In Caenorhabditis elegans, the terminal differentiation of the hypodermal cells occurs at the larval-to-adult molt, and is characterized in part by the formation of a morphologically distinct adult cuticle. The timing of this event is controlled by a pathway of heterochronic genes that includes the relatively direct regulatory gene, lin-29, and upstream genes lin-4, lin-14 and lin-28. Using northern analysis to detect endogenous collagen mRNA levels and collagen/lacZ reporter constructs to monitor collagen transcriptional activity, we show that the stage-specific switch from larval cuticle to adult cuticle correlates with the transcriptional activation of adult-specific collagen genes and repression of larval-specific collagen genes. Heterochronic mutations that cause precocious formation of adult cuticle also cause precocious transcription of the adult-specific collagen genes, col-7 and col-19; heterochronic mutations that prevent the switch to adult cuticle cause continued expression of the larval collagen gene, col-17, in adults and prevent adult-specific activation of col-7 or col-19. A 235 bp segment of col-19 5′ sequences is sufficient to direct the adult-specific expression of a col-19/lacZ reporter gene in hypodermal cells. These findings indicate that the heterochronic gene pathway regulates the timing of hypodermal cell terminal differentiation by regulating larval- and adult-specific gene expression, perhaps by the direct action of lin-29.

A genetic regulatory system controlling the timing of diverse postembryonic developmental events in C. elegans has been identified by mutations in the heterochronic genes lin-4, lin-14, lin-28 and lin-29 (Chalfie et al., 1981; Ambros and Horvitz, 1984). The events controlled by this regulatory pathway include the expression of stage-specific patterns of cell division (Ambros and Horvitz, 1984), dauer larva developmental arrest (Liu and Ambros, 1989), progress through the cell cycle (Ambros and Horvitz, 1984; Euling and Ambros, unpublished) and adult-specific terminal differentiation of hypodermal cells (Ambros, 1989). The action of the heterochronic genes in controlling these events must ultimately affect stage-specific gene expression in particular cells. An essential step in analyzing how heterochronic genes specify the timing of specific developmental programs is the identification of target genes whose stage-specific expression is affected by heterochronic mutations.

Previous studies have focused on the effects of heterochronic genes on a stage-specific event named the larva-to-adult switch or ‘L/A switch’ (Ambros, 1989). In the wild type, lateral hypodermal stem cells (or ‘seam cells’) divide at least once in each larval stage and synthesize morphologically larval cuticles at the end of the L1, L2 and L3 stages. However, at the end of the fourth larval stage (at the L4-to-adult molt), the seam cells cease cell division, fuse with each other and synthesize a morphologically distinct adult cuticle (Sulston and Horvitz, 1977; Singh and Sulston, 1978). The coordinate morphological and behavioral changes in hypodermal cells that constitute the L/A switch (cell cycle exit, cell fusion, switch in cuticle morphology) signify the terminal differentiation of these cells.

In animals homozygous for loss-of-function (lf) mutations of lin-29, the L/A switch never occurs, resulting in reiterated larval fates: the animals undergo extra larval molts beyond the L4 stage, characterized by supernumerary seam cell divisions and the formation of morphologically larval cuticle, indicating that lin-29 is essential for this event. Genetic epistasis has revealed that lin-29 is a relatively direct regulator of the L/A switch and that lin-4, lin-14 and lin-28 affect the timing of the L/A switch by regulating lin-29 (Ambros, 1989). The predicted product of lin-29 is a transcription factor (Rougvie and Ambros, 1995), which may act directly, or through more specialized regulators, to control the transcription of genes involved in various aspects of hypodermal cell differentiation. Identifying target genes of lin-29 is essential for understanding the molecular mechanisms by which lin-29, and hence the heterochronic gene pathway, regulates the L/A switch.

Candidate target genes for lin-29 would include genes involved in cell cycle control, cell fusion and stage-specific cuticle morphology. Collagens are major components of C. elegans extracellular cuticles (Cox et al., 1981a,b) and genes encoding collagens are a family of between 50 and 150 members dispersed throughout the genome (Kramer et al., 1982; Cox et al., 1984, 1985, 1989). Collagen genes whose mRNAs accumulate stage-specifically have been identified by high stringency hybridization using stage-specific cDNA probes (Cox et al., 1984). The stage-specific expression of these collagen genes at larval and adult molts may be responsible for some of the morphologically distinct features of larval and adult cuticles.

In work reported here, we show that mutations of heterochronic genes can affect the transcription of stage-specific collagen genes. We show that the precocious and retarded morphological defects exhibited by heterochronic mutants correlate with precocious and retarded transcription of the stage-specific collagen genes col-7, col-17 and col-19. These findings provide the first evidence that heterochronic genes function by controlling the stage-specific transcriptional activity of differentiated cell genes.

Strains and cloned DNA

lin-4, lin-14, lin-28 and lin-29 alleles are described in Chalfie et al. (1981), Ambros and Horvitz (1984, 1987), and Ambros (1989). Strains used in this work are N2, (wild-type C. elegans var. Bristol), n1270e187 (containing a rol-6 putative null mutation which renders the strain more sensitive than wild type to expression of the su1006 transgene; Mello et al., 1991 and Kramer et al., 1990), MT1397 lin-14(n179ts), VT284 lin-14(ma135)/sZT1, MT1176 lin-29(n546)/mnC1, MT1155 lin-4(e912)/mnC1, MT1388 lin-14(n355n679ts), VT292 lin-14(ma135), MT1524 lin-28(n719), MT1848 lin-14(n360), CB912 lin-4(e912); and VT181 lin-14(n536) dyp-6(e14)/szT1(X;I). Plasmids and lambda clones containing collagen genes were obtained from J. Kramer. β-galactosidase reporter vectors pPD22.04 and pPD35.41 were obtained from A. Fire (Fire et al., 1990).

