Masaki Shirayama et al. (2002) East Coast Worm Meeting "A Gene targeting strategy using the yeast endonuclease I-SceI in C.elegans"

Craig C Mello, Masaki Shirayama, Jennifer A McCullough

[East Coast Worm Meeting, 2002]

Although the nematode, C. elegans has proven to be powerful model organism to study a variety of biological problems, it is still not possible to (routinely) induce homologous recombination between introduced DNA and genomic loci. In an effort to solve this problem we are developing a method similar to that recently described for Drosophila (Rong and Golic, Science, 288, 2013, 2000). The plan is to initiate gene targeting by introducing double strand breaks into extrachromosomal transgenes, in vivo, in germline nuclei. To do this we have expressed the yeast endonuclease, I-SCE1, that recognizes an 18 bp target sequence absent in the worm genome but present in flanking regions of homology within a transgene. To express the I-SCEI gene in the germline, we subcloned the mes-6/cks-1 operon and replaced the mes-6 gene with the I-SCEI gene. Since the mes-6 and cks-1 genes are in the same operon, both mRNAs are simultaneously processed from a single polycistronic preRNA and expressed in the germline of larvae and adults. We injected the I-SCEI/cks-1 construct into cks-1 maternal effect temperature sensitive mutants and were able to rescue the ts phenotype of cks-1. I-SCEI expressing lines can now be maintained by keeping cks-1 rescued animals at 25 C for subsequent generations.

Masaki Shirayama et al. (2004) East Coast Worm Meeting "Three conserved protein kinases, DYRK, CDC2 and GSK3 promote OMA-1 degradation to establish proper cell fate and cell division polarity in early C. elegans embryos"

Takao Ishidate, Masaki Shirayama, Kuniaki Nakamura, Craig C Mello

[East Coast Worm Meeting, 2004]

The proper establishment of cell polarity and identity underlies development in all multicellular animals. In C. elegans, the 4-cell stage blastomere, EMS receives inductive signals that orient its division axis and specify the endoderm fate. GSK-3 is positively required for Wnt-mediated EMS cell fate determination and mitotic spindle alignment in EMS. Paradoxically, despite lacking E-derived endoderm, gsk-3 mutants frequently produce ectopic endoderm from the C blastomere. Thus GSK-3 functions to promote endoderm development in E and simultaneously functions to repress endoderm development in C. In our screens for temperature-sensitive mutants, we isolated 5 mutants that (like gsk-3 mutants) produce ectopic endoderm derived from the C blastomere and are defective in EMS cell polarity. These mutations include one allele of cdk-1, one allele of its binding partner, cks-1, one allele of the dyrk-family kinase member mbk-2, and two alleles of the zinc finger gene oma-1. Previous work has linked OMA-1 destruction to C-cell fate specification and EMS polarity1), therefore, we tested whether cdk-1, mbk-2 and gsk-3 mutants exhibit defects in OMA-1 degradation. We found that OMA-1 protein persists ectopically in all of these mutants. We have also found that MBK-2 directly phosphorylates OMA-1 protein and this phosphorylation is abolished in the ne411and zu405 OMA-1 mutant protein due to a mutation in a consensus MBK-2 site. We are currently testing the model that sequential phosphorylation of OMA-1 by MBK-2, CDK-1 and GSK-3 kinases is required for the timely degradation of OMA-1 protein. 1) Lin, R. Dev. Biol. 258 (2003) 226

Masaki Shirayama et al. (2005) International Worm Meeting "Four conserved protein kinases, DYRK, CDC2/Cyclin B3, CKI and GSK3 promote OMA-1 degradation to initiate ECS/ZIF-1 dependent proteolysis of cell fate determinants in C. elegans"

Craig Mello, Takao Ishidate, Masaki Shirayama

[International Worm Meeting, 2005]

