Final Report Summary - WORMPROLOC (The subnuclear relocation of promoters during development, and following stress induction in C. elegans)
The regulation of gene expression depends on a wide range of mechanisms that increase or decrease the production of specific gene products (RNA or protein) depending on cell type and environmental stimuli or stress. The formation of heterochromatin (the condensed compartment) in eukaryotic genomes, which accumulates as cells differentiate, represents a universally conserved means to silence genes and repetitive DNA elements found in it. Regulation of gene expression through heterochromatin is critical for the maintenance of developmental pathways and for ensuring appropriate response to environmental stimuli, so in other words is critical for life. To date, transcriptional repression is believed to be the main -if not the only - mechanism responsible for ensuring stable heterochromatin silencing. However, there are several indications that additional mechanisms that silence heterochromatic genes at the co- or post-transcriptional level may exist. In this project we took advantage of an unexpected recent finding and available resources produced by our laboratory in order to get a better understanding of these potential co- or post-transcriptional novel mechanisms leading to heterochromatin silencing. This unique opportunity and advances in research in between the proposal writing and the start of the project prompted the slight difference in the aspect of gene regulation investigated from the one proposed originally. So, in a genome-wide derepression screen of a silent, heterochromatic gene array expressing GFP in C. elegans, among the expected chromatin and other transcription regulators, Towbin et al. identified few unexpected hits able to derepress this heterochromatic array, such as three Lsm proteins, which are not involved apparently in transcription but in RNA processing. Lsm proteins bind RNA and are evolutionarily conserved with C. elegans Lsm proteins having between 60 to 90 percent identity with human Lsm proteins. Lsm proteins assemble in two major heptameric complexes, the Lsm1-7 and Lsm2-8 complexes. To date, it is known that the Lsm1-7 complex is involved in cytoplasmic mRNA decay and the Lsm2-8 complex regulates both nuclear pre-mRNA decay and U6 snRNA stability (Beggs et al., 2005; Tharun et al., 2009). In our attempt to study the role of Lsm proteins in heterochromatic gene silencing, we asked the following questions: What the Lsm proteins are silencing? Which of the Lsm proteins are involved in the heterochromatic gene silencing, and which other co-factors? How does the Lsm complex of interest silence heterochromatic transgenes?
Answering the first question, we found that the derepression caused by the down-regulation of the three Lsm hits (Gut-2/Lsm-2, Lsm-5 and Lsm-6) could be observed and quantified (using microscopy and an adapted method using a worm sorter machine) in old embryos and in every other developmental stage of the worm in every somatic cell but not in the germ cells. This suggests that the Lsm proteins are involved in a somatic silencing mechanism. We also found that the heterochromatic array derepression was found at the protein level as quantified by the GFP intensity but also at the RNA level, as expected, tested by qPCR. Additionally and importantly, we found that down-regulation by RNAi of Lsm proteins derepressed all of the four different heterochromatic reporter transgenes we have tested whereas the two tested ‘non-heterochromatic’ transgenes expression was not affected, which indicates that those Lsm proteins are involved specifically in silencing reporter transgene with heterochromatic features, and possibly this will be also true for endogenous heterochromatic regions, which we are currently investigating. Those results also show that the silencing of a heterochromatic reporter by the Lsm proteins is independent of the location of the transgene in the genome, and of its sequence since the promoter, ORF, 3’UTR and integration site in the genome are different in the four heterochromatic reporters that we have tested, and all of them could be derepressed. Therefore it raised the question whether heterochromatic-Lsm related silencing could affect heterochromatic marks such as H3K9me3 and H3K27me3 important for the transcriptional repression on the heterochromatic array Lsm derepressed, and we found no major changes in the presence of those marks on the heterochromatic array itself after its derepression caused by the downregulation of Lsm-7. This result strengthens the possibility that the heterochromatic-Lsm related silencing can be a novel mechanism at the (co- or) post-transcriptional level of regulation. Next, we asked whether heterochromatic-Lsm related silencing depends on those heterochromatic marks, such as H3K9me3, Hpl-2 (HP1 homologs), H3K27me3 or other marks. We have tested so far the H3K9me3 and HPL-2 and found that heterochromatic-Lsm related silencing does not depend on the presence of those marks on the genome, but very possibly depends on other features of heterochromatin such as epigenetic marks or different binding proteins or different structure of the chromatin regions that Lsm silences. Interestingly, when we combined strains lacking both H3K9me3 and Lsm-7 and Hpl-2 and Lsm-7 respectively, we observed an additional derepression of the heterochromatic arrays to the one observed after deletion of H3K9me3 and Hpl-2 alone. This synergistic effect for heterochromatic transgene derepression and the premature synthetic lethality clearly observed in the case of the hpl-2 knock-out strain suggest the existence of an additional mechanism involving the Lsm proteins which is independent of the H3K9me3-related transcriptional repression mechanism. It is then tempting to hypothesize that this proposed mechanism could serve as a back-up mechanism especially in abnormal conditions, for example, when transcriptional repression is impaired.
