Human Molecular Genetics

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (35) Request Permissions Google Scholar Articles by Lang, D. Articles by Epstein, J. A. Search for Related Content PubMed PubMed Citation Articles by Lang, D. Articles by Epstein, J. A. Social Bookmarking          What's this?

Human Molecular Genetics, 2003, Vol. 12, No. 8 937-945
DOI: 10.1093/hmg/ddg107
© 2003 Oxford University Press

Sox10 and Pax3 physically interact to mediate activation of a conserved c-RET enhancer

Deborah Lang and Jonathan A. Epstein^*

Cardiovascular Division, Department of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA

Received January 23, 2003; Accepted February 15, 2003

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
The neurocristopathies encompass a spectrum of developmental disorders characterized by abnormalities of neural crest-derived structures. Neural crest cells are pluripotent progenitors and the mechanisms by which specific cell-fate decisions are regulated have emerged as an important field of study. Many neurocristopathies are characterized by defects in melanocyte differentiation that can result in pigmentation abnormalities and deafness. One example is Waardenburg syndrome that can be caused by mutations in the PAX3, SOX10 or MITF genes. Other neural crest-related disorders are associated with enteric ganglia defects, such as those caused by mutations in the SOX10 or c-RET genes. The Pax3 and Sox10 transcription factors can directly regulate both MITF and c-RET. Here, we show that Pax3 and Sox10 can physically interact and we map the interaction domains. We show that this interaction contributes to Pax3 and Sox10 synergistic activation of a conserved c-RET enhancer and it explains why Sox10 mutants that cannot bind to DNA retain the ability to activate this enhancer in the presence of Pax3. However, in the context of the MITF gene, Pax3 and Sox10 must each bind independently to DNA in order to achieve synergy. This difference is consistent with the different structures of the c-RET and MITF enhancers, and the different mechanisms by which Pax3 binds to these enhancers. These observations explain the phenotype in the mild form of Yemenite deaf–blind syndrome caused by specific SOX10 mutations in the HMG box that abrogate DNA binding without disrupting association with Pax3.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
The neural crest is a pluripotent cell population arising from the dorsal neural tube. Neural crest cells migrate throughout the embryo and participate in the development of a variety of tissues including the skeletal system, pigment cells, circulatory system, endocrine glands and the peripheral nervous system. The neurocristopathies encompass a spectrum of human disorders characterized by abnormalities of neural crest-derived structures. Many neurocristopathies are characterized by melanocyte defects, which can result in pigmentation abnormalities and deafness (due to melanocyte contribution to the inner ear). One example is Waardenburg syndrome that can be caused by mutations in the PAX3, SOX10 or MITF genes.

Pax3 is a transcription factor that possesses both a homeo- and a paired DNA binding domain (1). Homozygous mutations are lethal in both mouse and man. Mutations in the Pax3 gene results in the Splotch mouse phenotype (2). Heterozygous mice are characterized by a white belly spot, while homozygotes die during embryogenesis with neural tube and cardiac abnormalities. In melanoblasts, Pax3 functions to activate expression of MITF (3–5), a transcription factor that is itself required for subsequent melanocyte differentiation (6).

Another class of neural crest-related disorders includes dysfunction or absence of enteric ganglia and associated gastrointestinal motility defects. This syndrome is commonly referred to as Hirschsprung disease and is most often caused by mutations in the c-RET gene. The Ret protein is expressed during embryogenesis throughout the peripheral nervous system including the enteric nervous system (7–9). Mice homozygous for targeted mutation in the c-ret gene present with profound loss of the enteric ganglia throughout the intestines (10). SOX10 is also essential for the development of enteric ganglia and is disrupted in some patients with Hirschsprung disease. Sox10 is a member of the Sry-related family of high mobility group (HMG) containing transcription factors that play a role in cell lineage specification (11).

Interestingly, recent work has demonstrated that Pax3 and Sox10 can directly regulate both MITF and c-RET in a synergistic fashion (3,4,12). However, the mechanism of synergistic activation is distinct. The mode by which Pax3 binds to each promoter is different, and the spacing between the Pax3 and Sox10 binding sites is not conserved. Here, we show that Pax3 and Sox10 activation of a conserved c-RET enhancer is dependent upon intact Pax3 DNA binding and Sox10 transactivation domains. However, whereas Sox10 DNA binding is required for MITF activation, it is not as necessary for c-RET activation. This observation explains the phenotype caused by specific SOX10 mutations including those that cause Waardenburg–Shah syndrome and Yemenite deaf–blind hypopigmentation syndrome (YDBS) (13).

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Homology between human and mouse c-ret promoters
In order to identify potential enhancer regions within the murine c-ret upstream genomic region, 4.8 kb of upstream genomic sequence (GenBank AY255629) was identified by BAC screening and comparative sequence analysis was performed with the homologous human sequence (AF032124). One region of homology located 3.3 kb upstream of the translational start site contains binding sites for the transcription factors Pax3 and Sox10. We have previously described this enhancer as part of our analysis of the human c-RET gene. The 36 bp island of homology is 86% conserved between mouse and human. Outside of this small region of homology, the mouse and human sequences diverge (Fig. 1A). Strikingly, the previously identified Pax3 and Sox10 binding sites are identical. Sequence conservation across species is strong evidence supporting an important physiologic role for this regulatory region (12).

View larger version (42K): [in this window] [in a new window]   Figure 1. A conserved Sox10/Pax3-responsive enhancer is present in the c-ret gene. (A) Comparison of upstream genomic sequence derived from mouse (top) and human (bottom) c-ret genes reveals 100% conservation of Sox10 and Pax3 binding sites located within a region 3.3 kb upstream of the translational start site. The sequence adjacent to the Sox10/Pax3 enhancer is not conserved. (B) Pax3 and Sox10 together induce synergistic activation of the murine c-ret gene. Cotransfection of 293T cells with a luciferase reporter construct including 4.8 kb of murine c-ret genomic sequence (including the Sox10/Pax3 enhancer) with Pax3, Sox10 or both. Results of three experiments, each performed in triplicate, are expressed as fold activation compared with transfections without Pax3 or Sox10 and are normalized for transfection efficiency. Error bars represent standard deviation.