Early embryonic development in C. elegans requires the proper temporal and spatial activation of numerous maternally-encoded cell-fate determinants. This regulation relies on the destruction of the zinc-finger protein OMA-1 in one-cell stage. OMA-1 is abundantly expressed in oocytes and is essential for proper oogenesis. OMA-1 is rapidly degraded after fertilization in wild-type embryos, however, in oma-1 gain-of-function (gf) allele, OMA-1 protein fails to be degraded resulting in the mis-localization of numerous maternal determinants1). We isolated several ts mutants with phenotypes similar to that of oma-1(gf). These mutations include one allele of cdk-1, one allele of its small binding partner cks-1, one allele of the dyrk-family kinase mbk-2, and two new gain-of-function (gf) alleles of the oma-1. Previous work had implicated MBK-2 in OMA-1 destruction2), and because of their similar phenotypes, we tested whether cdk-1, cks-1 and gsk-3 mutants might also exhibit defects in OMA-1 degradation. We found that OMA-1 protein persists ectopically in all of these mutants. We also found that OMA-1 is stabilized in embryos after RNAi targeting cyclin B3 (cyb-3) and casein kinase I (kin-19). Interestingly all three of the oma-1(gf) alleles alter the same proline residue within a consensus MBK-2 phosphorylation site in OMA-1, raising the possibility that MBK-2 directly regulates OMA-1. Consistent with this idea, we found that MBK-2 phosphorylates OMA-1 protein in-vitro and that this phosphorylation is abolished in OMA-1(gf) mutant proteins. Furthermore, we found that CKI can also phosphorylate OMA-1, but only after pre-phosphorylation by MBK-2. Similarly, GSK-3 phosphorylates OMA-1 but only after pre-phosphorylation by both MBK-2 and CKI. Since OMA-1 destruction commences during the first mitosis of embryogenesis, we speculate that destruction may be triggered by a CDK-1/cyclin B3 complex. However, we have not detected any direct phosphorylation of OMA-1 by CDK-1 kinase. Instead we found a highly conserved CDK-1 site located in the AXIN-binding domain of GSK-3. Altering this CDK-1 site does not abolish GSK-3 auto-phosphorylation but appears to prevent phosphorylation of OMA-1, suggesting that CDK-1 may directly regulate the substrate specificity of GSK-3. 1) Lin, R. Dev. Biol. 258, 226 (2003); 2) Pellettieri et al., Dev. Cell, 5, 451 (2003)

Ghanta, Krishna et al. (2015) International Worm Meeting "VET-2 interacts with CSR-1 in C.elegans early embryos."

Shirayama, Masaki, Ghanta, Krishna, Mello, Craig

[International Worm Meeting, 2015]

Argonaute proteins are central to small RNA pathways that control mRNA turnover, translational regulation, viral defense, and transposable element suppression. In Caenorhabditis elegans, 22 nucleotide RNAs (22G RNAs) produced by the RNA Dependent RNA Polymerases (RDRPs), EGO-1 and RRF-1, bind to the argonaute CSR-1 and members of the Worm specific Argonautes (WAGOs) respectively. CSR-1 associated 22G RNAs target thousands of germline mRNAs (Claycomb et al., 2009). However, unlike other argonautes, CSR-1 promotes the expression of its mRNA targets instead of silencing them (Claycomb et al., 2009 and Cecere et al., 2014). Genetic studies suggest that CSR-1 counteracts piRNA and WAGO dependent silencing (Shirayama et al., 2012 and Seth et al., 2013). CSR-1 is both maternally and paternally supplied to zygotes in C. elegans (Conine et al., 2013). In embryos, GFP::CSR-1 is diffusely cytoplasmic until the 4 cell stage. From 8-cell stage, it starts to aggregate into cytoplasmic foci in somatic cells. These foci are brightest at the 25-30-cell stage and completely disappear by the 60-cell stage. After the 60 cell stage GFP::CSR-1 is primarily located in the P-granules of the P-lineage cells.The vet-2 gene was originally identified as a very early zygotically transcribed gene, whose expression, beginning at the 4-cell stage, was restricted to somatic lineages by PIE-1 repression (Sedoux et al., 1996). Here we show that VET-2 protein interacts with CSR-1 both in vivo and in the yeast 2 hybrid assay. Interestingly, CRISPR-induced loss-of-function mutations in vet-2 result in the failure of CSR-1 to form aggregates in somatic blastomeres. Moreover, a VET-2::mCherry fusion protein (knocked in by CRISPR) is also expressed in cytoplasmic foci, where it accumulates with CSR-1 beginning at the 60-cell stage just prior to the disappearance of CSR-1 protein. It has been proposed that VET-2 might act as a linker protein in the autophagy pathway (Zhang et al., 2009). In future, we plan to characterize the nature of the CSR-1/VET-2 foci and explore what other factors localize there. It will be interesting to determine if VET-2 is required for the autophagy of CSR-1 protein complexes.

