Whereas Notch signalling is essential for somitogenesis, mice deficient for the basic helix-loop-helix (bHLH) genes Hes1 and Hes5, downstream Notch effectors, display normal somite formation, indicating that there may be an as-yet unidentified Hes1-related bHLH gene.
We identified a novel bHLH gene, designated Hes7, from mouse embryos. Hes7 has a conserved bHLH domain in the amino-terminal region and the WRPW domain at the carboxy-terminal end, like Hes1. The mouse Hes7 gene is located next to Aloxe3, which is mapped to a position 37.0 cM from the centromere on chromosome 11. In a transfection analysis, Hes7 represses transcription from the N box- and E box-containing promoters. In addition, Hes7 suppresses the E47-induced transcriptional activation. Promoter analysis indicated that Hes7 expression is controlled by Notch signalling. Strikingly, Hes7 is specifically expressed in the presomitic mesoderm in a dynamic manner. We also identified two related bHLH genes from human: one is closely related to mouse Hes7 and therefore designated hHes7 and the other designated hHes4.
The structure, transcriptional activity and expression pattern in the presomitic mesoderm of Hes7 are very similar to those of Hes1, suggesting that Hes7, together with Hes1, may play a role in somite formation under the control of Notch signalling.
Here, to identify a novel Hes1-related bHLH gene expressed in the PSM, we performed a polymerase chain reaction (PCR) with degenerate primers. We isolated a bHLH gene, designated Hes7, which exhibits a high structural homology with Hes1. We also found two related human bHLH genes, designated hHes4 and hHes7. Hes7 encodes a factor that represses transcription, like Hes1. Furthermore, Hes7 is specifically expressed in the PSM and the expression is controlled by Notch signalling. These results suggest that Hes7 plays a role in somite formation, in collaboration with Hes1 under the control of Notch signalling.
Structural analysis of new Hes factors
To identify a novel Hes1-related bHLH gene, we performed reverse transcription-mediated PCR with degenerate primers corresponding to the bHLH domain. When RNA from mouse embryos of day 8.5 (E8.5) were used as a template, a fragment with a novel bHLH sequence was obtained, and this fragment was used as a probe to screen the cDNA and genomic library. We obtained cDNA and the genomic clones for a novel Hes1-related bHLH factor, designated Hes7. Mouse Hes7 consists of 225 amino acid residues (Fig. 1A) and shows a 50.0% identity to mouse Hes1 in the bHLH domain (Fig. 1B). Hes7 contains a proline residue in the basic region which is conserved among mammalian Hes and Drosophila Hairy and Enhancer of split. In addition, Hes7 has a conserved WRPW sequence at the carboxy-terminal region, which is known to interact with the co-repressor Groucho (Paroush et al. 1994; Fisher et al. 1996; Grabavec & Stifani 1996), as well as an orange domain, which may be involved in functions specific to each member of the Hes family (Dawson et al. 1995). We also searched the databases for human bHLH genes and found two related genes: one is closely related to mouse Hes7 (designated human Hes7) and the other is less homologous to Hes7 but more similar to Hes1 (designated human Hes4) (Fig. 1A). Regarding human Hes7, a part of the coding region which was not included in the databases was amplified by PCR and the nucleotide sequence of the whole coding region was determined. Human Hes7 exhibits a complete identity of the amino acid sequence in the bHLH domain and 90.7% identity in the whole region to mouse Hes7. In contrast, human Hes4 and Hes7 display only a 51.8% identity in the bHLH domain and 33.3% identity in the whole region.
Among the Hes members, Hes4 exhibits a significant homology to Hes1, Hes2 and Hes3, while Hes7 is the most similar to Hes5 (Fig. 1E). In contrast, Hes6 has only a distant homology to the other Hes factors. Thus, Hes-related factors can be classified into three subgroups according to the sequence homology in the bHLH domain: the Hes1–4 subgroup, Hes5/7 subgroup, and Hes6 subgroup (Fig. 1E). This classification is also applicable to the orange domain, which exhibits a higher sequence similarity within the same subgroups (Fig. 1D). When compared with other related factors, mouse Hes7 exhibits approximately 60% identity to Xenopus and zebrafish Hes-related factors, xESR4/5 and zHer1, and 50% identity to chick Hairy1/2 and zebrafish Her6 in the bHLH domain (Fig. 1C,E). In contrast, mouse Hes1 exhibits a 96–98% identity to chick Hairy1/2 and zebrafish Her6 but only a 41–48% identity to ESR4/5 and Her1 in the bHLH domain. Because Hes7 is highly expressed in the PSM (see below), we focused on an analysis of Hes7. The characterization of Hes4 will be described elsewhere.
