Wnt signaling and CEH-22/tinman/Nkx2.5 specify a stem cell niche in C. elegans

Summary

Wnt signaling regulates many aspects of metazoan development, including stem cells [1-3]. In C. elegans, Wnt/MAPK signaling controls asymmetric divisions [4, 5]. A recent model proposes that the POP-1/TCF DNA binding protein works together with SYS-1/β-catenin to activate transcription of target genes in response to Wnt/MAPK signaling (Figure 1A) [6]. The somatic gonadal precursor (SGP) divides asymmetrically to generate distal and proximal daughters of distinct fates: only its distal daughter generates a distal tip cell (DTC), which is required for stem cell maintenance (Figure 1B) [7]. No DTCs are produced in the absence of POP-1/TCF or SYS-1/β-catenin, and extra DTCs are made upon overexpression of SYS-1/β-catenin [6, 8, 9]. Here we report that POP-1/TCF and SYS-1/β-catenin directly activate transcription of ceh-22/nkx2.5 isoforms in SGP distal daughters, a finding that confirms the proposed model of Wnt/MAPK signaling. In addition, we demonstrate that the CEH-22/Nkx2.5 homeodomain transcription factor is a key regulator of DTC specification. We speculate that these conserved molecular regulators of the DTC niche in nematodes may provide insight into specification of stem cell niches more broadly.

Results and Discussion

When Wnt/MAPK signaling is compromised, both SGP daughters adopt proximal cell fates, the Sys (for symmetric sister) phenotype [9, 10]. sys-3(q632)mutants display typical Sys defects [10]. Whereas wild-type hermaphrodites possess two DTCs and two gonadal arms, most sys-3(q632) mutants lack one or both DTCs and also lack one or both gonadal arms (Fig. 1D, top line) [10]. In hermaphrodites, the presence of a gonadal arm serves as a simple morphological readout for the presence of a DTC, because DTCs control formation of the elongate arm (“leader function”) as well as germline stem cell maintenance (“niche function”) [7]. Wild-type males have two DTCs that function solely to maintain germline stem cells [7], but most sys-3(q632) males have no DTCs and consequently possess little or no germ line. Genetic analyses indicated that sys-3 acts in parallel to or downstream of POP-1/TCF [10].

Figure 1.

Control of SGP asymmetric division and molecular identification of q632 as an unusual ceh-22 allele. (A) Model for control of target genes by POP-1/TCF and SYS-1/β-catenin. In this model, the amount of available SYS-1 determines whether POP-1 functions as a transcriptional activator or repressor. In the distal daughter (yellow), the abundance of nuclear POP-1 is reduced so that most POP-1 is associated with SYS-1 and activates transcription. In the proximal daughter (blue), abundance of nuclear POP-1 is high, and therefore most POP-1 cannot associate with SYS-1 and cannot activate transcription.(B) SGP asymmetric division. Top diagram, gonadal primordium containing SGPs (Z1 and Z4) at poles and primordial germ cells centrally. Bottom diagram, SGP daughters: yellow, distal daughters; blue, proximal daughters. D, distal; P, proximal. The distal daughter either differentiates as a distal tip cell (males) or generates a distal tip cell in its next division (hermaphrodites). The male DTC (mDTC) functions solely as a stem cell niche [7]; the hermaphrodite DTC has “leader” function in addition to niche function; therefore, the hermaphrodite DTC not only controls germline stem cells, but also controls formation of the elongate U-shaped gonadal arm [7].(C) ceh-22 gene structure. Boxes, exons; white, non-coding sequence; grey, coding sequence; black, coding for homeodomain (HD). M, predicted initiation codon for each isoform. Bent lines, introns. The regions deleted by cc8266 and q632 are shown below. The extent of the q632 deletion is represented in red in each isoform. PBS, POP-1 binding site. SL1, trans-spliced leader. D. Both the first intron and the ceh-22 coding region and POP-1 binding sites are required for rescue of q632 mutants. Transgenes (see Supplemental Methods for details), conventions as in Fig. 2A; numbers of DTC were scored by counting numbers of elongated gonadal arms. The ceh-22(genomic) fragment contained ∼2 kb of 5’ flanking sequence, all exons and introns plus 90 bp of 3’ flanking sequence.