Makeyeva, Yekaterina et al. (2021) International Worm Meeting "mRNA Splicing Promotes Polyadenylation and Counteracts Novel Default Argonaute Silencing in the Germline of Caenorhabditis elegans"

Shirayama, Masaki, Makeyeva, Yekaterina, Mello, Craig

[International Worm Meeting, 2021]

Mechanisms that safeguard the germline genome are essential for species preservation. For example, in metazoans the PIWI Argonaute pathway promotes genome integrity by silencing transposons and newly acquired foreign genes. PIWI Argonautes engage genomically-encoded piRNAs to detect their targets and to trigger gene regulation at both the mRNA and transcriptional level. However, the question of how foreign or aberrant gene expression is detected in the germline is still poorly understood. In yeast, genes with suboptimal mRNA splicing signals can trigger Argonaute silencing, and a similar connection has been made in the silencing of C. elegans early embryo transcripts with weak and abnormal splicing by the ERGO-1 ERI small RNA pathway. Here we show that the removal of introns from a gene and replacement with a cDNA-derived open reading frame results in complete silencing of the gene in the C. elegans germline. For example, a transgenic copy of gfp-tagged cdk-1 gene (cdk-1::gfp) yields robust germline and somatic expression. However, an intronless cdk-1::gfp transgene fails to express in the germline and can also transitively silence other intron-containing gfp transgenes. Transitive silencing depends on the WAGO Argonaute pathway and is correlated with robust production of RNA-dependent RNA Polymerase (RdRP)-derived antisense small RNAs from the unspliced RNA. Silencing and small RNA production were also observed when introns were removed from the endogenous oma-1 gene. Surprisingly, although transitive silencing requires the WAGO Argonaute pathway, the intronless genes (themselves) remain silenced in WAGO-pathway mutants, suggesting that an additional, cis-mode, of silencing occurs on intronless genes. The known primary Argonautes including prg-1, ergo-1, alg-3/4, and rde-1 are not required for the initiation of "intronless" silencing. Northern blot analysis revealed that intronless genes yield only a shorter mRNA isoform that appears to lack polyadenylation. These findings suggest mRNA splicing serves as a licensing step in germline gene expression promoting both polyadenylation and protein translation in cis, while preventing mRNAs from becoming default templates for RdRP-dependent transitive Argonaute silencing.

Lee, Heng-Chi et al. (2011) International Worm Meeting "Induction, maintenance and transvection of heritable germline gene silencing in C. elegans."

Shirayama, Masaki, Mello, Craig, Lee, Heng-Chi

[International Worm Meeting, 2011]

Organisms exhibit a fascinating array of gene-silencing pathways, which have evolved, in part, to confront a constant onslaught of parasitic RNA and DNA invaders. One such mechanism, which has also been a vexing problem for C. elegans researchers attempting to transform the germline, is a potent germline silencing pathway that silences multi-copy transgenes. Although transient co-suppression of a homologous chromosomal locus is sometimes observed, the primary mode of silencing within a multicopy array is thought to act in cis. Once established, silencing may be inherited for multiple generations without evidence of reversion. Methods such as biolistic transformation and Mos1-mediated Single Copy Insertion (MosSCI) can bypass germline silencing of transgenes, presumably by avoiding an unknown mechanism that senses DNA copy-number. We were therefore surprised to observe a high frequency of apparently intact but never-the-less silent, single-copy insertions obtained by MosSCI. In one MosSCI experiment using an extrachromosomal, high-copy array as the donor for the inserted single-copy DNA, all of the resulting single-copy insertion lines were completely silent in the germ line. The silent state of these single-copy transgenes might reflect persistence of copy-number-dependent silencing that was initiated during assembly of the high-copy array. To test this idea, we generated a donor array with fewer copies of the transgene by diluting only the targeting vector in our injection cocktail. Strikingly, nearly all of the resulting single-copy insertions exhibited robust germline expression. Next we explored the consequences of placing a silent locus in trans to an active locus. Remarkably, the silent allele exhibited complete dominance over the active allele; all of the F1 cross progeny failed to exhibit germ-line expression of the transgene. Furthermore, the silent state was transmitted at 100% penetrance in subsequent generations. Because silencing does not appear to spread to the endogenous chromosomal copy of the corresponding gene (co-suppression), we suspect that pairing of the parental chromosomes is necessary for transmission of the silent state, a phenomenon known as transvection. Our findings imply that initiation of silencing is sensitive to copy number, while persistence of silencing is mediated by a cis-acting mechanism that can be transferred to an active allele by transvection. We are currently investigating the genetic requirements for the establishment, maintenance and transvection of this remarkable gene-silencing phenomenon.