Structural analysis of mouse Hes7 gene
We next examined the structure of the mouse Hes7 gene. Hes7 consists of four exons, and all the introns are located within the coding region (Fig. 2A). Furthermore, the positions of the introns are well conserved compared to other Hes genes (Takebayashi et al. 1994, 1995; Sakagami et al. 1994; Nishimura et al. 1998; Hirata et al. 2000) (see Fig. 1A, arrowheads), suggesting that all Hes genes originate from a common ancestor. 5′-Rapid amplification of cDNA ends (RACE) analysis (see Experimental procedures) indicated that transcription starts at the nucleotide position 1 (Fig. 2A). In the upstream region, there are several regulatory sequences: a putative Sp1-binding site at −18, an N box (CACNAG) at −60 and two E boxes (CANNTG) at −118 and −194 (Fig. 2A). Furthermore, there are two putative binding sites of RBP-J (Fig. 2A, double underlined), which forms a complex with the intracellular domain of Notch, and up-regulates Hes1 and Hes5 expression (Jarriault et al. 1995; Honjo 1996; Nishimura et al. 1998; Ohtsuka et al. 1999), raising the possibility that Hes7 expression is controlled by Notch signalling. The polyadenylation site is located at 2862, and there is a polyadenylation signal (AATAAA) at 2845 (Fig. 2A). The total mRNA size should be approximately 1 kb, which agrees well with the size demonstrated by Northern blot analysis (Fig. 3).
Interestingly, a sequence analysis showed that Per1 (mouse homologue of Drosophila period gene) and Aloxe3 (epidermis-type lipoxygenase-3 gene) are located upstream and downstream, respectively, of Hes7 (Fig. 2B). The Aloxe3 locus is assigned to a position 37.0 cM from the centromere on Chromosome (Chr) 11 (Kinzig et al. 1999). Thus, these data indicated that Hes7 and Per1 are also assigned to the same position on Chr 11 (Fig. 2C). This Hes7 locus is different from other Hes loci: Hes1 on Chr 16 and Hes2, Hes3, and Hes5 on Chr 4 (Sakagami et al. 1994; Takebayashi et al. 1994; Nishimura et al. 1998).
Transcriptional analysis of Hes7: Hes7 is a transcriptional repressor
To analyse the transcriptional activity of Hes7, we carried out a transient transfection analysis. We have previously shown that Hes1 represses transcription from N box- or E box-containing promoters (Sasai et al. 1992). In addition, Hes1 suppresses E47-induced transcription (Sasai et al. 1992). Here, we examined whether Hes7 has a similar transcriptional activity to Hes1. The luciferase reporter gene under the control of the N box- or E box-containing promoter was co-expressed with the Hes7 expression vector in C3H10T1/2 fibroblast cells. Hes7 efficiently repressed transcription from both N box- and E box-containing promoters, like Hes1 (Fig. 4A,B). In addition, whereas the bHLH factor E47 efficiently up-regulated expression from the E box-containing promoter, Hes7 suppressed this E47-induced expression, like Hes1 (Fig. 4C). These results demonstrated that Hes7 has a very similar transcriptional activity to Hes1.
Promoter analysis of Hes7: Hes7 expression is controlled by Notch signalling
To determine whether Hes7 expression is controlled by Notch signalling, we next performed a promoter assay with the expression vector of a truncated form of Notch, which lacked the extracellular domain. This truncated Notch is known to function as a constitutively active form, independent of the Notch ligands (Jarriault et al. 1995; Nishimura et al. 1998). The luciferase reporter under the control of Hes7 promoter was co-expressed with the Notch vector. The truncated Notch exhibited about a sixfold up-regulation of Hes7 promoter activity (Fig. 5, lane 2), suggesting that Hes7 expression is controlled by Notch signalling. Although this up-regulation was not as dramatic as for the Hes1 promoter, which exhibited about a 50–60-fold up-regulation (Fig. 5, lane 8), it was still significant compared with the Hes3 promoter, which was not regulated by Notch signalling (lane 10) (Nishimura et al. 1998). This induction of Hes7 promoter activity is probably controlled by RBP-J, which mediates Notch signalling (Honjo 1996), because there are two putative RBP-J-binding sites in Hes7 promoter (Fig. 2A). In addition, deletion of these sites significantly reduced the Notch-induced up-regulation (compare Fig. 5, lanes 2, 4 and 6).
Although the Hes7 promoter contains an N box (Fig. 2A), Hes7 did not repress transcription from the Hes7 promoter (data not shown) unlike Hes1, which negatively regulates its own expression through the multiple N boxes (Takebayashi et al. 1994).