We cloned the sys-3 gene to further understand the molecular regulation of the SGP asymmetric division. To this end, we first mapped sys-3(q632) to a site near the ceh-22 locus (see Supplemental Methods). Three lines of evidence suggest that sys-3(q632) is allelic to ceh-22. First, sys-3(q632) and the ceh-22(cc8266) internal deletion [11] (Fig. 1C) failed to complement each other for the Sys phenotype. Second, reduction of ceh-22 by RNA interference (RNAi) resulted in loss of gonadal arms (20% 0 arm, 31% 1 arm, 49% 2 arms, n=64), a typical Sys defect. Finally, ceh-22 genomic DNA rescued the q632 Sys phenotype: most q632 mutants carrying genomic ceh-22 made two DTCs, two gonadal arms and were fertile (Fig. 1D, ceh-22(genomic)). To identify the q632 molecular lesion, we sequenced ceh-22 genomic DNA from sys-3(q632) homozygotes. We found one lesion, a deletion of 400 bp within the first intron (Fig. 1C). To explore the importance of the DNA deleted by sys-3(q632), we generated a transgene in which the first intron was deleted, but the ceh-22 coding region was intact (Fig. 1D, ceh-22(Δ1stIntron)). This transgene failed to rescue q632 (Fig. 1D), a result consistent with a previous study showing that ceh-22 cDNA rescued pharyngeal, but not gonadal, defects of cc8266 mutants [12]. We conclude that the first intron of ceh-22 is critical for gonadal development.

The ceh-22 gene generates transcripts of at least two sizes [13]. The exon/intron composition of the longer mRNA, which we dub ceh-22a, is well established [13] (Fig. 1C). Our finding that the first intron is critical for gonadal development suggested to us that the first intron might act as a promoter to drive transcription of a shorter ceh-22 mRNA; by this model, the shorter mRNA is predicted to lack the first exon. To test this idea, we performed RT-PCR with primers designed to identify ceh-22 cDNAs lacking the first exon and found two ceh-22 isoforms, which we call ceh-22b and ceh-22c (Fig. 1C; see Methods). The ceh-22b transcript contains exons 2 to 7 and carries SL1 trans-spliced directly to exon 2; its first methionine codon in-frame with the homeodomain occurs in exon 4 (Fig. 1C). The ceh-22c transcript includes a fragment of the first intron plus exons 2 to 7; this isoform harbors a methionine codon and potential initiation codon within the first intron that occurs in-frame with the ceh-22 coding region of the second exon (Fig. 1C, blue M). We did not detect transplicing of SL1 to the ceh-22c isoform. To ask whether ceh-22b and ceh-22c mRNAs are functional, we generated two transgenes. The first, called ceh-22b(genomic) (Fig. 1D), contained the first intron plus exons 2 through 7 of the ceh-22 coding region, but lacked exon 1 and other upstream sequences. The second, called ceh-22b(cDNA), was similar, but also lacked all introns except the first one (Fig. 1D). Both ceh-22b(genomic) and ceh-22b(cDNA) transgenes rescued q632 Sys defects as efficiently as the full genomic region (Fig. 1D). Importantly, the ceh-22 coding sequence was required for rescue: a ceh-22b(ΔC) transgene lacking exons 6 and 7 lost rescuing activity (Fig. 1D). To ask if the ceh-22c-specific methionine in the first intron might be employed, we mutated the methionine codon to a stop codon in the ceh-22c(mutMet) transgene; this mutant diminished, but did not abolish, rescue (Fig 1D, ceh-22c(mutMet)). Therefore, it seems likely that both ceh-22b and ceh-22c isoforms are used. For simplicity, we refer to both isoforms collectively as ceh-22b.