Synchronized worm populations

Starvation-arrested newly hatched L1 animals were obtained by treating gravid adult hermaphrodites with 0.5% hypochlorite/0.5 N NaOH solution (Emmons et al., 1979) and allowing the eggs to hatch in sterile M9 solution for 12-24 hours. After feeding with Escherichia coli OP50 (Brenner, 1974; Sulston and Brenner, 1974) synchronized L3-to-L4 molt and L4-to-Adult molt animals were harvested after 32-36 hours and 42-46 hours, respectively. Dauer larvae were selected from starved asynchronous cultures with 1% SDS (Cassada and Russell, 1975), and then fed to allow simultaneous resumption of development. Animals at the postdauer L3-to-L4 or postdauer L4-to-adult molt were harvested at about 18-20 and 28-30 hours after feeding, respectively. A sample of each population was monitored for molting behaviour, particularly cessation of pharyngeal pumping. lin-14(ma135), lin-28(n719) and lin-4(e912) populations were generally less synchronous than wild-type or other mutant strains.

RNA isolation

5 ml of LETS buffer (100 mM LiCl, 10 mM EDTA, 10 mM Tris-HCl (pH 7.8 adjusted with LiOH), 1% SDS) were added to 1 ml of frozen a worm pellet (kept on ice). An equal volume of phenol/ chloroform and 2 ml of baked glass beads (VWR) were added to the mixture, which was then vortexed to break the worms. RNA was precipitated from the aqueous phase with LiCl and ethanol (Igo and Losick, 1986).

Poly(A)+ RNA was isolated by a modification of Maniatis et al. (1982) (G. Ruvkun, personal communication). 1 mg of total RNA was dissolved in 160 μl of TES (10 mM Tris pH 7.5, 0.1 mM EDTA, 1% SDS), heated to 65°C for 5 minutes and chilled on ice. 40 μl 5× binding buffer (5× is 2 M NaCl, 50 mM HEPES pH 7.5, 1% Sarkosyl, 5 mM EDTA) was added, followed by 0.5 mg oligo(dT)-cellulose (Collaborative Research Inc.) in 150 μl 1× binding buffer. The mixture was shaken at 37°C for 30 minutes, then at room temperature for 30 minutes. The oligo(dT)-cellulose was pelleted at 3000 revs/minute in a microcentrifuge for 2 minutes, washed twice in 15 ml 1× binding buffer, resuspended in 1 ml 1× binding buffer and loaded into a 1 ml pipette tip plugged with glass wool. The loaded pipette tip was washed first with 0.5 ml 1× binding buffer and eluted with 600 μl elution buffer (10 mM Hepes pH 7.5, 0.02% Sarkosyl). The poly(A)+ RNA was precipitated with ethanol and dissolved in DEPC-treated H2O.

Northern blot hybridization

Northern blots were performed using Gene Screen (DuPont) and kinased oligonucleotide probes (Ausubel et al., 1987). Oligonu-cleotide sequences were chosen from regions of the collagen transcription units that are not conserved among C. elegans collagen genes (Cox et al., 1989). The sequences of the col-17 and col-7 20-base gene-specific oligonucleotide probes are: col-17: 5′ AAACTTTTGGAGAAGTTCGC 3′ (antisense strand immediately following the stop codon TAA; GenBank accession number U18384); col-7: 5′ TCTACCAGGTGAATGACTGT 3′ (antisense strand immediately preceding the start codon; Cox et al., 1989).

Hybridization conditions (in buffer containing 50% formamide; at 37°C for 16-48 hours using the col-17 20-base oligonucleotide probe or 42°C for 16-48 hours using the col-7 20-base oligonucleotide probe) and wash conditions (twice for 15 minutes at room temperature with 6× SSPE followed by 30 minutes at a temperature 5°C higher than the hybridization temperature; Sugita et al., 1987) were determined from the base composition of the probe and adjusted empirically so as to result in hybridization to a homogeneous transcript of the collagen gene family. Control probes were labeled using the random primed DNA labeling kit (Boehringer Mannheim) and hybridized to the northern filters at low stringency, followed by a low stringency wash (0.2× SSPE, 0.2% SDS for 1 hour at room temperature).

Collagen/lacZ reporter constructs

pZC19.61 (col-19): A 2.7 kb XbaI/BamHI fragment of col-19 containing 21 bp of coding sequence was fused in frame with the lacZ gene at the XbaI/BamHI sites of the pPD22.04 vector (Fire et al., 1990; Fig. 2). pZC19Δ846 (col-19): A HindIII fragment (from a site in col-19 to a site in the vector) was deleted from pZC19.61 to leave 846 bp of 5′ sequences. pZC7.21 (col-7): A 2.5 kb XbaI/BamHI fragment of col-7 containing 602 bp of coding sequence and 1900 bp of 5′ non-coding sequences was fused in frame with the lacZ gene at the XbaI/BamHI sites of the pPD35.41 vector (Fire et al., 1990; Fig. 2). pZC17.6 (col-17): A 1.1 kb HindIII/BamHI fragment of col-17 containing about 800 bp of 5′ non-coding sequences and 274 bp of coding sequence was fused in frame at the HindIII/BamHI sites of the pPD35.41 vector (Fire et al., 1990; Fig. 2).