Expression of Hes7: Hes7 is specifically expressed in the PSM
Hes7 expression was examined by in situ hybridization. At E8.5, Hes7 was specifically expressed in the PSM (Fig. 6A-D). In the PSM, Hes7 expression was observed in two bilateral domains: rostral and caudal stripes (Fig. 6A–D). The rostral bands were located just posterior to the newly formed somite (Fig. 6C,D), while the caudal bands were extended to the most caudal tip (Fig. 6A–D). During the E9.0–E12.0, Hes7 was again specifically expressed in two domains of the PSM (Fig. 6E–H, arrowheads in F, G).
If Hes7 expression oscillates in the PSM like Hes1 expression, different patterns of Hes7 expression should be detectable, even in embryos of the same stages. We therefore examined Hes7 expression in many embryos of E10.5. In some embryos, the caudal bands were more intense while the rostral bands were weak (Fig. 6I, L). In contrast, in others the rostral bands were more intense while the caudal bands were weak (Fig. 6J, K). Thus, there are several patterns of Hes7 expression, even in embryos of the same stages, suggesting that Hes7 expression may oscillate in the PSM. Furthermore, this expression pattern is very similar to that of Hes1, raising the possibility that Hes1 and Hes7 cooperatively regulate somite formation.
Hes7 is specifically expressed in the PSM
Although Notch signalling plays an essential role in somitogenesis, mutation of Hes1 and Hes5, two essential Notch effectors, does not display any abnormality in somite formation (Ishibashi et al. 1995; Jouve et al. 2000; Ohtsuka et al. 1999), suggesting that there may be an as-yet unidentified Hes1-related bHLH gene. By searching for new bHLH genes, we identified two Hes1-related genes, Hes4 and Hes7. Interestingly, Hes7 expression is specific to the PSM and is controlled by Notch signalling like Hes1 expression. Furthermore, Hes7 has a Hes1-like transcriptional activity. These results suggest that Hes1 and Hes7 cooperatively regulate somite formation in the PSM.
The expression of Hes7 is striking: it is specific to the PSM and not detectable in any other regions. Like Hes1, Hes7 is expressed in two bilateral bands in the PSM. It has been shown that Hes1 expression oscillates in the PSM with a 90 min cycle correlated with somite formation (Jouve et al. 2000). Therefore, the patterns of Hes1 expression domains are variable, even in embryo at the same stages: while in some embryos Hes1 is expressed in two bilateral stripes, the narrower rostral and broader caudal, in others the caudal stripe becomes narrower. This dynamic Hes1 expression is dependent on Notch signalling (Jouve et al. 2000). Since Hes7 expression is also controlled by Notch signalling, it may oscillate in the PSM. In accordance with this idea, Hes7 expression in the PSM is dynamic and variable, even in embryos at the same stage. Thus, it is likely that Hes7 expression is cyclic in the PSM, although it remains to be determined whether it oscillates in a 90 min cycle as does Hes1 expression.
Although the expression pattern in the PSM is very similar between Hes1 and Hes7, there is some difference in promoter activity. The Hes1 promoter displays a 50–60-fold up-regulation by Notch, whereas the Hes7 promoter displays only a sixfold up-regulation. This difference could be partly due to the different basal promoter activities: the basal activity of Hes7 promoter is already high (our unpublished observation) compared to that of Hes1 promoter, which is kept at a low level by negative autoregulation through the multiple N box elements (Takebayashi et al. 1994). Since the RBP-J site and the N box are overlapped in the Hes1 promoter, Notch signalling may not only activate transcription through the RBP-J site but may also suppress the Hes1-induced negative autoregulation by inhibiting Hes1 from binding to the N box.
The Notch ligand that may induce Hes7 expression in the PSM remains to be determined. Dll1 is expressed in the caudal domain of the prospective somite, while Dll3 is in the rostral domain, and thus it is likely that either Dll1 or Dll3 may regulate Hes7 expression in the PSM.
Hes7 is structurally and functionally similar to Hes1
The present study demonstrated that there are at least seven members of the Hes family. All Hes factors function as a transcriptional repressor, except for Hes6, which does not repress transcription. Hes6 suppresses Hes1 from inhibiting positive regulators and thereby up-regulates gene expression (Bae et al. 2000). Interestingly, the loop of Hes6 is four or five amino acid residues shorter than the others, and this loop region is important for Hes-specific activities (Bae et al. 2000). The loop of Hes7 is five amino acid residues longer than Hes6 and we found that Hes7 not only represses transcription from N box- and E box-containing promoters, but also antagonizes E47-induced transcriptional activation, like Hes1. Thus, Hes7 is structurally and functionally similar to Hes1.