 
Pax3 and Sox10 function together to activate the murine c-ret gene
In cultured cells, Pax3 and Sox10 individually are able to induce modest activation of a reporter construct containing the 4.8 kb murine c-ret promoter/enhancer region (Fig. 1B). Synergistic activation occurs when Pax3 and Sox10 are co-transfected. This is similar to results previously reported using regulatory sequences derived from the human c-RET gene (12). Hence, the ability of Pax3 and Sox10 to activate the c-ret promoter/enhancer is conserved across species, as are the sequence and relative positioning of Pax3 and Sox10 binding sites.

Pax3 and Sox10 physically interact
Given the proximity and conservation of the Pax3 and Sox10 binding sites in the c-ret enhancer, we tested the ability of Pax3 and Sox10 to physically interact. Immunoprecipitation experiments using radioactively labeled, epitope-tagged Pax3 and/or Sox10 proteins confirm a protein–protein interaction. Figure 2A demonstrates the ability of epitope tagged Sox10 to immunoprecipitate Pax3 (lane 2). An antibody directed against Sox10 is also able to precipitate Pax3 (lane 9), but only if Sox10 is present in the reaction (lane 6).

View larger version (67K): [in this window] [in a new window]   Figure 2. Pax3 and Sox10 physically interact. Pax3 is 479 kDa (1), and Sox10 is 466 kDa (14) (A) Immunoprecipitations of in vitro translated Pax3 and Sox10 protein with -Pax3, -Sox10 or -HA antibodies. Top panels (lanes 1–10), [35S]methionine labeled Pax3 is precipitated with -HA Ab when HA-tagged Sox10 is present in the reaction (lanes 1–5), and can also be precipitated by -Sox10 Ab if Sox10 is present in the reaction (lanes 6–10). [35S]methionine-labeled Sox10 is immunoprecipitated with -HA antibody when HA-tagged Pax3 is present in the reaction (lanes 11–15), and it is also immunoprecipitated with -Pax3 antibody when Pax3 is present (lanes 16–20). (B) Immunoprecipitations of cell lysates from 293T cells transfected with Pax3 and/or Sox10 constructs. Immunoprecipitations utilize either -Pax3 or -Sox10 antibodies, followed by western analysis with -HA antibody. (C) Immunoprecipitations of in vitro translated HA-tagged Pax3 truncated proteins and full-length Sox10 protein using either -Sox10 or -HA antibodies. A schematic for the proteins made from each deletion construct is shown on the top, with the paired DNA binding domain, homeo-DNA binding domain, and the C terminal transactivation domain shown. For each deletion construct, three lanes are shown. The first lane is an experimental lane (lanes 1, 4, 7, 10, 13 and 16) the middle lane is a negative control (lanes 2, 5, 8, 11, 14 and 17) and the last lane is a positive control (lanes 3, 6, 9, 12, 15, 18). (D) Immunoprecipitations of in vitro translated HA-tagged Sox10 truncated proteins and full length Pax3 protein using either -Pax3 or -HA antibodies. A schematic for the proteins made from each deletion construct is shown at the top, with the N terminal, HMG DNA binding and C terminal transactivation domains shown. The data are presented as in (C) with experimental, negative controls and positive controls shown in that order.

 
We confirmed the physical interaction between Pax3 and Sox10 by performing additional assays using epitope tagged Pax3 and radioactive Sox10. An -HA antibody results in precipitation of Sox10 (Fig. 2A, lane 14) only when HA-tagged Pax3 is present. An antibody directed against Pax3 is also able to precipitate Sox10 (lane 19). Hence, using four distinct approaches, we confirmed that Pax3 and Sox10 physically interact in vitro.

Pax3 and Sox10 also interact in vivo, as shown in Figure 2B. 293T fibroblasts were transfected with epitope tagged or untagged Pax3 and/or Sox10 followed by immunoprecipitation and Western blotting. Lanes 1–6 (Fig. 2B) demonstrate the specificity of the antibodies used. When cells are co-transfected with Pax3 and epitope tagged Sox10 followed by immunoprecipitation with an -Pax3 antibody, Sox10 is detected in the precipitate (lane 7). Conversely, when untagged Sox10 is co-transfected with epitope tagged Pax3 followed by immunoprecipitation with -Sox10 antibody, Pax3 is detected in the precipitate. These results confirm the ability of Pax3 and Sox10 to interact in vivo.

The Pax3 paired domain and the Sox10 HMG domain are required for physical interaction
We defined the specific regions within Pax3 and Sox10 that are required for physical interaction by constructing a series of deletion mutants and performing co-immunoprecipitation studies. As shown in Figure 2C, deletion of the Pax3 paired domain (PD) results in the loss of interaction with Sox10 (lanes 4 and 7). Pax3 deletion constructs that contain the paired box encode proteins capable of interacting with Sox10 (lanes 10, 13, 16) and the PD alone is sufficient to mediate this interaction (lane 16).

Similar experiments were performed to identify the interaction domain within Sox10. As shown in Figure 2D, the HMG domain is required for interaction with Pax3 (lanes 1 and 7), while the regions C- and N-terminal to the HMG domain were insufficient to mediate this interaction (lanes 4 and 10, respectively). Therefore, the PD of Pax3 and the HMG domain of Sox10 are required for protein–protein interaction.

