|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 279, Issue 28, 29270-29277, July 9, 2004
The Caenorhabditis elegans Ortholog of TRAP240, CeTRAP240/let-19, Selectively Modulates Gene Expression and Is Essential for Embryogenesis*![]() ![]() ||
From the
Received for publication, February 4, 2004 , and in revised form, April 2, 2004.
Mediator complexes are large multiprotein assemblies that function in the regulation of eukaryotic gene transcription. In yeast, certain mediator subunits appear to comprise a subcomplex that acts in the regulation of a specific subset of genes. We investigated in a metazoan, Caenorhabditis elegans, the roles and interactions of two of those subunits, CeTRAP240/let-19 and CeTRAP230/dpy-22. We found that CeTRAP240/let-19 contains four domains that are conserved in the human TRAP240 protein and that one of those domains displays intrinsic transcriptional repression activity. Using RNA interference, we found that reduced expression of CeTRAP240/let-19 displayed a high penetrance of embryonic lethality in F1 progeny; animals that escaped embryonic arrest showed mutant phenotypes such as burst vulva and molting defects. CeTRAP240/let-19 appeared to affect specific genes, as CeTRAP240/let-19(RNAi) led to selectively reduced expression of a subset of reporter genes examined. Genetic experiments supported the view that CeTRAP240/let-19 and CeTRAP230/dpy-22, like their Drosophila and yeast counterparts, can operate on common pathways. Thus, a male tail phenotype caused by the pal-1(e2091) mutation was suppressed not only by CeTRAP230/dpy-22 mutants, as reported previously, but also by reduced expression of CeTRAP240/let-19. Additionally, CeTRAP240/let-19(RNAi) in a CeTRAP230/dpy-22 mutant background produced a strong synthetic lethal phenotype. Overall, our results establish specific roles of CeTRAP240/let-19 in C. elegans embryonic development and a functional interaction between CeTRAP240/let-19 and CeTRAP230/dpy-22. Interestingly, whereas this interaction has been conserved from yeast to mammals, the subcomplex modulates metazoan-specific genetic pathways, likely in addition to those also controlled in yeast.
The Mediator complex is a multiprotein assembly that participates in both positive and negative regulation of transcription by RNA polymerase II in eukaryotes; it was first identified in Saccharomyces cerevisiae using biochemical and genetic approaches (14). The Mediator complex serves as a vehicle for physical and functional communication between DNA-bound regulatory factors and the transcription initiation apparatus; this communication can include enzymatic modifications as well as simple contacts (511). Although still a matter of some debate (12, 13), the Mediator complex can be conceptualized as a component of many functional regulatory complexes that assemble at response elements and modulate initiation by RNA polymerase II. Biochemical approaches have yielded numerous distinguishable mammalian Mediator-related complexes. For example, CRSP was isolated by conventional chromatography based on its ability to potentiate activation by Sp1 (14). Other complexes were identified by their association with various transcriptional regulatory factors, such as thyroid hormone receptor (TRAP), vitamin D3 receptor (DRIP), SREBP-1a (ARC), and E1A (1519). Still others, such as SMCC (20) and NAT (21), were isolated by affinity purification using antibodies directed against mammalian orthologs of yeast Mediator components. Many of these complexes activate transcription from naked DNA or chromatin templates in vitro (14, 18, 22, 23), and SMCC and NAT have been shown to mediate repression in vitro (8, 21); ARC has been fractionated into two complexes, one of which, ARC-L, lacks apparent activation activity in vitro (24). Whether the compositional or functional differences among mammalian Mediator-related complexes observed in vitro reflect the existence of multiple complexes in vivo remains to be determined. However, in yeast, biochemical evidence for certain subcomplexes of Mediator components is underpinned by genetic data, and it appears that some of those subcomplexes mediate the transcriptional regulation of distinct subsets of genes (8, 9). Interestingly, the twenty identified mammalian Mediator subunits are conserved across the eukaryotes (25), and certain inferred mammalian Mediator subcomplexes parallel those previously validated in yeast. For example, TRAP230 (also called ARC240/DRIP240), TRAP240 (also called ARC250/DRIP250), cyclin-dependent kinase 8, and cyclin C are found in ARC-L, but are absent from the smaller CRSP complex (24), reminiscent of the subcomplex comprised of their respective yeast orthologs, Srb8, Srb9, Srb10, Srb11 (26). Similarly, a Mediator complex purified from murine cells lacking TRAP100 is also devoid of TRAP95 and SUR2 (27), suggesting that these three subunits may form a subcomplex; consistent with this view, the Mediator isolated from murine cells lacking SUR2 correspondingly lacks TRAP95 and TRAP100 (28). However, although TRAP95 and SUR2 are conserved from yeast to mammals, TRAP100 appears to be a metazoan-specific Mediator subunit (25). It is intriguing to speculate that Mediator complexes of differing composition may have acquired new roles during evolution, acting in concert on subsets of genes that accommodate the specialized needs of metazoans while maintaining many of the functions evolved earlier. To examine this possibility, we have undertaken an analysis of Mediator complexes, subcomplexes, and component subunits in Caenorhabditis elegans, with its genetic manipulability, defined cellular lineage, and limited tissue complexity. In this report, we focused on the C. elegans ortholog of TRAP240 and Srb9, CeTRAP240/let-19.1 We set out to isolate the full-length CeTRAP240/let-19 cDNA and to characterize its expression pattern. We also sought to assess the in vivo functions and potential genetic interactions of CeTRAP240 with C. elegans TRAP230 and Srb8 ortholog, CeTRAP230/dpy-22.
