Jose, Antony M. et al. (2011) International Worm Meeting "Two classes of silencing RNAs move between C. elegans tissues."

Hunter, Craig P., Jose, Antony M., Garcia, Giancarlo

[International Worm Meeting, 2011]

Organism-wide gene silencing is readily observed in plants and in some animals even when only a few cells encounter double-stranded RNA (dsRNA). Such systemic RNA interference (RNAi) triggered by dsRNA is due to the transport of mobile silencing signals throughout the organism. In animals, these mobile silencing signals are likely RNA since their import into C. elegans cells requires the RNA transporter SID-1. Here we present genetic evidence that both the initial dsRNA and at least one dsRNA intermediate produced during RNAi can act as or generate mobile RNA in C. elegans. This dsRNA intermediate is likely a modified form of double-stranded short-interfering RNA (ds-siRNA) since its production requires the dsRNA-binding protein RDE-4, which recruits the endonuclease Dicer (DCR-1) to cleave long dsRNA, and the putative nucleotidyltransferase MUT-2/RDE-3. However, single-stranded siRNA and downstream secondary siRNA produced upon amplification by RNA-dependent RNA Polymerases (RdRPs) do not generate detectable mobile RNA. Restricting inter-tissue transport to long dsRNA and directly processed siRNA intermediates rather than amplified siRNA may serve to modulate the extent of systemic silencing in proportion to available dsRNA. Furthermore, because a SID-1 homolog can import ds-siRNA into mammalian cells, and because mammals have counterparts of Dicer, RDE-4, and MUT-2, our results suggest that mammals may be capable of generating mobile silencing RNA.

Jose, Antony M. et al. (2009) International Worm Meeting "C. elegans tissues do not require RNA interference to generate and export silencing signals derived from expressed RNAs."

Hunter, Craig P., Garcia, Carlo, Jose, Antony M., Smith, Jessica J.

[International Worm Meeting, 2009]

Double-stranded RNA (dsRNA) triggers RNA interference (RNAi) to silence genes of matching sequence. RNAi-like mechanisms underlie key cellular processes such as heterochromatin formation, regulation of gene expression, and silencing of repetitive DNA. In some animals, when exogenous dsRNA is introduced into a cell, the resultant silencing is transported to other cells. However, numerous dsRNAs and hairpin RNAs (hpRNAs) are transcribed from animal genomes but whether expressed RNAs and endogenous RNAi-like mechanisms generate mobile silencing signals is unknown. Studies in C. elegans have shown that the conserved dsRNA channel Systemic RNAi Defective-1 (SID-1) is required for the import of transported silencing signals. We found that the efficient silencing of multicopy transgenes requires SID-1, suggesting that transgene silencing in one cell produces mobile silencing signals that function to initiate and/or maintain transgene silencing in another cell. Further, the tissue-specific expression of RNAi triggers resulted in the transport of silencing from all tested tissues to other tissues, consistent with expressed RNAi triggers generating mobile silencing signals. Although import through SID-1 suggests that mobile silencing signals are likely forms of dsRNA, their precise identity and biogenesis are unknown. We reasoned that proteins that act on dsRNA for RNAi within a cell may modify dsRNA or other dsRNA-derived RNAs to generate mobile silencing signals. Therefore, we examined the role of RNAi Defective-1 (RDE-1), an Argonaute protein required for RNAi in C. elegans, in the export of mobile silencing signals. In rde-1(-) animals, when gfp-dsRNA was expressed in the pharynx, no silencing of GFP expression was detected in either the pharynx or the body wall muscles. However, when rde-1(+) was expressed specifically in the body wall muscles of these animals, gfp expression was silenced only in that tissue. Consistent with the export of mobile silencing signals from an rde-1(-) pharynx, GFP expression in the pharynx remained unaffected. Thus, RNAi-mediated degradation of target mRNA is not required for the generation and export of mobile silencing signals derived from transgenically expressed RNAi triggers. We are systematically testing additional components and regulators of the RNAi pathway for possible roles in the generation of mobile silencing signals.

Antony Jose et al. (2007) International Worm Meeting "Sid-3 mutants are defective in the export of a silencing signal during RNA interference."