The physical interaction between Pax3 and Sox10 contributes to synergistic activation of c-RET
We examined the relevance of the physical interaction between Pax3 and Sox10 in the context of the c-RET enhancer by engineering mutations in the Pax3 and/or Sox10 binding sites and asking whether protein–protein interactions were sufficient for recruitment and synergistic activation. As shown previously (12), Pax3 and Sox10 induce synergistic activation of a reporter construct containing the Pax/Sox enhancer (Fig. 3A and B). Mutation of the Pax3 PD binding site (‘P’, Fig. 3A and B) abolishes Pax3 binding (12) and transactivation (Fig. 3B). Modest activation induced by Sox10 remains unaffected, and activation is not enhanced by the addition of Pax3 and Sox10 together. Hence, Sox10 is not capable of recruiting Pax3 to this enhancer in the absence of direct Pax3 binding to DNA.

View larger version (28K): [in this window] [in a new window]   Figure 3. Pax3 DNA binding and Sox10 transactivation are needed for synergistic expression from the c-RET enhancer. (A) Schematic of luciferase reporter construct used in (B) and (C). This vector includes 1 kb proximal upstream sequence from the human c-RET gene that is insufficient to respond to Pax3 or Sox10 (12). A 280 bp upstream sequence (corresponding to 1404–1695 of GenBank AF032124) containing the 36 bp Pax3/Sox10 enhancer is included. The Sox10 and Pax3 binding sites are denoted by an ‘S’ and ‘P’, respectively. (B) Cotransfection studies using Pax3 and/or Sox10 expression constructs with the reporter construct shown in (A) are expressed as fold activation compared to transfections lacking Pax3 or Sox10. The 280 bp upstream sequence of the reporter constructs are shown schematically, either with both binding sites wild-type (top set), with the Pax3 site mutated, shown as an ‘X’ (second set), with the Sox10 site mutated (third set), or both mutated (bottom set). An intact Pax3 binding site is required for Pax3 activation and for synergy with Sox10. (C) Similar cotransfection experiments using the reporter construct shown in (A) were performed with wild-type Pax3 and Sox10 expression constructs or with those encoding truncated proteins lacking transcriptional activation domains (Pax3TA or Sox10TA). The Sox10 transcriptional activation domain is required for synergy with Pax3. All data points in B,C represent average of three experiments performed in triplicate, normalized for transfection efficiency, ±SD.

 
Conversely, mutation of the Sox10 binding site (‘S’, Fig. 3A and B) destroys Sox10 binding (12) and transactivation (Fig. 3B). Modest activation induced by Pax3 remains unchanged. Importantly, addition of Sox10 with Pax3 results in further activation, albeit to levels below that seen with intact Pax3 and Sox10 binding sites. This enhanced activation in the presence of Sox10 occurs despite the fact that Sox10 has no effect in the absence of Pax3. This result suggests that Pax3 bound to DNA is able to recruit Sox10, resulting in enhanced transcriptional activation.

The transcriptional activation domain of Sox10 is required for synergistic activation of c-RET
Additional transfection experiments demonstrate that the C-terminal transactivation domains of both Pax3 and Sox10 contribute to activation and that this region of Sox10 is required for synergistic activation in co-transfection assays. We transfected a reporter construct containing the c-RET enhancer element (Fig. 3A) with Pax3, Sox10 or both, or with vectors that encode proteins lacking the transactivation domains. Pax3 and Sox10 cooperate to produce a synergistic effect (Fig. 3C). Pax3 or Sox10 proteins devoid of their transactivation domains (Pax3TA or Sox10TA) are unable to activate reporter gene expression (Fig. 3C). The synergistic effect of adding Sox10 to Pax3 is entirely lost when the transactivation domain of Sox10 is removed (Pax3+Sox10TA, Fig. 3C). However, deletion of the transactivation domain of Pax3 only partially disrupts the synergistic activation seen in the presence of Pax3 and Sox10 (Sox10+Pax3TA, Fig. 3C). In addition to the importance of Pax3/Sox10 physical interactions, these data suggest that the DNA binding properties of Pax3 mediated by the PD together with the transactivation properties of Sox10 mediated by the C-terminal domain are critical for synergistic activation of the c-RET enhancer.

Human genotype–phenotype correlation is explained by an understanding of Pax3/Sox10 interactions
In humans, various SOX10 mutations can result in distinct phenotypes. Mutations that result in truncation of the protein such that the transactivation domain is lost (such as E189X depicted in Fig. 4A) cause Waardenburg–Shah (Waardenburg–Hirschsprung) syndrome (OMIM 277580) characterized by both enteric ganglia and melanocyte defects (14–17). In the mouse, the DOM mutation (shown schematically in Fig. 4A) results in loss of the transactivation domain, and affected mice have coat color abnormalities and megacolon (18–20). In some Waardenburg–Shah patients, a 6 bp insertion in the HMG domain leads to a duplication of Arg161Leu162 and alters the spacing between two highly conserved amino acids (Sox10 482ins6, Fig. 4A) (15). This mutation disrupts Sox10 DNA binding (21). In contrast, a missense mutation in the HMG box of Sox10 (S135T, Fig. 4A) that also results in the loss of DNA binding capacity by the mutated protein causes YDBS (OMIM 601706). These patients demonstrate melanocyte deficiencies including pigmentation and hearing defects but have normal enteric ganglia (13,15).