Answering the second question, we found that in C. elegans not only the three Lsm proteins (Gut-2/Lsm-2, Lsm-5 and Lsm-6) found in the original genome-wide screen were required for gene silencing within heterochromatic regions, but also Lsm-4 and Lsm-7, as shown by downregulation through RNAi of those proteins. RNAi constructs able to down-regulate Lsm-3 was not available in order to test its effect on heterochromatic array derepression. Nevertheless, together those results indicate that heterochromatic-Lsm related silencing is most likely through one of the two Lsm1-7 or Lsm2-8 complexes characterized in other species, and not functioning through single subunits of those complexes. In order to identify which of those two complexes is the one responsible for the heterochromatic silencing, we tested the effect of Lsm-1 and Lsm-8 downregulation, which are respectively specific for the cytoplasmic and nuclear complexes. We found that Lsm-1 and Xrn-1, which is the exonuclease responsible for cytoplasmic RNA degradation promoted by the Lsm1-7 complex, did not have an effect on the regulation of expression of the heterochromatic array. In contrast RNAi of Xrn-2, the exonuclease expected to be responsible for nuclear RNA degradation promoted by the Lsm2-8 complex, did derepress the heterochromatic array in a similar range as we observed for the downregulation of the proteins Lsm-2, -4 -5 -6 and -7. Lsm-8 downregulation by RNAi was inefficient therefore we could not test its effect by this method so we are currently testing its effect with a deletion mutant strain that we have generated. Altogether, the Lsm2-8 complex is most likely the one responsible for heterochromatic-Lsm related silencing by promoting heterochromatic mRNA degradation in the nucleus, through Xrn-2. Moreover, we have performed a heterochromatic derepression targeted RNAi screen to identify other co-factors, predicted to be co-regulated or functionally or physically interacting with the Lsm proteins and our results with other sets of experiments reinforce our hypothesis, and gave us several directions to pursue a better understanding of the mechanism leading to the derepression.
Finally, answering the question “How does the Lsm complex of interest silence heterochromatic transgenes?, we are currently in a set of experiments which should prove directly the direct binding of the Lsm2-8 complex to the heterochromatic transcripts which are derepressed, and assess the degradation efficiency in the presence and absence of a functional Lsm2-8 complex. We also plan to test if there is a cross-talk effect of Lsm depletion on transcription and splicing of the misregulated heterochromatic regions and finally to test the conservation of the heterochromatic-Lsm related silencing in mammalian cells.
Our research provides important knowledge on a very novel mechanism involved in gene silencing in heterochromatic regions which will be continued and will broaden our understanding of gene regulation in general.