StrainsC. elegans (N2 wild type and derived mutant strains) was grown at room temperature. The following integrated GFP2 reporter strains were used: nhr-2::GFP (29), pha-4::GFP (30), hlh-1::GFP (31), pes-10::GFP (32), med-1::GFP (33), end-1::GFP (34), elt-5::GFP (35), and sur-5::GFP (36). Mutant alleles sop-1(bx93) and pal-1(e2091) was described previously (37).
Isolation of CeTRAP240/let-19 cDNAThe full-length CeTRAP240/let-19 cDNA was isolated by reverse transcription (RT)-PCR using mix-staged N2 total RNA. RT-PCR was performed according to the manual of the SuperScript one-step RT-PCR for long template kit (Invitrogen). Total RNA was isolated using Tri-reagent (Molecular Research Inc.) according to manufacturer's protocol. The CeTRAP240/let-19 cDNA was isolated as three fragments. (i) The 5' end was isolated by RT-PCR using an SL1 primer (GGTTTAATTACCCAAGTTTGAG) and a primer that encompasses a KpnI site in CeTRAP240 (AAGGTACCAAATGTTCGATGC); this RT-PCR product was amplified again by primers GGGGTACCATGTCCTCCGCCAAGGATAG and AAGGTACCAAATGTTCGATGC. (ii) The middle region was amplified by RT-PCR using primers TTGGTACCTTGGCTCAGAAAG and TTCTGCAGCAAATGTTCGATGCGGCGCC. (iii) The 3' end was amplified by RT-PCR using primers ACCTGCAGAAACTCCATTTGA and AACTGCAGCTAGGCAAGTAGACGAGCC. The three fragments were digested by KpnI, KpnI/PstI, and PstI, respectively, and then sub-cloned into pBluescript KS (+) plasmid (Stratagene) for sequence analysis. Identification of CeTRAP240/let-19 Expression PatternThe CeTRAP240/let-19 promoter to first exon was amplified by PCR of cosmid K08F8 (provided by the Sanger Center) using primers AATCTAGAATTACGTTGGTTAAAAG and CGCGGATCCGCCAAAATGTATTTGGAGAATTATATT. This genomic fragment was digested by XbaI and BamHI and subcloned in-frame into the pPD95.75 plasmid. This construct (50 ng/µl) was then injected into N2 hermaphrodites along with pRF4 (rol-6, 50 ng/µl) as a marker. Multiple extrachromosomal transgenic arrays were generated; multiple animals were examined in each transgenic strain, and all displayed similar GFP expression patterns. RNA InterferenceFor CeTRAP240/let-19(RNAi), sequences corresponding to nucleotides 41055386 of CeTRAP240/let-19 cDNA were amplified by PCR and subcloned into pBluescript KS (+). This plasmid was digested by PvuII, and single strand RNAs were synthesized using the Ambion MEGA-script T3 and T7 in vitro transcription kit. To make dsRNA, equal amounts of sense and antisense strand RNA were mixed well and incubated at 95 °C for 5 min followed by 65 °C for 5 min, 50 °C for 5 min, 37 °C for 5 min, and 25 °C for 5 min. The dsRNA was then injected into N2 hermaphrodites. The injected worms were placed on plates to recover for 1824 h. After the recovery period the worms were moved to fresh plates for embryo collection. Worms were allowed to lay eggs for 812-h intervals and then transferred to a new plate. Embryonic arrest was measured 1824 h after the end of the egg-laying interval. Larval growth was monitored daily, and larvae that did not progress to the adult stage by day 4 post-hatch were scored as larval arrest. To confirm our CeTRAP240/let-19(RNAi) results, an additional dsRNA that corresponds to the C terminus of CeTRAP240/let-19 cDNA (yk398g7) was also used for RNA interference (RNAi) experiments. Although of lower potency, the latter dsRNA produced results consistent with those reported here. For CeTRAP230/dpy-22(RNAi), the EST clone