View larger version (30K): [in this window] [in a new window]   Figure 4. Analysis of Sox10 mutants found in Waardenburg–Shah and YDBS. (A) Schematic of wild-type, S135T, E189X, 482ins6 and DOM mutant Sox10 proteins. The HMG DNA binding domain and the C-terminal transactivation (TA) domains are shown. S135T is a missense mutation in the HMG domain. E189X results in premature stop after the HMG domain. 482ins6 is caused by a 6 bp insertion (GTCTCCT) that results in two extra amino acids near the end of the HMG domain. The murine DOM mutation results in a frame-shift (shaded) and premature stop. A western blot against HA antigen demonstrates efficient expression of wild-type and mutant constructs in 293T cells. Equal volumes of cell lysates are loaded in each lane after transfection with the indicated expression construct. (B) Immunoprecipitations of in-vitro translated [35S]methionine-labeled Sox10 proteins (wild-type or mutant) with -Pax3 Ab indicates that Pax3 can associate with S135T or E189X mutant Sox10 proteins (lanes 1, 4 and 7) but not Sox10 482ins6 (lane 9). Negative controls (lanes 2, 5, 8 and 11) lack Pax3 protein. Positive controls (lanes 3, 6, 9 and 12) represent direct immunoprecipitation of Sox10HA proteins with -HA Ab. (C) Cotransfection experiments (performed as in Fig. 3) using the c-RET reporter (as in Fig. 3A) indicate that Sox10S135T induces synergistic activation when cotransfected with Pax3 (hatched bar), while Sox10E189X and Sox10 482ins6 do not. Neither mutant form of Sox10 induces activation on its own. (D) Cotransfection experiments performed using the MITF reporter that is activated by Pax3 and Sox10 reveal that Sox10S135T is unable to induce synergistic activation with Pax3 (hatched bar). Sox10E189X and Sox10 482ins6 are also unable to activate expression alone or when co-transfected with Pax3. Note the difference in response of the c-RET and MITF reporters to Sox10S135T when cotransfected with Pax3 (hatched bars, C and D).

 
We examined the effects of these distinct SOX10 mutations on gene expression in enteric ganglia and in melanocytes. Accumulated data suggests that Sox10 activation of c-RET is a critical step during enteric ganglia development (12), while Sox10 activation of MITF is required for proper melanocyte formation (3,4,22). Pax3 functions with Sox10 to activate both c-RET and MITF (3–5). Therefore, we tested the effects of adding mutant forms of Sox10 to Pax3 in the context of c-RET and MITF reporter constructs. Constructs expressing these proteins are efficiently expressed in mammalian cells (Fig. 4A).

Both the truncation mutant (E189X) and the missense mutant (S135T) are able to interact with Pax3 (Fig. 4B). While it is not surprising that the E189X mutation does not affect the interaction with Pax3 (since the HMG box is unaffected), it is important to emphasize that the S135T mutation specifically interrupts DNA binding by the HMG domain (13) without destroying the ability of this domain to interact with Pax3 (Fig. 4B). In contrast, the Sox10 482ins6 mutant disrupts DNA binding (21) and also results in the inability to interact with Pax3 (Fig. 4B).

Transfection assays using the c-RET reporter construct indicate that the S135T, E189X and 482ins6 mutations all eliminate the ability of Sox10 to directly activate transcription (Fig. 4C). However, the S135T mutant protein is able to function with Pax3 to induce synergistic activation (hatched bar, Fig. 4C). This is consistent with our finding that a specific Sox10 DNA binding site is not required (Fig. 3B), and with our suggestion that Pax3 can recruit Sox10. Both the E189X mutation that eliminates the transactivation domain of Sox10 and the 482ins6 mutation that fails to bind to Pax3 produce proteins that are unable to function synergistically with Pax3 (Fig. 4C).

Identical experiments performed using a previously described MITF reporter construct (3) yielded distinct results. This reporter contains adjacent Pax3 and Sox10 binding sites, but the nature and spacing of these sites is distinct from those described in the c-RET gene. Pax3 binds to the MITF enhancer via the homeodomain (HD) as opposed to the PD and it is unclear if Pax3 bound to DNA in this fashion is capable of recruiting Sox10.

Pax3 and Sox10 are each able to induce modest activation of the MITF reporter (Fig. 4D). None of the Sox10 mutants is able to induce activation individually (Fig. 4D). Pax3 and Sox10 induce synergistic activation when added together, as has been previously described (3,4). The E189X and 482ins6 mutants are unable to enhance Pax3 activation of the MITF reporter. The S135T mutant is also unable to enhance activation induced by Pax3 alone (hatched bar, Fig. 4D), in contrast to what is observed with the c-RET reporter (hatched bar, Fig. 4C). These results are consistent with a model in which Pax3 is unable to recruit Sox10 in the context of the MITF enhancer.

These data provide a molecular explanation for the phenotype caused by S135T mutations. Our results suggest that enteric ganglia development is unaffected due to the retained ability of Pax3 to recruit the mutant form of Sox10 to the c-RET enhancer, while melanocyte development is perturbed because MITF expression requires intact Sox10 DNA binding. In contrast, SOX10 mutations that result in loss of the transactivation domain or failure to interact with Pax3 affect both c-RET and MITF expression, and hence result in enteric ganglia and melanocyte defects.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
In this report, we describe the ability of Pax3 and Sox10 to physically interact in vitro and in vivo. Previously, we have described synergistic activation of the human c-RET gene by Pax3 and Sox10, and we have defined binding sites for these transcription factors located close to one another in an upstream enhancer (12). Subsequently, others have demonstrated that Pax3 and Sox10 cooperate to activate the MITF gene (3,4), suggesting that Pax3 and Sox10 may function together generally to regulate neural crest developmental pathways. The enhancer that we have identified in the c-ret gene is highly conserved through evolution, but is not sufficient to initiate expression in vivo in transgenic mice (our unpublished observations) and likely functions in the context of additional regulatory elements to maintain c-ret expression (12,23).

Pax3 and Sox10 proteins interact through contacts mediated by the PD and HMG domain, respectively. This is in accord with recent data derived from the study of Pax6 and Sox2 function during ocular development. Pax6 and Sox2 physically interact via the PD and HMG domain and mediate synergistic transcriptional activation of -crystallin gene expression in the developing lens (24). Interestingly, several members of the Pax and Sox families are co-expressed in specific organs or tissues during development, suggesting that additional examples of physical interaction and synergistic gene activation may be identified. For example, Pax5 and Sox4 are co-expressed in developing B-cells and inactivation of either gene in the mouse leads to similar B-cell developmental defects (25,26).

