Figures
Abstract
Organisms used as model genomics systems are maintained as isogenic strains, yet evidence of sequence differences between independently maintained wild-type stocks has been substantiated by whole-genome resequencing data and strain-specific phenotypes. Sequence differences may arise from replication errors, transposon mobilization, meiotic gene conversion, or environmental or chemical assault on the genome. Low frequency alleles or mutations with modest effects on phenotypes can contribute to natural variation, and it has proven possible for such sequences to become fixed by adapted evolutionary enrichment and identified by resequencing. Our objective was to identify and analyze single locus genetic defects leading to RNAi resistance in isogenic strains of Caenorhabditis elegans. In so doing, we uncovered a mutation that arose de novo in an existing strain, which initially frustrated our phenotypic analysis. We also report experimental, environmental, and genetic conditions that can complicate phenotypic analysis of RNAi pathway defects. These observations highlight the potential for unanticipated mutations, coupled with genetic and environmental phenomena, to enhance or suppress the effects of known mutations and cause variation between wild-type strains.
Citation: Asad N, Aw WY, Timmons L (2012) Natural and Unanticipated Modifiers of RNAi Activity in Caenorhabditis elegans. PLoS ONE 7(11): e50191. https://doi.org/10.1371/journal.pone.0050191
Editor: Jo Anne Powell-Coffman, Iowa State University, United States of America
Received: July 16, 2012; Accepted: October 22, 2012; Published: November 28, 2012
Copyright: © 2012 Asad et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by a grant from the National Science Foundation (www.nsf.gov), Grant No. 0951296 to LT; by the United States Educational Foundation in Pakistan/Fulbright Scholarship Program (http://www.iie.org/en/Fulbright/ and http://www.usefpakistan.org/) to NA; and the KU Honors Program (http://www.kuub.ku.edu/honors/) to WYA. Publication costs were supported by the University of Kansas General Research Fund allocation #2301303-906. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
RNA silencing is a term commonly used to describe the multiple, overlapping, RNA-directed pathways that are used by cells to regulate gene expression and chromatin functions. RNA silencing mechanisms serve as cellular sentinels, providing fundamental anti-foreign genome defense activities that protect cells against transposon mobilization and organisms against systemic viral infection. RNA silencing pathways include RNA interference (RNAi) mechanisms that are triggered when dsRNA is provided to cells. Hundreds of genes have been implicated in RNA silencing using forward genetic and RNAi-based screens, which is a testament to the scope of these mechanisms in cells. Genetic defects in some members of these pathways can lead to developmental or environmentally-influenced phenotypes in Caenorhabditis elegans. Considering the vital and fundamental nature of RNA silencing functions, it is surprising that relatively few RNA silencing genes are required for viability or fertility, suggesting a functional overlap of RNA silencing activities.
One practical consequence of the interwoven nature of RNA silencing networks is that mutations in these pathways are often tolerated by the organism. Indeed, several unexpected mutations that affect RNA silencing have been identified in laboratory strains as well as in feral isolates [1], [2]. The mutations found in laboratory strains are considered background mutations as the ancestry of the mutations can often be traced back to a common source. The lack of a selective disadvantage in the laboratory environment may contribute to the perpetuation of such mutations, especially considering that C. elegans strains are normally cultivated in the laboratory in a manner that does not typically select for animals with “wild-type” levels of RNA silencing activity. Alternatively, a selective advantage under some genetic or environmental conditions might contribute to fixation of RNA silencing mutations.
During the course of our studies on RNA silencing mechanisms, we observed that some of our strains harbored unanticipated mutations that affect RNAi. Unlike previously reported background mutations, these have a more recent genesis. Here we report novel mutations in genes that encode Dicer-interacting proteins, as well as long-lasting maternal effects, that allow for RNAi activity in adults. The recurrent observations of background mutations affecting RNA silencing mechanisms, as well as environmental influences and maternal effects that can influence the robustness of RNAi and thereby obscure underlying RNAi-related mutations, highlights the importance of comparing multiple alleles or differently outcrossed strains, as such unexpected mutations may have pleiotropic effects that could complicate phenotypic analyses.
Materials and Methods
C. elegans strains
Mapping and other strains (from A. Fire): PD2119 [dpy-1(e1) ncl-1(e1865) III], PD2039 [dpy-5(e61) unc-54(e1092) I, PD2014 [dpy-10(e128) unc-52(e669)] II, PD2027 [unc-17(e245) dpy-4(e1166)] IV, PD9064 [dpy-11(e224) unc-60(e723)] V: pPD4300 [ccIs4251 (myo-3::GFP) I; ayIs2 (egl-15::GFP) IV; ayIs6 (hlh-8::GFP) X]; pPD5063 [ccIs4251 (myo-3::GFP) I]; PD8175 [rde-1(ne219)].