To test whether the first ceh-22 intron has promoter activity and to learn whether it drives expression in the somatic gonad, we created ceh-22b::VENUS, an integrated reporter transgene that links the ceh-22 first intron to the Venus coding sequence (see Supplemental Methods); Venus is a bright variant of YFP [14]. Two independent lines displayed the same expression pattern (Fig. 2 and Supplemental Methods). In first stage larvae (L1) of both sexes, ceh-22b::VENUS expression was not detected in SGPs at hatching, but became visible midway through the first larval stage (L1) (Fig. 2A,B; data not shown). We note that Z1 and Z4 refer to the anterior and posterior SGPs, respectively (Fig. 1A). After the SGP divided, the intensity of the reporter began to increase in distal SGP daughters (Z1.a and Z4.p) and began to diminish from proximal SGP daughters (Z1.p and Z4.a) (Fig. 2A-C). In hermaphrodites, both progeny of the distal SGP daughter retained robust ceh-22b::VENUS expression through L2 or early L3 (Fig. 2A,C). In males, the distal SGP daughter, which does not divide further, retained strong expression until L3 (Fig. 2B); the expression decreased during L4. We conclude that the first intron can function as a promoter to drive expression asymmetrically in the SGP lineage. We term this intronic region the ceh-22b promoter.

Figure 2.

ceh-22b::VENUS is expressed asymmetrically in the SGP lineage. In both (A) and (B), each row shows immunofluorescent (left), Nomarski (middle) and merged (right) images. Z1, anterior SGP; Z4, posterior SGP. Z1.a, distal Z1 daughter; Z1.p, proximal Z1 daughter. Images were taken at different exposures, so levels of expression are not comparable from stage to stage. (A) ceh-22b::VENUS expression in hermaphrodites (XX) gonads. During L1, expression is detected in both SGPs, and remains high in the distal SGP daughter (Z1.a) and is reduced in the proximal SGP daughter (Z1.p) (arrow). Expression is increased and maintained in the distal SGP daughter and its progeny, Z1.aa and Z1.ap, throughout L2, but disappears from the proximal SGP daughter and its descendants. (B) ceh-22b::VENUS expression in male (XO) gonads. Expression was detected in the distal SGP daughters, but is not visible in proximal SGP daughters at this exposure. (C) Summary of ceh-22b::VENUS expression in L1 and L2 SGP lineage. Only Z1 lineage shown, although Z4 lineage was similar. Each vertical line represents a cell and each horizontal line represents a division. Black line, no detectable reporter expression; green line, reporter is expressed at level represented by line thickness: thin lines indicate low expression and thick lines indicate high expression; dotted line, diminished expression. (D) ceh-22b::VENUS expression in non-gonadal tissues. Top, expression in pharynx (bracket), head neuron (arrow) and ventral nerve cord (arrowheads). Middle, lower exposure to show distinct areas of pharyngeal expression. Bottom, expression in posterior intestinal cells.

The ceh-22b::VENUS reporter was also expressed in the pharynx, intestine, and ventral nerve cord as well as in unidentified neurons in the head and tail (Fig. 2D, not shown). Expression in the pharynx and intestine was sustained throughout larval development into adulthood; expression in the ventral nerve cord was visible until L3. The significance of the nongonadal expression remains unknown, because no obvious nongonadal defect was seen in q632 mutants.

The finding that ceh-22b::VENUS is expressed more strongly in distal than proximal SGP daughters suggested that ceh-22 transcription might be controlled by Wnt/MAPK signaling. To ask if ceh-22b might be a direct target of transcriptional activation by POP-1/TCF and SYS-1/β-catenin, we first tested POP-1 binding to the ceh-22b promoter and identified two POP-1/TCF binding sites (PBS1 and PBS2) by a combination of sequence analysis and DNA footprinting (Fig. 3A; see Methods). Both sites had a similar sequence and a comparable POP-1-binding affinity to that of the consensus TCF binding site (TTCAAAG) (Fig. 3A and not shown) [15, 16]. Remarkably, both sites were located within the q632 deletion (Fig. 1C). Using a gel electrophoretic mobility assay, recombinant POP-1/TCF bound specifically to both sites, but not to a mutated probe in which the sequence of the core TCF consensus element had been altered (Fig. 3A,B).