Although Hes7 is similar to Hes1 in the bHLH domain (50%), the former is more similar to Xenopus ESR4/5 and zebrafish Her1 (57–66%) while the latter is more similar to chick Hairy1/2 and zebrafish Her6 (96–98%). These results indicate that Hes1 and Hes7 belong to two different subclasses of Hes-related factors: the highly homologous Hes1 subclass (Hes1, cHairy1/2, zHer6) and the less homologous Hes7 subclass (Hes7, ESR4/5, zHer1). Members of the Hes1 subclass also contain a highly conserved orange domain, in contrast to those of the Hes7 subclass, which have a less conserved orange domain. Interestingly, members of both subclasses are expressed in the PSM. Since the important structures, such as the bHLH and orange domains are less conserved between the Hes1 and Hes7 subclasses, it remains to be determined whether these two Hes-related groups of genes are functionally redundant or whether they have distinct roles in somite formation.
Chromosomal locus of Hes7
Since Hes7 is next to Aloxe3, mouse Hes7 is assigned to a position 37.0 cM from the centromere on Chr 11, which is different from other Hes loci. At the same locus, Alox12b (Kinzig et al. 1999) as well as Per1 are located in the mouse. Similarly, in the human, Alox12b and Per1 are located at very close positions: Chr 17p13.1-p12 and Chr 17p13, respectively (Sun et al. 1997; Sun et al. 1998), indicating that Per1 and Alox12b are located close to each other, both in the mouse and human. Since the relative locus positions of these genes are conserved between the mouse and human, it is likely that Hes7 is also located at the same position on Chr 17 in the human genome. Many genetic disorders associated with this locus have been reported in the human and mouse, but none of them have been implicated in defects of somitogenesis. Therefore, gene knock-out experiments should be carried out to demonstrate the exact functions of Hes7 in embryos. Further characterization of Hes7 will help to understand the molecular mechanism for somitogenesis through the Notch signalling.
Isolation of Hes7 and Hes4 cDNAs
RT-PCR was performed with RNA of E8.5 whole mouse embryos. Degenerate primers used for RT-PCR corresponded to the amino acid sequences conserved among Hes factors, KP(I/L)(M/L/V)EK(R/K/M)RR(A/D)RIN and KLE(K/N)A(D/E)(I/V)LE. We obtained a 110 bp PCR fragment, which encoded a novel bHLH factor, designated Hes7. The 129SVJ mouse genomic library (Stratagene) was screened with a radiolabelled PCR fragment, and a genomic clone was isolated from 106 plaques. Hes7 cDNA was next obtained by screening a cDNA library and by RT-PCR with primers designed from the genomic sequence of Hes7 gene. Following a sequence search of human genome databases, a BAC clone (Accession no. AC015734) that contained a sequence similar to mouse Hes7 was identified. A missing part (corresponding to the second exon) was obtained by PCR. For Hes4, we searched the database of expressed sequence tags (ESTs) and obtained clones (AW148795, AW161193) with sequences closely related to Hes1.
To determine the transcription initiation site of Hes7, we carried out 5′-RACE using Marathon-Ready™ cDNA (Clontech) of mouse E7 embryo. Nested PCR was performed with the specific primers, 5′-ggggccgtccctattctcagctcgctc-3′ and 5′-cgggggctccacccggcttcgctccct-3′.
Northern blot analysis and whole mount in situ hybridization
Ten micrograms of poly(A)+ RNAs were fractionated on a formaldehyde-containing 1.2% agarose gel and transferred on to a nylon membrane (NEN). Hybridization was carried out with a radiolabelled DNA probe corresponding to the entire coding region of the Hes7 cDNA.
Whole mount in situ hybridization of mouse embryos were performed as previously described (Bae et al. 2000) with a digoxygenin-labelled RNA probe corresponding to the entire coding region of the Hes7 cDNA.
Transient transfection analysis
For transcriptional analysis, luciferase reporters under the control of β-actin promoter with six repeats of N boxes or seven repeats of E boxes (0.1 µg) were transfected into C3H10T1/2 cells, which were plated in 12-multiwell plates at the density of 5 × 104 cells, with or without expression vectors (pCI, Promega) for Hes7, Hes1 or E47, as previously described (Hirata et al. 2000). For promoter analysis, luciferase reporters (0.1 µg) under the control of the Hes7, Hes1 (−467 to +46) or Hes3 (−270 to +260) promoter were transfected into NIH3T3 cells, which were plated in 12-multiwell plates at the density of 8 × 104 cells, with or without 0.2 µg of the expression vector for the constitutively active form of Notch, as previously described (Nishimura et al. 1998). The vector for Renilla luciferase gene under the control of the SV40 promoter (1 ng) was co-transfected as an internal standard to normalize the transfection efficiency. After 40–48 h, the cells were harvested and luciferase activities were measured.
This work was supported by Special Coordination Funds for Promoting Science and Technology and research grants from the Ministry of Education, Science, Sports and Culture of Japan and Japan Society for the Promotion of Science.