Sox proteins are weak activators of transcription (14,27) and are likely to function as components of larger transcriptional complexes. The HMG domain binds in the minor groove of DNA and Sox proteins can function to bend target DNA thus altering the architecture of adjacent protein docking sites. Hence, Sox proteins can potentially recruit partners to enhancer sequences by direct protein–protein interaction or by altering protein–DNA interactions. Cofactor binding sites are often located in close proximity to the HMG binding sequence, as is seen in the case of Sox2 and Pax6 binding to the -crystallin promoter (24), and Sox2 and Oct3/4 in the FGF4 (28) and UTF1 (29) promoters.

Pax3 is capable of interacting with DNA in diverse ways. The Pax3 protein includes two distinct DNA binding domains (the PD and HD) that can interact with DNA independently or together (30). The PD itself is composed of two subdomains that bind to DNA in complex fashions (31). It is likely that Pax3 adopts different conformations depending upon the precise DNA binding site and the mechanism of DNA protein interaction. Hence, it is of interest that Pax3 interacts with the c-RET enhancer via PD–DNA interactions, while it binds to the MITF promoter via HD–DNA interactions.

The phenotypes produced by mutations in SOX10 are sensitive to gene dosage. In both humans and mice, homozygous deletion of Sox10 produces a far more severe phenotype than that seen in heterozygotes. However, it is unlikely that the differences in phenotypes produced by the Sox10 S135T, E189X and 482ins6 mutations are simply related to dosage effects. The mutant proteins are efficiently expressed in mammalian cells (Fig. 4A and B), and none are capable of activating transcription on their own. In the context of the MITF enhancer, the mutant proteins are inert, either alone or in association with Pax3. On the other hand, in the context of the c-RET enhancer, the S135T mutant behaves similar to wild-type Sox10 in its ability to cooperate with Pax3.

Our results comparing Sox10 and Pax3 activation of c-RET and MITF enhancers are consistent with a model in which Sox10 binding to DNA, and resulting alteration of DNA architecture, is required for synergy with Pax3 in the context of the MITF enhancer. Perhaps relatively weak Pax3 HD-DNA interactions are stabilized by Sox10 binding to an adjacent site and resulting conformational changes. In contrast, Pax3 PD binding to the c-RET enhancer is relatively less affected by adjacent DNA binding by Sox10. Protein–protein interactions between Sox10 and Pax3 predominate to recruit the transactivation activity of Sox10 to the transcriptional complex.

Previous reports have examined the effect of Pax3 and Sox10 activation of MITF. In melanoma and NIH-3T3 cells, Verastegui et al. did not observe synergistic activation (22). On the other hand, Bondurand et al. reported synergistic activation of a MITF promoter/enhancer construct in Hela cells (3). Synergy was reduced by greater than 20-fold when putative Sox10 binding sites were mutated. However, residual Sox10 binding to cryptic sites was not assessed, and at least one residual Sox10 binding site was not mutated (sx2, -282 to -272 upstream of the translational start) (4). The residual level of synergy that these authors observed could be cell type-specific or could relate to residual DNA binding by Sox10. Our result using the S135T mutation suggests that DNA binding by Sox10 is required for synergy with Pax3 in the context of the MITF promoter/enhancer. While direct experimental verification and analysis of DNA structure will be required to validate or refute our model, it is reasonable to conclude that Pax3 binding to the c-RET enhancer is capable of recruiting Sox10, while Pax3 binding to the MITF enhancer is not. This result in reminiscent of the manner in which Pax5 recruits ETS family transcription factors to the mb-1 enhancer when bound to DNA via its PD (32).

In summary, we have demonstrated the ability of Pax3 and Sox10 to interact both in vitro and in vivo, and we have explored the physiologic significance of this interaction in the context of two enhancer regions derived from downstream regulated genes. Our results underscore the complexity of protein–protein and protein–DNA interactions in the context of transcriptional control. The detailed analysis of these interactions in the case of Pax3, Sox10, MITF and c-RET provide a compelling model for explaining the distinct phenotypes produced in humans with SOX10 mutations. Specifically, we suggest that the lack of enteric ganglia defects in patients with YDBS is due to the retained ability of Sox10 mutants in these patients to interact with Pax3 and to activate expression of c-RET, despite defective DNA binding capabilities.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Isolation of murine c-ret upstream genomic region
Exon 2 of the mouse c-ret cDNA was utilized as a probe for a BAC screen (Incyte Genomics, Newark, DE, USA). The upstream sequence, 4.8 kb, was obtained from BAC no. 25180, and sequenced. A luciferase reporter construct was engineered as a derivative of pGL2-basic (Promega, Madison, WI, USA).

Cell culture and transfection
293T cells (American Type Culture Collection, Manassas, VA, USA) were maintained in DME with 10% FBS (Invitrogen Life Technologies, Carlsbad, CA, USA) and transfected at 50% confluency. A total of 0.5 µg of DNA was mixed with 10 µl of Effectene reagent (Qiagen Inc, Valencia, CA, USA). Luciferase activity (Luciferase assay kit, Promega) was normalized for transfection efficiency using pCMVß (BD biosciences/Clontech, Palo Alto, CA, USA).

Generation of constructs
pcDNA3-Sox10HA was created by PCR using primers 5'-GCC CTC GAG CGA GGT CGG GAT AGA GTC GTA TAT AC (forward) and 5'-ACA CTA TAG AAG GTA CGC CTG C (reverse) and pCMV-Sox10 (derived from dcgs10-1, provided by M. Southard-Smith and W. Pavan) as template. The MITF reporter construct pGL3-MITF was provided by M. Girard and M. Goossens (3). Pax3 and Sox10 deletions were generated by PCR, and verified by sequencing. Using the numbering system from GenBank accession no. X59358 for Pax3, deletions were generated using forward primers at bp 290, 690, 860 or 1040, or with reverse primers at bp 310, 680, 860 or 1060. For Sox10, using the numbering system from GenBank accession no. XM_128139, deletions were generated using forward primers at bp 340 or 560 with a reverse primer at bp 1440, or with a reverse primer at bp 560 or 340.