Deletion strains used to map rde-4(ne309) (from the Caenorhabditis Genetics Center): BC4637 [sDf130(s2427) unc-32(e189) III; sDp3 (III;f)], TY1353 [yDf10 unc-32(e189)/unc-93(e1500) dpy-17(e164) III], BC4638 [sDf127(s2428) dpy-17(e164) unc-32(e189)III; sDp3 (III;f)], BC4697 [sDf121(s2098) unc-32(e189) III; sDp3 (III;f)], MT1928 [nDf16/unc-36(e251) dpy-19(e1259) III], MT5491 [nDf40 dpy-18(e364) III/eT1 (III;V)], DG801 [unc-32(e189) tnDf2/sma-2(e502) ced-7(n1892) unc-69(e587) III], CB4118 [unc-32(e189) ooc-4(e2078)/eDf20 III], GE2158 [tDf2/qC1 dpy-19(e1259) glp-1(q339) III], GE2180 [unc-32(e189) tDf7/qC1 dpy-19(e1259) glp-1(q339) III; him-3(e1147) IV], BW1535 [dpy-18(e364) ctDf3 unc-25(e156)/qC1 dpy-19(e1259) glp-1(q339) III], BW1369 [unc-32(e189) dpy-18(e364) ctDf2/qC1 dpy-19(e1259) glp-1(q339) III], NG2618 [yDf10 unc-32(e189)/qC1 dpy-19(e1259) glp-1(q339) III], CX2914 [nDf16/dpy-17(e164) unc-32(e189)III]
Other strains (from the Caenorhabditis Genetics Center): CB879 [him-1(e879)], AZ244 [unc-119(ed3)III; ruIs57], BB1 [dcr-1(ok247)/unc-32(e189) III], NL1820 [mut-7(pk720) III]
New strains: XX636 [rde-4(ne309); him-1(e879) I], XX637 [dpy-1(e1) unc-119(ed3) III], XX47 [rde-4(ne309)], XX1849 [him-1(e879) I; rde-4(ne309) III], XX1076 [ccIsPD4251(myo-3::GFP)] I; yyEx1.1(myo-3::gfp hairpin, rol-6(su1006))]; XX528 [haf-6(ne335) ccIsPD4251(myo-3::GFP)] I; yyEx1.1(myo-3::gfp hairpin, rol-6(su1006))], XX1151 [ccIs4251(myo-3:: GFP) I; him-8(me4)IV; rde-1(yy11) V, XX1278 [ccIs4251(myo-3:: GFP) I; him-8(me4)IV; rde-1(yy11) V; XX1835 [rde-4(ne299); ccIs4251(myo3::GFP); yyEx1.1(myo-3 promoter::gfp hairpin ; rol-6(su1006)], XX1823 [him-5(e1490) V; ccIs4251(myo3 promoter::GFP)]; Strains XX103, XX172, XX183, XX193, XX194, XX195, XX364, and XX526 are RNAi defective strains that are homozygous for the haf-6(ne335) mutation.
Transgenic strains: The ccIs4251 and yyEx1.1 transgenes used here have been described [3], [4]. The two transgenes were introduced together in wild-type and mutant backgrounds using standard genetics. Both transgenes utilize a myo-3 promoter, which is transcriptionally active in muscle from late embryogenesis through adulthood. ccIs4251 expresses a GFP reporter, and was obtained by inducing an extrachromosomal array to integrate into the middle of chromosome I near the rde-2 gene [3]. The yyEx1.1 transgene contained a myo-3 promoter that drives gfp dsRNA expression in muscle from repetitive extrachromosomal DNA sequences that were assembled by the worm from injected plasmids. The ccIs4251 transgene allows for muscle-specific accumulation of gfp mRNA and of GFP reporter protein and is the target of RNAi in this double transgene RNAi system.
Plasmids
The pLT494 and pLT495 plasmids contain rde-1(yy11) genomic region defined by primers 164 and 207 subcloned into pCR2.1 for sequencing. Amplification of this transposon insertion allele typically produced ∼2300 bp and ∼700 bp PCR products, respectively, presumably due to slipped strand mispairing caused by the Tc1 inverted repeats.