Figure 3.

ceh-22b is a direct target of POP-1 and SYS-1 transcriptional activation. (A) Sequence of POP-1 binding sites (PBS) in ceh-22b promoter. Capital letters indicate core binding region with sequence similarity to canonical TCF binding site of TTCAAAG. Red nucleotides show sequences mutated to abolish POP-1 binding and used as controls in the following experiments. (B) Electrophoretic mobility assay. Above, diagram of probe; each is a 254 bp DNA fragment containing both PBS1 and PBS2 (1 and 2). Mutant sites indicated by asterisks. For each probe, 0, 0.3, 1, 3, 10, or 30 ng of recombinant POP-1 was added to the reaction. Arrow: unbound probes. (C) Transcriptional activation in NCI-H28 cells. In each experiment, the luciferase coding region was driven by one of several promoters (see key in figure and text). Cells were transfected with reporters either alone (left experiments), with a plasmid expressing POP-1 (middle experiments), or with plasmids expressing POP-1 and SYS-1 (right experiments) (see Methods). Luciferase activity in the absence of POP-1 and SYS-1 was set at 1, and fold induction measured in other experiments. Results are the average of three independent experiments, each done in duplicate. (D) Comparison of ceh-22b::VENUS and ceh-22b(mutPBS)::VENUS expression in hermaphrodite and male SGPs and their descendants. Conventions as in Fig. 3C. (E) Comparison of ceh-22b::VENUS in wild-type and mutant SGP daughters. The same transgene was used in all three genotypes, and all animals were at the same stage, just after the division of the SGP distal daughter, Z1.a. All images were obtained with the same exposure and were treated identically. In sys-1 and pop-1 mutants, VENUS levels were similar in distal and proximal SGP descendents. The two daughters of Z1.a are indicated by cell division diagram, and one daughter of Z1.p is marked by white arrow. Wild-type expression is shown in the top row for comparison; in wild-type gonads, expression in the distal SGP lineage is so bright that any faint expression in proximal SGP lineage could not be determined at this exposure.

To assay the function of the POP-1/TCF binding sites in the ceh-22b promoter, we first used a reporter assay in tissue culture cells. A previous study showed that POP-1/TCF and SYS-1/β-catenin activate transcription from a promoter harboring eight copies of the consensus TCF binding site upstream of the luciferase coding region (8xTOPFlash) [6]. Here we replaced the 8xTOPFlash promoter with either the wild-type ceh-22b promoter or one of two control promoters. The ceh-22b(q632)::luciferase reporter harbors the q632 deletion and ceh-22b(mutPBS)::luciferase carries mutated versions of PBS1 and PBS2 (see Supplemental Methods). POP-1/TCF alone did not activate transcription from any of the reporters (Fig. 3C, middle), but POP-1/TCF and SYS-1/β-catenin together enhanced ceh-22b::luciferase expression by 3-5 fold, a level comparable to that of the 8xTOPFlash reporter transgene (Fig. 3C, right). Furthermore, POP-1/TCF and SYS-1/β-catenin did not enhance expression of either ceh-22b(mutPBS)::luciferase or ceh-22b(q632)::luciferase (Fig. 3C). We conclude that POP-1/TCF and SYS-1/β-catenin transcriptionally activate the ceh-22b promoter via the PBS1 and PBS2 sites.

We next asked if POP-1/TCF and SYS-1/β-catenin control expression of the ceh-22b promoter in nematodes. To this end, we compared expression of a ceh-22b::VENUS reporter transgene to that of a mutated transgene, ceh-22b(mutPBS)::VENUS, which carries mutations in PBS1 and PBS2, but otherwise is identical to ceh-22b::VENUS. Our results are summarized in Fig. 3D. In non-gonadal tissues, both transgenes expressed similarly in pharynx and neurons, but ceh-22b(mutPBS)::VENUS was not expressed in the intestine (not shown). In the SGPs, both transgenes were initially expressed midway through L1 as normal. Therefore, PBS1 and PBS2 have no apparent effect on initiation of ceh-22b expression in SGPs. By contrast, expression of the two transgenes was dramatically different in SGP daughters. Whereas ceh-22b::VENUS expression intensified in distal SGP daughters and was maintained at a high level in the distal SGP lineage through L2, ceh-22b(mutPBS)::VENUS was expressed similarly only at a low level in both distal and proximal SGP daughters, and usually disappeared by the time the SGP daughters began their division (65%, n=14). No ceh-22b(mutPBS)::VENUS expression was seen in the late L1 gonads (n=45). Therefore, PBS1 and PBS2 appear to be required specifically for the robust and sustained ceh-22b expression in the distal SGP daughters and their progeny as well for intestinal expression.

