulate are not yet clear. For example, it is not known how or when NuRD is recruited to chromatin neither is the nature of the redundancy between the class A and class B SynMuv genes. One problem in understanding how this system works is that additional SynMuv genes remain to be identified. We screened our RNAi feeding library that encompasses 86% of the C. elegans genome for new SynMuv components. We chose lin-15A(n767) and lin-15B(n744) as representative of class A and B genes, respectively. We identified a total of 25 new synthetic multiple vulvae genes: 20 SynMuvB, 3 SynMuv A and 2 SynMuvAB. Of the 25 genes, all but 6 have a putative human homologue. Based on their sequence, 10 genes appear to have a role in transcription regulation. Our data suggest that regulation of the vulval fate involves coordination between a number of different complexes, including CCR4-NOT, Mediator, and Polycomb. We are currently conducting an all against all two-hybrid test of all the new and previously identified SynMuv proteins to test for direct interaction between the different complexes. We also found that the sumoylation pathway is involved in controlling vulval fate since the sumo-1 homologue and some of the sumoylation enzymes were identified as Muv or SynMuv genes. We are further classifying the new SynMuv genes through characterisation of vulval markers and the ability to affect the Muv phenotype of a let-60/Ras gain of function mutant. This systematic approach should help our understanding of links between cellular signalling and transcriptional regulation.

Gino Poulin et al. (2004) East Coast Worm Meeting "Genome wide RNAi screen for new synMuv genes"

Yan Dong, Andrew Fraser, Gino Poulin, Julie Ahringer, Neil Hopper

[East Coast Worm Meeting, 2004]

Ras signalling promotes vulval development, and is antagonized by the functionally redundant class A and class B synthetic multivulval genes. Previous work showed that a Retinoblastoma protein and components of the NuRD histone deacetylase complex are encoded by some of the synthetic multivulval genes, indicating that negative transcriptional regulation of vulval development genes underlies the process of Ras antagonism. We carried out a genome-wide RNAi screen to identify systematically other components of this regulatory mechanism. After RNAi of 16,757 genes, we found 19 synMuv genes, 9 of which are new; 80% of previously known synMuv genes in the library were identified in the screen. Based on the sequence of the 9 newly identified genes, 7 appear to have a role in transcription regulation, and these have human homologs. For example, we identified a histone methytransferase and a homolog of lethal (3) malignant brain tumor, a tumor suppressor in Drosophila. We also found a requirement for components of a TRRAP/TIP60 histone acetyltransferase complex, suggesting that both acetlyation and deacetylation of histone tails may have roles in repression of vulval development. Ceol et al. (2004) have also recently identified this complex as part of a new synMuv C class. In addition, we show involvement of several components of the sumoylation pathway and find that the histone deacetylase HDA-1 is sumoylated. These data indicate that antagonism of Ras induced vulval development involves unexpectedly complex coordination among numerous chromatin regulators. As most of the genes identified in this screen are conserved in humans, we suggest that similar interactions are relevant for generating a repressed chromatin state to inhibit Ras signalling in mammalian cells.

Gino Poulin et al. (2005) International Worm Meeting "Chromatin regulation and sumoylation in the inhibition of Rasinduced vulval development in C. elegans"

Yan Dong, Gino Poulin, Julie Ahringer, Andrew Fraser, Neil Hopper

[International Worm Meeting, 2005]

During vulval development, Ras signalling induces vulval precursor cells to have a vulval fate. Ras signalling in the VPCs is antagonised by three genetically redundant groups of genes, the class A, B and C synthetic multivulval (synMuv) genes. Single mutants have a wild-type vulva (except for class C) and double mutants between each class, but not within the same class are multivulval (Muv). Many synMuv genes encode chromatin-associated proteins involved in transcriptional repression, including an ortholog of Rb and components of the NuRD histone deacetylase complex. To identify other components of this regulatory mechanism, we carried out a genome-wide RNAi screen. After RNAi of 16,757 genes, we found 19 synMuv genes, 9 of which are new. Based on the sequence of these genes and genetic epistasis experiments, we propose that at least four post-translational modifications converge to inhibit Ras-stimulated vulval development: sumoylation, histone tail deacetylation, methylation, and acetylation. In further investigating the role of sumoylation during vulva development, we found that sumoylation positively modulates synMuv activity, but inhibits LIN-12/Notch signalling. We also show that many of the synMuv genes are involved in gene regulation outside the vulva, negatively regulating the expression of a lag-2::gfp reporter. Taken together, our results suggest that the same group of chromatin regulators function together in multiple developmental events and that sumoylation modulates key signalling pathways during vulva development.