Primers
rde-1 primers: 164: ggaagaaatgcaaaaaagtac; 165: ttagggtattttctttgtag;
170: tatcaatccaggtggaactat; 171: ttcatatcctcctcccattgtt; 207: atcccgttccgacatgaattg; 717: atatatatgtcctcgaattttcccga; 718: atatattgcgaacgacattccagg;
1142: gatctctctagctggccaccaaacatt; 1143: gtttcctgtaaccaaaatatagtcttc;
1144: ggtatttagtgtacaatttttgtagaa; 1145: caccgaactaattggtggttgcaagtt;
1249: gcgatcacaccatcggtgtagctaa; 1250: tcaccagaagaaaaagaaagacgga
Tc1 primers: 1333: tcgataatcatgtaatgtttggtc; 1334: atccatttgacttgaatttttccgt;
1335: gaaacttcaccacagtgttcacaaaatct; 1336: ttcgtattgaaaacgatccataatgcttt;
1337: gacaacccataggatggatcgcaacatcc
rde-4 primers: 209: gaaggatccatggatttaaccaaactaacg; 719: atatatatccgtgaaatcataggtg
PCR- based detection of rde-1(yy-11) allele in mixed populations
Animals were washed with water and suspended in NET buffer (100 mM NaCl, 20 mM EDTA, 50 mM Tris-HCl, pH8.0) and frozen. 1∶200 volume of 10% SDS and 1 ug proteinase K were added to thawed pellets, which were incubated at 65°C for 2–4 hours, then phenol extracted, and ethanol precipitated. Primer set 1249/207 anneals to the rde-1 gene, producing an amplified product of 640 bp in wild-type animals. (This primer set preferentially amplifies the wild-type allele in mixed populations of rde-1(yy11) and wild-type animals.) Primer set 1249/1333 specifically amplifies the rde-1(yy11) insertion, yielding an amplification product of 279 bp.
dsRNA delivery by “feeding”
Standard C. elegans plates supplemented with cholesterol, ampicillin, tetracycline, and IPTG were seeded with the Escherichia coli strain HT115(DE3) harboring feeding plasmids as described [5], [6], [7], [8]. Transformed bacteria harboring pop-1 cDNA in L4440 vector transcribe dsRNA that targets the TCF/LEF1 transcription factor, inducing sterility in young animals reared on this food, while the dpy-11 food targets a protein that affects body shape, eliciting a shorter and fatter appearance in comparison to wild type. Control experiments utilized wild-type worms and empty L4440 vectors. Experiments were performed at 15°C, 20°C and 25°C.
Genetic analysis of RDE-4 maternal effects
To assess the extent of RDE-4 maternal effects, we first generated strains such as XX1835, which harbors the rde-4(299) mutation, the yyEx1.1 gfp dsRNA-expressing extra-chromosomal array, and the integrated ccIs4251 GFP reporter. Homozygous rde-4 animals (XX1835, as well as a similar strain harboring the rde-4(ne309) mutation) were mated with XX1823 males, which harbor the GFP reporter. F1 progeny harboring the yyEx1.1 gfp dsRNA-expressing transgene array marked with a dominant “roller” transformation marker, as well as the GFP reporter, were individually placed on standard C. elegans culture plates. F2 animals harboring the “roller” phenotype were scored for RNAi activity as young adults and subsequently genotyped by DNA sequencing of the rde-4 gene. 25% of such F2 progeny are predicted to be homozygous for the rde-4 mutation. Before sequence analysis, some F2 adults were again placed as individuals on standard culture plates and allowed to produce F3 animals. The transgenic F3 animals were scored for RNAi activity and some were also sacrificed for DNA sequencing. 25% of the plates with such F3 progeny are predicted to be homozygous for the rde-4 mutation. Animals were tracked, and the genotype was correlated with phenotype at the end of each experiment. We scored animals as RNAi+ if they displayed GFP fluorescence in fewer than 30% of muscle cells and RNAi- when GFP fluorescence in was observed in 70–90% of muscle cells. (All RNAi- animals in these assays displayed bright GFP fluorescence in >90% of muscle cells.) Experiments were conducted at 25°C, 22°C and 15°C.
Results and Discussion
A novel allele of rde-1 arose spontaneously in a transgenic stock
As an extension of experiments designed to uncover roles for the Caenorhabditis elegans ABC transporter gene haf-6 in RNA interference [9], [10], we built strains to determine if haf-6 is required for RNAi when dsRNA is transcribed from transgenes. The strains harbored two distinct configurations of transgenic DNA such that gfp sequences served as both trigger and target of RNAi, allowing for direct visual inspection of RNAi activity. In a wild-type background, this dual transgene system allows for easy visualization of an RNAi response against gfp. Our experiments using these transgenic strains led to observations of RNAi defects (Figure 1). The RNAi defects were not rescued by the introduction of wild-type haf-6 sequences, and we ascertained that the transgene-mediated RNAi defects were due to an unknown mutation, which we termed yy11, present in the background of the transgenic haf-6(ne335) strain.
The yy11 allele, present as a background mutation in a haf-6(ne335) transgenic stock, was genetically dissected from the haf-6(ne335) mutation and analyzed separately for RNAi activity. (A) yy11 provides resistance to transgene-delivered dsRNAs, as evidenced by lack of gene silencing of the GFP reporter in homozygous animals. By contrast, haf-6(ne335) single mutants display wild-type RNAi activity in this transgene assay. (B). Animals homozygous for the yy11 allele are RNAi resistant when dsRNAs are delivered by ingestion of bacteria that express dsRNAs. Thus, the rde-1(yy11) mutants displayed strong RNAi phenotypes irrespective of delivery method (A, B), while, as expected, the haf-6(ne335) single mutant was RNAi defective by ingestion only (A, B). Bacteria expressing dsRNA targeting the pop-1 and dpy-11 genes were used in feeding experiments. (C) Genetic and molecular analysis revealed the nature of the yy11 mutation—a transposon insertion in exon 9 of rde-1. DNA sequences flanking the insertion site are indicated.