yk222f3 was digested with either PvuI or PvuII, and single-stranded RNAs were synthesized using the Ambion MEGA-script T3 and T7 in vitro transcription kit. Production of dsRNA and RNAi experiments were performed as described above. To confirm the results from CeTRAP230/dpy-22(RNAi), dsRNA synthesized from the PvuII fragment of another EST clone, yk294f12, was used and yielded results similar to those using yk222f3 dsRNA. For bacterial feeding RNAi experiments, cDNA corresponding to the nucleotide sequence 35415386 of CeTRAP240 cDNA was amplified by PCR and subcloned into the KpnI and PstI sites of plasmid pAD12 (provided by Andy Dillon, University of California, San Francisco), which contains two T7 promoters in inverted orientation. This construct was then transformed into an IPTG-inducible Escherichia coli strain, HT115, which lacks double-strand-specific RNase III. Single colonies of HT115 containing pAD12 were picked and grown in culture for 1824 h with 120 µg/ml carbenicillin and tetracycline and seeded on NGM-lite plates containing 1 mM IPTG and 50 µg/ml ampicillin. L1 hermaphrodites were placed on these plates for 2 days; L4 or young adults were then transferred to new plates seeded with the same bacteria. The worms were allowed to lay eggs for 24 h. Embryonic arrest was assessed 24 h later, and larval growth was monitored daily. ImmunostainingImmunostaining of C. elegans embryos using antibody against Ser(P)-2 (H5) of polymerase II C-terminal domain (CTD) was as previously described (38).
Transient Transfection, Luciferase Assay, and Cell CultureTo construct an expression plasmid that encodes the C terminus of CeTRAP240/LET-19, the nucleotide sequences of CeTRAP240/LET-19 corresponding to amino acids 22292863 were amplified by PCR using the primer pairs CGGGATCCGTTGGAAACAGCGCGACACTC and ACTCTAGACTAGGCAAGTAGACGAGCC. This PCR fragment was then digested by BamHI and XbaI and subcloned into pSG424 plasmid (39). Construction of GAL4-TLE1 and (GAL4)5E1BLuc reporters (40, 41) and assays for
CeTRAP240 cDNA and Expression PatternA Blast search identified a single C. elegans ortholog of hTRAP240 protein (C. elegans clone K08F8.6). We amplified the full-length CeTRAP240 cDNA by RT-PCR using total RNA prepared from mixed-stage animals. To confirm our results, we sequenced an EST clone (yk288e10) encoding the 3' end of CeTRAP240/let-19. Amino acid sequence alignment of CeTRAP240 and hTRAP240 revealed four homologous domains, as shown in Fig. 1, AE. Interestingly, our RT-PCR also identified a shorter CeTRAP240/let-19 cDNA that lacks the fourth and most C-terminal homologous domain. These results suggest that the CeTRAP240 transcription unit may be subject to alternative splicing in a manner that yields two protein isoforms.
To visualize the expression pattern of CeTRAP240/let-19, a DNA fragment extending 2.0 kilobases upstream of the translation start site was cloned in-frame into a GFP-encoding plasmid, pPD95.75. This construct was microinjected into N2 worms along with the pRF4 (rol-6) plasmid as a transformation marker. Ubiquitous GFP expression was observed in the embryo (Fig. 2A, panel a). In larvae and adults, GFP expression was observed in most tissues, including vulva (Fig. 2A, panel b), tail neurons (Fig. 2A, panel c), skeletal muscle (Fig. 2A, panel c, arrow), hypodermal cells (Fig. 2A, panel c, yellow arrow), the H-shaped excretory canal (Fig. 2A, panel d) and pharynx (data not shown).