Fig. 1.

Northern analysis of col-17 and col-7 gene expression in heterochronic mutants. Precocious mutants, lin-14(ma135), lin-28(n719) and lin-14(n360), were analysed at the L3-to-L4 molt; retarded mutants, lin-4(e912), lin-29(n546) and lin-14(n536) dyp-6(e14)/lin-14(+)szT1(X;I) were analysed at the L4-to-adult molt. 5 μg total RNA of each mutant was analysed using gene-specific col-17 or col-7 oligonucleotide probes (Materials and Methods). (A) col-17 probe. (B) A separate blot hybridized with col-7 probe. As a control for the total amount of collagen mRNA in each sample (col), filters were stripped and hybridized with a stage-nonspecific col-3 probe at low stringency.

Fig. 1.

Northern analysis of col-17 and col-7 gene expression in heterochronic mutants. Precocious mutants, lin-14(ma135), lin-28(n719) and lin-14(n360), were analysed at the L3-to-L4 molt; retarded mutants, lin-4(e912), lin-29(n546) and lin-14(n536) dyp-6(e14)/lin-14(+)szT1(X;I) were analysed at the L4-to-adult molt. 5 μg total RNA of each mutant was analysed using gene-specific col-17 or col-7 oligonucleotide probes (Materials and Methods). (A) col-17 probe. (B) A separate blot hybridized with col-7 probe. As a control for the total amount of collagen mRNA in each sample (col), filters were stripped and hybridized with a stage-nonspecific col-3 probe at low stringency.

Fig. 2.

Diagrams of col-19/lacZ, col-7/lacZ and col-17/lacZ reporter gene fusions (Materials and Methods). The nuclear localization signal (NLS) targets β-galactosidase to nuclei (Fire et al., 1990). The transmembrane domain (TM), which blocks secretion of the fusion protein (Fire et al., 1990), was critical for the detection of col-7/lacZ expression. Thick vertical bar marks the position of the translation start site of each fusion gene. X, XbaI; B, BamHI; H, HindIII.

Fig. 2.

Diagrams of col-19/lacZ, col-7/lacZ and col-17/lacZ reporter gene fusions (Materials and Methods). The nuclear localization signal (NLS) targets β-galactosidase to nuclei (Fire et al., 1990). The transmembrane domain (TM), which blocks secretion of the fusion protein (Fire et al., 1990), was critical for the detection of col-7/lacZ expression. Thick vertical bar marks the position of the translation start site of each fusion gene. X, XbaI; B, BamHI; H, HindIII.

Fig. 3.

β-galactosidase expression in transgenic wild-type and heterochronic mutant animals transformed with col-19/lacZ (pZC19.61) and col-7/lacZ (pZC7.21) reporter genes (Materials and Methods; Fig. 2). The rolling phenotype conferred by the rol-6(su1006) transformation marker (Mello et al., 1991) is recognizable in fixed animals by a twist in the worm’s body. (A,B) Wild-type hermaphrodites at the L4-to-adult molt and adult stages, respectively, transformed with the col-19/lacZ reporter. The nuclear localization signal included in the reporter gene product (Fig. 2) directs β-galactosidase to the nuclei of expressing seam (se) cells and syncytial (sy) cells. β-galactosidase staining appeared first in vulva cells at the L4-to-adult molt; then, in seam cells of young adults; finally, in syncytial cells and non-vulval ventral hypodermal cells of older adults. (C) β-galactosidase expression in the hypodermal seam of a wild-type adult hermaphrodite transformed with the col-7/lacZ reporter. β-galactosidase staining was adult-specific and restricted to the seam cells. The fusion protein produced from this construct lacks a nuclear localization signal, so much of the β-galactosidase activity is extranuclear. (D) Wild-type larvae (L) and adult (A) hermaphrodites transformed with the col-19/lacZ fusion. (E) lin-28(n719) larvae transformed with col-19/lacZ. (F) lin-29(n546) adult transformed with col-19/lacZ. Note the staining near the protruding vulva. (G) Wild-type larva (L) and adult (A) hermaphrodites transformed with the col-7/lacZ. (The non-expressing adult at left was not a roller). (H) lin-14(n360) larva (L) and adult (A) hermaphrodites transformed with col-7/lacZ. (I)lin-4(e912) adult transformed with col-19/lacZ. Bar for A-C, 20 μm; for D-I, 100 μm.

Fig. 3.