Answering the first question, we found that the derepression caused by the down-regulation of the three Lsm hits (Gut-2/Lsm-2, Lsm-5 and Lsm-6) could be observed and quantified (using microscopy and an adapted method using a worm sorter machine) in old embryos and in every other developmental stage of the worm in every somatic cell but not in the germ cells. This suggests that the Lsm proteins are involved in a somatic silencing mechanism. We also found that the heterochromatic array derepression was found at the protein level as quantified by the GFP intensity but also at the RNA level, as expected, tested by qPCR. Additionally and importantly, we found that down-regulation by RNAi of Lsm proteins derepressed all of the four different heterochromatic reporter transgenes we have tested whereas the two tested ‘non-heterochromatic’ transgenes expression was not affected, which indicates that those Lsm proteins are involved specifically in silencing reporter transgene with heterochromatic features, and possibly this will be also true for endogenous heterochromatic regions, which we are currently investigating. Those results also show that the silencing of a heterochromatic reporter by the Lsm proteins is independent of the location of the transgene in the genome, and of its sequence since the promoter, ORF, 3’UTR and integration site in the genome are different in the four heterochromatic reporters that we have tested, and all of them could be derepressed. Therefore it raised the question whether heterochromatic-Lsm related silencing could affect heterochromatic marks such as H3K9me3 and H3K27me3 important for the transcriptional repression on the heterochromatic array Lsm derepressed, and we found no major changes in the presence of those marks on the heterochromatic array itself after its derepression caused by the downregulation of Lsm-7. This result strengthens the possibility that the heterochromatic-Lsm related silencing can be a novel mechanism at the (co- or) post-transcriptional level of regulation. Next, we asked whether heterochromatic-Lsm related silencing depends on those heterochromatic marks, such as H3K9me3, Hpl-2 (HP1 homologs), H3K27me3 or other marks. We have tested so far the H3K9me3 and HPL-2 and found that heterochromatic-Lsm related silencing does not depend on the presence of those marks on the genome, but very possibly depends on other features of heterochromatin such as epigenetic marks or different binding proteins or different structure of the chromatin regions that Lsm silences. Interestingly, when we combined strains lacking both H3K9me3 and Lsm-7 and Hpl-2 and Lsm-7 respectively, we observed an additional derepression of the heterochromatic arrays to the one observed after deletion of H3K9me3 and Hpl-2 alone. This synergistic effect for heterochromatic transgene derepression and the premature synthetic lethality clearly observed in the case of the hpl-2 knock-out strain suggest the existence of an additional mechanism involving the Lsm proteins which is independent of the H3K9me3-related transcriptional repression mechanism. It is then tempting to hypothesize that this proposed mechanism could serve as a back-up mechanism especially in abnormal conditions, for example, when transcriptional repression is impaired.
Answering the second question, we found that in C. elegans not only the three Lsm proteins (Gut-2/Lsm-2, Lsm-5 and Lsm-6) found in the original genome-wide screen were required for gene silencing within heterochromatic regions, but also Lsm-4 and Lsm-7, as shown by downregulation through RNAi of those proteins. RNAi constructs able to down-regulate Lsm-3 was not available in order to test its effect on heterochromatic array derepression. Nevertheless, together those results indicate that heterochromatic-Lsm related silencing is most likely through one of the two Lsm1-7 or Lsm2-8 complexes characterized in other species, and not functioning through single subunits of those complexes. In order to identify which of those two complexes is the one responsible for the heterochromatic silencing, we tested the effect of Lsm-1 and Lsm-8 downregulation, which are respectively specific for the cytoplasmic and nuclear complexes. We found that Lsm-1 and Xrn-1, which is the exonuclease responsible for cytoplasmic RNA degradation promoted by the Lsm1-7 complex, did not have an effect on the regulation of expression of the heterochromatic array. In contrast RNAi of Xrn-2, the exonuclease expected to be responsible for nuclear RNA degradation promoted by the Lsm2-8 complex, did derepress the heterochromatic array in a similar range as we observed for the downregulation of the proteins Lsm-2, -4 -5 -6 and -7. Lsm-8 downregulation by RNAi was inefficient therefore we could not test its effect by this method so we are currently testing its effect with a deletion mutant strain that we have generated. Altogether, the Lsm2-8 complex is most likely the one responsible for heterochromatic-Lsm related silencing by promoting heterochromatic mRNA degradation in the nucleus, through Xrn-2. Moreover, we have performed a heterochromatic derepression targeted RNAi screen to identify other co-factors, predicted to be co-regulated or functionally or physically interacting with the Lsm proteins and our results with other sets of experiments reinforce our hypothesis, and gave us several directions to pursue a better understanding of the mechanism leading to the derepression.
Finally, answering the question “How does the Lsm complex of interest silence heterochromatic transgenes?, we are currently in a set of experiments which should prove directly the direct binding of the Lsm2-8 complex to the heterochromatic transcripts which are derepressed, and assess the degradation efficiency in the presence and absence of a functional Lsm2-8 complex. We also plan to test if there is a cross-talk effect of Lsm depletion on transcription and splicing of the misregulated heterochromatic regions and finally to test the conservation of the heterochromatic-Lsm related silencing in mammalian cells.
Our research provides important knowledge on a very novel mechanism involved in gene silencing in heterochromatic regions which will be continued and will broaden our understanding of gene regulation in general.