CeTRAP240/let-19 RNAi PhenotypesWe used RNAi to compare the developmental consequences of down-regulating CeTRAP240/let-19; phenotypes were scored in F1 animals. Injection of a CeTRAP240 double-stranded RNA (dsRNA, 2 µg/µl) that corresponds to the middle region of the CeTRAP240/let-19-coding region resulted in a >90% embryonic lethality. Using RT-PCR, we confirmed that accumulation of CeTRAP240/let-19 transcripts was selectively affected, because levels of CeRgr1 mRNA, for example, were unchanged (data not shown). Most embryos were arrested at the 200300 cell stage, and thus, lacked differentiated cells (Fig. 2B, panel a); a small proportion of the embryos were arrested at later stages, such as the comma, 2- or 3-fold stages. As a control, injection of full-length rat glucocorticoid receptor (GR) dsRNA did not produce significant mutant phenotypes. When we reduced the concentration of injected CeTRAP240/let-19 dsRNA 10-fold (0.2 µg/µl), a greater proportion of animals escaped arrest and displayed a range of mutant adult phenotypes, including abnormal body morphology (Fig. 2B, panel b and c), withered tail (Fig. 1C, panel d), defective cuticle shedding (Fig. 2B, panel d), multivulva (Fig. 2B, panel e), and dumpy and burst vulva (data not shown). All of these mutant phenotypes are characteristic of defects in hypodermal development. Importantly, CeTRAP240/let-19(RNAi) caused a more severe phenotype (embryonic versus larval lethality) than the let-19 homozygous allele, which also showed molting defects and multivulva phenotypes (data not shown).3 Thus, these results suggest that the CeTRAP240/let-19 homozygous allele is not a null mutant. The CeTRAP240/let-19 mutant lacks the conserved hTRAP240 C-terminal domain (Fig. 1, A and E). We fused the C-terminal domain to the GAL4 DNA binding domain to test whether this domain contains intrinsic transcriptional regulatory activity. Interestingly, this GAL4 DNA binding domain fusion protein (GAL4-LET-19/C) inhibited transcription from a (GAL4)5E1BLuc reporter in A549 human lung adenocarcinoma cells (Fig. 2C); the repression activity of GAL4-LET-19/C was as potent as a fusion of GAL4 to TLE1, a human ortholog of the Drosophila Groucho corepressor (Fig. 2C). How the repression activity of the CeTRAP240/LET-19 C-terminal domain relates to the phenotypes of CeTRAP240/let-19 has not yet been explored.
Effects of Mediator Components on Expression from Reporter Gene PromotersTo investigate whether the requirement of CeTRAP240/let-19 in embryogenesis reflects a role in general embryonic gene transcription, we first analyzed phosphorylation of the RNA polymerase II (pol II) large subunit CTD, which consists of heptapeptide repeats, YSPTSPS; full polymerase activity requires phosphorylation of Ser-2 and Ser-5 within these repeats. As shown in Fig. 2A, depletion of the pol II large subunit (AMA-1) by RNAi leads to a dramatic reduction in detection of transcriptionally active pol II by an antibody specific for phosphorylated Ser-2 (38). In contrast, although CeTRAP240/let-19(RNAi) caused an embryonic lethal phenotype, like ama-1(RNAi), reduced expression of CeTRAP240/let-19 did not influence the phosphorylation status of CTD Ser-2 in embryos (Fig. 3A); similarly, CeTRAP230/dpy-22(RNAi), which resulted in a 2030% of embryonic lethal phenotype in F1 progeny, did not influence CTD Ser-2 phosphorylation (Fig. 3A). These results suggest that CeTRAP240/let-19 and CeTRAP230/dpy-22 act selectively in gene expression rather than being required for general transcription.
