|Home | About | Journals | Submit | Contact Us | Français|
In the nematode Caenorhabditis elegans, the let-7 microRNA (miRNA) and its family members control the timing of key developmental events in part by directly regulating expression of hunchback-like-1 (hbl-1). C. elegans hbl-1 mutants display multiple developmental timing deficiencies, including cell cycle defects during larval development. While hbl-1 is predicted to encode a transcriptional regulator, downstream targets of HBL-1 have not been fully elucidated. Here we report using microarray analysis to uncover genes downstream of HBL-1. We established a transgenic strain that overexpresses hbl-1 under the control of a heat shock promoter. Heat shock-induced hbl-1 overexpression led to retarded hypodermal structures at the adult stage, opposite to the effect seen in loss of function (lf) hbl-1 mutants. The microarray screen identified numerous potential genes that are upregulated or downregulated by HBL-1, including sym-1, which encodes a leucine-rich repeat protein with a signal sequence. We found an increase in sym-1 transcription in the heat shock-induced hbl-1 overexpression strain, while loss of hbl-1 function caused a decrease in sym-1 expression levels. Furthermore, we found that sym-1(lf) modified the hypodermal abnormalities in hbl-1 mutants. Given that SYM-1 is a protein secreted from hypodermal cells to the surrounding cuticle, we propose that the adult-specific cuticular structures may be under the temporal control of HBL-1 through regulation of sym-1 transcription.
The development of multicellular organisms occurs in four dimensions: the three axes of space and a fourth axis of time. Spatial pattern formation in development, such as morphogenesis and organogenesis, is regulated with precise timing during development. Although a large number of studies have identified genes playing important regulatory roles in spatial pattern formation, much less is known about temporal patterning of development. The genes controlling developmental timing have been studied most extensively in the nematode Caenorhabditis elegans.1-4 C. elegans develops into the adult through four larval stages, designated L1-L4. Each larval stage is characterized by invariant, stage-specific patterns of cell fate decisions and stage-specific cuticle formation.5 Some genes, designated heterochronic genes, control the timing of cell fate determination during postembryonic development.6 Mutations in heterochronic genes result in either precocious or retarded development in which stage-specific schedules of cell division and differentiation are either skipped or reiterated.
Recent work has shown that several heterochronic genes are evolutionarily conserved in sequence from C. elegans to mammals, epitomized by the developmental timing microRNA (miRNA) called let-7, which is 100% conserved in humans.3,7-9 MiRNAs are ~22 nucleotide, single-stranded RNA molecules that posttranscriptionally regulate their messenger RNA (mRNA) targets by binding with imperfect complementarity to sites in their 3′ untranslated regions.10 One role of the let-7 miRNA is to promote the transition from L4 to adulthood (L/A switch) by downregulating several downstream targets.11,12 In addition to their nucleotide sequences, the temporal expression pattern of let-7 as well as several let-7 target genes are highly conserved during animal evolution.8,9 For example, it has been shown that the let-7 miRNA regulates developmental events necessary for the switch from pupal to adult stages in the fruit fly Drosophila melanogaster.13,14 These data imply that let-7 is a core component of temporal switches during development in multiple animal phyla.
Among the identified let-7 targets, hbl-1 is a critical regulator in controlling the L/A switch in C. elegans.15,16 hbl-1 encodes a Cys2-His2 zinc-finger transcription factor, homologous to the segmentation gene hunchback (hb) of Drosophila. In C. elegans, hbl-1 acts to inhibit precocious expression of later fates during the L2 to L3 transition and also the L4 to adult transition. Both of these developmental transitions require the proper downregulation of hbl-1 at the posttranscriptional level by let-7 or other let-7 miRNA family members, mir-48, mir-84 and mir-241.15-18 In Drosophila, Hb regulates not only the development of the syncytial embryo but also the temporal identity of neuroblasts,19,20 suggesting that the hb/hbl-1 family shares a conserved role in regulating developmental timing across species.3 From Drosophila research, several Hb targets have been identified and a consensus binding site has been elucidated. In C. elegans, a recent study has revealed that the let-7 miRNA itself is a direct target of HBL-1.21 However, no direct protein coding transcriptional targets have yet been identified for HBL-1.