In order to identify the nature of the gene with the yy11 mutation, we employed standard genetic methodology to separate the recessive haf-6(ne335) and yy11 mutations (Figure 1A,B). The resulting single mutant yy11 homozygotes displayed RNAi defects in response to transgene-delivered dsRNAs, as well as defects in response to ingestion of dsRNAs targeting genes expressed in the germ line or soma. By contrast, the haf-6(ne335) single mutants were RNAi-defective in response to ingested dsRNA, as expected, but did not display RNAi defects in response to transgenes. Thus, the transgene-mediated RNAi defect was solely due to the unknown yy11 mutation.
We mapped the yy11 mutation to chromosome V. Considering that the yy11 mutation elicits a strong RNAi defect, we reasoned that the corresponding gene might already have been uncovered in screens for RNAi-defective mutants. We therefore performed complementation tests between yy11 strains and RNAi-defective worms harboring mutations in genes located on chromosome V. The rde-1(ne219) mutation failed to complement the RNAi defect in yy11, suggesting that yy11 is an allele of rde-1. We then sequenced the rde-1 locus in our strain and found a novel Tc1 transposon inserted in exon 9 (Figure 1C). Tc1 elements are members of the large mariner superfamily of simple transposable elements that harbor a single transposase gene flanked by inverted repeats [11]. The rde-1(yy11) allele is a novel and unusual mutation in that most of the documented alleles of rde-1 are single base substitutions.
We reasoned that the sequence of the Tc1 transposon insertion in rde-1(yy11) might help illuminate how the mutation arose in our lab. The rde-1(yy11) Tc1 element has almost 100% identity with a Tc1 element that resides within an intron of T07D3.3, a gene on chromosome 2. This sequence identity provides an indication as to the source of mobilized DNA, while the presence of a T insertion after base 204 in the rde-1(yy11) element (the only sequence difference between the two transposons) may reflect the recent nature of the transposition event. A potentially active transposon in rde-1, coupled with the ability to select for rde-1 activity using RNAi methodology, raises the possibility of using the rde-1(yy11) allele as a screening or assessment tool for transposon silencing or RNAi activities.
The rde-1(yy11) allele was attained serendipitously–it was first observed in a haf-6(ne335); rde-1(yy11) double mutant (Figure 1). Two of the ancestors of this stock had been exposed to mutagens, which may have resulted in the transposition event that gave rise to the rde-1(yy11) allele [12]. One of these ancestral strains was the original haf-6(ne335) isolate; a second ancestral strain was exposed to mutagen in order to induce integration of the ccIs4251 GFP reporter sequence [3]. We therefore hypothesized that the transposition event leading to the rde-1(yy11) mutation might have originated in such ancestral strains and thereafter lingered as a low frequency mutation. We designed a PCR-based assay to test for the presence of the Tc1 insertion using PCR and amplification conditions that would allow for the detection of the rde-1(yy11) allele in a mixed population of worms (Figure 2A). We applied this assay to progenitor strains, and strains later derived from them, for the presence of the rde-1(yy11) allele. We detected the allele in a frozen archived stock from the original, un-outcrossed haf-6(ne335) isolate (Figure 2B). haf-6 mutants display mutator activity [10]; therefore, the rde-1(yy11) insertion may be a reflection of the transposon silencing defect in haf-6 mutants.
A PCR-based assay was developed in order to detect the presence of the rde-1(yy11) allele in a mixed population of animals. (A) A single rde-1(yy11) homozygous animal was introduced into a tube with a number of hand picked, wild type adults. PCR reactions were performed using rde-1(yy11) allele-specific primers (1249/1333, see Materials and Methods). The ratios of rde-1(yy11): wild-type animals in each PCR assay were 1∶10 (lane 1), 1∶100 (lane 2), 1∶500 (lane 3), 1∶1000 (lane 4), and 1∶5000 (lane 5). The PCR-based assay was also performed on individual wild-type (lane 6) and rde-1(yy11) (lane 8) worms, as well as genomic DNA isolated from mixed-stage worms harvested from twenty-five 10 cm plates of mixed stage worms of wild-type (lane 7) or rde-1(yy11) (lane 9) genotype. (B) Frozen, archived stocks of mixed stage worms were thawed and sacrificed for genomic DNA isolation in order to determine the possible origin of the rde-1(yy11) allele. Each lane represents a PCR reaction from a different stock; for each strain the oldest tube from the archive was selected. The strains assayed harbored haf-6(ne335) alleles (lanes 11–18) or the ccIS4251 transgene insertion (lanes 10 and 19). Top panel in B: control PCR reaction using primers designed to amplify wild-type rde-1 alleles (640 bp product, see Materials and Methods). Bottom panel: PCR reaction using primers specific for the rde-1(yy11) insertion allele (279 bp product, see Materials and Methods). The MW marker is 500 bp in size. The XX103 stock analyzed in lane 11 is an archived stock of the original, un-outcrossed haf-6(ne335) mutant stock; the other haf-6(ne335) stocks analyzed (lanes 12–18) are outcrossed or derivative strains.