β-galactosidase expression in transgenic wild-type and heterochronic mutant animals transformed with col-19/lacZ (pZC19.61) and col-7/lacZ (pZC7.21) reporter genes (Materials and Methods; Fig. 2). The rolling phenotype conferred by the rol-6(su1006) transformation marker (Mello et al., 1991) is recognizable in fixed animals by a twist in the worm’s body. (A,B) Wild-type hermaphrodites at the L4-to-adult molt and adult stages, respectively, transformed with the col-19/lacZ reporter. The nuclear localization signal included in the reporter gene product (Fig. 2) directs β-galactosidase to the nuclei of expressing seam (se) cells and syncytial (sy) cells. β-galactosidase staining appeared first in vulva cells at the L4-to-adult molt; then, in seam cells of young adults; finally, in syncytial cells and non-vulval ventral hypodermal cells of older adults. (C) β-galactosidase expression in the hypodermal seam of a wild-type adult hermaphrodite transformed with the col-7/lacZ reporter. β-galactosidase staining was adult-specific and restricted to the seam cells. The fusion protein produced from this construct lacks a nuclear localization signal, so much of the β-galactosidase activity is extranuclear. (D) Wild-type larvae (L) and adult (A) hermaphrodites transformed with the col-19/lacZ fusion. (E) lin-28(n719) larvae transformed with col-19/lacZ. (F) lin-29(n546) adult transformed with col-19/lacZ. Note the staining near the protruding vulva. (G) Wild-type larva (L) and adult (A) hermaphrodites transformed with the col-7/lacZ. (The non-expressing adult at left was not a roller). (H) lin-14(n360) larva (L) and adult (A) hermaphrodites transformed with col-7/lacZ. (I)lin-4(e912) adult transformed with col-19/lacZ. Bar for A-C, 20 μm; for D-I, 100 μm.

Plasmids were microinjected into C. elegans hermaphrodite syncytial gonads at a concentration of 80 μg/ml (Fire, 1986; Mello et al., 1991). Plasmid pRF4 (Kramer et al., 1990) containing a dominant mutant rol-6(su1006) gene was co-injected at 80 μg/ml as a behavioral marker for transformation (Mello et al., 1991). β-galactosidase activity was detected in fixed worms as described (Fire et al., 1990).

col-19 sequence analysis

pZC19Δ846, and clones derived from pZC19Δ846 by digestion with Bal31 (Ausubel et al., 1987), were sequenced by the Sanger method (1977) using M13 reverse primer and primer EGM1 (CCAAGGGTC-CTCCTGAAA), which is complementary to sequences of pPD22.04 to the right of the col-19 insert, and the Sequenase kit (United States Biochemicals). Portions of the sequence derived from analysis of pZC19Δ846 were confirmed by sequence analysis of DNA amplified by the Polymerase Chain Reaction (PCR) from phage clone lambdaCG34 (from which the insert in pZC19Δ846 was derived; Cox et al., 1984) and from genomic DNA (see legend to Fig. 4).

Fig. 4.

5′ sequences of col-19 contained in pZC19Δ846 (Materials and Methods). Underlined sequence was not reported in the previously published col-19 sequence (Cox et al., 1989). ATG (bold), probable initiator methionine codon of col-19. The position of the first col-19 nucleotide of each Bal31 derivative (Table 2) is indicated by bold face type. A putative TATA box is positioned at −47 bp from the ATG. The Genbank accession number of this sequence is U18385.

Fig. 4.

5′ sequences of col-19 contained in pZC19Δ846 (Materials and Methods). Underlined sequence was not reported in the previously published col-19 sequence (Cox et al., 1989). ATG (bold), probable initiator methionine codon of col-19. The position of the first col-19 nucleotide of each Bal31 derivative (Table 2) is indicated by bold face type. A putative TATA box is positioned at −47 bp from the ATG. The Genbank accession number of this sequence is U18385.

Stage-specific collagen gene transcripts in wild-type C. elegans

Stage-specific collagen transcripts had been identified previously by dot-blot analysis using cDNA probes reverse-transcribed from mRNAs isolated from developmentally synchronized worm populations (Cox and Hirsh, 1985). In those experiments, worms were synchronized by dauer larva arrest (Cassada and Russell, 1975). We repeated the dot-blot analysis of Cox and Hirsh (1985), but used populations of animals that had developed continuously from the L1 stage without dauer larva formation (see Materials and Methods). Our dot-blot results (data not shown) essentially confirm those of Cox and Hirsh (1985) with one exception, col-17, described below. Three major classes of collagens were detected based on their expression patterns, (1) stage non-specific, (2) adult-specific (expressed at the L4-to-adult molt only) and (3) larval-specific (expressed at the L3-to-L4 molt but not the L4-to-adult molt).

Collagen genes of the adult-specific class had been previously sequenced by Cox et al. (1989). We sequenced genomic DNA of col-17, one of a novel larval-specific class of collagen genes, from 159 bp 5′ of the translation start to 186 bp 3′ of the translational termination signal (sequence not shown; Genbank accession number U18384). The larval-specific expression of col-17 and the adult-specific expression of col-7 were confirmed by northern blot analysis (Fig. 1). The adult-specific expression of col-19 was confirmed by col-19/lacZ reporter gene expression (Liu and Ambros, 1991) described below. The expression patterns of col-7, col-17 and col-19 during postdauer development were the same as during continuous development, as determined by northern blot analysis (data not shown) and lacZ reporter studies (Liu and Ambros, 1991).