To unravel downstream targets of HBL-1, we performed a microarray analysis using a transgenic strain for conditional over-expression of hbl-1. We show that forced expression of hbl-1 under the control of a heat shock promoter is sufficient to drive developmental timing defects, including retarded seam cell differentiation and molting defects. A microarray analysis to compare gene expression profiles between control and hbl-1 overexpression animals identified potential downstream targets of HBL-1, including sym-1, which encodes a secreted protein containing a signal sequence and 15 contiguous leucine-rich repeats. Lastly, we demonstrate that sym-1 shows a genetic interaction with hbl-1 in controlling temporal hypodermal differentiation programs.
For the identification of HBL-1 downstream targets, we generated transgenic C. elegans that conditionally overexpress hbl-1 under the heat shock promoter hsp-16. Inducible versions of transcription factors have been widely used for the identification of downstream targets in several model organisms including C. elegans.22 We constructed a plasmid in which a DNA fragment of the hbl-1 open reading frame (ORF) was fused with the promoter from the heat shock protein 16 (hsp-16) gene. For control experiments, we constructed another plasmid carrying a gene encoding the sea anemone red fluorescence protein DsRed2 fused with the hsp16 promoter. We obtained stable lines containing the hbl-1 transgene and the control DsRed2 transgene, respectively, both of which were integrated into the host chromosomes. The heat shock-dependent induction of hbl-1 was confirmed in two independent integrated lines as judged by reverse transcription (RT)-PCR (Fig. 1A). Heat shock treatment also induced DsRed2 fluorescence as compared to the non-heat shock condition (Fig. S1). These results demonstrated that the established integrated transgenic lines could be used for the conditional induction of hbl-1 and DsRed2. Hereafter we refer to these transgenic genotypes as ‘hsp::hbl-1’ or ‘hsp::DsRed2’ for simplicity.
Depletion of function of either lin-41 or hbl-1, both of which are targeted by the let-7 miRNA family, causes precocious seam cell terminal differentiation in the L4 stage.15,16,23 It has been shown that the animals containing a higher than normal dose of lin-41 show retarded adult differentiation, which is opposite to loss of lin-41 function.23 Similar to the lin-41-overexpression phenotype, we found that forced hbl-1 expression induced by heat shock at the L4 stage caused retarded adult differentiation defects. The heat shocked hbl-1 overexpression strain often produced gapped alae in the adult stage (Fig. 1B and Table 1). The gapped, patchy alae phenotype has also been observed for other retarded heterochronic mutants.16,17,24,25 In addition, animals induced at the L4 stage to overexpress hbl-1 often had trouble executing the molt into the adult stage, and remain stuck in the L4 stage cuticle (Fig. 1C). In conjunction with the fact that molting is known to be affected in hbl-1 hypomorphic mutants to some degree,15,26 this result suggests that a proper level of HBL-1 activity is required for completing the molting process. The effects of hbl-1 were specific to the time of overexpression, as neither gapped alae nor molting defects were observed in hsp::hbl-1 integrated animals treated by heat shock during the L1 molt or in adult stages (Table 1; data not shown).
Identical phenotypes were observed in two independent transgenic lines that carry the hsp::hbl-1 transgene, excluding the possibility that the phenotypes were due to an effect of transgenic insertions in the genome (Table 1). This phenotype was specific to hbl-1 overexpression, as the hbl-1 overexpression phenotypes were completely suppressed by hbl-1(RNAi) (Table 1). In addition, such gapped alae were not observed in heat shock-induced DsRed2 animals (Table 1). Since a heterochronic developmental abnormality is caused by both loss-of-function and gain-of-function of hbl-1, these results show that hbl-1 is a critical heterochronic gene.
We utilized the hsp::hbl-1 and hsp::DsRed2 strains to identify genes that can be induced by the HBL-1 transcription factor. These animals were heat shocked at the L4 stage for just 1.5 hours to conditionally induce hbl-1 in a short time window. This allowed us to focus on immediate-early genes, most likely to be directly downstream of hbl-1. One hour after the end of the heat shock, we harvested the animals and extracted RNAs for microarray analysis. For this purpose, the hsp::hbl-1 or hsp::DsRed2 animals possessed the temperature sensitive glp-4 mutation, which lacks the germline cell lineages in animals at restrictive temperature to allow us to eliminate the highly variable gene expression in germline cells.27 Using microarrays containing approximately 90% of predicted C. elegans genes,28 we performed at least three hybridizations comparing gene expression profiles in the hbl-1 overexpression animals versus the DsRed2 overexpression animals. In this microarray analysis, the hbl-1 transcript itself exhibited an approximately 16-fold increase as compared to the DsRed2 overexpressing lines (Table 2), confirming that heat shock induction of hbl-1 was robust.