The rde-1(yy11) mutation was present at low frequency in the background of the original haf-6(ne335) strain, as we did not detect the yy11 allele PCR assays on smaller numbers of animals from this stock nor from genomic DNA isolated from all worms in similar tubes of archived stocks of various haf-6 strains (our unpublished observations). Thus, even though the haf-6(ne335) strain used to build the transgenic strain in Figure 1A had been out-crossed to wild-type animals, the rde-1(yy11) mutation was not eliminated from the population. It is likely that the strategy we used to produce the transgenic haf-6 strains resulted in the selected enrichment of rde-1(yy11). A recombination event was required to place the integrated ccIs4251 GFP reporter transgene onto the same chromatid as haf-6 (both elements reside in chromosome I). To facilitate the identification of the infrequent GFP-expressing haf-6 recombinants, populations of progeny animals were selected for RNAi defectiveness using dsRNA-expressing bacteria, a strategy which may have contributed to the selection of the strong RNAi-defective rde-1(yy11) background mutation. It is interesting that GFP/haf-6 recombinant animals that arose from this selection were also homozygous for rde-1(yy11). We have not detected the rde-1(yy11) allele in other strains or in other batches of the same archived strain depicted in Figure 2; thus, this allele was eliminated from subsequent derivative strains, or lost from the population before the derivative stocks were frozen.
The novel rde-1(yy11) allele should prove to be a useful tool, facilitating analysis of the RNAi pathway in C. elegans. For example, an insertion allele such as rde-1(yy11) affords particular molecular advantages over existing rde-1 alleles, all of which are single base substitutions, in that the insertion can be identified quickly using PCR. This strategy should allow for more expeditious molecular screening for rde-1 mutations in situations where multiple mutations in the RNAi pathway must be brought together in the same strain. Additionally, the transposon insertion helps to confirm the null phenotype, allowing for a more complete phenotypic analysis. Existing rde-1 alleles with experimentally established RNAi defects harbor small alterations to the DNA and/or protein sequence. These alleles are amino acid substitutions or disruptions that are downstream of the yy11 insertion: for example, a E414K substitution in the rde-1(ne219) allele; G485E in rde-1(ne4086); I562L in rde-1(ne4085); G1016E in rde-1(ne297); and a Q825STOP mutation in rde-1(ne300). The Tc1 insertion in yy11 occurs after amino acid G685, and is a more severe disruption than existing point mutations, meaning that it is also likely to be a null mutation.
Maternal effects can mask the presence of mutations in genes with RNAi function
We previously described a strain with an RNAi-defective allele, ne309, that does not mount an RNAi response to ingested dsRNAs [4]. We mapped the ne309 mutation to chromosome III, and we performed complementation tests using large deletions to localize the ne309 mutation and to help confirm that the strain harbored a single RNAi-defective allele (Figure 3A). Deletions lacking DNA sequences near the center of chromosome III failed to complement ne309, which led us to perform additional complementation tests between ne309 and RNAi-defective strains with mutations in specific genes that map to this region. Among these genes was rde-4, which failed to complement ne309. Upon sequencing of the rde-4 gene in ne309 mutants, we found a mutation consisting of a single T insertion in exon 1. The rde-4(ne299) mutation also has a non-identical T insertion within the same codon for Ala123 [13](Figure 3C).
(A) Strains with the indicated chromosomal deficiencies were used in complementation tests in order to localize the ne309 mutation. The deleted regions highlighted in gray failed to complement ne309, and the relative location of breakpoints is indicated on the genetic map. Breakpoint information is collected by the C. elegans community. The data reflects the information posted on Wormbase, version WS231. (B) The RNAi defect in ne309 mutants is due to a single base insertion of a T residue in the first exon of rde-4. This mutation is predicted to lead to a truncated protein lacking the second of two double-stranded binding domains. The first dsRNA binding domain alone is unable to bind dsRNA [28]. (C) A similar T insertion in rde-4(ne299) affects the same codon.
Our preliminary analyses of rde-4(ne309) homozygotes revealed that the mutants were RNAi deficient in their response to ingested dsRNA, yet they displayed a systemic RNAi response to transgene-delivered dsRNA [4]. These results are inconsistent with reports of fairly strong RNAi defects in rde-4 mutants, irrespective of dsRNA delivery method [13]. We therefore re-investigated the RNAi response to transgene-delivered dsRNA in rde-4(ne309) mutants. We first ensured that the strains in this study did not harbor previously described background mutations that affect RNAi activity; for example, we used strains that had been extensively outcrossed in order to remove any incidental mutations that might affect RNAi. We again used the two transgene system to target silencing of a gfp reporter to assess RNAi activity (as in Figure 1). In these experiments, we observed strong RNAi defects in rde-4(ne309) and rde-4(ne299) mutants, which is consistent with previous reports of strong RNAi defects in rde-4. By contrast, we also observed RNAi competency in rde-4 homozygotes due to maternal effects (Figure 4A and Tables 1 and 2).