To examine whether CeTRAP240/let-19 is only required for the expression of specific genes, we injected CeTRAP240/let-19 dsRNA into C. elegans strains bearing a series of integrated reporter::GFP fusion genes. We first tested the requirement of CeTRAP240/let-19 on pes-10::GFP, which is expressed at the onset of embryonic transcription (Fig. 3B, panel 1) and ubiquitously expressed sur-5::GFP (Fig. 3B, panel 2). Interestingly, reduction of CeTRAP240/let-19 activity by RNAi did not influence the expression of these two reporters. We next tested the effect of CeTRAP240/let-19(RNAi) on reporter constructs containing the promoters from genes that are expressed in certain cell types or developmental stages. Reduction of CeTRAP240/let-19 diminished hlh-1::GFP (a muscle cell marker, Fig. 3B, panel 3), end-1::GFP (an endodermal marker, Fig. 3B, panel 4), and elt-5::GFP expression (an epidermis/seam cell marker, Fig. 3B, panel 5) but not nhr-2::GFP (an early developmental maker that begins to express at 16-cell stage but disappears during the initial stages of morphogenesis; Fig. 3B, panel 6), pha-4::GFP (a digestive tract maker, Fig. 3B, panel 7), and med-1::GFP (a mesendodermal marker; Fig. 3B, panel 8). Although we have not determined if these genes are direct targets of CeTRAP240/let-19 activity, our results suggest that CeTRAP240/let-19 is responsible for the expression of a specific subset of genes rather than for general embryonic gene expression. Together, our results suggest that CeTRAP240/let-19 is selectively involved in multiple developmental pathways. Suppression of the pal-1(2091) Phenotype by Inhibition of Mediator ExpressionCeTRAP240/let-19 is the C. elegans ortholog of Srb9, which in yeast forms a subcomplex with Srb8, -10, and -11 (26). Thus, we sought to determine whether CeTRAP240/let-19 interacts during C. elegans development with other components of its orthologous subcomplex. Previous work had been done on the C. elegans ortholog of Srb8, CeTRAP230/dpy-22; several mutant alleles of that gene, such as sop-1, suppress a mutant male tail phenotype of pal-1(e2091) (37). Animals carrying the pal-1(e2091) mutation fail to develop V6 rays in the adult male tail (Ref. 37, Fig. 4A). We first confirmed that CeTRAP230/dpy-22(RNAi) in the pal-1(e2091) strain results in 8090% rescue of the male tail phenotype (Ref. 37, Fig. 4D). We then tested the effect of CeTRAP240/let-19(RNAi) on the mutant male tail phenotype of pal-1(e2091). Indeed, CeTRAP240/let-19(RNAi) yielded 57% of F1 male pal-1(e2091) worms with fully (Fig. 4B) or partially (Fig. 4C) recovered tails (Fig. 4D). As a control, GR dsRNA injected into pal-1(e2091) produced little or no suppression of the male tail phenotype, similar to that seen in uninjected pal-1(e2091) mutants (Fig. 4D). Thus, both CeTRAP240/let-19 and CeTRAP230/dpy-22 are likely involved in pal-1 gene expression.
Notably, sop-1 mutants alone do not exhibit significant phenotypes (37), whereas the dpy-22 mutant is sick and displays some larval lethality (43). Injection of a low concentration (0.2 µg/µl) of CeTRAP230/dpy-22 dsRNA into N2 worms did not result in any mutant phenotype, whereas a higher concentration (2 µg/µl) produced 15% embryonic and larval arrest and mutant adult phenotypes in about half of the remainder. The predominant mutant phenotype from CeTRAP230/dpy-22(RNAi) was protruding/burst vulva; some animals also displayed dumpy, abnormal body morphology and molting defects (data not shown). All of these mutant phenotypes are characteristic of defects in hypodermal development and overlap with those observed in CeTRAP240(RNAi) animals, suggesting that both of these Mediator subunits function in that pathway.