We focused on 79 genes, of which 26 and 53 are genes whose expression was up or downregulated, respectively, in the heat-shocked hsp::hbl-1 animals relative to the heat-shocked hsp::DsRed2 animals, exceeding a two-fold difference at p < 0.05 (Table 2; see Materials and Methods). We classified the biological functions of these identified genes using a database of Clusters of Orthologous Groups (COGs), a functional and phylogenetic annotation of the proteins encoded in numerous complete genomes across kingdoms.29 Although many genes were not classified in specific COGs (Table 2), some COGs were more predominantly enriched in identified genes than others. For example, 17% of the upregulated genes were expected to affect carbohydrate transport and metabolism, whereas 13% of the downregulated genes are predicted to be involved in lipid transport and metabolism (Table S1). These data imply that HBL-1 may transcriptionally control these biological functions.
To identify candidate regulatory sequence motifs for the genes upregulated and downregulated by HBL-1, we examined the sequence upstream of each ORF (maximum 1 kb) using the MEME program, which identifies over-represented sequences de novo.30 At least one and sometimes two significant motifs were found upstream of all of the 26 upregulated genes and the 53 downregulated genes, respectively (Fig. 2A-C and Tables S2-S4). A property of all three motifs is an A-enriched stretch, which bears a similarity to the characteristics of the known Drosophila Hb binding consensus site (Fig. 2D),31 suggesting that these motifs are likely to be direct targets of the HBL-1 transcription factor.
If a gene identified from the microarray analysis has an essential function in the hbl-1 regulation of developmental timing, one would expect that loss of such a gene's function might modulate the heterochronic phenotype of hbl-1 mutants, namely a precocious phenotype at the L4 stage, or a retarded phenotype at the adult stage. We therefore examined whether feeding RNAi for each of the identified genes enhances or suppresses the precocious alae formation of a loss-of-function mutant of hbl-1, hbl-1(ve18).15 We found that the precocious alae caused by loss of hbl-1 function was significantly suppressed by RNAi for sym-1 (Table 3), which encodes a secreted protein containing 15 contiguous leucine-rich repeats.32 As RNAi for other genes than sym-1 had no effect on the precocious alae formation of the hbl-1 mutant (data not shown), we focused on sym-1 for further analysis.
sym-1(RNAi) suppressed the precocious alae formation caused by hbl-1(ve18) as well as hbl-1(RNAi) with statistical significance (Table 3). We also confirmed that this suppression was specific to sym-1 RNAi and was not an off-target effect, as precocious alae formation in the hbl-1 mutant was also suppressed in the sym-1(mn601) mutant, a complete loss-of-function allele of sym-1,32 (Table 3). On the other hand, we did not detect any obvious modification of the bursting phenotype of let-7(n2853) mutant (Table S5).
Besides the hbl-1 function for inhibiting alae formation in the L4 stage, it has been known that hbl-1 plays a role in promoting seam cell differentiation in the adult stage.15,16 We also examined whether loss of sym-1 function modulated the adult alae phenotype in hbl-1(RNAi) animals that frequently display incomplete, gapped alae. We found that either sym-1(RNAi) or sym-1(mn601) enhanced, but did not suppress, the hbl-1(RNAi) adult alae phenotype, as a greater number of adults exhibited the gapped alae structures (Table 3). These results show that sym-1 is involved in controlling adult cell fate determination negatively in the L4 stage and positively in the adult stage, consistent with the known, dual role of HBL-1 in seam cell development.15,16
The microarray analysis showed a 2.67-fold higher expression of the sym-1 gene in the heat shock-treated L4 animals of hsp::hbl-1 than those of hsp::DsRed2 (Table 2), suggesting that HBL-1 positively regulates sym-1 expression. It is therefore expected that knocking down hbl-1 function would decrease the sym-1 expression level. This was found to be the case, as a quantitative RT-PCR analysis revealed 0.23-fold expression of the sym-1 gene in hbl-1(RNAi) L4 animals as compared to control RNAi animals fed by an empty vector pL4440. In addition, the fact that hbl-1 plays an indispensable role in regulating differentiation in seam cells15-17,21,33 led us to consider whether hbl-1 is involved in regulating sym-1 expression in seam cells. It has previously been shown that sym-1 is expressed in hypodermal cells including seam cells during larval stages.32 When nematodes carrying a translational sym-1::gfp transgene were fed on bacteria expressing a hbl-1 RNAi construct, the frequency of larvae expressing GFP in the seam cells significantly decreased (Table 4). All of these data imply that HBL-1 promotes sym-1 expression during development.