Christina Molodowitch, Craig Hunter, Antony Jose, Marie Sutherlin

[International Worm Meeting, 2007]

Double-stranded RNA (dsRNA) triggers the silencing of genes with a matching sequence through a process called RNA interference (RNAi). Surprisingly, this sequence-specific silencing information is transported between cells, sometimes causing most tissues to be silenced even when only a single cell encounters dsRNA (systemic RNAi). Screens for systemic RNAi defective (sid) mutants identified the widely-expressed dsRNA channel SID-1 that is required for the import of dsRNA or the transported silencing signal into most tissues. Further, a small transmembrane protein SID-2 required for the import of dsRNA into the gut, and an additional complementation group, sid-3, comprised of 15 alleles were identified. We found that cells expressing a hairpin dsRNA can export a silencing signal independent of SID-1. In wild-type animals, gfp dsRNA made in pharyngeal muscle cells can silence gfp in the pharynx and, due to the transport of a silencing signal, also in the body-wall muscles. In contrast, sid-1 mutants, silence gfp only in the pharynx (the source of the dsRNA). However, when SID-1 was expressed only in body-wall muscles, transport from pharynx to muscle was restored and gfp was silenced in body-wall muscles. This suggests that pharyngeal cells lacking sid-1 are able to export a dsRNA-derived gfp-specific silencing signal that is then imported into body-wall muscles through SID-1. Thus, SID-1 is not required for export of the silencing signal during RNAi. In contrast, sid-3 appears to be required for the export of a dsRNA-derived silencing signal. When we injected single gut cells of sid-3(qt14) mutants with gfp dsRNA, gfp was silenced through out the gut but not in other tissues, suggesting that sid-3 mutants fail to export a silencing signal from the gut to other tissues. sid-3 mutant cells showed enhanced silencing when dsRNA was either expressed in them or was supplied in their environment. This suggests that inhibition of export enhances cell autonomous RNAi, perhaps due to increased availability of a silencing signal. Interestingly, when placed over a deficiency (def) that covers sid-3, the resultant hemizygous sid-3(qt14)/def mutants showed a more severe defect in the transport of silencing information and had lower brood sizes than sid-3(qt14)/sid-3(qt14) mutants. These results suggest that the isolated sid-3 alleles are partial loss of function alleles and that sid-3 null mutants may be lethal.

Antony Jose et al. (2004) East Coast Worm Meeting "Biochemical and structure/function analysis of DGK-1, a neuronal diacylglycerol kinase and putative downstream effector of Gaphao signaling."

Michael Koelle, Antony Jose

[East Coast Worm Meeting, 2004]

G a o , the most abundant G protein in the brain, mediates signaling by many neurotransmitter receptors, yet its signaling mechanism is not understood. GOA-1, the C. elegans ortholog of G a o , is thought to inhibit neurotransmission by decreasing the levels of the second messenger diacylglycerol (DAG)(1). Diacylglycerol kinases are enzymes that inactivate DAG by phosphorylating it. The diacylglycerol kinase DGK-1 is a potential downstream component of GOA-1 signaling since dgk-1 mutants are phenotypically similar to goa-1(lf) mutants and dgk-1(lf) mutations are epistatic to goa-1(gf) transgenes. We have used biochemical approaches to test if GOA-1 signaling activates DGK-1 to reduce DAG levels and thereby inhibit neurotransmission. We also carried out a structure/function analysis of DGK-1. To test for a possible direct interaction between GOA-1 and DGK-1, the proteins were expressed recombinantly and purified. Purified DGK-1 phosphorylated diacylglycerol in vitro . However, the instability of purified DGK-1 precluded further biochemical analysis, and we thus have looked for effects of GOA-1 on native DGK-1 protein in C. elegans extracts. Fractionation of C. elegans extracts and western blotting indicated that DGK-1, which acts on the membrane-bound substrate diacylglycerol, is predominantly cytosolic. One hypothesis is that GOA-1 recruits the soluble DGK-1 to the membrane, and thus activates it. However, we were not able to detect any translocation of DGK-1 in mutant backgrounds that activate GOA-1 signaling, or after activating GOA-1 in wild-type extracts by treatment with GTP analogs. We were also unable to pull down DGK-1 from C. elegans extracts using purified recombinant GST::GOA-1 fusion protein. Our results have thus failed to demonstrate any effect of GOA-1 on DGK-1 protein. It is possible that DGK-1 acts constitutively to shorten the half-life of DAG, and that GOA-1 acts by a mechanism independent and parallel to DGK-1. Diacylglycerol kinases have several conserved sequence motifs of unknown function. To understand the functional importance of these conserved sequence motifs in DGK-1, we identified the molecular lesions in 20 alleles of dgk-1 that have been generated by genetic screens [(2) and our lab]. We quantitated the behavioral defects caused by these mutations and also examined effect of these mutations on DGK-1 protein stability using western blotting. We identified missense mutations in conserved residues in the kinase domain and the cysteine-rich domains that dramatically reduce DGK-1 function but produce stable protein, suggesting that these domains are essential for the physiological function of DGK-1. We also identified an early stop codon mutation that disrupts all previously known dgk-1 isoforms yet retains significant function, and our preliminary analysis suggests that novel splice variants that skip the stop codon supply DGK-1 function in this mutant. Finally, we found a missense mutation that changes a serine to a valine in the kinase domain that resulted in a protein with altered mobility on SDS-PAGE. We are testing the possibility that phosphorylation of this serine residue affects the function of DGK-1 and alters its mobility on SDS-PAGE. (1). Nurrish S. et al ., Neuron Vol. 24, 231-242. (2). Hajdu-Cronin Y. M. et al. , Genes and Development Vol. 13, No. 14, 1780-1793.