Ishidate, Takao et al. (2021) International Worm Meeting "Proteolysis dependent gene silencing in C.elegans germline"

Makeyeva, Yekaterina, Ishidate, Takao, Shirayama, Masaki, Mello, Craig

[International Worm Meeting, 2021]

Animal germ cells employ small RNAs and Argonaute proteins to surveil transcripts and silence transposons and foreign genetic elements. Here we report that, in C.elegans germline, expression of a transgene encoding codon optimized Cas9 gene fused to GFP is silenced via small RNA dependent and small RNA independent mechanisms. Interestingly, small RNA-dependent silencing was initiated and small RNAs targeting Cas9::GFP sequence were generated even in the absence of prg-1 activity. Furthermore, disarming the small RNA pathway led to Cas9::GFP expression only in the most distal region of the gonad, suggesting that a second small RNA-independent silencing mechanism maintains the silencing throughout the pachytene region and in oocytes. A forward genetic screen revealed that inactivation of the ubiquitin-proteasome pathway causes full de-silencing of the CAS9::GFP throughout the germline, suggesting that the second silencing mechanism is mediated by protein degradation. Remarkably, insertion of virus derived T2A sequence that causes ribosome skipping, between Cas9 and GFP, lead to the expression of the downstream GFP moiety throughout the gonad. This observation suggests that the upstream Cas9 ORF is triggering protein degradation and that the mRNA is not degraded and can still encode downstream GFP. Thus ribosome subunits are likely not dissociated as in the RQC response to ribosome stalling. We propose that some modification within the upstream Cas9 mRNA signals for co-translational ubiquitylation and degradation of the peptide. Interestingly, a Cas9 transgene that contains multiple extended AT rich introns has been shown to escape this mode of silencing, pointing toward mRNA splicing as the triggering event and linking the phenomenon to the recently described intronless gene silencing. We are currently investigating the nature of the triggering signal and whether our findings extend to the regulation of endogenous targets.

Ishidate, Takao et al. (2009) International Worm Meeting "Regulation of WRM-1/b-catenin by the cell-cycle regulator CDK-1 during asymmetric cell division in C.elegans."

Ishidate, Takao, Sharma, Rita, Kim, Soyoung, Mello, Craig, Shirayama, Masaki

[International Worm Meeting, 2009]

In the 4-cell stage C.elegans embryo, the b-catenin related protein WRM-1 responds to a polarizing signal that induces the EMS cell to divide asymmetrically and accumulates in the nucleus of posterior daughter E. WRM-1 localizes uniformly to the cell cortex in EMS during the interphase, however, at the onset of M-phase, it dissociates from the cortex proximal to the signaling cell P2 in a Wnt-signaling dependent manner. We noticed a conserved region in WRM-1 that contains potential phosphorylation sites for CDK-1 and GSK-3. Through in vitro studies we found that CDK-1 can phosphorylate WRM-1 directly at this site priming WRM-1 for subsequent phosphorylation by GSK-3. WRM-1 proteins with mutations at either the CDK-1 and GSK-3 sites failed to be released from the cortex and its downstream asymmetric nuclear accumulation was also diminished. In a rare temperature-sensitive mutant of cdk-1(ne2257) that exhibits a partial defect in Wnt signaling, WRM-1 dissociation from the cortex was found to be defective. Consistent with these findings, mutant CDK-1 isolated from the embryonic extract showed reduced kinase activity toward WRM-1 recombinant protein. To investigate the physiological significance of these observations in vivo, we have used mosSCI to introduce single-copy insertions of wild-type and mutant WRM-1 constructs. While our interpretations of these preliminary studies are still complicated by a lack of full rescue from the wild-type construct, we were encouraged to find that for each different construct, Western-blot analysis revealed that independent transgene lines exhibited very reproducible levels of expression. We were surprised, however, to find that the mutation of the CDK-1 site resulted in a significantly stronger rescue of the wrm-1 mutant strain, and also resulted in much higher levels of protein expression in each of the 2 lines examined. Similar evidence for increased activity from this mutant WRM-1(DCDK) construct was observed from a bombardment derived transgenic line. The GSK-3-site deletion construct did not exhibit this enhanced level of rescue and expression. These findings support the notion that WRM-1 is a direct regulatory target of CDK-1, and call into question the importance of its cortical localization during endoderm induction. We are currently trying to improve the transgene rescue utilizing different WRM-1 constructs. Through these studies, we would like to shed more light on the potential involvement of cell cycle regulator in developmental decisions.