It is well known that the L/A switch in C. elegans requires proper regulation by heterochronic genes encoding transcription factors, such as hbl-1 and lin-29.3 In addition, large-scale analyses identifying let-7 targets have indicated a bias of let-7 toward regulating transcription factors in many different tissues.11 While a study identified downstream targets of the heterochronic transcription factor LIN-14 that are responsible for the L1/L2 transition,34 the transcription factors involved in the L/A switch have not been the subject of such research so far. In this study, we took advantage of a technology of conditionally-inducible versions of transcription factors, which have been applied for the identification of downstream targets of several transcription factors in Drosophila melanogaster,35 Arabidoposis thaliana,36 as well as C. elegans.22 While the genes we identified by microarray hybridization as hbl-1-responsive might include direct targets as well as indirect targets, this study demonstrates the power of such conditional expression tools to uncover the transcription network of the L/A switch. Similar molecular biological techniques can be applied to investigate downstream gene of other heterochronic transcription factors.
This study confirms that sym-1 gene expression is positively regulated by HBL-1. This idea is supported by data from a previous microarray analysis,37 which reveals that sym-1 expression was detected at an approximately 8-fold higher level in animals around the L3 molt compared to around the L4 molt (http://www.wormbase.org/db/microarray/results?name=Aff_C44H4.3;reference=WBPaper00004386). This temporal expression profile was similar to the expression level of HBL-1, as the protein level of HBL-1 is postulated to decrease through downregulation by let-7 miRNAs gradually from larval to adult stages.15,16 Besides HBL-1 regulation of sym-1 gene expression, we also show a genetic interaction between sym-1 and hbl-1, as loss of sym-1 function modulates the alae formation defects in hbl-1 mutants (Table 3). However, the functional relationship between sym-1 and hbl-1 is not simply explained by the regulation of sym-1 expression by HBL-1. Our results suggest that hbl-1 function is not only antagonized by sym-1 in the L4 stage but also is promoted by the same gene in the adult stage (Table 3). Therefore, it is likely that sym-1 may play roles in both promoting and inhibiting hbl-1-dependent adult differentiation programs. Thus sym-1 itself may not be a canonical heterochronic gene. This idea is consistent with our recent finding that the evolutionarily conserved nuclear receptor gene nhr-25 exhibits a complicated genetic function to control developmental timing (Hada K, Hasegawa H, Kanaho Y, Asahina M, Slack FJ and Niwa R, a submitted manuscript).
While the proper regulation of sym-1 expression appears critical for the control of the hbl-1-dependent adult differentiation program in seam cells, the actual physiological function of SYM-1 for adult determination is not revealed from our study, nor from previous studies.32,38 sym-1 was originally identified as a gene causing a synthetic lethal phenotype with loss-of-function mutations of mec-8, which encodes a regulator of alternative RNA splicing.32 mec-8 null mutants have defects in sensory neurons and body muscle attachment, but are generally viable and fertile. The phenotypes of the sym-1 single mutant are also essentially wild type, which is reproducibly consistent with our observation (Table 3). In contrast, mec-8; sym-1 embryos arrest during embryonic elongation and exhibit defects in the attachment of body muscle to the extracellular cuticle. In addition to embryonic expression, the sym-1 gene is also expressed in larval and adult stages.32,37 While an in vivo function of sym-1 has not been elucidated, a previous study demonstrates that sym-1 is expressed in seam cells and that the SYM-1 protein is a leucine-rich repeat protein secreted into C. elegans hypodermal cuticles throughout C. elegans development.32 We therefore speculate that sym-1 is involved in the creation of adult-specific cuticular structures, namely alae, but redundantly cooperates with other unknown gene(s) that might be regulated by HBL-1. It would therefore be important to examine which genes genetically interact with sym-1 and especially whether other HBL-1 downstream genes identified in this study are involved in hbl-1- and sym-1-dependent temporal regulation of alae formation.
It is noteworthy that sym-1 has also been identified as a potential downstream target of the heterochronic transcription factor LIN-14, which is essential for the developmental transition from L1 to L2 stages.34 These microarray analyses indicated that sym-1 transcripts are reduced in lin-14 loss-of-function mutants, suggesting that LIN-14 also positively regulates sym-1 expression. This tendency is consistent with the endogenous temporal change of sym-1 expression, as published microarray analysis showed that sym-1 is found to be more abundant in L1 larvae than in L2 larvae.37 In conjunction with data demonstrated in this study, in the future it would be intriguing to verify the interesting possibility that sym-1 is involved in regulating timing of developmental transitions at multiple time points in C. elegans.