We next compared expression of ceh-22b::VENUS in wild-type animals to that in pop-1 and sys-1 mutants. Specifically, we employed pop-1(q645) and sys-1(q544) mutants, which have fully penetrant Sys gonadal defects [8, 9]. In both mutants, ceh-22b::VENUS was initially expressed in the SGPs at a low level as normal; however, that expression did not intensify in distal SGP daughters and was not maintained in the distal SGP lineage (pop-1(q645), n=30; sys-1(q544), n=14) (Fig. 3E). A similar effect was seen after sys-1(RNAi) (data not shown). We conclude that POP-1/TCF and SYS-1/β-catenin are both required for the robust and sustained ceh-22b::VENUS expression in the distal SGP lineage.

Given the fact that the ceh-22b promoter lacking POP-1 binding sites was able to drive low level expression in the SGPs, we wondered if the sites were critical for q632 rescue. We therefore compared the rescuing activities of ceh-22b(genomic) transgenes with wild-type or mutated POP-1 binding sites. Whereas ceh-22b(genomic) efficiently rescued q632 mutants, the ceh-22b(mutPBS,genomic) transgene failed to rescue q632 mutants (Fig. 1C, ceh-22b(mutPBS,genomic)). Therefore, the POP-1 binding sites in the ceh-22b promoter are indeed crucial and the low level of expression that is POP-1-independent does not appear to be sufficient for rescue.

Our experiments demonstrate that POP-1/TCF and SYS-1/β-catenin control ceh-22b expression via POP-1 binding elements, and that POP-1/TCF and SYS-1/β-catenin achieve a high level of ceh-22b expression specifically in one daughter of an asymmetric division. These results provide the first example of a direct downstream target controlled by both POP-1/TCF and SYS-1/β-catenin, and confirm the hypothesis that POP-1/TCF and SYS-1/β-catenin can transcriptionally activate target genes in nuclei with lowered POP-1/TCF abundance (Fig. 1A) [6]. The idea that POP-1/TCF can transcriptionally activate target genes, rather than simply derepressing them in cells with lowered nuclear levels of POP-1, has also received support from experiments in the early embryo [17, 18].

The model depicted in Fig. 1A also predicts that POP-1/TCF represses ceh-22b expression in proximal SGP daughters. We have not been able to see that POP-1/TCF repression. The VENUS reporter remains detectable at a low level in the proximal SGP daughters (Fig 3A,C), but that expression could either reflect perdurance of the reporter protein or a low level of transcription. VENUS disappears from the proximal SGP descendants at about the same stage in animals carrying either ceh-22b::VENUS (2 independent lines) or ceh-22b(mutPBS)::VENUS (3 independent lines), but a subtle difference might have been missed.

Although repression of ceh-22b may occur in the proximal daughters of SGP, one should note that POP-1 loss-of-function mutations have no effect on SGP proximal daughter fate [9], suggesting that POP-1 repression of target genes is not critical for proximal fate determination. This contrasts with EMS asymmetric divisions in which the repression of E-specific genes by high nuclear levels of POP-1/TCF is critical for the MS fate [17, 19, 20]. Therefore, both activation and repression of target genes by POP-1 are critical for determination of the E as well as MS fates [17, 19, 20]. We suggest that this difference between SGP and EMS divisions might be determined by strength of promoter activity of POP-1 target genes.

The control of ceh-22 by POP-1/TCF and SYS-1/β-catenin by Wnt/MAPK signaling has intriguing similarities with the control of its vertebrate homolog, nkx2.5, by Wnt signaling. Thus, Wnt signaling is required for specification of cardiac progenitors, and the effect of Wnt signaling is commonly assayed by Nkx2.5 expression [21-23]. Therefore, a regulatory link between Wnt signaling and Nkx2.5 transcription factors has been conserved. In C. elegans, that link is direct, but in vertebrates, a direct link has not been demonstrated to date.