RET promoter constructs were prepared from RETD15 (12). The RET 280 deletion construct (Fig. 3A) contains the proximal promoter (bp 4050–5100, GenBank AF032124) in addition to PCR generated upstream sequence (bp 1404–1695). Mutations in the Pax and/or Sox binding sites were made by changing the consensus sequences to EcoRI restriction sites (Stratagene, Quickchange kit, catalog no. 200518-5). The S135T and E189X Sox10HA mutant constructs were also generated by site directed mutagenesis, changing the codons as described (13,15). The Sox10 482ins6 expression construct with or without an HA tag was made from a modified from pECE Sox10 482ins6 (provided by M. Goossens and N. Bondurand) (15,21).

Immunoprecipitation
Proteins were obtained from transfected cells or from in vitro translation/transcription (TNT, Promega). Full length Pax3 was generated from pBH 3.2 (1). Sox10 was generated from pCMV-Sox10 (33). Cell protein was generated by transfecting pCMV-Pax3 (34), pCMV-Pax3HA (34), pCMV-Sox10, pCMV-Sox10HA and/or deletion constructs described above. Cell lysates or in vitro produced proteins were incubated at 4°C for 1 h in 40 mM HEPES, 100 mM KCl, 40% glycerol, 2 mM ß-ME, 0–0.5% NP-40. Antibody [either Pax3 pAb (35), Sox10 pAb, catalog no. AB5774, Chemicon International, Temecula, CA, USA, or HA mAb 12CA5, catalog no. 1583 816, Roche, Indianapolis, IN, USA] and Protein A/G agarose beads (catalog no. IP10, Oncogene Research Products, Cambridge, MA, USA) were incubated for 1 h and separated on a 10% acrylamide Bis/Tris gel. Western blotting (Western Breeze, Invitrogen) utilized HA mAb.

	ACKNOWLEDGEMENTS

 
We want to thank M. Goossens, N. Bondurand, M. Girard, M. Southard-Smith and W. Pavan for constructs. This work was supported by grants from the WW Smith Foundation and the NIH (DK59176, HL61475) to J.A.E. and from the American Heart Association to D.L. and J.A.E.

	FOOTNOTES

 
^* To whom correspondence should be addressed at: 954 BRB II, 421 Curie Boulevard, Philadelphia, PA 19104, USA. Tel: +1 2158988731; Fax: +1 2155732094; Email: epsteinj{at}mail.med.upenn.edu

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 

1. Goulding, M.D., Chalepakis, G., Deutsch, U., Erselius, J.R. and Gruss, P. (1991) Pax-3, a novel murine DNA binding protein expressed during early neurogenesis. EMBO J., 10, 1135–1147.[Web of Science][Medline]

2. Epstein, D.J., Vogan, K.J., Trasler, D.G. and Gros, P. (1993) A mutation within intron 3 of the Pax-3 gene produces aberrantly spliced mRNA transcripts in the splotch (Sp) mouse mutant. Proc. Natl Acad. Sci. USA, 90, 532–536.[Abstract/Free Full Text]

3. Bondurand, N., Pingault, V., Goerich, D.E., Lemort, N., Sock, E., Caignec, C.L., Wegner, M. and Goossens, M. (2000) Interaction among SOX10, PAX3 and MITF, three genes altered in Waardenburg syndrome. Hum. Mol. Genet., 9, 1907–1917.[Abstract/Free Full Text]

4. Potterf, S.B., Furumura, M., Dunn, K.J., Arnheiter, H. and Pavan, W.J. (2000) Transcription factor hierarchy in Waardenburg syndrome: regulation of MITF expression by SOX10 and PAX3. Hum. Genet., 107, 1–6.[CrossRef][Web of Science][Medline]

5. Watanabe, A., Takeda, K., Ploplis, B. and Tachibana, M. (1998) Epistatic relationship between Waardenburg syndrome genes MITF and PAX3. Nat. Genet., 18, 283–286.[CrossRef][Web of Science][Medline]

6. Tassabehji, M., Newton, V. and Read, A. (1994) Waardenburg syndrome type 2 caused by mutations in the human microphthalmia (MITF) gene. Nat. Genet., 8, 251–255.[CrossRef][Web of Science][Medline]

7. Avantaggiato, V., Dathan, N.A., Grieco, M., Fabien, N., Lazzaro, D., Fusco, A., Simeone, A. and Santoro, M. (1994) Developmental expression of the RET protooncogene. Cell Growth Diff., 5, 305–311.[Abstract]

8. Durbec, P.L., Larsson-Blomberg, L.B., Schuchardt, A., Costantini, F. and Pachnis, V. (1996) Common origin and developmental dependence on c-ret of subsets of enteric and sympathetic neuroblasts. Development, 122, 349–358.[Abstract]

9. Pachnis, V., Mankoo, B. and Costantini, F. (1993) Expression of the c-ret proto-oncogene during mouse embryogenesis. Development, 119, 1005–1017.[Abstract]

10. Schuchardt, A., D'Agati, V., Pachnis, V. and Costantini, F. (1996) Renal agenesis and hypodysplasia in ret-k-mutant mice result from defects in ureteric bud development. Development, 122, 1919–1929.[Abstract]

11. Wegner, M. (1999) From head to toes: the multiple facets of Sox proteins. Nucl. Acids Res., 27, 1409–1420.[Abstract/Free Full Text]