It is quite likely that we have missed some hbl-1 targets in our microarray screens for hbl-1-responsive genes. For example, this study did not pick up Alzheimer's amyloid precursor protein-like-1 (apl-1) gene, whose expression is affected by HBL-1 in seam cells.26 It is conceivable that apl-1 is not an immediate-early downstream gene of hbl-1. Related to this point, it is possible that a certain concentration of HBL-1 in nuclei is critical for transcriptional regulation of a gene such as apl-1, as it has been proposed that at low concentrations Hb acts as an activator, while at high concentrations it acts as an inhibitor of the transcription of its regulatory targets.39 Alternatively, as a target might be expressed only in a subset of cells, the microarray analysis with RNAs extracted from the whole body of C. elegans might not be able to detect a transcriptional change in a small number of cells. Cell-type specific RNA selection methods that have been used for recent, improved microarray analysis40 would overcome this limitation for more precise study of developmental timing in the future.
C. elegans strains were grown at 20°C for all experiments unless indicated. The mutant strains used in this study were as follows: wild-type N2 Bristol, glp-4(bn2ts),27 hbl-1(ve18),15 hbl-1(mg285),16 let-7(n2853ts),7 sym-1(mn601)32 and Is[apl-1::gfp::unc-54].26
Heat shock-driven overexpression of Discosoma sp. DsRed2,41 (Clontech) and C. elegans hbl-1 genes was achieved using expression vectors pPD49.78 and pPD49.83 (Andrew Fire), which have the hsp16-2 and hsp16-41 small heat shock promoters, respectively, without any enhancer. The 2,949 bp hbl-1 ORF region was amplified by PCR with an expressed-sequence tag clone containing hbl-1 cDNA, yk568f2.5, which was a kind gift from Yuji Kohara. The following primers were used to amplify the hbl-1 ORF and attach NheI and EcoRV sites at 5′ and 3′ ends, respectively: HBLF-NHE (5′-GCT AGC ATG GTG CAA TCC GAT AGT CCA GAA G-3′) and HBL-RV (5′-GAT ATC TTA TTG GTG TCT GGC TTG GTA CAT G-3′). The PCR product was subcloned into pCR2.1-TOPO plasmid (Invitrogen). The pCR2.1-TOPO containing the hbl-1 ORF was digested with NheI and EcoRI, followed by a ligation with NheI/EcoRI-digested pPD49.78 and pPD49.83. A DNA fragment containing an ORF region of DsRed2 gene was amplified by PCR from the tph-1::dsRed2 construct (a gift from Jessica Tanis and Michael Koelle) with the following primers: NHE-RED+ (5′-CTA GCT AGC ATG GCC TCC TCC GAG AA-3′) and KPN-RED-(5′-GAC TGG TAC CTC ACA GGA ACA GGT GGT-3′). The PCR product was sub-cloned into the pCR4-TOPO plasmid (Invitrogen). pCR4-TOPO containing the DsRed2 ORF was digested with NheI and KpnI, followed by a ligation with NheI/KpnI-digested pPD49.78 and pPD49.83. The sequences of all of the resulting cloned plasmids were confirmed by sequencing. The sym-1::gfp translational fusion construct32 was kindly provided by Robert K. Herman. DNA transformation in wild type was achieved by inserting plasmid DNA into the distal arm of the hermaphrodite gonad as described.42,43 For establishing transgenic lines carrying extrachromosomal arrays of each of DsRed2 and hbl-1, a mixture of 50 ng/μl of pPD49.78-based plasmid and 50 ng/μl pPD49.83-based plasmid was injected into N2 animals with myo-2::gfp::unc-54 plasmid (a gift from Michael Stern) as a co-injection marker at 5 ng/μl. The sym-1::gfp construct was injected into N2 animals at ~10 ng/μl with rol-6(su1006) as a co-injection marker at 80 ng/μl. Chromosomal integration of extrachromosomal transgenes was performed using trimethylpsoralen (Sigma-Aldrich) and subsequent UV-irradiation, essentially as described (http://18.104.22.168/tmp_uv_integration.htm). Two strains with stably integrated chromosomal transgenes were isolated. These integrated males were backcrossed with wild-type hermaphrodites at least 5 times to exclude the rare possibility of random mutations.