col-17 and col-7 mRNA accumulation in heterochronic mutants

We used northern blot analysis to examine larva-specific col-17 and adult-specific col-7 mRNA accumulation in precocious mutants of lin-14 and lin-28 at the L3-to-L4 molt and in retarded mutants of lin-4, lin-14 and lin-29 at the L4-to-adult molt (Fig. 1). In wild-type animals, col-17 mRNA was abundant at the L3-to-L4 molt and was not detectable at the L4-to-adult molt (Fig. 1A). The precocious mutants lin-14(ma135), lin-14(n360) and lin-28(n719) showed reduced col-17 expression at the L3-to-L4 molt (Fig. 1A). Although none of these precocious mutants exhibited a complete shut-off of col-17 expression at the L3-to-L4 molt, the lin-14(ma135) and lin-28(n719) mutations, which cause a more severely precocious defect than lin-14(n360) by morphological criteria (Ambros, 1989), caused a more severely reduced col-17 expression at the L3-to-L4 molt. In contrast, the retarded mutants lin-4(e912), lin-14(n536/+) and lin-29(n546), which reiterate larval cuticle synthesis at the L4-to-adult molt, inappropriately expressed the larva-specific col-17 mRNA at the L4-to-adult molt at levels comparable with the wild type at the L3-to-L4 molt (Fig. 1A).

In the wild type, the adult-specific col-7 RNA was undetectable at the L3-to-L4 molt and was abundant at the L4-to-adult molt (Fig. 1B). The retarded mutants lin-4(e912), lin-14(n536/+) and lin-29(n546), which synthesize morphologically larval cuticles at the L4-to-adult molt, failed to express col-7 at the L4-to-adult molt (Fig. 1B). Surprisingly, the precocious mutants lin-14(ma135), lin-14(n360) and lin-28(n719), which synthesize a morphologically adult-like cuticle at the L3-to-L4 molt, did not exhibit detectible levels of col-7 RNA at the L3-to-L4 molt.

The expression of collagen/lacZ reporter genes in the wild type

We used lacZ reporter genes and DNA transformation (Fire et al., 1990) as an independent assay of transcriptional activity that does not depend on detection of stable transcripts. A col-7/lacZ translational fusion (pZC7.21) and a col-19/lacZ translational fusion (pZC19.61) were constructed (Materials and Methods; Fig. 2). Stably transformed lines were established carrying pZC7.21 or pZC19.61 and transgenic animals were stained with X-gal. Transformed worms containing either the col-7/lacZ or the col-19/lacZ reporter genes consistently expressed β-galactosidase adult-specifically, beginning at the L4-to-adult molt (Fig. 3; Table 1). The col-19/lacZ and col-7/lacZ reporter genes were expressed only in hypodermal cells, although not in precisely the same sets of hypodermal cells (Fig. 3). The stage-specific expression of these col-7/lacZ and col-19/lacZ reporter genes in wild-type animals parallels the results of northern blot and dot-blot analysis. Further, the hypodermal expression of these constructs is consistent with the expected tissue-specificity of cuticle collagen synthesis. Like the col-19/lacZ reporter (Liu and Ambros, 1991), the col-7/lacZ reporter was expressed adult-specifically whether animals had developed continuously or through a period of dauer larva arrest (data not shown). The expression of a col-17/lacZ reporter was not consistent with the northern blot analysis. Worms containing a col-17/lacZ translational fusion (pZC17.6; Fig. 2) displayed variable levels of staining at all stages (data not shown), in contrast with the larval-specific expression of the endogenous col-17 mRNA (Fig. 1). Because the col-17/lacZ construct did not express β-galactosidase activity stage-specifically, it was not used in the analysis of heterochronic mutants described below.

Table 1.

Stage-specificity of collagen/lacZ reporters in transgenic worms

Stage-specificity of collagen/lacZ reporters in transgenic worms
Stage-specificity of collagen/lacZ reporters in transgenic worms

col-7/lacZ and col-19/lacZ reporter gene expression in heterochronic mutants

col-19/ lacZ reporter gene expression was analysed in heterochronic mutant animals to test whether the reporter gene expression correlates with the precocious or retarded execution of the L/A switch. When retarded mutants (lin-4(e912) or lin-29(n546)) were transformed with a col-19/lacZ reporter gene, β-galactosidase expression was not detected in lateral hypodermal cells (Fig. 3; Table 1). The lack of col-19/lacZ expression is consistent with the formation of a morphologically larval cuticle at the L4-to-adult stage in these retarded animals. Therefore, lin-29 and lin-4 wild-type gene products are required for the activation of col-19 transcription at the L4-to-adult molt. Interestingly, we observed β-galactosidase staining near the vulva of lin-29(n546) and lin-29(n1368) adults carrying col-19/lacZ constructs (Fig. 3F), indicating that col-19 expression in those (presumably vulval) cells does not require lin-29. The col-7/lacZ chimeric gene was not expressed at any stage in lin-29(n546) animals (Table 1).

The col-19/lacZ reporter was expressed at an abnormally early stage, the L3-to-L4 molt, in precocious mutants of lin-28(n719) (Fig. 3E; Table 1), lin-14(n360) (Table 1) and lin-14(n179ts) (Liu and Ambros, 1991; Table 1) correlating with the abnormally early adult cuticle formation in these mutants. Similarly, the col-7/lacZ reporter was expressed precociously in the lin-14(n360) mutants (Fig. 3H, Table 1). As described earlier, northern blot analysis failed to reveal precocious col-7 mRNA in these mutants (Fig. 1). This result suggests that expression of col-7/lacZ reporter may be a more sensitive indicator of heterochronic transcriptional activity than northern analysis.