Antony Jose et al. (2001) International C. elegans Meeting "Does GOA-1, the C. elegans ortholog of the major human brain G alpha protein, signal through diacylglycerol kinase-1?"

Michael Koelle, Antony Jose

[International C. elegans Meeting, 2001]

The most abundant G protein of the human brain, G a o , signals through an unknown mechanism. The nematode C. elegans has a highly similar (~80% identical) ortholog of human G a o called GOA-1. GOA-1 signaling modulates a variety of behaviors in C. elegans , including locomotion and egg-laying. dgk-1 is the predominant gene identified in genetic screens for mutants with a goa-1 like phenotype. Further, genetic screens for mutations that suppress the defects of constitutively active GOA-1 result in many alleles of dgk-1 (1). Genetic analysis suggests that dgk-1 , which encodes a diacylglycerol kinase, acts downstream of or in parallel to GOA-1(1). DGK-1 is 38% identical to DGK q , an enzyme expressed in the human brain (2). We decided to test the possibility that GOA-1 signals through DGK-1 by directly activating it. The GOA-1 and DGK-1 proteins were expressed recombinantly and purified as untagged (GOA-1) or affinity tagged fusion proteins (DGK-1). An in vitro assay for the activity of DGK-1 was developed. In this assay, the substrate, diacylglycerol (DAG) along with the lipid cofactor phosphatidylserine is presented as vesicles and as a monolayer at the air/aqueous interface. Diacylglycerol kinase that is present in the aqueous phase then uses g 32 P-ATP to phosphorylate DAG, producing radioactive phosphatidic acid (PA). This can be extracted into an organic phase and quantitated. The recombinant DGK-1 was active in this assay. The in vitro assay is linear over a significant range with respect to time, amount of enzyme and substrate (DAG). These characteristics are essential to detect possible effects of GOA-1 on the Km or Vmax of DGK-1. However, as the protein preparation of DGK-1 is prone to aggregation, we are optimizing purification strategies in order to obtain soluble and stable protein. We then plan to use the purified proteins in the in vitro assay to test if GOA-1 activates DGK-1. In C. elegans there are two G alpha proteins , G a o (GOA-1) and G a q (EGL-30), that have opposing effects on egg-laying and locomotion. GOA-1 signals to reduce the rates of egg-laying and locomotion, while EGL-30 signals to increase the rates of egg-laying and locomotion. EGL-30 activates a PLC b (EGL-8) which results in the production of the second messenger DAG, which may activate neurotransmission through UNC-13 (2). If GOA-1 directly activates DGK-1, then this provides a molecular mechanism to explain the antagonism between the two G alpha signaling pathways in C. elegans. The opposing effects of EGL-30 and GOA-1 on neurotransmission can be explained by EGL-30 activating EGL-8 to produce more DAG and GOA-1 activating DGK-1 to convert the DAG to inactive PA. 1. Hajdu-Cronin et al. GenesDev,1999. 13 : 1780-93 2. Nurrish et al. Neuron, 1999. 24 : 231-42