RNA extraction and DNase treatment were performed as described.26 Single-stranded cDNA was synthesized from total RNA (1 μg for each sample) with oligo(dT) primers and RETROscript (Applied Biosystems) according to the manufacture's instructions. The primers used were: F13D11(23) (5′-CCC GAT GAG CAA CGA CAG TGC AAT GGA G-3′) and F13D11(24) (5′-GGT GTG TGG AAT CTC ATA TGA CTG TCC-3′) for hbl-1; SYM-1F (5′-ATG CTT CTC CGA CTT TGT GTG GC-3′) and SYM-1R (5′-TCT CTG ACA GAG AAA GCT CTT GCA TG-3′) for sym-1. These primers were chosen such that any remaining potential genomic DNA contamination would have resulted in larger amplification products. These products would have been easily distinguishable from RT products on agarose gels; however, no contaminating bands were observed in any sample. The control primers against the eft-2 housekeeping gene were as previously described.11 PCR conditions were: 94°C for 3 min; one cycle, followed by 94°C for 30 sec; 55°C for 30 sec; 72°C for 40 sec; 30 cycles. Equal volumes of the PCR reactions were analyzed by agarose gel electrophoresis. Quantitative RT-PCR was performed essentially as described.44
Is[hsp::DsRed2 myo-2::gfp];glp-4(bn2ts) and Is[hsp::hbl-1 myo-2::gfp];glp-4(bn2ts) transgenic lines were staged by bleaching gravid adults to collect eggs, which were then hatched in M9 solution in the absence of food. Starved L1 larvae were cultured with food (bacterial strain OP50) at 25°C and harvested 32 hours later. The animals, which were early L4 larvae, were heat shocked at 32°C for 90 min, then recovered at 25°C for 60 min. After rotating the animals in M9 buffer to wash off bacteria, washed animals were transferred into an Eppendorf tube. Total RNA from each sample was isolated using Trizol (Invitrogen). About 1 μg of the total RNAs were amplified with the Ambion MessageAmp™ II aRNA Amplification Kit (Applied Biosystems) according to the manufacturer's protocol. Three independent samples of each overexpressor were collected. Hybridization and detection for microarray analysis were performed as described.45 All data have been deposited in GEO. The resulting data were analyzed using both an average fold-difference and a Z-test [Z = (observed − expected)/S.E.] with correction for multiple testing. A moderate correction for multiple testing (~17,600 genes) was performed by multiplying the calculated p-value by 10,000. After this correction, all genes with up or downregulation greater than two-fold, p < 0.05 in any given mutant were selected and shown in Table 2.
To identify candidate regulatory sequences in the 5′ non-coding regions of target genes, the online program MEME (Mutiple Em for Motif Elicitation)30 was applied to sequences upstream taken from the start codon of each target gene to the neighboring gene, up to 1 kb. The motif sequence logo was created using the WebLogo3 program.46
Gene knockdown was achieved through RNAi by feeding as described.47-49 Synchronized populations of L1 larvae were fed bacteria expressing dsRNA corresponding to the target genes. RNAi plasmids for 76 out of the 79 identified genes were available (Table S6) in Julie Ahringer's RNAi library.48,49 We used lin-29 RNAi as a positive control for the RNAi experiments. We confirmed that lin-29(RNAi) adults almost completely suppressed the precocious alae phenotype in hbl-1(If) mutants (98%; n = 25) as expected.
Microscopy images were acquired using an Axioplan II microscope (Carl Zeiss) equipped with an AxioCam MRm CCD camera (Carl Zeiss).
We are grateful to Andrew Fire, Michael Stern, Michael Koelle, Yuji Kohara, Robert K. Herman, Jessica Tanis and Caenorhabditis Genetics Center for stocks and reagents. We also deeply thank Yusuke Kato, Osamu Numata, Katsuo Furukubo-Tokunaga, Yoshihiro Shiraiwa, and their lab members for allowing K.H., K.M. and R.N. to share their space and equipment. R.N. was supported by a fellowship from the Human Frontier Science Program Organization. This work was supported in part by Special Coordination Funds for Promoting Science and Technology of the Ministry of Education, Culture, Sports, Science of the Japanese government. F.J.S. was supported by a grant from the NIH (GM64701).
Supplementary materials can be found at: www.landesbioscience.com/supplement/NiwaCC8-24-Sup.pdf