Gino Poulin et al. (2002) European Worm Meeting "The human and C.elegans p66 NuRD chromatin remodelling components: transcriptional repression and SynMuv activities"

Gino Poulin, Cara Neades, Julie Ahringer

[European Worm Meeting, 2002]

We are currently using the C. elegans vulva as a model to study chromatin remodelling activity. An important chromatin remodelling complex for vulva specification is NuRD. Studies in mammalian cells has shown that the NuRD complex contains ATP-dependant nucleosomal remodelling activity and histone deacetylase activity and it is thought to be involved in transcriptional repression. Vulval development requires positive and negative signals: three of the six vulval precursor cells (VPCs) are positively induced to have a vulval fate by Ras signalling, and an antagonising activity from the SynMuv (synthetic multiple vulvae) genes prevents vulval cell determination of the remaining three VPCs. There are two groups of SynMuv genes, SynMuvA and SynMuvB, which act redundantly to repress vulval development. Simultaneous inactivation of one gene in each class results in ectopic vulval development through inappropriate Ras activation. Inactivation by RNAi of the NuRD complex components in a SynMuvA or SynMuvB mutated genetic background has been shown to cause a Muv phenotype. The NuRD components that have been tested so far appear to be part of either the SynMuvA or the SynMuvB pathway and in some cases both. One component of the NuRD complex, EGL-27, is a SynMuvA gene and related to human MTA1 (metastasis-associated gene 1). To try identifying EGL-27 interacting partners, a two-hybrid screen was undertaken and the C.elegans homologue of the NuRD component p66 was found. This gene has a predicted GATAlike zinc- finger at its C-terminal end. RNAi by injection results in lethality, so to analyse vulva development, RNAi by feeding was used. We found that p66 inactivation in wild type animals results in a weak Muv phenotype (3%) and that this phenotype is strongly enhanced (65%) in a SynMuvB background. We have also tested the transcriptional activity of two human p66 homologues by transfection assay in human 293T kidney cells and have found that both efficiently repress transcription when fused to GAL4 DNA binding domain. Testing of C.elegans p66 is in progress. Therefore, p66 is a new SynMuvA gene and its human counterparts can repress transcription, supporting the idea that NuRD is involved in repression of vulval development genes. Finally, because the vulva is a good paradigm for studying interplay between the Ras-signalling pathway and signals antagonising its effects, we are using our nearly comprehensive RNAi library to screen for new SynMuv genes.

Gino Poulin et al. (2006) European Worm Meeting "A Candidate-Based Approach for Genetic Interaction with gap-1 identifies a WD40 Encoding Gene"

Julie Ahringer, Gino Poulin

[European Worm Meeting, 2006]

Gino Poulin and Julie Ahringer. Vulval fate induction is an established readout to identify regulators of multiple signalling pathways including the canonical RTK/RAS/RAF/MPK signalling pathway. Vulval fate adoption requires an inductive signal (LIN-3 EGF) that emanates from the anchor cell (AC) and activates Ras signalling in three of the six vulval precursor cells (VPCs). The VPC closest to the source of LIN-3 EGF adopts the primary fate and produces the lateral signal that activates LIN-12 NOTCH signalling in the neighbouring secondary VPCs. In these cells, negative regulators of Ras ensure down regulation of Ras signalling in a Notch dependant manner, an important process to maintain vulval cell patterning. The last three VPCs adopt a non-vulval fate, because of the Ras inhibiting activity of the genetically redundant synMuv genes. GAP-1 is a GTPase activating protein that directly regulates LET-60 RAS by stimulating its GTPase activity and therefore increases the rate of LET-60 RAS bound to GDP, which is the inactive form. In absence of gap-1 animals develop a normal vulva, however vulval fate adoption becomes more susceptible to elevated Ras signalling. Hence, additional loss of a negative regulator of Ras causes extra vulval precursor cells to adopt the vulval fate and produces the Muv phenotype. We have previously performed genome-wide RNAi screens for new synMuv genes. RNAi of about half of our initial candidates caused a synMuv independent low percentage Muv phenotype. We reassessed these and the synMuv interacting genes (100 genes) in a gap-1 background and found seven where RNAi caused a high percentage Muv phenotype. We identified the members of the sumoylation pathway, two components of the NuRD complex, sys-1, and a novel WD40 encoding gene, that we named gin-1, for gap interactor-1. We focused our analysis on gin-1 and found that it inhibits Ras signalling upstream of sem-5 (Grb2). We also found that RNAi of gin-1 suppresses a let-23 EGFR loss-of-function mutant, suggesting that the inhibition occurs at or downstream of the receptor. In addition, using egl-17::cfp as a marker for primary cells, we found that gin-1(RNAi) animals show persistent expression in the normally non-expressing secondary cells, showing that gin-1 plays a role in Ras signalling inhibition during vulval fate adoption. To rule out the possibility that anchor cell duplication is responsible for our observations, we used zmp-1::yfp as an anchor cell marker and found no defect at this level. Taken together, our candidate-based approach has identify seven novel gap-1 interactors; one of these, gin-1, appears to regulate Ras signaling between the receptor and sem-5.