Antony M Jose et al. (2003) International Worm Meeting "Biochemical and structure function analysis of the diacylglycerol kinase, DGK-1, a putative effector of Go signaling in neurons"

Antony M Jose, Michael R Koelle

[International Worm Meeting, 2003]

The most abundant G protein of the human brain, Go, signals through an unknown mechanism. GOA-1, the C. elegans ortholog of the human Go, is ~80% identical to its mammalian counterpart and is expressed throughout the C. elegans nervous system. GOA-1 signaling modulates a variety of behaviors in C. elegans, including locomotion and egg laying. Mutations in the downstream components of the GOA-1 signaling pathway would likely result in defects similar to those seen in goa-1 mutants. Genetic screens designed to identify such mutations have been carried out in our laboratory. dgk-1, encoding a diacylglycerol kinase, is the predominant gene identified in such screens. Epistasis analysis suggests that DGK-1 acts downstream of or parallel to GOA-1 (1). DGK-1 is 38% identical to DGK, an enzyme expressed in brain (1). DGK-1, like GOA-1, is expressed throughout the C. elegans nervous system (1). Guided by these observations, we have started a biochemical analysis of DGK-1 to test the possibility that it is a direct downstream effector of GOA-1. To this end, the GOA-1 and DGK-1 proteins were expressed recombinantly and purified. An in vitro assay for the activity of DGK-1 was developed. Purified DGK-1 can phosphorylate diacylglycerol, and we can also detect DGK-1 activity in crude lysates of C. elegans. We have generated antibodies against DGK-1. Fractionation of C. elegans extracts and western blotting indicate that DGK-1, which acts on the membrane-bound substrate diacylglycerol, is predominantly cytosolic.GOA-1 may recruit the soluble DGK-1 to the membrane, and thus activate it. We are looking for effects of GOA-1 on the enzymatic activity and subcellular localization of DGK-1. DGKs have several conserved sequence motifs of unknown function. To understand the structure function relationship of DGK-1, we identified the molecular lesions in the numerous alleles of dgk-1 that have been generated by genetic screens [(2) and our lab]. We are measuring the behavioral defects caused by these mutations and are also measuring the effect of these mutations on DGK-1 protein stability and enzymatic activity. All DGKs contain conserved cysteine rich domains (CRDs) of unknown function. Our analyses suggest that the CRDs of DGK-1 could be sufficient to provide significant DGK-1 function in vivo.(1). Nurrish S. et al., Neuron 24, 231-42. (2). Hajdu-Cronin Y. M. et al., Genes and Dev. 13:14, 1780-9.

Antony M Jose et al. (2002) East Coast Worm Meeting "Identifying the downstream effector(s) of GOA-1, the C. elegans ortholog of the major human brain G alpha protein."

Antony M Jose, Michael R Koelle

[East Coast Worm Meeting, 2002]