All animals harbor an integrated transgene expressing GFP in muscle, as well as an extrachromosomal array expressing gfp dsRNA in muscle. (A) F2 generation rde-4 mutants, derived from heterozygous parents, display RNAi activity due to maternal effect (third row). rde-4 F3 animals, derived from homozygous rde-4 parents, display RNAi defects, as evidenced by the lack of GFP reporter silencing (fourth row). Each generation of animals was derived from a series of genetic crosses, with genotypes depicted on the right. Similar results were observed for both rde-4(ne299) and rde-4(ne309) alleles. The animals depicted in (A) are adults. (B) Early stage wild-type L1 larvae harboring both gfp dsRNA trigger and gfp mRNA target sequences do not display RNAi. In later stages, RNAi activity in these wild-type animals becomes apparent, as depicted in (A), top left. 40× magnification.
An RNAi response was observed in the adult somatic tissues of rde-4 homozygotes born from a heterozygous parent, which is surprisingly long-lasting for a maternal effect (Figure 4A). Such long-lasting maternal effects have also been observed for dcr-1 mutants [14], [15]; however, unlike rde-4 mutants, dcr-1 animals are sterile. There are several factors which may contribute to the perdurance of RNAi activity in rde-4 mutants displaying such maternal effects. C. elegans harbors dsRNA amplification mechanisms that, in part, involve the activity of RNA-dependent RNA Polymerases and the production of secondary siRNAs [16]. Thus, any maternally deposited silencing RNAs, along with maternally deposited RDE-4 could contribute to secondary siRNA production early in development, allowing for later RNAi activity, long after the maternally deposited RDE-4 and dsRNA are eliminated. Another factor that could contribute to RNAi activity in rde-4 adults might involve early modifications to chromatin, triggered by maternally deposited molecules, and persistence of the modifications through larval development [17], perhaps maintained by later exposure to dsRNA. Maternal effects have previously been observed for rde-4, yet such experiments made use of dsRNAs that were injected into heterozygotes, and these injected molecules may have persisted in the progeny animals for the duration of the experiment [18], [19]. By contrast, our experiments utilize a dual transgene system, which affords particular advantages in analyzing RDE-4 maternal effects. In this dual transgene system, the myo-3 promoters that drive transcription of dsRNA are activated during late embryogenesis in the body wall muscles of the developing embryo [20]. Furthermore, the dsRNA sequences are expressed from a repetitive extra-chromosomal array, a DNA configuration that is normally transcriptionally silenced in the germ line [21]. The relatively late-acting nature of the promoter in this system provides additional insurance that dsRNA is not present in early developmental stages, and potentially allows time for any maternally deposited RDE-4 or dsRNA molecules to be cleared before new gfp dsRNA is transcribed. Indeed, 100% of the newly hatched L1 larvae from this dual transgene system do not display an RNAi phenocopy even in wild-type animals (as depicted in Figure 4B), an indication that any dsRNA molecules or chromatin modifications remaining at this stage have been eliminated or are functionally ineffective. Thus, the observation of RNAi activity in rde-4 homozygotes at adult stage (Figure 4A) provides an indication that maternally-deposited RDE-4 persists until later stages, when it then acts upon dsRNA that is newly transcribed.
In addition to maternal effects, there are other conditions that can influence the level of RNAi activity observed in rde-4 mutants. For example, RNAi activity has been observed in homozygous rde-4 mutants when the dosage of dsRNA is high. Indeed, we previously observed an RNAi response in rde-4(ne309) mutants to injected dsRNA [4]. Earlier reports also demonstrated that rde-4 mutants can display an RNAi response when large amounts of dsRNA are delivered by transgenes [22]. We can then infer several factors that may influence the level of dsRNA accumulation from transgenes: the copy number of dsRNA-expressing cassettes in transgene arrays, the promoter strength, and the expression status of the transgene–transgenes configured as extrachromosomal arrays are targets of epigenetic mechanisms that can lead to increased levels of transcriptional repression in a generational fashion, especially in the germ line [21]. Our earlier observations that rde-4(ne309) mutants were capable of mounting a systemic RNAi response to dsRNA expressed in muscle were likely influenced by the rde-4 maternal effects, the young age of the dsRNA-expressing array (newly formed arrays are metastable with respect to transcriptional status and may be capable of higher levels of transcription than an array maintained for many generations [21]), and by the relatively low abundance of the gfp mRNA target used at that time [4]. Finally, environmental influences, such as temperature, have also been observed to affect the RNAi activity in wild-type as well as rde-4 and other RNAi-related mutants [22], [23]. Taken together, these considerations highlight the possibility that dsRNA concentrations and environmental conditions might mask the presence of an underlying and unanticipated RNAi defect caused by a background mutation in a non-essential RNAi pathway component.