5′ sequences of col-19 required for stage- and tissue-specific expression

A col-19/lacZ reporter (pZC19Δ846) containing 846 bp of col-19 5′ sequences was constructed from pZC19.61 (Materials and Methods; Figs 2, 4). Worms transformed with pZC19Δ846 expressed the reporter gene adult-specifically (at the L4-to-adult molt), indicating that this 846 bp of col-19 5′ sequence contains regulatory sequences sufficient for the stage- and tissue-specific expression of col-19. To further delineate essential col-19 promoter sequences, 5′ deletion derivatives of pZC19Δ846 were generated (Materials and Methods; Fig. 4) and tested for the ability to specify adultspecific lacZ expression in transgenic worms. As shown in Table 2, two col-19 deletion constructs (Δ122 and Δ153) with only 122 or 153 bp of sequences 5′ of the col-19 ATG, failed to direct the expression of lacZ in transgenic worms. In contrast, three col-19 deletion derivatives (Δ235, Δ315 and Δ448) that retain at least 235 bp of 5′ sequences expressed lacZ adult-specifically. One of these constructs, Δ235, expressed lacZ relatively weakly, but still adult-specifically. These results indicate that sequences between −153 and −235 may contain elements that control the adult-specificity of col-19 transcription, and sequences between −235 and −315 contain elements that affect the level of transcription.

In the course of sequencing the above col-19 deletion constructs, we found that the col-19 5′ promoter region contains 296 bp that were not reported in the published col-19 sequence (Cox et al., 1989). These additional sequences (the underlined nucleotides from positions −337 through −632 in Fig. 4) are present in a col-19 genomic lambda clone and in wild-type genomic DNA, as revealed by sequence analysis of DNA amplified by Polymerase Chain Reaction (PCR).

Transcriptional regulation of collagen genes by the heterochronic gene pathway

We employed three methods to monitor the stage-specific expression of the collagen genes col-17, col-7 and col-19 during development of wild-type and heterochronic mutants. Dot-blot analysis of cloned collagen genes using cDNA probes representing staged mRNA fractions suggested that these genes were expressed stage-specifically during continuous development (this study) and postdauer development (Cox and Hirsh, 1985). Northern analysis with gene-specific oligonucleotide probes detected stage-specific accumulation of endogenous col-7 and col-17 transcripts. Finally, extrachromosomal arrays of transgenic DNA containing col-7/lacZ or col-19/lacZ reporter genes expressed β-galactosidase in a temporal pattern similar to that of the endogenous col-7 and col-19 genes.

Table 2.

Expression of col-19 promoter deletions in transgenic worm strains

Expression of col-19 promoter deletions in transgenic worm strains
Expression of col-19 promoter deletions in transgenic worm strains

The adult-specific expression of col-7 and col-19 (assayed using lacZ reporter constructs) and the larval-specific expression of col-17 (assayed using northern blots) were altered in accordance with the precocious or retarded expression of cuticle phenotypes in heterochronic mutants. This result indicates that lin-4, lin-14, lin-28 and lin-29 regulate, directly or indirectly, the stage-specific expression of these collagen genes. This regulation is likely to occur on the level of transcriptional initiation and not, for example, mRNA stability, since our col-19/lacZ constructs contain only 21 bp of col-19 protein coding sequence and only approximately 40 bp of 5′ untranslated sequence.

Our results identify col-17 transcripts as larval-specific by dot-blot (data not shown) and northern blot hybridization (Fig. 1), suggesting that col-17 represents a novel class of stage-specific collagen gene. Although the col-17 gene-specific oligonucleotide probe detected a larval-specific mRNA on northern blots, transgenic animals carrying the col-17/lacZ construct tested here showed β-galactosidase staining at both larval and adult stages. We nevertheless believe that the northern analysis is more likely to reflect accurately activity of the endogenous gene than is reporter gene expression, since the transcriptional regulatory sequences on the transgenic construct may be incomplete. Dot-blot analysis reported by Cox and Hirsh (1985) showed that sequences hybridizing to col-17 were not expressed larval-specifically, but were detectable in larval and adult RNA samples. It is possible that differences in the dot-blot hybridization conditions detected low levels of col-17, or related sequences, in the adult RNA.

In addition to being adult-specific, the expression of these col-7/lacZ and col-19/lacZ reporter genes is restricted to hypodermal cells, consistent with the presumed role of these collagens in the assembly of the external adult cuticle. There were some anatomical differences between the the adultspecific expression patterns of the col-7/lacZ and col-19/lacZ reporters (Fig. 3). These differences may either be specific to the transgenes, or reflect anatomical differences in endogenous gene activity between col-7 and col-19.

Stage-specific collagen gene expression and cuticle morphology

Northern blot analysis indicated that, in some situations, there is not a complete correlation between the level of expression of certain stage-specific collagen genes and the stage-specific cuticle morphology of heterochronic mutants. For example, our northern blots of lin-14 and lin-28 precocious mutants (Fig. 1) indicated an apparently incomplete precocious shut-off of larval collagen (col-17) and an incomplete precocious activation of adult collagen (col-7). This result suggests that the morphological features of adult cuticle expressed precociously in these mutants (lateral alae and specialized ultrastructural layering; Ambros and Horvitz, 1987) are not disturbed by col17 expression and also that abundant col-7 is not essential for their formation.