The identification of ceh-22/nkx2.5 as a target of Wnt signaling suggested that this homeodomain transcription factor might be an essential regulator in the specification of the distal tip cell fate. Indeed, ceh-22(q632) loss-of-function mutants fail to make DTCs [10]. To ask whether ceh-22b is sufficient to specify DTCs, we used the heat shock promoter to drive the ectopic expression of CEH-22B (Fig. 4A). As a marker of DTCs, we employed lag-2::GFP [24]. Without heat shock, all animals carrying the hs::CEH-22b transgene survived to adulthood, and all hermaphrodites and males contained two distal tip cells, the number typical of wild-type animals (Fig. 4B,D). When hs::CEH-22b transgenic animals were heat-shocked soon after the SGP divided (see Methods), over half of the surviving adults possessed extra DTCs (XX, 14/19; XO, 6/8) (Fig. 4C,E). In hermaphrodites, extra DTCs led to formation of extra gonadal arms (Fig. 4C). A vulva was missing in half of the hermaphrodites that had four total DTCs (n=8), indicating loss of the anchor cell which is normally produced by a proximal SGP daughter [25, 26]. In males, extra DTCs were always found in a disorganized gonad and a linker cell was usually not observed (Fig. 4E), indicating defects in proximal SGP daughters. We conclude that CEH-22B is sufficient, when overexpressed, to specify the proximal daughter of SGP to the DTC fate in both hermaphrodites and males.

Figure 4.

CEH-22B is sufficient to specify distal cell fate.Animals carrying hs::ceh-22b were heat-shocked or not. The number of distal tip cells was then scored by morphology and expression of the distal tip cell marker lag-2::GFP. (A) Structure of the hs::ceh-22b transgene. (B) Adult hermaphrodite without heat shock. Two distal tip cells are made as normal. (C) Adult hermaphrodite after heat shock. In the animal shown, four distal tip cells are made. (D) Male without heat shock; two distal tip cells and a gonad arm are made. (E) Male after heat shock. In the animal shown, four distal tip cells are made. The gonad is disorganized; no gonad arm is formed. Images showing distal tip cells are composed from multiple pictures taken from different focal planes. Nomarski pictures in (D) and (E) show one focal plane of the gonad. Gonad arms are delineated by dotted yellow lines. DTCs are marked with yellow arrows.

We have found that Wnt signaling and ceh-22/nkx2.5 work together to specify the DTC fate. The common function of DTCs in hermaphrodites and males is that of a stem cell niche [7]. Wnt signaling has emerged as a key regulator of stem cells in many tissues and in many organisms, and that role relies on transcriptional activation by TCF/LEF and β-catenin transcription factors [27]. Our work suggests that one role of Wnt signaling may be to control the stem cell niche. A similar suggestion was recently put forward with respect to osteoblasts, which provide a niche for hematopoietic stem cells [28-30]. CEH-22/Nkx2.5 and its homologs have not previously been implicated in the control of stem cells. Indeed, the fly and vertebrate homologs, which are called tinman and Nkx2.5 respectively, are best known for their roles in heart specification and differentiation [31]. Nematodes have no heart, but CEH-22 controls development of the rhythmically contracting musculature of the pharynx [11], and zebrafish Nkx2.5 can functionally replace CEH-22 [12]. Therefore, the CEH-22/Nkx2.5 class of homeodomain transcription factors has broadly conserved functions in animal development.

A remaining question is whether CEH-22 control of the DTC fate reflects a conserved role for this class of homeodomain transcription factors in regulating stem cell niches. Mouse mutants deleted for Nkx2.5 die with a broad spectrum of defects, including severe defects in vasculogenesis and angiogenesis as well as hematopoiesis in the yolk sac [32]. Intriguingly, endothelial cells appear to function as stem cell niches [33-35]. It is tempting to speculate that the severe vasculature defects in Nkx2.5 mutants may reflect some role of this conserved regulator in control of a vertebrate niche, much as CEH-22 controls the DTC. Two important challenges for the future are to learn how CEH-22 specifies the DTC niche in C. elegans and to learn whether its homologs specify an analogous stem cell niche in flies and vertebrates.

Methods

Strains and genetics

Standard protocols were used for culturing C. elegans strains. Strains were derived from the Bristol strain N2 and maintained at 20°C unless otherwise noted. The following mutations were used for this work: LG I: sys-1(q544) [8], pop-1(q645) [9]. LG V: ceh-22(cc8266) [11], sys-3(q632) [10], him-5(e1490). The integrated transgene qIs56 [lag-2::GFP] was used.