Gee, Fiona et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "A candidate-based approach to identify genetic interaction with SUMO identifies a bromodomain-encoding gene"

Poulin, Gino, Gee, Fiona

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

The post-translational modifier SUMO (small ubiquitin-like modifier) has a key role in regulation of transcription. The C. elegans SUMO, smo-1 , was found in a synMuv screen, and may inhibit Ras signalling through interactions with chromatin-remodelling factors. The development of the C. elegans vulva provides an ideal model system for the identification of new interactors of SUMO. Ras signalling leads to adoption of the vulval cell fate by vulval precursor cells, and raised levels of this signal lead to the multivulva (Muv) phenotype. On this basis, we have performed a small-scale, candidate-based RNAi screen in a smo-1 balanced background, in order to identify novel chromatin factor interactors of SUMO. We identified expected and novel interactors of smo-1 , amongst which was the double bromodomain-encoding gene bet-1 . Using egl-17::cfp as a marker for high levels of Ras signalling in the vulval precursor cells, we found that animals depleted of bet-1 show ectopic expression of egl-17::cfp , suggesting that bet-1 has a role in Ras signalling inhibition. As we suspected that loss of bet-1 leads to hyperactivation of the Ras signalling cascade, we tested global levels of phosphorylated MPK-1 by western blotting. We found increased levels of phospho-MPK-1 in bet-1 depleted animals. Since bet-1 was identified in a smo-1 synthetic screen, we constructed the double mutant bet-1;smo-1 , and observed an unexpected and striking synthetic Clear (synClr) phenotype. The Clr phenotype has been previously documented to be due to muscle degradation associated with elevated Ras signalling through the FGFR pathway. Taken together, this shows that bet-1 plays an important role in the inhibition of Ras signalling in a range of tissues. As the human homolo gues of bet-1 are implicated in epigenetic memory (due to their retention on mitotic chromosomes), further investigations could shed light upon the workings of this process.

Lang, Jiaqing et al. (2021) International Worm Meeting "Role of Stress Granules in Stress Responses and Ageing"

Whitmarsh, Alan, Poulin, Gino, Sfakianos, Aristeidis, Lang, Jiaqing

[International Worm Meeting, 2021]

Stress granules (SGs) are cytoplasmic ribonucleoprotein condensates that help reprogramme cells to adapt to and survive stress. Their dysregulation has been implicated in neurodegenerative diseases, cancer and ageing. SGs are conserved amongst eukaryotes and are composed predominantly of untranslated mRNAs, translation initiation complexes and RNA-binding proteins. They also interact with cellular signalling pathways to regulate changes in mRNA translation underpinning altered cell fate. Much of our understanding of SG function and behaviour is derived from cell-based studies, so it is important to address their role in an animal model. Our group has shown that key pathways regulating translation are important for SG assembly and function in human cell lines and we are now determining the in vivosignificance of this regulation in C. elegans. For example, we have found that the mTOR-S6 kinase pathway regulates the assembly and maintenance of SGs in both human cells and in C. elegans in response to heat shock. We have also found that specific translation initiation factors play key regulatory roles in SG assembly and aim to uncover the relationship between stress-induced translation inhibition, SG dynamics and organism adaption over the lifespan of C. elegans. This research will provide new insights into the role SGs play in the integrated organismal response to both acute environmental insult and longer-term stress.

Elvin, Mark et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "A food consumption assay to enable super high throughput RNAi"

Frejno, Martin, Snoek, Basten, Kammenga, Jan, Poulin, Gino, Elvin, Mark

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

Performing RNAi by feeding was an important step that made loss of function analysis truly high throughput. However, the bottleneck of this technology remains the scoring. Different approaches have been used to bypass the tedious task of sitting at the microscope, but many of these are either expensive or of a lower throughput. Here we will present a strategy that is based on how much bacteria is eaten by the worms in function of the time, which is an indication of the worms fitness. The RNAi is performed in liquid, in a 96-well format and analysed using a plate reader with a 600 nm filter. Readings are taken everyday for a week and a curve of OD in function of the time is derived for every experiment. Here, we s how that this method can be used to distinguish between different worm strains, e.g. lin-15A, and lin-15B; and detect synthetic interactions e.g. lin-15AB. We have also tested whether the lack of RNAi sensitivity observed in CB4856 (Hawaii) compared to N2 (Bristol) could be detected. This effect has been largely attributed to polymorphisms in ppw-1. We have used a panel of 12 RNAi feeding clones, and found that Hawaii is rather resilient to most RNAi clones, but not so to sumo-1. Therefore, we can distinguish Bristol from Hawaii using this assay. Then, using 50 Recombinant Inbred Lines (RILs) between Hawaii and Bristol, we tested which of these are sensitive to RNAi and performed QTL analysis. We have found that even though most can fall into one or the other category a number of RILs are hypersensitive or nearly totally resistant to RNAi. Interestingly, we have found a RIL that phenotypically behaves as Hawaii (i.e. is RNAi insensitive), but is genotypically Bristol for ppw -1. Therefore, we have established a new method that significantly improves the throughput of RNAi in liquid, since reading a 96-well plate takes seconds. Additionally, we demonstrate that RNAi sensitivity is a complex trait that can be quantified using our food consumption assay.