Conclusions
The Dicer double-stranded RNA-specific endonuclease plays a central role in many RNA silencing pathways and is an essential gene in Caenorhabditis elegans and other organisms. The DCR-1 protein, in part, catalyzes the conversion of dsRNAs into small siRNAs, facilitates the formation of the RISC complex, and processes miRNA genes in C. elegans. DCR-1 is directed towards particular functions in the context of the associated proteins. For example, the ERI/DCR complex contains DCR-1 as well as ERI-1, RRF-3, ERI-3, and ERI-5 [24]. The ERI/DCR complex is important for the proper function of endogenous siRNAs, a pathway that is relevant to proper sperm development [25]. In another distinct DCR-1 complex can be found the RDE-1 and RDE-4 proteins. Among other activities, this complex is required for an RNAi response to experimentally-delivered dsRNA. RDE-4 is a dsRNA binding protein required for production of siRNAs from long dsRNAs [26], while the RDE-1 Argonaute protein facilitates removal of the passenger strand from the siRNA, which is necessary in order for proper function of the guide strand [27]. Mutations in genes encoding the eri class of DCR-1 interacting proteins enhance RNAi activity, while mutations in rde-1 and rde-4 strongly abrogate RNAi. Despite the important endogenous roles played by Dicer and associated proteins, most of the proteins associated with Dicer are not essential. Thus, when such genes are unknowingly mutated, there is potential for genetic complication and misinterpretation of data.
Here we report observations of rde-4 and rde-1 mutations that highlight the potential for maternal effects, environmental conditions, dsRNA dosage and recently derived mutations to complicate phenotypic analyses. The rde-1(yy11) mutation, a recently derived transposon insertion, spontaneously appeared in a strain that harbored a known RNAi-defective mutation. As a result, the effects of the background rde-1 mutation were at first mistakenly attributed to the known mutation. Fortunately, the rde-1(yy11) allele was localized to only one set of experiments, and the mutation was identified before further complications arose. The phenotypic complications related to rde-1(yy11) were opposite to those observed for rde-4(ne309). In this case, the systemic RNAi defects in a known, but not yet identified, RNAi-defective mutant were masked by maternal effects and a high ratio of dsRNA trigger molecules to mRNA target. There are many non-essential RNAi pathway mutants with RNAi defects that are more subtle than those observed in rde-4, and many of these mutants are dosage-sensitive with respect to the amount of dsRNA as well. Therefore, unanticipated background mutations in such genes have the potential to complicate phenotypic analysis in a similar manner.
It has become increasingly apparent in the C. elegans RNAi field that the possibility of background mutations must be considered. There are several steps that can be taken in order to avoid complications from unanticipated background mutations, such as performing experiments using multiple, independently derived alleles. If multiple alleles are not available, then different strains can be derived from the same mutant using different outcrossing strategies. Mutant analysis can be also compared to data derived from dsRNA-treated animals. Finally, trans-heterozygotes animals derived from crossing together strains homozygous for different alleles or for the same allele obtained from two different outcrossing strategies can be used to confirm experimental results.
Acknowledgments
We thank Dr. Andrew Fire for PD strains and the ccIs4251 integrated transgene, Dr. Craig Mello for the ne309 and ne299 alleles, the Caenorhabditis elegans Genetics Center for strains, and Juliana Metichecchia for manuscript review.
Author Contributions
Conceived and designed the experiments: LT NA. Performed the experiments: NA WYA LT. Analyzed the data: NA WYA LT. Contributed reagents/materials/analysis tools: NA WYA LT. Wrote the paper: NA LT.
References
- 1. Tijsterman M, Okihara KL, Thijssen K, Plasterk RH (2002) PPW-1, a PAZ/PIWI protein required for efficient germline RNAi, is defective in a natural isolate of C. elegans. Current biology: CB 12: 1535–1540.
- 2. Zhang C, Montgomery TA, Gabel HW, Fischer SE, Phillips CM, et al. (2011) mut-16 and other mutator class genes modulate 22G and 26G siRNA pathways in Caenorhabditis elegans. Proceedings of the National Academy of Sciences of the United States of America 108: 1201–1208.
- 3. Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, et al. (1998) Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391: 806–811.
- 4. Timmons L, Tabara H, Mello CC, Fire AZ (2003) Inducible systemic RNA silencing in Caenorhabditis elegans. Molecular biology of the cell 14: 2972–2983.
- 5. Timmons L, Court DL, Fire A (2001) Ingestion of bacterially expressed dsRNAs can produce specific and potent genetic interference in Caenorhabditis elegans. Gene 263: 103–112.