The most abundant G protein of the human brain, G o, signals through an unknown mechanism. GOA-1, the C. elegans ortholog of the human G o, is highly similar (~80% identical) to its mammalian counterpart, and is expressed throughout the C. elegans nervous system. GOA-1 signaling modulates a variety of behaviors in C. elegans, including locomotion and egg laying. It is expected that mutations in the downstream components of the goa-1 signaling pathway will result in defects in these behaviors similar to those seen in goa-1 mutants. Genetic screens designed to identify such mutations have been carried out in our laboratory. These have identified alleles of some known genes (e.g. vs24, an allele of dgk-1, encoding a diacylglycerol kinase) as well as alleles of novel genes (e.g. vs39) (Amy Bany, pers. comm.). dgk-1 is the predominant gene identified in such screens. Further, genetic screens for mutations that suppress the defects of constitutively active GOA-1 result in many alleles of dgk-1 (1). Epistasis analysis suggests that DGK-1 acts downstream of or parallel to GOA-1 (2). DGK-1 is 38% identical to DGK , an enzyme expressed in the human brain (2). DGK-1, like GOA-1, is expressed throughout the C. elegans nervous system (2). Guided by these observations, we decided to test the possibility that DGK-1 is the direct downstream effector of GOA-1. To this end, the GOA-1 and DGK-1 proteins were expressed recombinantly and purified. An in vitro assay for the activity of DGK-1 was developed. Purified DGK-1 shows diacylglycerol kinase activity, but biochemical analysis of this protein is complicated by its tendency to aggregate. We are refining purification strategies to obtain aggregation-free protein. As a complementary approach, we plan to use immunoprecipitation to test for direct interaction between GOA-1 and DGK-1 in worm extracts. We have generated antibodies against DGK-1. Preliminary western blotting experiments indicate that a significant amount of DGK-1 is soluble. While this is in agreement with the fact that the mammalian homologue, DGK , was purified from rat brains as a soluble protein (3), it is surprising since the enzyme acts on a substrate, diacylglycerol, that is present in the cell membrane. It is possible that GOA-1 recruits the soluble DGK-1 to the membrane. We plan to use immunofluorescence to look for changes in the localization of DGK-1 in goa-1 mutants. We would also like to detect any changes in DGK-1 activity in the goa-1 mutant. Hence, we are currently fine-tuning an assay that detects DGK activity in worm extracts. The defects observed in goa-1 mutants appear to be more severe than those seen in the dgk-1 mutants (2), raising the possibility that GOA-1 has more than one effector in C. elegans. Hence, we are also mapping and molecularly cloning the gene corresponding to the lesion vs39, identified in our lab in a screen for mutants that phenocopy goa-1 mutants. References:

Blumenfeld, Andrew et al. (2015) International Worm Meeting "Abundant small RNAs in C. elegans have predictable features and include a class of 1-nt-staggered clusters of RNAs."

Blumenfeld, Andrew, Jose, Antony

[International Worm Meeting, 2015]

Small interfering RNAs regulate gene expression and map to many genes in C. elegans. The most abundant of these small RNAs are generated by RNA-dependent RNA polymerases (RdRPs) and have a 5'-triphosphate. Many labs have used a 5'-monophosphate-independent ligation of 18-nt to 28-nt RNAs followed by RNA-Seq to identify these RNAs, which are typically called "22G RNAs" because most of them are 22-nt long and begin with a 5' G. However, it remains unclear if such RNAs form a single class and if the sets of RNAs identified by different labs differ in their composition. Here, we analyze RNA-Seq data from multiple labs to discover reliable features of abundant small RNAs in C. elegans. We find that these datasets have a characteristic and reproducible distribution of 15-nt to 26-nt RNA. These reproducibly detected RNAs are derived from at least two distinct classes: 20-nt to 22-nt RNAs (class A) and 15-nt and longer RNAs that are found as 1-nt-staggered clusters (class B). The 5' end of class A RNAs are associated with a YGW motif where G is the 5'-nt of the RNA, Y = C or U, and W = A or U. Furthermore, RNAs with a 5' A or U have a strong preference for an upstream C and RNAs with a 5' C have a strong preference for an upstream U. Consistent with earlier proposals, these biases associated with class A RNAs likely reflect the requirements for initiation of synthesis by RdRPs. Class B RNAs are independent of all known RdRPs and their 5' ends have no detectable nucleotide bias. Preliminary analyses suggest that class A and class B RNAs originate from distinct sets of genes. The reproducible detection of RNAs that are as short as 15-nt suggests that current methods that typically select for RNAs longer than 18-nt underestimate stable and potentially functional shorter RNAs. Given the size of the C. elegans genome, sequences as small as 13-nt can potentially be used for sequence-specific regulation and even shorter RNAs may be used for gene regulation if a meaningful fraction of the genome (e.g. the exome) is selected for targeting through other mechanisms.

Shugarts, Nathan et al. (2017) International Worm Meeting "Towards establishing when double-stranded RNA can move between cells in C. elegans."