Collagen/lacZ reporter genes may be a particularly sensitive assay for precocious heterochronic phenotypes. For example, in mutants that did not exhibit obvious precocious expression of the endogenous col-7 gene by northern analysis, precocious col-7 promoter activity was easily detected using a col-7/ lacZ reporter. Further, we have observed that animals displaying a ‘leaky’ precocious defect, such as lin-14(n179ts) at 20°C, can express detectable col-7/lacZ expression in hypodermal seam cells not expressing precocious adult lateral alae (data not shown). The precocious reporter gene expression in these mutants may parallel a low level of precocious expression of the endogenous col-7 gene, or may be a characteristic of the the high copy col-7 transgenes used here.

lin-29 regulates stage-specific collagen gene transcription

Genetic epistasis tests with lin-4, lin-14 and lin-28 showed lin-29 to be the most direct regulator of the L/A switch (Ambros, 1989). Exhaustive genetic screening for suppressor mutations of lin-29 failed to identify any regulatory genes acting between lin-29 and the L/A switch (Papp, Euling and Ambros, unpublished; Bettinger and Rougvie, personal communication). This suggests that there are few, if any, regulatory steps between lin-29 and the genes that encode adult and larval-specific proteins of hypodermal cells. Our finding that lin-29 is required to repress col-17 and activate col-7 and col-19 transcription at the L4-to-adult molt is consistent with the finding that lin-29 encodes a probable zinc-finger transcription factor (Rougvie and Ambros, 1995). According to one model, lin-14, lin-4 and lin-28 may control the stage-specific activation of the Lin-29 protein, a transcription factor that triggers the hypodermal cell terminal differentiation process (Fig. 5). Recent findings that lin-29 protein synthesized in E. coli binds in vitro to the promoter sequences of col-19 and col-17 supports the view that lin-29 directly regulates collagen gene transcription and perhaps also the transcription of other hypodermal cell differentiation genes (Rougvie and Ambros, 1995). The roles of lin-29 protein as either a repressor or activator of stage-specific transcription could reflect the precise nature of its binding to particular promoters. Alternatively, lin-29 may activate adultspecific transcription in conjunction with certain specific partners and could repress larva-specific transcription when acting together with other partners.

Fig. 5.

A model for the regulation of stage-specific transcription of hypodermal cell genes by the heterochronic gene pathway. Epistasis relationships (Ambros, 1989) indicate that lin-29 acts most directly on the adult-specific terminal differentiation program of hypodermal cells. Since lin-29 encodes a probable transcription factor (Rougvie and Ambros, 1995), the lin-29 gene product may directly repress and activate, respectively, the transcription of larval- and adult-specific genes. According to this model, lin-4, lin-14 and lin-28 act to ensure the proper stage-specific activation of lin-29.

Fig. 5.

A model for the regulation of stage-specific transcription of hypodermal cell genes by the heterochronic gene pathway. Epistasis relationships (Ambros, 1989) indicate that lin-29 acts most directly on the adult-specific terminal differentiation program of hypodermal cells. Since lin-29 encodes a probable transcription factor (Rougvie and Ambros, 1995), the lin-29 gene product may directly repress and activate, respectively, the transcription of larval- and adult-specific genes. According to this model, lin-4, lin-14 and lin-28 act to ensure the proper stage-specific activation of lin-29.

Using the col-19/lacZ reporter, we delineated col-19 promoter sequences sufficient for abundant adult-specific hypodermal expression. This region, extending 315 bp 5′ from the start of translation, likely includes sequences through which lin-29 acts. However, a simple comparison of col-7 and col-19 upstream sequences does not reveal any strikingly conserved sequence motifs that might be Lin-29 binding sites. Detailed analysis of promoter sequences that mediate Lin-29 activity may lead to identification of other direct targets of Lin-29. The activation and repression of stage-specific transcription by the Lin-29 gene product may serve as a model for how developmental regulatory genes control cell differentiation.

We thank Craig Mello for advice on microinjection, Andrew Fire for β-galactosidase reporter vectors and James Kramer for the collagen clones and valuable advice. We also thank Michael Finney, Gary Ruvkun and Rosalind Lee for oligonucleotides, and Rosalind Lee and Ann Rougvie for help with sequencing. This work was supported by US Public Health Service research grant GM34028 and March of Dimes Birth Defects Foundation research grant 5-445.