Genetic Interaction between CeTRAP230/dpy-22 and CeTRAP240/let-19To further investigate the potential relationship between CeTRAP230/dpy-22 and CeTRAP240/let-19 functions, we tested for a synthetic lethal relationship between these two genes. We used the bacterial feeding RNAi method for these experiments, in which CeTRAP240/let-19 dsRNA expression was dependent upon IPTG (44, 45). Approximately 20% of wild type N2 worms grown in bacteria producing CeTRAP240/let-19 dsRNA displayed embryonic and larval arrest (Table I). In the absence of CeTRAP240/let-19 dsRNA (no IPTG control), the rate of embryonic or larval arrest was 4% (Table I). Thus, the effect of bacterial feeding RNAi is significant, but it provides a weaker effect than the direct injection of CeTRAP240/let-19 dsRNA into worms. We then tested the effect of CeTRAP240/let-19(RNAi) on sop-1(bx93) mutants. When sop-1(bx93) was fed with uninduced CeTRAP240/let-19 dsRNA bacteria,
In this study, we demonstrated important roles for CeTRAP240/let-19 in C. elegans embryonic development and showed that it is likely involved in regulating the expression of a specific subset of genes. The reduced expression of CeTRAP240/let-19 by RNAi is unlikely to disrupt the integrity of the Mediator complex because in vitro studies of mammalian and yeast mediator complexes have demonstrated the presence of complexes lacking cyclin-dependent kinase 8-cyline c-TRAP230-TRAP240. However, we have not excluded the possibility that decreased expression of CeTRAP240/let-19 may influence the association of cyclin-dependent kinase 8, cyclin C, and CeTRAP230/DPY-22 within the Mediator complex. Two other C. elegans Mediator components, CeMed6 and CeMed7, are similarly required for the expression of certain developmental but not ubiquitously expressed genes (46), whereas CeRgr1 functions more globally in gene transcription (38). Thus, it appears that two broad classes of components may reside in Mediator complexes, one represented by CeRgr1, which affects the expression or regulation of all or most RNA polymerase II genes, and a second, represented by CeTRAP240/let-19, CeTRAP230/dpy-22, CeMed6, CeMed7, and sur-2, which acts more selectively in gene regulation.
This concept is consistent with genetic data from yeast and mammalian Mediator complexes. In yeast, various Mediator subunits seem to function selectively on distinct subsets of genes, whereas Srb4 and Srb6 appear to participate in the expression of almost all genes (9). In mammalian cells TRAP100 is broadly required (27), whereas TRAP220 and SUR-2 play specific roles in transcriptional regulation. Primary fibroblast cells cultures derived from TRAP220 knock-out mouse embryos show an impaired thyroid hormone response and are defective in peroxisome proliferator-activated receptor How can we rationalize and begin to understand this apparent complexity of Mediator composition and function? What is apparent is that the Mediator evolved as a device of eukaryotes to accommodate a substantially increased demand for complex networks of transcriptional regulation. This requirement can be seen even in simple eukaryotes, such as yeast, but is most stringent in metazoans, which produce terminally differentiated cells with clearly distinct expression patterns, and must integrate myriad cell signals to produce physiologically rational responses at the level of the whole organism. Strong evolutionary conservation of Mediator components suggest that C. elegans, with its powerful genetics, fully sequenced genome, simple anatomical system, and known cell lineage offers multiple approaches to analyzing mechanisms of differential transcriptional regulation that will likely be relevant to all metazoans. It is well established that the yeast Mediator complex is composed of several subcomplexes. Genetic disruption of the components in distinct subcomplexes leads to common or overlapping phenotypes (5, 9). In metazoans, subcomplexes are implied but not proven. The yeast ortholog of CeTRAP240/let-19, Srb9, is part of a module that also contains Srb8, Srb10, and Srb11 (26). Interestingly, recent genetic studies in Drosophila reveal that dTRAP240 and dTRAP230, orthologs of yeast Srb9 and Srb8 respectively, act in concert in eye-antennal disc development (48). Our results here suggest that CeTRAP240/let-19 and CeTRAP230/dpy-22 likely act together in some biological pathways. Thus, both CeTRAP240/let-19(RNAi) and CeTRAP230/dpy-22(RNAi) suppressed the pal-1(e2091) male tail phenotype (Fig. 3). Notably, these results also suggest the functions of CeTRAP240/let-19 and CeTRAP230/dpy-22 are not redundant because we can reduce the expression of either one to suppress the pal-1(e2091) phenotype. Furthermore, we demonstrated a genetic interaction between CeTRAP240/let-19 and CeTRAP230/dpy-22 (Table I). Finally, low levels of CeTRAP240/let-19(RNAi) yielded defects similar to those seen with CeTRAP230/dpy-22(RNAi) (Fig. 1C and data not shown). Although we cannot discern whether the effects of CeTRAP240/let-19(RNAi) and CeTRAP230/dpy-22(RNAi) reflect qualitative or only quantitative differences, the simplest interpretation of current results is that these two Mediator components can act together on