Shugarts, Nathan, Chan, Winnie, Jose, Antony

[International Worm Meeting, 2017]

Extracellular RNAs are found in the circulation of plants and animals. While RNAs that move between cells can regulate physiology and development in plants, identities and roles of such RNAs in animals are unknown. Evidence for the movement of RNA between cells has been obtained in C. elegans through transgenic expression of double-stranded RNAs (dsRNAs) within tissues. Specifically, when a transgene is used to express dsRNA in one tissue, silencing of a matching gene can be detected in other tissues. This silencing requires the dsRNA-selective importer SID-1, suggesting movement of dsRNA between cells. However, multiple models can be envisioned to explain the observed silencing. For example, dsRNA within a cell could be packaged during early embryogenesis into intracellular vesicles that are then inherited through cell divisions, such that dsRNA can later enter into the cytosol and cause silencing. In this model, dsRNA is produced in one cell and causes silencing in another without becoming extracellular. Alternatively, an embryonic cell could deliver dsRNA into an adjacent embryonic cell. In this model, dsRNA enters the extracellular space before the formation of the pseudocoelom. Finally, cells in mature tissues could release dsRNA into circulation and distant tissues could then import dsRNA. In this model, dsRNA transits through the pseudocoelom. Because each model requires a different developmental setting, determining when the SID-1-dependent movement of dsRNA is supported in C. elegans could constrain possibilities. To restrict production, export, and import of dsRNA to specific periods during development, we are implementing optogenetic methods that can enable precise control of these processes. To control export, we will damage neurons expressing dsRNA at different time points during development using the photosensitizer protein miniSOG. Expressing miniSOG in neurons and exposing the worm to blue light will generate reactive oxygen species, causing neuronal damage. Evaluating silencing after the resultant damage will indicate when in development neurons producing dsRNA must be functional to observe gene silencing in a non-neuronal tissue. Preliminary data suggest that the extent of silencing differs depending on the targeted gene and the developmental time point when worms are exposed to light. Finally, to control production of dsRNA and its import, we will use a light-activated gene expression system to express dsRNA in neurons and the dsRNA-selective importer SID-1 in non-neuronal tissues of sid-1(-) worms, respectively. Progress towards implementing these optogenetic systems will be presented at the meeting.

Choi, Y.S. et al. (2017) International Worm Meeting "Investigating the role of MUT-2 in transgenerational gene silencing triggered by neuronal double-stranded RNAs."

Choi, Y.S., Edwards, Lanelle, DiBello, Aubrey, Jose, Antony

[International Worm Meeting, 2017]

Studies in C. elegans have shown that dsRNA expressed in neurons can transmit a signal to the germline of an animal to cause silencing of a germline gene in that animal and in its descendants. The molecular details of how mobile RNA is processed, exported, and transmitted from neurons to the germline are poorly understood. Tissue-specific rescue experiments using repetitive transgenes suggest that MUT-2, a putative RNA nucleotidyltransferase, is sufficient in the sending cells for gene silencing by mobile dsRNA. One hypothesis suggested by this finding is that forms of RNA that are modified by MUT-2 move between cells. Alternatively, use of repetitive transgenes could have led to misexpression in receiving cells and MUT-2 might function only in receiving cells. Distinguishing between these possibilities requires the development of better reagents and methods for the analysis of MUT-2 and small RNAs. In the absence of MUT-2, transposons are activated, leading to progressive mutagenesis in the background, which could complicate interpretations. Indeed, when we generated new deletions of mut-2 using CRISPR-based genome editing and analyzed three independently propagated mut-2(-) deletion lines, we did not detect the previously reported reduced brood size. To examine possible changes in small RNAs that result from loss of mut-2, we have developed a sensitive northern blotting approach that can detect RNAs as short as 14 nt and thus effectively complement RNA-seq (1). Finally, we are defining the neurons that can export dsRNA to elicit the most effective and enduring transgenerational gene silencing through a large-scale effort called the Transgenerational Brain Initiative using cohorts of ~35 undergraduates each year. This effort will create and characterize a collection of strains expressing dsRNA, reporter genes, and MUT-2 from single copy transgenes in different subsets of neurons. Collectively, with these tools, we will establish where MUT-2 acts to initiate transgenerational gene silencing triggered by neuronal dsRNA. 1. Choi et al., Nucleic Acids Research, 2017.