- 6. Hull D, Timmons L (2004) Methods for delivery of double-stranded RNA into Caenorhabditis elegans. Methods in molecular biology 265: 23–58.
- 7. Timmons L, Fire A (1998) Specific interference by ingested dsRNA. Nature 395: 854.
- 8. Timmons L (2006) Delivery methods for RNA interference in C. elegans. Methods in molecular biology 351: 119–125.
- 9. Sundaram P, Echalier B, Han W, Hull D, Timmons L (2006) ATP-binding cassette transporters are required for efficient RNA interference in Caenorhabditis elegans.. Mol Biol Cell 17: 3678–3688.
- 10. Sundaram P, Han W, Cohen N, Echalier B, Albin J, et al. (2008) Caenorhabditis elegans ABCRNAi transporters interact genetically with rde-2 and mut-7. Genetics 178: 801–814.
- 11. Plasterk RH, Izsvak Z, Ivics Z (1999) Resident aliens: the Tc1/mariner superfamily of transposable elements. Trends in Genetics: TIG 15: 326–332.
- 12. Staleva L, Venkov P (2001) Activation of Ty transposition by mutagens. Mutation Research 474: 93–103.
- 13. Tabara H, Yigit E, Siomi H, Mello CC (2002) The dsRNA binding protein RDE-4 interacts with RDE-1, DCR-1, and a DExH-box helicase to direct RNAi in C. elegans. Cell 109: 861–871.
- 14. Ketting RF, Fischer SE, Bernstein E, Sijen T, Hannon GJ, et al. (2001) Dicer functions in RNA interference and in synthesis of small RNA involved in developmental timing in C. elegans.. Genes & development 15: 2654–2659.
- 15. Knight SW, Bass BL (2001) A role for the RNase III enzyme DCR-1 in RNA interference and germ line development in Caenorhabditis elegans.. Science 293: 2269–2271.
- 16. Sijen T, Fleenor J, Simmer F, Thijssen KL, Parrish S, et al. (2001) On the role of RNA amplification in dsRNA-triggered gene silencing. Cell 107: 465–476.
- 17. Gu SG, Pak J, Guang S, Maniar JM, Kennedy S, et al. (2012) Amplification of siRNA in Caenorhabditis elegans generates a transgenerational sequence-targeted histone H3 lysine 9 methylation footprint. Nature genetics 44: 157–164.
- 18. Grishok A, Tabara H, Mello CC (2000) Genetic requirements for inheritance of RNAi in C. elegans. Science 287: 2494–2497.
- 19. Blanchard D, Parameswaran P, Lopez-Molina J, Gent J, Saynuk JF, et al. (2011) On the nature of in vivo requirements for rde-4 in RNAi and developmental pathways in C. elegans. RNA biology 8: 458–467.
- 20. Ardizzi JP, Epstein HF (1987) Immunochemical localization of myosin heavy chain isoforms and paramyosin in developmentally and structurally diverse muscle cell types of the nematode Caenorhabditis elegans. The Journal of cell biology 105: 2763–2770.
- 21. Kelly WG, Xu S, Montgomery MK, Fire A (1997) Distinct requirements for somatic and germline expression of a generally expressed Caernorhabditis elegans gene. Genetics 146: 227–238.
- 22. Habig JW, Aruscavage PJ, Bass BL (2008) In C. elegans, high levels of dsRNA allow RNAi in the absence of RDE-4. PloS one 3: e4052.
- 23. Han W, Sundaram P, Kenjale H, Grantham J, Timmons L (2008) The Caenorhabditis elegans rsd-2 and rsd-6 genes are required for chromosome functions during exposure to unfavorable environments. Genetics 178: 1875–1893.
- 24. Duchaine TF, Wohlschlegel JA, Kennedy S, Bei Y, Conte D Jr, et al. (2006) Functional proteomics reveals the biochemical niche of C. elegans DCR-1 in multiple small-RNA-mediated pathways. Cell 124: 343–354.
- 25. Pavelec DM, Lachowiec J, Duchaine TF, Smith HE, Kennedy S (2009) Requirement for the ERI/DICER complex in endogenous RNA interference and sperm development in Caenorhabditis elegans.. Genetics 183: 1283–1295.
- 26. Parrish S, Fire A (2001) Distinct roles for RDE-1 and RDE-4 during RNA interference in Caenorhabditis elegans.. RNA 7: 1397–1402.
- 27. Steiner FA, Okihara KL, Hoogstrate SW, Sijen T, Ketting RF (2009) RDE-1 slicer activity is required only for passenger-strand cleavage during RNAi in Caenorhabditis elegans. Nature Structural & Molecular Biology 16: 207–211.
- 28. Parker GS, Maity TS, Bass BL (2008) dsRNA binding properties of RDE-4 and TRBP reflect their distinct roles in RNAi. Journal of Molecular Biology 384: 967–979.