Ambros
,
V.
and
Horvitz
,
H. R.
(
1984
).
Heterochronic mutants of the nematode Caenorhabditis elegans
.
Science
226
,
409
416
.
Ambros
,
V.
and
Horvitz
,
H. R.
(
1987
).
The lin-14 locus of Caenorhabditis elegans controls the time of expression of specific post-embryonic developmental events
.
Genes Dev
.
1
,
398
414
.
Ambros
,
V.
(
1989
).
A hierarchy of regulatory genes controls a larva to adult switch in C
.
elegans. Cell
57
,
49
57
.
Ausubel
F. M.
,
Brent
,
R.
,
Kingston
,
R. E.
,
Moore
,
D. D.
,
Seidman
,
J. G.
,
Smith
,
J. G.
and
Struhl
,
K.
(
1987
).
Current Protocols in Molecular Biology
, volume
1.
New York
:
John Wiley & Sons
.
Brenner
,
S.
(
1974
).
The genetics of Caenorhabditis elegans
.
Genetics
77
,
71
94
.
Cassada
,
R. C.
and
Russell
,
R. L.
(
1975
).
The dauer larva, a post-embryonic developmental variant of the nematode Caenorhabditis elegans
.
Dev. Biol
.
46
,
326
342
.
Chalfie
,
M.
,
Horvitz
,
R. H.
and
Sulston
,
J. E.
(
1981
).
Mutations that lead to reiterations in the cell lineages of C
.
elegans. Cell
24
,
59
69
.
Cox
,
G. N.
,
Kusch
,
M.
and
Edgar
,
R. S.
(
1981a
).
The cuticle of C. elegans: its isolation and partial characterization
.
J. Cell Biol
.
90
,
7
17
.
Cox
,
G. N.
,
Staprans
,
S.
and
Edgar
,
R. S.
(
1981b
).
The cuticle of C. elegans. II. Stage-specific changes in ultrastructure and protein composition during post-embryonic development
.
Dev. Biol
.
86
,
456
470
.
Cox
,
G. N.
,
Kramer
,
J. M.
and
Hirsh
D.
(
1984
).
Number and organization of collagen genes in Caenorhabditis elegans
.
Mol. Cell. Biol
.
4
,
2389
2395
.
Cox
,
G. N.
and
Hirsh
,
D.
(
1985
).
Stage-specific patterns of collagen gene expression during development of Caenorhabditis elegans
.
Mol. Cell. Biol
.
5
,
363
372
.
Cox
,
G. N.
,
Fields
,
C.
,
Kramer
,
J. M.
,
Rosenzweig
,
B.
and
Hirsh
,
D.
(
1989
).
Sequence comparisons of developmentally regulated collagen genes of Caenorhabditis elegans
.
Gene
76
,
331
344
.
Emmons
,
S.
,
Klass
,
M. R.
and
Hirsh
,
D.
(
1979
).
Analysis of the constancy of DNA sequences during development and evolution of the nematode Caenorhabditis elegans
.
Proc. Natn. Acad. Sci. USA
76
,
1333
1337
.
Fire
,
A.
(
1986
).
Integrative transformation of C
.
elegans. EMBO J
.
5
,
2673
2680
.
Fire
,
A.
,
Harrison
,
S.
and
Dixon
,
D.
(
1990
)
A modular set of lacZ fusion vectors for studying gene expression in C
.
elegans. Gene
93
(
2
),
189
-
198
.
Igo
,
M. M.
and
Losick
,
R.
(
1986
).
Regulation of a promoter that is utilized by minor forms of RNA polymerase holoenzymes in Bacillus subtilis
.
J. Mol. Biol
.
191
,
615
624
.
Kramer
,
J. M.
,
French
,
R.P.
,
Park
,
E.
and
Johnson
,
J. J.
(
1990
).
The C. elegans rol-6 gene, which interacts with the sqt-1 collagen gene to determine organismal morphology, encodes a collagen
.
Mol. Cell. Biol
.
10
(
5
),
2081
-
2089
.
Kramer
,
J. M.
,
Cox
,
G. N.
and
Hirsh
,
D.
(
1982
).
Comparisons of the complete sequences of two collagen genes from Caenorhabditis elegans
.
Cell
30
,
599
606
.
Liu
,
Z.
and
Ambros
,
V.
(
1989
).
Heterochronic genes control stage-specific initiation and expression of the dauer larva developmental program in Caenorhabditis elegans
.
Genes Dev
.
3
,
2039
2049
.
Liu
,
Z.
and
Ambros
,
V.
(
1991
).
Alternative temporal control systems for hypodermal cell differentiation in C
.
elegans. Nature
350
,
162
165
.
Maniatis
,
T.
,
Fritsch
,
E. F.
and
Sambrook
,
J.
(
1982
).
Molecular Cloning: A Laboratory Manual
.
Cold Spring Harbor, NY
:
Cold Spring Harbor Laboratory
.
Mello
,
C. C.
,
Kramer
,
J. M.
,
Stinchcomb
,
D.
and
Ambros
,
V.
(
1991
).
Efficient DNA transfer in C. elegans: Extrachromosomal maintenance and integration of transforming sequences
.
EMBO J
.
10
,
3959
3970
.
Rougvie
,
A. R.
and
Ambros
,
V. R.
(
1995
).
The heterochronic gene lin-29 encodes a zinc-finger protein that triggers a terminal differentiation event in C
.
elegans. Development
121
,
2492
2500
.
Sanger
,
F.
,
Nicklen
,
S.
and
Coulson
,
A. R.
(
1977
).
DNA sequencing with chain terminating inhibitors
.
Proc. Natl. Acad. Sci. USA
74
,
5463
5467
.
Singh
,
R.
and
Sulston
,
J.E.
(
1978
).
Some observations on molting in Caenorhabditis elegans
.
Nematologica
24
,
63
71
.
Sugita
,
M.
,
T.
Manzara
,
E.
Pichersky
,
A.
Cashmore
and
Gruissem
,
W.
(
1987
).
Genomic organization, sequence analysis and expression of all five genes encoding the small subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase from tomato
.
Mol. Gen. Genet
.
209
,
247
256
.
Sulston
,
J. E.
and
Brenner
,
S.
(
1974
).
The DNA of Caenorhabditis elegans
.
Genetics
77
,
95
104
.
Sulston
,
J. E.
and
Horvitz
,
R.H.
(
1977
).
Post-embryonic cell lineages of the nematode Caenorhabditis elegans
.
Dev. Biol
.
56
,
110
156
.