12. Lang, D., Chen, F., Milewski, R., Li, J., Lu, M.M. and Epstein, J.A. (2000) Pax3 is required for enteric ganglia formation and functions with Sox10 to modulate expression of c-ret. J. Clin. Invest., 106, 963–971.[Web of Science][Medline]

13. Bondurand, N., Kuhlbrodt, K., Pingault, V., Enderich, J., Sajus, M., Tommerup, N., Warburg, M., Hennekam, R.C., Read, A.P., Wegner, M. et al. (1999) A molecular analysis of the yemenite deaf–blind hypopigmentation syndrome: SOX10 dysfunction causes different neurocristopathies. Hum. Mol. Genet., 8, 1785–1789.[Abstract/Free Full Text]

14. Kuhlbrodt, K., Herbarth, B., Sock, E., Hermans-Borgmeyer, I. and Wegner, M. (1998) Sox10, a novel transcriptional modulator in glial cells. J. Neurosci., 18, 237–250.[Abstract/Free Full Text]

15. Pingault, V., Bondurand, N., Kuhlbrodt, K., Goerich, D.E., Prehu, M.O., Puliti, A., Herbarth, B., Hermans-Borgmeyer, I., Legius, E., Matthijs, G. et al. (1998) SOX10 mutations in patients with Waardenburg–Hirschsprung disease. Nat. Genet., 18, 171–173.[CrossRef][Web of Science][Medline]

16. Shah, K.N., Dalal, S.J., Desai, M.P., Sheth, P.N., Joshi, N.C. and Ambani, L.M. (1981) White forelock, pigmentary disorder of irides, and long segment Hirschsprung disease: possible variant of Waardenburg syndrome. J. Pediatr., 99, 432–435.[Web of Science][Medline]

17. Sham, M.H., Lui, V.C., Chen, B.L., Fu, M. and Tam, P.K. (2001) Novel mutations of SOX10 suggest a dominant negative role in Waardenburg–Shah syndrome. J. Med. Genet., 38, E30–E34.[CrossRef][Medline]

18. Southard-Smith, E.M., Kos, L. and Pavan, W.J. (1998) Sox10 mutation disrupts neural crest development in Dom Hirschsprung mouse model. Nat. Genet., 18, 60–64.[CrossRef][Web of Science][Medline]

19. Herbarth, B., Pingault, V., Bondurand, N., Kuhlbrodt, K., Hermans-Borgmeyer, I., Puliti, A., Lemort, N., Goossens, M. and Wegner, M. (1998) Mutation of the Sry-related Sox10 gene in Dominant megacolon, a mouse model for human Hirschsprung disease. Proc. Natl Acad. Sci. USA, 95, 5161–5165.[Abstract/Free Full Text]

20. Lane, P.W. and Liu, H.M. (1984) Association of megacolon with a new dominant spotting gene (Dom) in the mouse. J. Hered., 75, 435–439.[Abstract/Free Full Text]

21. Kuhlbrodt, K., Schmidt, C., Sock, E., Pingault, V., Bondurand, N., Goossens, M. and Wegner, M. (1998) Functional analysis of Sox10 mutations found in human Waardenburg–Hirschsprung patients. J. Biol. Chem., 273, 23033–23038.[Abstract/Free Full Text]

22. Verastegui, C., Bille, K., Ortonne, J.P. and Ballotti, R. (2000) Regulation of the microphthalmia-associated transcription factor gene by the Waardenburg syndrome type 4 gene, SOX10. J. Biol. Chem., 275, 30757–30760.[Abstract/Free Full Text]

23. Sukumaran, M., Waxman, S.G., Wood, J.N. and Pachnis, V. (2001) Flanking regulatory sequences of the locus encoding the murine GDNF receptor, c-ret, directs lac Z (beta-galactosidase) expression in developing somatosensory system. Dev. Dyn., 222, 389–402.[CrossRef][Web of Science][Medline]

24. Kamachi, Y., Uchikawa, M., Tanouchi, A., Sekido, R. and Kondoh, H. (2001) Pax6 and SOX2 form a co-DNA-binding partner complex that regulates initiation of lens development. Genes Dev., 15, 1272–1286.[Abstract/Free Full Text]

25. Urbanek, P., Wang, X.Q., Fetka, I., Wagner, E.F. and Busslinger, M. (1994) Complete block of early B cell differentiation and altered patterning of the posterior midbrain in mice lacking Pax5/BSAP. Cell, 79, 901–912.[CrossRef][Web of Science][Medline]

26. Schilham, M., Oosterwegel, M., Moerer, P., Ya, J., de Boer, P., van de Wetering, M., Verbeek, S., Lamers, W., Kruisbeek, A., Cumano, A. et al. (1996) Defects in cardiac outflow tract formation and pro-B-lymphocye expansion in mice lacking Sox-4. Nature, 380, 711–714.[CrossRef][Medline]

27. Kamachi, Y., Uchikawa, M. and Kondoh, H. (2000) Pairing SOX off: with partners in the regulation of embryonic development. Trends Genet., 16, 182–187.[CrossRef][Web of Science][Medline]

28. Yuan, H., Corbi, N., Basilico, C. and Dailey, L. (1995) Developmental-specific activity of the FGF-4 enhancer requires the synergistic action of Sox2 and Oct-3. Genes Dev., 9, 2635–2645.[Abstract/Free Full Text]

29. Nishimoto, M., Fukushima, A., Okuda, A. and Muramatsu, M. (1999) The gene for the embryonic stem cell coactivator UTF1 carries a regulatory element which selectively interacts with a complex composed of Oct-3/4 and Sox-2. Mol. Cell. Biol., 19, 5453–5465.[Abstract/Free Full Text]

30. Jun, S. and Desplan, C. (1996) Cooperative interactions between paired domain and homeodomain. Development, 122, 2639–2650.[Abstract]