Ishidate, Takao et al. (2009) International Worm Meeting "cdk-1(ne2257) suppressor screen and the developmental function of CDK-1/CYB-3 complex in polarity generation during early embryogenesis."

Shirayama, Masaki, Kim, Soyoung, Mello, Craig, Ishidate, Takao, Sharma, Rita

[International Worm Meeting, 2009]

We have isolated a rare temperature-sensitive allele of cdk-1(ne2257) that does not affect cell cycle progression but instead causes very specific developmental defects that appear to stem from a failure to degrade the CCCH-type zinc finger proteins OMA-1 and OMA-2 prior to the onset of the first mitosis. Persistence of the OMA-1/2 proteins leads to defects in polarity generation and in the expression of other developmental regulators. OMA-1/2 degradation depends on direct phosphorylations by a set of conserved kinases, including MBK-2, KIN-19 and GSK-3. However, although cdk-1 and cyclin B3 lesions both lead to a dramatic stabilization of OMA-1/2 proteins, their role in promoting OMA-1/2 degradation is not clear since unlike the other kinases, CDK-1/CYB-3 does not phosphorylate the OMA-1/2 proteins directly. In order to study the developmental function of the CDK-1/CYB-3 complex in generating embryonic polarity, we have undertaken a suppressor screen, looking for mutants that rescue cdk-1(ne2257) embryonic lethality at restrictive temperature. From this powerful screen we have isolated multiple rare suppressing mutations including reverting mutations in cdk-1 that restore the altered amino acid to wild-type, compensatory mutations in cyb-3 that accommodate the altered cdk-1-mutant interface, as well apparently three novel suppressors that map to LGI (2 alleles), LGII (6 alleles) and LGIV (1 allele). As expected, the cdk-1 and cyb-3 suppressor mutations rescue the OMA-1 degradation defect seen in cdk-1(ne2257). However, in contrast the novel suppressors on LGI, II and IV all continue to exhibit persistence of OMA-1, suggesting that they function downstream of the OMA proteins. We are currently attempting to clone these suppressors using a deep sequencing approach. Characterizing these suppressors will likely shed light on the role of the cell cycle machinery in coupling the onset of mitosis to polarity generation during C.elegans early embryogenesis.

Elewa, Ahmed et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "gld-3 functions in blastomere specification"

Shirayama, Masaki, Elewa, Ahmed, Mello, Craig, Miliaras, Nicholas, Shin, Tae-Ho

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

The CCCH zinc finger proteins PIE-1, POS-1, MEX-1/5/6, OMA-1/2 and the KH domain proteins MEX-3 and GLD-1 have been implicated in blastomere identity determination. A yeast-two hybrid screen in our lab revealed GLD-3 as an interactor with PIE-1. GLD-3 is a KH domain protein homologous to Drosophila Bicaudal-C and is required for identity determination in the germ-line, as it participates in mitosis/meiosis and sperm/oocyte decisions. In gld-3(RNAi) embryos PIE-1 protein is mislocalized and persists transiently in each germ-line sister cell, and ultimately the P4 cell appear to adopt a D-like fate resulting in sterility. Furthermore, we have found that gld-3 interacts genetically with pos-1 , gld-1 and mex-1 . pos-1 and gld-1 embryos exhibit an Apx-like ABp-to-ABa transformation and lack gut, however gut development is restored in pos-1; gld-3 and gld-1; gld-3 double mutants. These double mutants resemble pos-1(ne51) embryos, which have a point mutation in their second zinc finger (CCCH -> YCCH) and arrest as Apx-like embryos with gut. mex-1 embryos arrest with excess muscle in the anterior but exhibit proper gut specification. However mex-1; gld-3 double mutants frequently arrest as gutless Mex embryos. pos-1 and mex-1 exhibit similar genetic interactions with gld-2 and gls-1 (a poly A polymerase and a poly A binding factor) which were previously shown to interact with gld-3 . GLD-3 has been proposed to act as an RNA-binding cofactor that recruits GLD-2 to stimulate cytoplasmic poly adenylation of targets important for germline specification. The complex genetic interactions with pos-1 and mex-1 and the apparent physical interaction with PIE-1 raise the possibility that GLD-3 engages members of this CCCH protein family to regulate diverse targets involved in embryonic cell fate specification.