subsets of genes. Thus, the functional interaction between these two subunits in the Mediator complex appears to be conserved during evolution. In contrast, a mutation in sur-2, an ortholog of yeast Mediator subunit Gal11, failed to suppress the pal-1(e2091) male tail phenotype (49), consistent with the finding that Gal11 interacts neither genetically nor biochemically with Srb8 and Srb9 (5, 9). It is intriguing that mutations of the yeast Srb8-Srb9-Srb 10-Srb11 subcomplex lead to increased transcription, implying that this subcomplex is inhibitory (5, 50). Furthermore, some human mediator complexes that include cyclin-dependent kinase 8-cyclin C·hTRAP230·hTRAP240 fail to activate transcription in vitro and inhibit an activation-competent mediator complex (24). Consistent with this view, we found that the C terminus of CeTRAP240/LET-19 protein confers transcriptional repression. Notably, this domain is conserved between CeTRAP240/LET-19 and hTRAP240, implying that the C-terminal domain of hTRAP240 may also repress transcription; we are currently testing this possibility. Interestingly, human cyclin-dependent kinase 8 can phosphorylate general transcription factor TFIIH and inhibit the transcriptional initiation (51). Thus, if the function of CeTRAP240/LET-19 is conserved during evolution, the metazoan cyclin-dependent kinase 8 subcomplex may inhibit transcription by two independent mechanisms. In principle, discrete Mediator subcomplexes could explain the differential effects of specific subunits on the regulation of particular genes despite the function of the Mediator as a unit in in vitro assays. One model is that different combinations of subcomplexes are recruited to different response elements through protein-protein interactions with regulatory factors. This could account for the multiple forms of Mediator complex that have been described in vitro and is supported by findings that regulatory factors interact selectively with specific Mediator components (16). However, there is no direct support for the existence of multiple Mediator complexes in vivo, i.e. the multiple forms observed in vitro might be artifacts of biochemical purification. It should be possible in future studies to determine by chromatin immunoprecipitation the presence or absence of Mediator subunits in vivo at specific genes at which they are either required or functionally dispensable. Subcomplexes appear to be neither functionally dedicated to general expression or specialized genes nor evolutionarily "sequestered" with respect to their subunit composition. Thus, TRAP100 plays a general role in gene regulation and is metazoan-specific, whereas SUR2, which resides in the same subcomplex, acts on a specific subset of genes and is conserved from yeast to mammals (27). This suggests that functions of the Mediator complex cannot be attributed entirely to the differential recruitment of different subcomplexes. Instead, different regulators may confer distinct conformations to one or more Mediator subunits (24), perhaps affecting their functions within the complex or creating novel functional surfaces. According to this model, a given subcomplex might reside in two different regulatory complexes and participate in different regulatory mechanisms. This would be analogous to the behavior of GR and GRIP1, two regulatory factors that can reside at different response elements and use different surfaces to either activate or repress transcription (52). In this study we have identified several apparent target genes and tissues for CeTRAP240/let-19 action. It should now be possible to determine whether CeTRAP230/dpy-22 acts in these same settings and whether the functional surfaces are similar or distinct in the different contexts. By carrying out this investigation in C. elegans, we can examine these alternatives throughout the course of development and in all tissues of this metazoan organism.
* This work was supported by grants from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|| To whom correspondence should be addressed: Dept. of Cellular and Molecular Pharmacology, University of California, S572D, Genentech Hall, 600 16th St., San Francisco, CA 94107-2280. Tel.: 415-476-3128; Fax: 415-476-6129; E-mail: yamamoto{at}cmp.ucsf.edu.
1 During the course of this work H. Sawa (RIKEN, Center for Developmental Biology, Kobe, Japan) showed that CeTRAP240 is identical to the let-19 gene and CeTRAP2308 and sop-1 are identical to the previously identified dpy-22 gene.
2 The abbreviations used are: GFP, green fluorescence protein; RT, reverse transcription; dsRNA, double-stranded RNA; IPTG, isopropyl-1-thio-
3 H. Sawa, personal communication.
We thank Dr. Hitoshi Sawa for communicating unpublished data. We thank Drs. Andy Dillon, Scott Emmons, Cynthia Kenyon, Mike Krause, Susan Mango, Joel Rothman, and Ann Sluder for providing C. elegans strains and plasmids and Drs. Hitoshi Sawa, Mike Krause, Marta Kostrouchova, Zedenek Kostrouchova, Rik Derynck, Daryl Granner, Brian Freeman, Hans Leucke, Marc Van Glist, Andy Dillon, Marc Meneghini, and Inez Rogatsky for reviewing the manuscript.
This article has been cited by other articles:
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||