Rescue experiments

To rescue q632, different ceh-22 DNAs (20 ng/μl) (Fig. 2, below) along with the coelomocyte marker unc-122::GFP (20 ng/μl) were injected into q632 homozygote animals. Numbers of gonad arms were scored in animals carrying the transgenes. For each ceh-22 DNA, at least three independent transgenic lines were analyzed.

RT-PCR

RNA (1 μg) from mixed staged worms was reverse transcribed using Oligo-dT primer and Superscript II reverse transcriptase (Invitrogen). ceh-22b was amplified by primers GGTTTAATTACCCAAGTTTGAG and GAGAAACGAGATGTATTCTGGGA. The 5’ primer anneals to the SL-1 splice leader. The 3’ primer anneals to the ceh-22 3’ UTR corresponding to nucleotides 20275 to 20297 of cosmid F29F11 (Genbank accession no. Z73974). ceh-22c was amplified by primers TGTCCGACTCCTTCACATTTCACC and GAGAAACGAGATGTATTCTGGGA. The 5’ primer anneals to the first intron of ceh-22 corresponding to nucleotides 18698 to 18721 of cosmid F29F11. The 3’ primer is the same used to amplify ceh-22b.

Transgenic animals

To make ceh-22b::VENUS and ceh-22b(mutPBS)::VENUS transgenic animals N2 (wild-type) worms were injected with ceh-22::VENUS (40 ng/μl) or ceh-22(mutPBS)::VENUS (50 ng/μl) together with influenza viral DNA (70 ng/μl). The transgenes were integrated by γ-irradiation. Animals having integrated transgenes were out-crossed five times. Male animals were obtained by crossing. To make hs::ceh-22b transgenic animals JK3131 (qIs56 him-5(e1490)) worms were injected with hs::ceh-22b (30 ng/μl) and a coinjection marker (ttx-3::DsRED, a gift from Josh Kaplan) (30 ng/μl). The transgene was maintained as extra-chromosomal arrays.

Identifying POP-1 binding sites in the ceh-22b promoter

We found several sites similar to the consensus TCF binding sites (TTCAAAG) in the ceh-22b promoter by scanning the sequence. To test whether POP-1 binds the ceh-22b promoter specifically, we performed gel electrophoretic mobility assays. The histidine-tagged HMG-box DNA-binding domain of POP-1 was purified from E. coli [15]. Four overlapping fragments encompassing the first intron were used as probes. To identify specific binding, a DNA fragment containing six copies of consensus TCF binding site (6×TOP) was used as a positive control; a DNA fragment containing eight copies of a mutated TCF binding site (8×FOP) was used as a negative control. A typical 20 μl binding reaction contained 100 fmol DNA probe, 0.1 to 30 ng purified POP-1, 10 ng/μl poly dI-dC, and 100 ng BSA in 1x buffer (20 mM HEPES pH7.6, 40 mM KCl, 1 mM MgCl₂, 1 mM DTT, 1 mM EDTA, 10% glycerol). The binding reactions were performed at 4°C for 20 min and separated by 4% native poly-acrylamide gel electrophoresis.

Two potential POP-1 binding sites were identified in a 254 bp fragment (nucleotides 18468 to 18721 of F29F11). To fine map the POP-1 binding site, we performed DNaseI footprinting using Core Footprint System (Promega). POP-1 protected two stretches of ∼ 20 nt, each including a predicted POP-1 binding site in the assay (not shown).

Luciferase reporter assay

NCI-H28 cells (1×10⁵) were transfected with 500 ng of luciferase reporters, 40 ng of TK-Renilla luciferase plasmid, 0 or 1 μg of POP-1 expression plasmid, and 0 or 1 μg of SYS-1 expression plasmid using Lipofectamine 2000 reagent (Invitrogen). Luciferase activities were measured using Dual luciferase system (Promega). Transfection efficiencies were normalized by Renilla luciferase activities.

Heat-shock

To test for phenotypes caused by overexpression of ceh-22b, L1 larvae (hs::ceh-22b qIs56 him-5(e1490)) at about 8.5 hour after hatching (25°C) were subjected to a 60 min heat shock (33°C). At the time of heat-shock, Z1 and Z4 had just divided in most of the animals. This treatment showed toxicity to the animals. 70% of the animals carrying the transgene arrested at L1 (n=106); the rest continued to develop to adulthood.