31. Epstein, J.A., Glaser, T., Cai, J., Jepeal, L., Walton, D.S. and Maas, R.L. (1994) Two independent and interactive DNA binding subdomains of the PAX6 paired domain are regulated by alternative splicing. Genes Dev., 8, 2022–2034.[Abstract/Free Full Text]

32. Fitzsimmons, D., Hodsdon, W., Wheat, W., Maira, S., Wasylyk, B. and Hagman, J. (1996) Pax-5 (BSAP) recruits Ets proto-oncogene family proteins to form functional ternary comlexes on a B-cell-specific promoter. Genes Dev., 10, 2198–2211.[Abstract/Free Full Text]

33. Southard-Smith, E.M., Angrist, M., Ellison, J.S., Agarwala, R., Baxevanis, A.D., Chakravarti, A. and Pavan, W.J. (1999) The Sox10(Dom) mouse: modeling the genetic variation of Waardenbur–Shah (WS4) syndrome. Genome Res., 9, 215–225.[Medline]

34. Epstein, J.A., Shapiro, D.N., Cheng, J., Lam, P.Y. and Maas, R.L. (1996) Pax3 modulates expression of the c-Met receptor during limb muscle development. Proc. Natl Acad. Sci. USA, 93, 4213–4218.[Abstract/Free Full Text]

35. Li, J., Liu, K.C., Jin, F., Lu, M.M. and Epstein, J.A. (1999) Transgenic rescue of congenital heart disease and spina bifida in Splotch mice. Development, 126, 2495–2503.[Abstract]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				A. J. Thomas and C. A. Erickson FOXD3 regulates the lineage switch between neural crest-derived glial cells and pigment cells by repressing MITF through a non-canonical mechanism Development, June 1, 2009; 136(11): 1849 - 1858. [Abstract] [Full Text] [PDF]

				Z. Mou, A. R. Tapper, and P. D. Gardner The Armadillo Repeat-containing Protein, ARMCX3, Physically and Functionally Interacts with the Developmental Regulatory Factor Sox10 J. Biol. Chem., May 15, 2009; 284(20): 13629 - 13640. [Abstract] [Full Text] [PDF]

				M.-M. Garcia-Barcelo, C. S.-m. Tang, E. S.-w. Ngan, V. C.-h. Lui, Y. Chen, M.-t. So, T. Y.-y. Leon, X.-p. Miao, C. K.-y. Shum, F.-q. Liu, et al. Genome-wide association study identifies NRG1 as a susceptibility locus for Hirschsprung's disease PNAS, February 24, 2009; 106(8): 2694 - 2699. [Abstract] [Full Text] [PDF]

				K. Kanno, M. K. Wu, D. S. Agate, B. J. Fanelli, N. Wagle, E. F. Scapa, C. Ukomadu, and D. E. Cohen Interacting Proteins Dictate Function of the Minimal START Domain Phosphatidylcholine Transfer Protein/StarD2 J. Biol. Chem., October 19, 2007; 282(42): 30728 - 30736. [Abstract] [Full Text] [PDF]

				S. Yokoyama, K. Takeda, and S. Shibahara Functional Difference of the SOX10 Mutant Proteins Responsible for the Phenotypic Variability in Auditory-Pigmentary Disorders J. Biochem., October 1, 2006; 140(4): 491 - 499. [Abstract] [Full Text] [PDF]

				C. S. K. Mayanil, A. Pool, H. Nakazaki, A. C. Reddy, B. Mania-Farnell, B. Yun, D. George, D. G. McLone, and E. G. Bremer Regulation of Murine TGFbeta2 by Pax3 during Early Embryonic Development J. Biol. Chem., August 25, 2006; 281(34): 24544 - 24552. [Abstract] [Full Text] [PDF]

				N. Bondurand, D. Natarajan, A. Barlow, N. Thapar, and V. Pachnis Maintenance of mammalian enteric nervous system progenitors by SOX10 and endothelin 3 signalling Development, May 15, 2006; 133(10): 2075 - 2086. [Abstract] [Full Text] [PDF]

				S. Wissmuller, T. Kosian, M. Wolf, M. Finzsch, and M. Wegner The high-mobility-group domain of Sox proteins interacts with DNA-binding domains of many transcription factors. Nucleic Acids Res., January 1, 2006; 34(6): 1735 - 1744. [Abstract] [Full Text] [PDF]

				F. Puppo, M. Musso, D. Pirulli, P. Griseri, T. Bachetti, S. Crovella, G. Patrone, I. Ceccherini, and R. Ravazzolo Comparative genomic sequence analysis coupled to chromatin immunoprecipitation: a screening procedure applied to search for regulatory elements at the RET locus Physiol Genomics, November 17, 2005; 23(3): 269 - 274. [Abstract] [Full Text] [PDF]

				S. E. Owens, K. W. Broman, T. Wiltshire, J. B. Elmore, K. M. Bradley, J. R. Smith, and E. M. Southard-Smith Genome-wide linkage identifies novel modifier loci of aganglionosis in the Sox10Dom model of Hirschsprung disease Hum. Mol. Genet., June 1, 2005; 14(11): 1549 - 1558. [Abstract] [Full Text] [PDF]

				M. Garcia-Barcelo, R. W. Ganster, V. C.H. Lui, T. Y.Y. Leon, M.-T. So, A. M.F. Lau, M. Fu, M.-H. Sham, J. Knight, M. S. Zannini, et al. TTF-1 and RET promoter SNPs: regulation of RET transcription in Hirschsprung's disease Hum. Mol. Genet., January 15, 2005; 14(2): 191 - 204. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (35) Request Permissions Google Scholar Articles by Lang, D. Articles by Epstein, J. A. Search for Related Content PubMed PubMed Citation Articles by Lang, D. Articles by Epstein, J. A. Social Bookmarking          What's this?

Online ISSN 1460-2083 - Print ISSN 0964-6906

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics