Human Connexin 32, a gap junction protein ... - Semantic Scholar

Published by Oxford University Press

Human Molecular Genetics, 2001, Vol. 10, No. 24 2783–2795

Human Connexin 32, a gap junction protein altered in the X-linked form of Charcot–Marie–Tooth disease, is directly regulated by the transcription factor SOX10 Nadége Bondurand, Mathilde Girard, Véronique Pingault, Nicole Lemort, Odile Dubourg1 and Michel Goossens* Génétique Moléculaire et Physiopathologie, INSERM U468, et Laboratoire de Biochimie et Génétique Moléculaire, AP-HP, Hôpital Henri Mondor, 94010 Créteil Cedex, France and 1INSERM U289, Hôpital de la Pitié-Salpétrière, Paris, France Received July 18, 2001; Revised and Accepted September 26, 2001

Mutations in SOX10, a transcription modulator crucial in the development of the enteric nervous system (ENS), melanocytes and glial cells, are found in Shah–Waardenburg syndrome (WS4), a neurocristopathy that associates intestinal aganglionosis, pigmentation defects and sensorineural deafness. Expression of MITF and RET, two genes that play important roles during melanocyte and ENS development, respectively, are controlled by SOX10. The observation that some WS4 patients present with myelination defects of the central and peripheral nervous systems correlates with the recent finding that P0, a major component of the peripheral myelin, is another transcriptional target of SOX10. These phenotypic features suggest that SOX10 could regulate expression of other genes involved in the myelination process as well. Thus, we tested the ability of SOX10 to regulate expression of MBP, PMP22 and Connexin 32, three major proteins of the peripheral myelin. Our study shows that this factor, in synergy with EGR2, strongly activates Cx32 expression in vitro by directly binding to its promoter. In agreement with this finding, SOX10 and EGR2 mutants identified in patients with peripheral myelin defects fail to transactivate the Cx32 promoter. Moreover, we show that a mutation of the Cx32 promoter previously described in a patient with the X-linked form of Charcot–Marie–Tooth (CMTX) disease impairs SOX10 function. In addition to providing new insights into the molecular mechanisms underlying some of the peripheral myelin defects observed in CMTX disease, these results further extend the spectrum of genes that are regulated by SOX10.

INTRODUCTION The SOX proteins belong to the HMG box superfamily of DNA-binding proteins and are highly conserved across evolution. These factors play key roles in decisions of cell fate during diverse developmental processes, and display properties of both classical transcription factors and architectural components of chromatin (reviewed in 1,2). Among them, SOX10 was initially shown to be predominantly expressed in glial cells of the nervous system (3), but further investigation demonstrated its broader implication during development. SOX10 is first expressed widely in cells of the neural crest at the time of their emergence, and later in neural crest cells that contribute to the melanocyte lineage and to the forming peripheral nervous system (PNS). Its expression is detected in the enteric, sensory and sympathetic ganglia as well as along nerves in a manner typical for the Schwann cell lineage (3–5). During late stages of embryogenesis, its expression seems to be confined to glial cells (6). In the central nervous system (CNS), SOX10 expression occurs later but progressively increases to reach maximal levels during adulthood, being mainly confined to glia of the oligodendrocyte type (3,5,7). The study of the Dom (dominant megacolon) mouse, an animal model of neurocristopathy in which the Sox10 gene is mutated, confirmed the crucial role of SOX10 during neural crest cells development (4,8,9). Heterozygous animals display intestinal aganglionosis and pigmentation defects such as a white belly spot and white paws. The homozygous Dom mice are characterized by a loss of glia and neurons in the PNS and a complete lack of the enteric nervous system (ENS). SOX10 mutations in humans also lead to a combination of neural crest defects that include pigmentation abnormalities, deafness and Hirschsprung disease (intestinal aganglionosis), a phenotype that closely resembles that of the Dom mouse and defines the so-called Waardenburg–Hirschsprung disease or Shah– Waardenburg syndrome (WS4) (10,11). The extensive neural crest dysfunction that affects both mice and humans with SOX10 defects is variable in its phenotypic expression.

*To whom correspondence should be addressed at: Laboratoire de Biochimie et Génétique Moléculaire, Hôpital Henri Mondor, 94010 Créteil Cedex, France. Tel: +33 1 49 81 28 61; Fax: +33 1 49 81 22 19; Email: [email protected] The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors

2784 Human Molecular Genetics, 2001, Vol. 10, No. 24

Indeed, several of the patients with SOX10 mutations present with various neurological phenotypes, in addition to WS4. A severe leukodystrophy compatible with Pelizaeus–Merzbacher disease and a peripheral neuropathy consistent with Charcot–Marie–Tooth (CMT) disease were initially observed in one patient presenting with WS4 (12). Subsequent reports of SOX10 defects associated with both WS4 and various neurological features suggestive of myelin deficiency further confirmed that SOX10 dysfunction could, in some cases, lead to deficiency of myelination in the CNS and PNS (7,13,14). The pleiotropic effects of SOX10 dysfunction can be explained by its role as a modulator of expression of various target genes, in cooperation with different cofactors, such as PAX3, POU3F1 and EGR2 (3). Recently, a few of these transcriptional targets have been identified, illuminating our understanding of the molecular mechanisms that underlie the various phenotypes observed. The demonstration that SOX10 and PAX3 interact directly to activate MITF expression in the melanocyte lineage shed some light on the molecular bases of the auditory–pigmentary abnormalities that are common to various forms of Waardenburg syndrome (15–18). SOX10 was also shown to function with PAX3 to activate transcription of c-RET, one of the major genes that controls morphogenesis and differentiation of the ENS (19,20). Finally, the finding that it regulates MPZ gene (encoding the myelin protein zero P0) expression established a missing link between this transcription factor and cell-specific activation of myelin genes expression in the PNS (21). However, the spectrum of neurological features associated with SOX10 defects makes likely that this transcription factor controls other genes. In particular, the genes encoding the few proteins involved in the myelination process are plausible other SOX10 transcriptional targets. The myelin sheath is a specialized organelle that contains a small number of highly abundant proteins. Among them, the myelin basic protein (MBP) is an integral part of myelin in both the CNS and PNS. In contrast, P0 and PMP22 (peripheral myelin protein 22) are preferentially expressed in Schwann cells of the PNS (22). Alterations of the P0 and PMP22 genes were identified in different forms of CMT disease, a group of inherited chronic progressive conditions fairly frequent (with a population incidence of 1 in 2500) and affecting peripheral nerves (reviewed in 23). CMT type 1 (CMT1), the autosomal dominant demyelinating CMT, is the most common form in Europe. CMT1 is most often associated with a 1.5 Mb duplication of the 17p11.2 chromosome region including the PMP22 gene (CMT1A) or with mutations in the P0 gene (CMT1B) (24). The second most common form of CMT, the X-linked form of CMT disease (CMTX), is caused by mutations in GJB1 (gap junction beta 1), a gene encoding an integral membrane protein, Connexin 32 (Cx32) (25,26). Like MBP, P0 and PMP22, Cx32 is a crucial protein expressed in Schwann cells (27). This gap junction protein belongs to a family of homologous integral membrane proteins that form functional channels allowing rapid transport of ions and small nutrients between coupled cells. In Schwann cells, Cx32 forms intracellular gap junctions between paranodal loops and Schmidt– Lanterman incisures that allow the diffusion of small molecules through the myelin sheath (reviewed in 28). As MBP, PMP22 and Cx32 are major components of the peripheral myelin, we undertook to test whether SOX10 is able to regulate expression of the genes encoding these proteins.

MBP is transcribed from a single promoter, from which the 300 most proximal bp are sufficient to drive MBP expression in glial cells of the CNS and PNS (29). On the contrary, Cx32 and PMP22 are both controlled by two tissue-specific alternate promoters. In glial cells, PMP22 expression is driven by its P1 promoter and Cx32 expression by its P2 promoter (30,31). In both cases, the most proximal part of these promoters is sufficient for PMP22 and Cx32 expression in Schwann cells (32,33). Thus, we tested the possible effect of SOX10 on the proximal promoters of these three genes. Here we show that this factor, in synergy with EGR2, indeed directly activates Cx32 expression in transfection assays, but not that of MBP and PMP22. In agreement with this finding, SOX10 and EGR2 mutants identified in patients having peripheral myelin defects fail to transactivate this promoter. Finally, a mutation of the Cx32 promoter previously described in a CMTX patient impairs SOX10 activation of transcription from this promoter. Therefore, this study provides new insights into the molecular mechanisms underlying the peripheral myelin defects observed in some cases of CMTX disease. RESULTS Regulation of MBP, PMP22 and Cx32 promoters by SOX10 To investigate the possible involvement of SOX10 in the regulation of MBP, PMP22 and Cx32 expression, we analysed the effect of this transcription factor on their promoter activity in transient co-transfection experiments in HeLa cells, a cell line chosen because of absence of endogenous Cx32, MBP, PMP22, SOX10 and its cofactors expression (15,34–36 and data not shown). Approximately 750–1000 bp of glial-specific MBP, PMP22 and Cx32 human promoter sequences were cloned upstream of the luciferase gene (pGL3 basic). These constructs were transfected in HeLa cells alone or together with an expression vector encoding human SOX10. When SOX10 was co-transfected with MBP and PMP22 promoter constructs, no increase in luciferase activity was detected. On the contrary, when SOX10 was co-transfected with the Cx32 promoter construct, an ∼50-fold increase in luciferase activity was detected. Thus, in our experimental conditions, SOX10 increased significantly Cx32 expression, but not that of MBP and PMP22 (Fig. 1). EGR2, PAX3 and POU3F1 act synergistically with SOX10 on the Cx32 glial-specific promoter During glial development and, in particular, certain stages of Schwann cell development, SOX10 is co-expressed with other transcription factors, including PAX3, POU3F1 and EGR2 (3). As SOX10 had previously been reported to modulate the effect of these transcription factors on a synthetic promoter containing binding sites for SOX10 and PAX3, POU3F1 or EGR2, we studied its ability to act in synergy with each of these transcription factors to stimulate MBP, PMP22 and Cx32 expression. This was not the case with MBP and PMP22, as neither addition of PAX3, POU3F1 or EGR2 alone, nor co-transfection of these factors with SOX10 allowed any induction of luciferase activity (data not shown). In contrast, a 15-fold induction in luciferase activity was observed upon co-transfection of EGR2

Human Molecular Genetics, 2001, Vol. 10, No. 24 2785

transcription was observed upon SOX10 or KROX20 induction, indicating that both factors are actually able to activate Cx32 expression in this system (Fig. 2B). SOX10- and EGR2-responsive regions in the Cx32 promoter

Figure 1. Effect of SOX10 on MBP, PMP22 and Cx32 promoters. Luciferase reporter plasmids containing 750–1000 bp of glial-specific MBP, PMP22 and Cx32 human promoters were transfected in HeLa cells with empty pECE vector or pECE/SOX10. Luciferase activity was normalized by measuring β-galactosidase activity. Data from all transfections are presented as SOX10 fold induction upon basal level determined from values obtained from transfections with each luciferase reporter and empty pECE plasmid. Fold induction obtained with empty luciferase reporter plasmid (pGL3 basic) is shown as a negative control. Data are means ± SE of three different experiments performed in triplicates.

with the Cx32 promoter construct, whereas PAX3 and POU3F1 had no effect on this promoter. Thus, EGR2 seems to be the sole of these three factors able to transactivate the Cx32 promoter by itself (Fig. 2A). We next tested the possible cooperative activation of the Cx32 promoter by SOX10 and each of these factors. The combined action of PAX3, POU3F1 or EGR2 with SOX10 resulted in a significant increase in luciferase activity as compared with SOX10 alone. In particular, the combined action of EGR2 and SOX10 resulted in an increase in luciferase activity by a factor of 2.6 as compared with activation of SOX10 alone, showing that the effect that EGR2 and SOX10 exert on the Cx32 promoter is synergistic and not solely additive (Fig. 2A). These results indicate that, in the in vitro system used, SOX10 acts in synergy with EGR2, POU3F1 and PAX3 to activate Cx32 expression. This mode of action is peculiar to the Cx32 glial-specific promoter, as SOX10 acts in synergy with PAX3 but not with POU3F1 and EGR2 on the MITF promoter (15,17 and data not shown). Whether each of these factors act in synergy with SOX10 to activate Cx32 expression in vivo remains to be determined. As EGR2 is the only one of the three tested cofactors that shows an autonomous function on the Cx32 glial-promoter, we further studied the regulatory function of SOX10 and this factor on this promoter. To confirm the results obtained in HeLa cells concerning the regulation of Cx32 by SOX10 and EGR2, we tested whether both factors are able to induce endogenous Cx32 expression in N2A neuroblastoma cells expressing SOX10 or KROX20 (the murine EGR2 homologue) in an inducible manner (21). We assessed Cx32 expression by RT–PCR experiments on N2A RNAs collected before or after induction of SOX10 or KROX20 expression. Cx32 was only weakly expressed in noninduced N2A cells, whereas a significant increase in Cx32

To identify the cis-acting elements accountable for the SOX10 and EGR2 response on the Cx32 promoter, we constructed a 5′ deletion panel of this promoter by deleting sequences of various lengths from the distal end. The deletion constructs were co-transfected with SOX10, EGR2 or both factors to test their ability to be activated by these factors. As shown in Figure 3, deletion of the most distal 204 or 369 bp influenced neither activation by SOX10 and EGR2, nor synergistic action of both factors. In contrast, all activations were completely lost on removal of 581 bp. These results indicate that the region between nucleotides –233 and –21 is responsible for SOX10 and EGR2 responsiveness of the Cx32 promoter. Identification of SOX10-binding sites within the Cx32 proximal promoter Although we had provided firm evidence that SOX10 regulates Cx32 expression in vitro, the possibility remained that requirement for SOX10 is indirect. Examination of the proximal Cx32 promoter sequence (from nucleotide –233 to –21) revealed several potential SOX protein binding sites (37,38). We focused our attention on two of these sites (named S1 and S2; Fig. 4A) because they consist of two non-consensus SOX sites with particular spacing and orientation that are reminiscent of C/C′ sites shown to play a major role in the regulation of the P0 promoter by SOX10 (21,39). Indeed, the S1 and S2 sites show single mismatches as compared to the SOX consensus-binding site sequence, are separated by 4 bp and are oriented towards each other. To confirm the importance of these binding sites for SOX10 function, we mutated each site or both by sitedirected mutagenesis, in such a way that they were replaced by a GC-rich element. The mutant Cx32 promoter constructs were co-transfected with SOX10, EGR2 or both (Fig. 4B). When the enhancing effect of SOX10 alone on the mutated S1, S2 or S1 plus S2 (named S1S2) was tested, SOX10-dependent stimulation of transcription was abolished, indicating that integrity of S1 and S2 is essential for this regulation. Mutations of each or the two SOX10-binding sites had no significant effect on EGR2 activation ability, but the synergistic cooperation between SOX10 and EGR2 was lost in each case. All these results show that integrity of both SOX10-binding sites is required for SOX10 activation, as well as for SOX10 and EGR2 synergistic action on the Cx32 promoter. To confirm that SOX10 binds directly to these nonconsensus SOX-binding sites and to evaluate its binding affinity, we designed electrophoretic mobility shift assays (EMSAs), making use of DNA fragments that correspond to the region of the promoter surrounding S1 and S2. SOX10 is able to bind both sites, but with a higher affinity for S2 (Fig. 4C). As Peirano et al. (39) showed that the C/C′ sites of the P0 promoter are bound by two SOX10 molecules, we tested whether the same phenomenon applies to the Cx32 promoter. To this end, we used DNA fragments corresponding to S1S2 in EMSAs. Both a significant increase in SOX10 binding and a decrease in the mobility of the SOX10–DNA complex were


Figure 2. Regulation of Cx32 expression by SOX10, EGR2, PAX3 and POU3F1. (A) PAX3, POU3F1 and EGR2 act synergistically with SOX10 on the Cx32 promoter. The luciferase reporter plasmid containing 750 bp of the Cx32 glial-specific promoter was transfected in combination with empty pECE vector, pECE/ SOX10 (SOX10), and/or pECE/PAX3 (PAX3), pECE/POU3F1 (POU3F1), pECE/EGR2 (EGR2), as indicated. Data from all transfections are presented as fold induction above basal levels as indicated in Figure 1 and are means ± SE of three different experiments performed in triplicates. (B) Upregulation of endogenous Cx32 expression by SOX10 and KROX20. RT–PCR analysis of cDNAs obtained from N2A Tet-On cell lines capable of inducibly expressing SOX10 or KROX20. Transcription factors induction (+) was through treatment with doxycycline. cDNAs from uninduced cells (–) served as a negative control. Cx32 expression was compared in the various transfectants by PCR with primers specific for Cx32 cDNA. PCRs with primers specific for β-actin, SOX10 and KROX20 were performed to monitor the experiment. cDNAs from mouse brain served as positive controls.

Figure 3. SOX10- and EGR2-responsive regions in the Cx32 promoter. Schematic representation of the various Cx32 promoter deletions used is shown on the left. Numbers above each construct indicate, from left to right, the 5′ boundary (expressed in bp upstream of the transcription initiation site), the transcription start site (+1) and the 3′ boundary (position +150), luc, luciferase reporter gene. The designation of each deletion mutant is also indicated. Each of these deletion constructs was transfected in combination with empty pECE vector, pECE/SOX10 (SOX10) and/or pECE/EGR2 (EGR2), as indicated. Data from all transfections are presented as fold induction above basal levels (Fig. 1) and are means of three different experiments performed in triplicate.

observed (Fig. 4C). These results suggest that SOX10 binds both sites as a dimer, a configuration that confers on the protein a much higher affinity for the promoter sequence. To confirm the importance of the S1- and S2-binding sites, EMSAs were carried out making use of S1S2 probes mutated in one or both sites (Fig. 4C). The mutation of S1 resulted in a dramatic decrease in SOX10 binding and in an increase in the mobility of the SOX10–DNA complex, indicative of a switch

from dimeric binding to monomeric binding. When S2 or both sites were mutated in the same probe, SOX10 binding was completely abolished. Thus, it appears that integrity of S1 is necessary for effective dimeric SOX10 binding, whereas the integrity of S2 is crucial for SOX10 binding on the Cx32 promoter, since S1 can be occupied at a significant level only in the presence of wild-type S2, as expected in case of cooperativity.


Figure 4. SOX10-binding sites and EGR2-responsive elements on the Cx32 promoter. (A) Schematic representation of part of the Cx32 promoter indicating sequence of potential SOX10 (S1 and S2) and EGR2 (E1, E2 and E3) binding sites and their relative localization in the relevant region of this promoter (–233 to –21). (B) pCx32del2 construct, and the same construct in which S1, S2 or both S1S2 SOX-binding sites are mutated, were transfected with empty pECE vector, pECE/ SOX10 (SOX10) and/or pECE/EGR2 (EGR2). Data from all transfections are presented as indicated in Figure 1. (C) Double-strand oligonucleotides containing S1, S2 or S1 and S2 (S1S2) potential SOX-binding sites were analysed for their ability to bind SOX10 in gelshift experiments. The S5-binding site from the MITF gene (15) and the C/C′ site from the P0 gene (21) are shown as the control for monomeric and dimeric binding, respectively. Experiments were performed using as protein source nuclear extracts from COS cells transfected with the SOX10 mutant E189X (+). Extracts from COS cells transfected with the empty pECE vector (–) served as the control. S1m, S2m: mutated S1 or S2 site. (D) pCx32del2 construct, and the same construct in which E1, E2 and E3 putative EGR2-binding sites are mutated were transfected with empty pECE vector, pECE/SOX10 (SOX10) and/or pECE/EGR2 (EGR2). Data from all transfections are presented as fold induction above basal levels (Fig. 1) and are means of three different experiments performed in triplicate.


EGR2-responsive elements in the Cx32 promoter As observed in Figure 3, the EGR2 responsive region in the Cx32 promoter lies between nucleotides –233 and –21 as well. Therefore, we searched for EGR2-binding sites in this proximal promoter region. Based on the EGR2-binding sites previously reported (40–42) and allowing 1 bp mismatch, we identified three putative EGR2-binding sites in the relevant region (named E1, E2, E3; Fig. 4A). To assess the importance of these binding sites for EGR2 function, the three sites were mutated simultaneously by site-directed mutagenesis. When the enhancing effect of SOX10 alone on this construct was tested, SOX10-dependent stimulation of the transcription was maintained. In contrast, EGR2-dependent stimulation was greatly reduced and the synergistic cooperation between the two factors was abolished, confirming that these EGR2 putativebinding sites are essential for effective EGR2 function and for its synergistic cooperation with SOX10 (Fig. 4D). To further investigate the role of these three sites in EGR2 function, we mutated each site individually and tested the mutant constructs in co-transfection experiments with SOX10 and EGR2. SOX10-dependant activation of the transcription was maintained in all cases. Mutation of the E1 site had no effect on EGR2-dependent stimulation, suggesting that this site does not play any role in EGR2 function on the Cx32 promoter. On the contrary, mutation of E2 and E3 resulted in a 50% decrease of EGR2-dependent activation (data not shown). These results indicate that these sites are essential for EGR2 function, but unlike the SOX10 sites, both sites have to be altered for a complete loss of EGR2-dependent activation of the promoter. Consequences of SOX10 and EGR2 mutations on Cx32 promoter induction Mutations in the SOX10 gene have been described in several cases of WS4, a neurocristopathy characterized by the association of Hirschsprung disease (intestinal aganglionosis) and Waardenburg syndrome (pigmentation defects and sensorineural deafness) (11). In addition to WS4 features, a few patients also presented with myelination defects of the CNS and PNS (7,12–14). On the other hand, mutations of EGR2 were found in patients with inherited peripheral neuropathies such as CMT1 or congenital hypomyelination neuropathy (43). Thus, it was interesting to test the effect of these various mutations on Cx32 transcription (Fig. 5). Four EGR2 mutants were generated and tested (Fig. 5A). (i) Three of these mutations (R359W, R381H, D383Y) are located within the zinc finger domain. Two (R359W and D383Y) have previously been characterized and prevent or lessen DNA binding. (ii) Mutation I268N alters the inhibitory R1 domain and prevents interaction of EGR2 with the NAB (NGFI-A binding) co-repressors. It was thereby described as having the capacity to increase EGR2 transcriptional activity (44). The first three mutants showed a complete or partial loss of the EGR2-mediated effect on Cx32 promoter activation. In contrast, in the in vitro conditions used, the I268N mutant behaved in a manner similar to wild-type EGR2 (Fig. 5B). On the other hand, the synergistic effects of R359W and R381H EGR2 mutants with wild-type SOX10 were reduced by half, whereas the synergistic effects of D383Y and I268N EGR2

mutants with SOX10 were partially or totally maintained, respectively (Fig. 5C). The five SOX10 mutants tested were as follows (Fig. 5D): (i) 482ins6 (p161_162insLR, according to den Dunnen and Antonarakis’ nomenclature) (45) is located in the HMG domain and prevents DNA binding (46); (ii) S251X and Y313X are stop mutants that truncate the protein before the transactivation domain; (iii) c795delG is a frameshift mutation that produces a premature stop codon and a loss of the transactivation domain; (iv) 1400del12 (c1400_1411del, according to den Dunnen and Antonarakis’ nomenclature) removes the stop codon and thus elongates the protein, adding 82 amino acids. For all mutants, the SOX10-dependent activation of the Cx32 promoter was significantly reduced (A variation was evaluated in a control population and found to be 50%, which is in accordance with a recently published report establishing this sequence variation as a frequent polymorphism in the German population (49). However, as this polymorphism is located in the promoter region sensitive to SOX10 and EGR2 regulation, it could alter the autonomous and synergistic effects of these factors, and thus influence the Cx32 expression level, possibly acting as a modifier of CMT disease phenotype. Moreover, two other Cx32 mutations were previously reported in non-coding regions (50). One alters the SOX10 S2binding site (g-528T>G) and the other is located downstream of the transcription start site (c-458C>T). We focused on the g-528T>G mutation because of its location within the region critical for SOX10 and EGR2 functions on the Cx32 promoter. To test whether the two sequence variations g-713G>A and g-528T>G impair SOX10 and EGR2 activation of the Cx32 promoter, we transfected wild-type or mutant pCx32del2


Figure 5. Consequences of SOX10 and EGR2 mutations on Cx32 promoter induction. (A) Schematic representation of the EGR2 factor. The R1 repressor domain and the zinc finger domain are indicated by boxes. The positions of the mutations studied in (B) and (C), I268N, R359W, R381H, D383Y are indicated by arrows. (B) The luciferase reporter plasmid containing 750 bp of the Cx32 promoter was transfected in combination with empty pECE vector, pECE/EGR2 (EGR2) or pECE/EGR2 mutants cited above. The residual activity of each mutant is represented as the ratio of EGR2 mutants over wild-type EGR2 factor. (C) The reporter plasmid was transfected in combination with empty pECE vector, pECE/EGR2 (EGR2) or pECE/EGR2 mutants (A) and pECE/SOX10 (SOX10). The residual synergistic effect of EGR2 mutants with SOX10 is represented as a ratio of EGR2 mutants and SOX10 co-activation over wild-type EGR2 and SOX10 co-activation. (D) Schematic representation of the SOX10 factor. The DNA-binding domain (HMG domain), the transactivation domain and a region that is conserved between group E SOX proteins are indicated by boxes. The positions of mutations studied in (E) and (F), 482ins6, S251X, 795delG, Y313X and 1400del12 are indicated by arrows. (E) The luciferase reporter vector was transfected in combination with empty pECE vector, pECE/SOX10 or pECE/SOX10 mutants cited above. The residual activity of each mutant is represented as the ratio of SOX10 mutants over wild-type SOX10 factor. (F) The reporter plasmid was transfected in combination with empty pECE vector, pECE/SOX10 (SOX10) or pECE/SOX10 mutants (D) and pECE/EGR2 (EGR2). The residual synergistic effect of SOX10 mutants with EGR2 is represented as a ratio of SOX10 mutants and EGR2 co-activation over wild-type SOX10 and EGR2 co-activation. Data in (B), (C), (E) and (F) are means ± SE of three different experiments performed in triplicates.


Figure 6. Functional effect of g-713G>A and g-528T>G Cx32 promoter mutations on EGR2 and SOX10 function. (A) Wild-type pCx32del2 construct or mutant forms of this construct containing g-713G>A or g-528T>G mutations were transfected in combination with empty pECE vector, pECE/SOX10 (SOX10) and/or pECE/EGR2 (EGR2), as indicated. The residual activity of SOX10, EGR2 or residual synergistic activity of both factors is represented as a ratio of activity of each or both factors on the mutant reporter constructs over activity of each or both factors on the wild-type reporter construct. (B) Double-strand oligonucleotides containing the potential SOX-binding sites S2 and S2 with g-528T>G mutation, S1S2 and S1S2 with g-528T>G mutation were analysed for their ability to bind SOX10 in gelshift experiments using nuclear extracts from COS cells transfected with the SOX10 mutant E189X (+). Extracts from COS cells transfected with the empty pECE vector (–) served as the control.

promoter constructs with these factors, alone or in combination (Fig. 6A). The g-713G>A polymorphism had no influence on EGR2 and SOX10 functions. In contrast, the g-528T>G mutation impaired SOX10 effect on Cx32 promoter activity, resulting in a 50% decrease in SOX10-dependent activation. On the other hand, EGR2 function was not altered. The 40% decrease in synergistic activation observed can thus be attributed to SOX10 partial loss of activity. To demonstrate that the SOX10 loss of activity was due to a reduced affinity of the protein for the mutated site, we performed EMSAs using DNA fragments corresponding to wild-type S2 or S1S2 sites, or to S2 or S1S2 sites carrying the g-528T>G mutation (Fig. 6B). When the mutated S2 probe was used, SOX10 binding was completely abolished. With the mutated S1S2 probe, we observed a dramatic decrease in SOX10 binding, but a faint dimeric binding could still be detected. These results are in total agreement with our transfection experiments, which did not show a complete loss of SOX10 activation. Thus, the g-528T>G mutation alters SOX10 function on the Cx32 promoter, whereas the g-713G>A polymorphism seems to have no effect on SOX10 and EGR2 function in the experimental conditions used. DISCUSSION SOX10 is a member of the SOX gene family, a class of genes coding for transcription factors highly conserved through evolution. SOX10 was identified as an essential factor in the ENS, melanocyte and glial cell development (3). Mutations in the SOX10 gene have been described in several cases of WS4, a neurocristopathy characterized by the association of Hirschsprung disease (intestinal aganglionosis) and Waardenburg syndrome (pigmentation defects and sensorineural deafness), and some patients present additional myelination defects of the CNS and/or

PNS (7,11–14). In accordance, it has been shown that SOX10 controls expression of P0, MITF and RET, three genes known to play important roles during glial cell, melanocyte and ENS development, respectively (15–19,21). The recent identification of P0 as a transcriptional target of SOX10 provided the first clues as to the role of this factor during the myelination process in the PNS. A small number of proteins is known to be implicated in this process. Some of them, like MBP, are an integral part of myelin in both the CNS and PNS, whereas others are essentially confined to one or the other. P0 and PMP22, for instance, are preferentially represented in Schwann cells (22). Another protein essential for Schwann cell integrity is Connexin 32, which forms gap junctions allowing the diffusion of small molecules through the myelin sheath (27). It is noteworthy that the genes encoding three of these proteins (P0, PMP22 and Cx32) are implicated in CMT disease (24). Characterization of their transcriptional regulation could improve our understanding of the pathogenesis of peripheral neuropathies. To examine the possible involvement of SOX10 in the control of MBP, PMP22 and Cx32 transcription, we analysed the effect of this transcription factor on each of these promoters in transient co-transfection experiments. In the experimental conditions used, the Cx32 promoter only was responsive to SOX10. Moreover, the three previously identified SOX10 cofactors, PAX3, POU3F1 and EGR2 (3), appeared to act in synergy with SOX10 to activate Cx32 transcription, but EGR2 is the sole factor that displayed an autonomous activating function. This observation is of particular interest, considering that EGR2 has also been described as being impaired in several CMT patients (43). Thus, the regulation of the Cx32 gene by EGR2 implies an epistatic relationship between two genes involved in the same disease. The influence of SOX10 and EGR2 on Cx32 expression was further confirmed by the


observation of Cx32 upregulation in cells of neural origin inducibly expressing SOX10 and KROX20. Using microarrays and quantitative RT–PCR techniques, Nagarajan et al. (51) recently reported that EGR2 was sufficient for induction of genes critical for myelin formation and maintenance, including Cx32 but also MBP and PMP22. Our finding that Cx32 is indeed a transcriptional target of EGR2 is in agreement with this observation. Furthermore, our study shows that this regulation requires the presence of EGR2 sites on the Cx32 promoter. On the contrary, we observed no influence of EGR2, alone or in combination with SOX10, on MBP and PMP22 expression in our experimental conditions. In their study, Nagarajan et al. (51) observed an increase in MBP expression 48 h after having introduced EGR2 in their cell culture system. In our study, luciferase activity was measured 24 h after transfection. This, together with the fact that we used a different cell type for our study, could explain why we did not observe any effect of EGR2 on the MBP promoter. Concerning PMP22, the discrepancies between the results of Nagarajan et al. (51) and ours could be explained by the use of different cell types (rat and mouse Schwann cells versus HeLa cells) and methods. Two hypotheses could be proposed to explain why we did not observe any effect of EGR2 on the PMP22 promoter in our in vitro conditions: (i) EGR2 does not activate PMP22 expression in a direct manner and the proteins involved in this regulatory pathway are not present in HeLa cells; (ii) EGR2 indeed activates PMP22 expression in a direct manner, but in association with other factors that are not expressed in HeLa cells. Our study clearly establishes that SOX10 and EGR2 interact directly by binding to the proximal region of the Cx32 promoter, which contains binding sites for both factors. We identified two SOX-binding sites (S1 and S2) that are involved in SOX10 regulation of transcription from this promoter. These sites consist of two non-consensus SOX sites with a particular spacing and orientation reminiscent of what was described for the P0 promoter (21,39). EMSAs suggested that SOX10 binds to these sites in a dimeric configuration. However, SOX10 does not seem to display an equal affinity for each site. Indeed, S2 seems to be crucial for binding of SOX10 to the promoter, whereas S1 seems necessary for effective dimerization only, as expected in case of cooperativity. These results are consistent with what was observed with the P0 promoter, where dimeric binding is mediated by a segment of the N-terminal region of SOX10 and requires the presence of two SOX-binding sites in the same configuration. We assume that a similar mechanism accounts for the dimeric binding of SOX10 on the Cx32 promoter. Furthermore, it was demonstrated that binding of two SOX10 proteins instead of one increased SOX-dependent DNA bending (39). This observation is compatible with the proposed role of SOX10 as both a transcriptional and architectural factor. Taken together, the work of Peirano et al. (39) and our present results suggest that SOX10 dimeric binding results in specific DNA bending, thus altering promoter topology and allowing formation of a stable multiprotein complex responsible for gene transcriptional regulation. This mechanism might provide a useful model of the way SOX10 regulates expression of some of its transcriptional targets. Our study also demonstrates that two EGR2-binding sites (E2 and E3) are involved in EGR2 regulation of transcription

from the Cx32 promoter. It is noteworthy that unlike the SOX10-binding sites, each EGR2 site is responsible for half of the EGR2-dependent stimulation of the transcription, EGR2 function being completely lost only when both sites are impaired. This observation suggests that there is no cooperation between the two EGR2-binding sites and that activation of the Cx32 promoter is the result of EGR2 independent binding on these sites. Using EMSA Musso et al. (52) recently reported that EGR2 was able to bind an element within the Cx32 promoter, which corresponds to our E3-binding site. Our results are in agreement with this observation, and further demonstrate the implication of a second EGR2-binding site (E2) in the regulation of Cx32 expression. To confirm the data obtained with the wild-type factors, we sought possible deleterious effects on Cx32 promoter activation of several natural SOX10 and EGR2 mutations. Warner et al. (44) previously showed that two EGR2 mutant factors (R359W and D383Y) fail to activate transcription of a luciferase reporter gene fused to a synthetic promoter containing two EGR2-binding sites. We confirmed these results in testing the effects of these mutant factors on the Cx32 promoter. Since the study of Warner et al., (44) additional mutations of EGR2 have been identified in patients with peripheral neuropathies. Here we describe for the first time the functional consequences of one of these mutations: R381H (53). In our in vitro model, this EGR2 mutant displayed a complete loss of function. As SOX10 and EGR2 act in synergy on the Cx32 promoter, we thought it interesting to see how EGR2 mutations affect this synergy. Surprisingly, we observed that the three mutants tested retained 50% of synergistic activity. Thus, these mutants, which have lost their autonomous effect on the Cx32 promoter, are still partially able to act in synergy with SOX10 in our experimental conditions. Warner et al. (44) also described a third mutation, I268N, that affects the inhibitory domain R1 of EGR2. They observed an increase in transcriptional activity of this mutant as compared with wild-type EGR2 in their transfection experiments. However, in our in vitro model this mutant behaves as the wild-type protein. This discrepancy could be explained by the use of different cell types (CV-1 versus HeLa) and of different reporter constructs. Nevertheless, our study describes the first functional assay of EGR2 mutants carried out on a natural transcriptional target of this factor. This assay will be useful for future studies of other EGR2 mutant proteins. With regard to SOX10, all five mutants tested resulted in a dramatic decrease in both autonomous and synergistic activations of the Cx32 promoter. The loss of SOX10 autonomous activity indicates that these mutations impair the regulatory function of SOX10 on the Cx32 promoter. Similarly, the loss of synergistic activation indicates that unlike EGR2, SOX10 mutant proteins are no longer able to act in synergy with the cofactor to enhance activation of the Cx32 promoter. Interestingly, four of the five SOX10 mutants were studied because of their association with neurological phenotypes. This failure of the mutant proteins to transactivate the Cx32 promoter could be one of the possible mechanisms underlying the defects in the PNS observed in patients carrying these mutations. Thus, it was of particular interest to test the c795delG mutation, which was not functionally characterized so far, because it had been identified in a patient presenting with particular peripheral myelin defects (14). Here we show that, in our experimental


conditions, the resulting mutant protein is unable to transactivate the Cx32 promoter, alone or in cooperation with EGR2. Impairment of Cx32 expression results in CMTX disease. More than 160 mutations of the Cx32 gene were described, including nonsense, missense and non-coding regions mutations, with varying clinical manifestation (25). The identification of SOX10 as a direct regulator of Cx32 expression provides new insights into the pathophysiology of CMTX. As SOX10 and EGR2 seem to be major regulators of Cx32 expression during development, it is likely that alteration of Cx32 regulation by SOX10 and/or EGR2 could lead to a similar phenotype. To test this hypothesis, we searched for sequence alteration in the region of the Cx32 promoter shown to be sensitive to SOX10 and EGR2 activation. Our results, in addition to confirming the high frequency of the g-713G>A polymorphism in the European Caucasian population, showed that this sequence variation has no functional consequences, neither on SOX10 or EGR2 effect, nor on synergistic transactivation of the Cx32 promoter. Whether this sequence variation could impair function of other factors remains to be determined. However, Bergmann et al. (49) showed no distorsion in the frequency of this polymorphism between a control population and CMT patients and no correlation between disease severity and genotype. Together with this observation, our results argue for a silent effect of this polymorphism. Two other Cx32 mutations in non-coding regions were previously reported (50). One alters the SOX10 S2-binding site (g-528T>G) and the other is located downstream of the transcription start site (c-458C>T). We focused on the g-528T>G mutation because it is located within the region critical for SOX10 and EGR2 function on the Cx32 promoter and because it alters the core of the S2 SOX10-binding site. This mutation leads to a dramatic decrease of affinity of SOX10 for the Cx32 promoter and thus results in a loss of transactivation. These results confirm the deleterious nature of this mutation. Although Cx32 promoter mutations do not seem to be very frequent in CMTX disease, the demonstration of the functional significance of the g-528T>G mutation confirms the physiologic and pathophysiologic relevance of our study. It is noteworthy that a Cx32 mutation was reported in a patient with CMTX disease and sensorineural deafness (54). The authors suggested that impairment of Cx32 expression both in oligodendrocytes and Schwann cells underlies the pathophysiology of deafness. Similarly, impaired expression of Cx32, together with that of MITF (15), could result in alteration of inner ear function, and thus in the sensorineural deafness observed in WS4 associated with SOX10 mutations. Finally, it is interesting to note that the two transcriptional targets of SOX10 identified in the PNS, P0 and Cx32, are involved in CMT disease (23,24). This suggests a broader role for SOX10 in this disease. Another interesting aetiologic candidate, Periaxin, was recently described as a new gene responsible for an autosomal form of demyelinating CMT disease (55,56). Two reports have shown that this gene is regulated by EGR2 (see discussion in 51,55), without determining whether this regulation is direct or not. As our results establish that EGR2 acts in cooperation with SOX10 to directly regulate Cx32 expression, it will be of great interest to test the possible involvement of these two factors in the control of Periaxin expression.

MATERIALS AND METHODS Plasmids Promoters isolation. Approximately 750–1000 bp of MBP, PMP22 and Cx32 human promoter sequences were amplified by PCR using the following couples of primers 5′-CCGGTACGCCAGATCTCAGAGGAGAAGCCAAGTCAAA-3′ and 5′-GACGTCCAGGAGATCTGAGTCAAGGATGCCCGTGTC-3′, 5′-CCGGTACGCCAGATCTAGCGGGGAGAAAGACACTGG-3′ and 5′-GACGTCCAGGAGATCTTGCCTGAAGCCTGTGATGCC-3′, and 5′-CCGGTACGCCAGATCTGCCTGCTGTAGAAAGACTT-3′ and 5′-GACGTCCAGGAGATCTCAAACTCAACAAAGCCCTCT-3′, respectively. These primers were designed from the previously reported sequences (GenBank accession nos M63599, AF059314, AL570962 and L47127). PCR products were cloned in the pGL3-basic luciferase reporter vector (Promega, Madison, WI) following BglII digestion. The nucleotide sequence was verified using BigDye Terminator cycle sequencing on a 373 ABI apparatus (PE Applied Biosystems, Les Ulis, France). Sequence divergences observed upon comparison with the reported sequence of GenBank were also found in PCR amplification from three different control DNAs. Cx32 constructs. Deletion constructs pCx32 del1, pCx32 del2, and pCx32 del3 were generated by introduction of enzyme restriction sites by directed mutagenesis using the Quick Change mutagenesis kit (Stratagene, Amsterdam, The Netherlands), followed by digestion and ligation. Constructs pCx32del2 S1, pCx32del2 S2, pCx32del2 S1S2, pCx32del2 E1, pCx32del2 E2, pCx32del2 E3, pCx32del2 E1E2E3, pCx32del2 g-713G>A and pCx32del2 g-528T>G were also generated by sequential site-directed mutagenesis steps, making use of the same kit. The nucleotide sequence of each construct was verified. Positions are indicated from the transcription start site described by Neuhaus et al. (31). Each construct corresponds to the pCx32 del2 sequence except as follows: (i) in pCx32del2 S1, the sequence between positions –53 and –27 is substituted with the mutant sequence GTTCAGAGCCCCACggAGGTCTCATTGTG; (ii) in pCx32del2 S2, the sequence between positions –44 and –16 is substituted with the mutant sequence CACAAAGGTCTCAccGTGCAGACACTGGG; (iii) pCx32del2 S1S2 includes the two sequence substitutions described above; (iv) in pCx32del2 E1, the sequence between –237 and –203 is substituted with ATCTCCCTCCCCCCGtCaCtCCGTGGCCATTCTTC; (v) in pCx32del2 E2, the sequence between –200 and –166 is substituted with GTGGGGCTATGGGGCaGtTaCGGCGATGGACCGGG; (vi) in pCx32del2 E3, the sequence between –115 and –79 is substituted with GTGCTCTGTGTGTAGGaTtGaCGGAAGTCAGGGCGTT; (vii) pCx32del2 E1E2E3 includes the three sequence substitutions described above. In pCx32del2 g-713G>A and pCx32del2 g-528T>G, sequence substitution mimics mutations of the non-coding region of the Cx32 gene found in patients presenting with CMTX neuropathy (48,50). PAX3, POU3F1 and SOX10 constructs. Constructs containing the human SOX10 and PAX3 cDNAs cloned in pECE vector were described previously (15,57). The previously identified SOX10 mutations, 482ins6, S251X, 795delG, Y313X and 1400del12 (7,11,12,14) were introduced by site-directed


mutagenesis as already described and sequence modifications were verified by direct sequencing. POU3F1 sequence contained in the previously described pCMV/Tst1 (58) was subcloned in pECE vector at XbaI and EcoRI sites. EGR2 constructs. The human EGR2 cDNA was amplified by RT–PCR using the following conditions: 1 µg of total brain RNA was reverse-transcribed using SuperscriptII-Reverse Transcriptase (GibcoBRL, LifeTechnologies, Gaithersburg, MD) according to the manufacturer’s protocol. Three microlitres of this reaction mixture was used for subsequent PCR analysis using primers 5′-TCAGTCCAACCCCTCTCCAA-3′ and 5′-GGTGGTAGTGTTTGTTGTGC-3′. The PCR product was then cloned in the pECE vector at the EcoRI site. Four mutations I268N, R359W, R381H and D383Y of EGR2 cDNA corresponding to mutations reported in patients presenting with peripheral neuropathies (43,53,59) were generated by site-directed mutagenesis using the Quick Change mutagenesis kit (Stratagene). The nucleotide sequence of each construct was verified by direct sequencing. RNA preparation, reverse transcription and PCR Total RNAs isolated from N2A stable transfectants inducibly expressing SOX10 and KROX20 were kindly provided by M.Wegner. One microgram of each RNA sample was reversetranscribed as described above. cDNA (1–6 µl) was amplified with primer pairs specific for Cx32, SOX10, KROX20 and βactin. The following primer pairs were used: Cx32 5′CTGCTCTACCCCGGCTATGC-3′ and 5′-CAGGCTGAGCATCGGTCGCTCTT-3′; SOX10 5′-GCTGAACGAGAGTGACAAGC-3′ and 5′-ATGGCTGATCTCCCCGATGT-3′; KROX20 5′-ACATGACCGGAGAGAAGAGGCCCTT-3′ and 5′-GCGTCTTGCTGGGCCTGTTGGGGTA-3′; β-actin 5′-CCAAGGCCAACCGCGAGAAGATGAC-3′ and 5′-AGGGTACATGGTGGTGCCGCCAGAC-3′. After denaturation (5 min at 94°C), PCR conditions were specific for each gene tested: Cx32, 1 min at 72°C, then 35 cycles with denaturation 30 s at 94°C, annealing 1 min 30 s at 64°C and elongation 2 min at 72°C; SOX10, 25 cycles with denaturation 30 s at 94°C, annealing 30 s at 62°C and elongation 45 s at 72°C; KROX20, 30 cycles with denaturation 1 min at 94°C, annealing 1 min at 65°C, and elongation 1 min at 72°C. Each PCR was followed by an elongation step of 7 additional min at 72°C. PCR products were then visualized on 2% agarose gel. Cell culture, transfection and reporter assays HeLa cells were grown in DMEM supplemented with 10% fetal calf serum and transfected using Lipofectamine PLUS reagents (GibcoBRL) in six-well plates. Cells were plated at 3 × 105/well and were transfected 1 day after with 0.350 µg of each effector and reporter plasmid. The total amount of plasmid was kept constant by addition of empty pECE vector. The plasmid pCH110, which contains the SV40 promoter driving expression of a LacZ reporter, was used as an internal control to assess transfection efficiency (0.300 µg per transfection) as previously described (60). Twenty-four hours after transfection, cells were washed twice with PBS, lysed and extracts were assayed for luciferase activity using the Luciferase Assay System (Promega).

EMSA A 32P-labelled probe (0.5 ng) was incubated with nuclear extracts for 20 min on ice in a 20 µl reaction mixture containing 10 mM HEPES pH 8.0, 5% glycerol, 50 mM NaCl, 5 mM MgCl2, 2 mM DTT, 0.1 mM EDTA, 3.5 µg of bovine serum albumin and 1 µg of poly(dGdC) as unspecific competitor. Nuclear extracts from COS cells transfected with the SOX10 mutant E189X or with pECE served as protein sources (2 µg/reaction). E189X is a shortened SOX10 version that shows DNA-binding characteristics identical to full-length SOX10 (21). As probes we used double-stranded oligonucleotides each containing one or two potential SOX10-binding sites from the region between –233 and –21 of the Cx32 promoter (for sequences see Fig. 4A). After incubation, samples were loaded onto native 4% polyacrylamide gels and electrophoresed in 0.5× TBE at 120 V for 1.5 h. Gels were dried and exposed for autoradiography. Patients and mutation screening of the Connexin 32 promoter sequence Thirty-four CMT patients were selected based on NCV of the median nerve (NCV < 30 m/s, four patients; 30 < NCV < 40 m/s, 19 patients; NCV > 40 m/s, 11 patients). The clinical criteria for CMT, the examination protocol and the electrophysiological studies were reported (47). This cohort includes 12 males and 22 females with familial transmission compatible with an X-linked dominant disorder. SSCP screening of the Cx32 coding sequence was negative for these patients (47). DNA fragments containing the partial sequence of the glialspecific P2 promoter region of the Cx32 gene was obtained by performing PCR amplification on genomic DNA of patients and controls using primers 5′-TCCCCTCTTCACATCCACCT-3′ and 5′-GTCTGTGCAGGGGGAGTGGT-3′. Direct sequencing of PCR products was performed in both strands using the ABI PRISM Dye Terminator Cycle DNA Sequencing kit. The frequency of the g-713G>A polymorphism in French Caucasian population was assayed on 55 control DNAs representing 83 X chromosomes, by restriction analysis of the PCR fragment using NcoI. ACKNOWLEDGEMENTS The expert technical assistance of Viviane Baral is acknowledged. This work was supported by the Institut National de la Santé et de la Recherche Médicale (INSERM) and by a grant from Association Française contre les Myopathies (AFM). REFERENCES 1. Pevny, L.H. and Lovell-Badge, R. (1997) Sox genes find their feet. Curr. Opin. Genet. Dev., 7, 338–344. 2. Wegner, M. (1999) From head to toes: the multiple facets of Sox proteins. Nucleic Acids Res., 27, 1409–1420. 3. 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. 4. 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. 5. Bondurand, N., Kobetz, A., Pingault, V., Lemort, N., Encha-Razavi, F., Couly, G., Goerich, D.E., Wegner, M., Abitbol, M. and Goossens, M.


6.

7.

8.

9. 10. 11.

12.

13.

14.

15.

16.

17.

18.

19.

20. 21.

22. 23. 24.

25.

26.

(1998) Expression of the SOX10 gene during human development. FEBS Lett., 432, 168–172. Britsch, S., Goerich, D.E., Riethmacher, D., Peirano, R.I., Rossner, M., Nave, K.A., Birchmeier, C. and Wegner, M. (2001) The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev., 15, 66–78. Touraine, R.L., Attie-Bitach, T., Manceau, E., Korsch, E., Sarda, P., Pingault, V., Encha-Razavi, F., Pelet, A., Auge, J., Nivelon-Chevallier, A. et al. (2000) Neurological phenotype in Waardenburg syndrome type 4 correlates with novel SOX10 truncating mutations and expression in developing brain. Am. J. Hum. Genet., 66, 1496–1503. Herbarth, B., Pingault, V., Bondurand, N., Kuhlbrodt, K., HermansBorgmeyer, 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. 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. Read, A.P. and Newton, V.E. (1997) Waardenburg syndrome. J. Med. Genet., 34, 656–665. 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. Inoue, K., Tanabe, Y. and Lupski, J.R. (1999) Myelin deficiencies in both the central and the peripheral nervous systems associated with a SOX10 mutation. Ann. Neurol., 46, 313–318. 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 Waardenburg–Shah (WS4) syndrome. Genome Res., 9, 215–225. Pingault, V., Guiochon-Mantel, A., Bondurand, N., Faure, C., Lacroix, C., Lyonnet, S., Goossens, M. and Landrieu, P. (2000) Peripheral neuropathy with hypomyelination, chronic intestinal pseudo-obstruction and deafness: a developmental ‘neural crest syndrome’ related to a SOX10 mutation. Ann. Neurol., 48, 671–676. Bondurand, N., Pingault, V., Goerich, D.E., Lemort, N., Sock, E., Le Caignec, C., Wegner, M. and Goossens, M. (2000) Interaction among SOX10, PAX3 and MITF, three genes altered in Waardenburg syndrome. Hum. Mol. Genet., 9, 1907–1917. Lee, M., Goodall, J., Verastegui, C., Ballotti, R. and Goding, C.R. (2000) Direct regulation of the Microphthalmia promoter by Sox10 links Waardenburg–Shah syndrome (WS4)-associated hypopigmentation and deafness to WS2. J. Biol. Chem., 275, 37978–37983. 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. 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. 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. Taraviras, S. and Pachnis, V. (1999) Development of the mammalian enteric nervous system. Curr. Opin. Genet. Dev., 9, 321–327. Peirano, R.I., Goerich, D.E., Riethmacher, D. and Wegner, M. (2000) Protein zero gene expression is regulated by the glial transcription factor Sox10. Mol. Cell. Biol., 20, 3198–3209. Snipes, G.J. and Suter, U. (1995) Molecular anatomy and genetics of myelin proteins in the peripheral nervous system. J. Anat., 186, 483–494. Keller, M.P. and Chance, P.F. (1999) Inherited neuropathies: from gene to disease. Brain Pathol., 9, 327–341. Nelis, E., Haites, N. and Van Broeckhoven, C. (1999) Mutations in the peripheral myelin genes and associated genes in inherited peripheral neuropathies. Hum. Mutat., 13, 11–28. Abrams, C.K., Oh, S., Ri, Y. and Bargiello, T.A. (2000) Mutations in connexin 32: the molecular and biophysical bases for the X-linked form of Charcot–Marie–Tooth disease. Brain Res. Brain Res. Rev., 32, 203–214. Ressot, C. and Bruzzone, R. (2000) Connexin channels in Schwann cells and the development of the X-linked form of Charcot–Marie–Tooth disease. Brain Res. Brain Res. Rev., 32, 192–202.

27. Scherer, S.S., Deschenes, S.M., Xu, Y.T., Grinspan, J.B., Fischbeck, K.H. and Paul, D.L. (1995) Connexin32 is a myelin-related protein in the PNS and CNS. J. Neurosci., 15, 8281–8294. 28. Krutovskikh, V. and Yamasaki, H. (2000) Connexin gene mutations in human genetic diseases. Mutat. Res., 462, 197–207. 29. Devine-Beach, K., Lashgari, M.S. and Khalili, K. (1990) Myelin basic protein gene transcription. Identification of proximal and distal cis-acting regulatory elements. J. Biol. Chem., 265, 13830–13835. 30. Suter, U., Snipes, G.J., Schoener-Scott, R., Welcher, A.A., Pareek, S., Lupski, J.R., Murphy, R.A., Shooter, E.M. and Patel, P.I. (1994) Regulation of tissue-specific expression of alternative peripheral myelin protein-22 (PMP22) gene transcripts by two promoters. J. Biol. Chem., 269, 25795–25808. 31. Neuhaus, I.M., Bone, L., Wang, S., Ionasescu, V. and Werner, R. (1996) The human connexin32 gene is transcribed from two tissue-specific promoters. Biosci. Rep., 16, 239–248. 32. Saberan-Djoneidi, D., Sanguedolce, V., Assouline, Z., Levy, N., Passage, E. and Fontes, M. (2000) Molecular dissection of the Schwann cell specific promoter of the PMP22 gene. Gene, 248, 223–231. 33. Neuhaus, I.M., Dahl, G. and Werner, R. (1995) Use of alternate promoters for tissue-specific expression of the gene coding for connexin32. Gene, 158, 257–262. 34. Elfgang, C., Eckert, R., Lichtenberg-Frate, H., Butterweck, A., Traub, O., Klein, R.A., Hulser, D.F. and Willecke, K. (1995) Specific permeability and selective formation of gap junction channels in connexin-transfected HeLa cells. J. Cell Biol., 129, 805–817. 35. Staugaitis, S.M., Smith, P.R. and Colman, D.R. (1990) Expression of myelin basic protein isoforms in nonglial cells. J. Cell Biol., 110, 1719–1727. 36. D’Urso, D., Prior, R., Greiner-Petter, R., Gabreels-Festen, A.A. and Muller, H.W. (1998) Overloaded endoplasmic reticulum-Golgi compartments, a possible pathomechanism of peripheral neuropathies caused by mutations of the peripheral myelin protein PMP22. J. Neurosci., 18, 731–740. 37. Harley, V.R., Lovell-Badge, R. and Goodfellow, P.N. (1994) Definition of a consensus DNA binding site for SRY. Nucleic Acids Res., 22, 1500–1501. 38. Mertin, S., McDowall, S.G. and Harley, V.R. (1999) The DNA-binding specificity of SOX9 and other SOX proteins. Nucleic Acids Res., 27, 1359–1364. 39. Peirano, R.I. and Wegner, M. (2000) The glial transcription factor Sox10 binds to DNA both as monomer and dimer with different functional consequences. Nucleic Acids Res., 28, 3047–3055. 40. Nardelli, J., Gibson, T.J., Vesque, C. and Charnay, P. (1991) Base sequence discrimination by zinc-finger DNA-binding domains. Nature, 349, 175–178. 41. Nardelli, J., Gibson, T. and Charnay, P. (1992) Zinc finger-DNA recognition: analysis of base specificity by site-directed mutagenesis. Nucleic Acids Res., 20, 4137–4144. 42. Swirnoff, A.H. and Milbrandt, J. (1995) DNA-binding specificity of NGFI-A and related zinc finger transcription factors. Mol. Cell. Biol., 15, 2275–2287. 43. Warner, L.E., Mancias, P., Butler, I.J., McDonald, C.M., Keppen, L., Koob, K.G. and Lupski, J.R. (1998) Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat. Genet., 18, 382–384. 44. Warner, L.E., Svaren, J., Milbrandt, J. and Lupski, J.R. (1999) Functional consequences of mutations in the early growth response 2 gene (EGR2) correlate with severity of human myelinopathies. Hum. Mol. Genet., 8, 1245–1251. 45. Dunnen, J.T. and Antonarakis, S.E. (2000) Mutation nomenclature extensions and suggestions to describe complex mutations: a discussion. Hum. Mutat., 15, 7–12. 46. 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. 47. Dubourg, O., Tardieu, S., Birouk, N., Gouider, R., Leger, J.M., Maisonobe, T., Brice, A., Bouche, P. and LeGuern, E. (2001) The frequency of 17p11.2 duplication and Connexin 32 mutations in 282 Charcot–Marie–Tooth families in relation to the mode of inheritance and motor nerve conduction velocity. Neuromusc. Disord., 11, 458–463. 48. Wang, H.L., Wu, T., Chang, W.T., Li, A.H., Chen, M.S., Wu, C.Y. and Fang, W. (2000) Point mutation associated with X-linked dominant Charcot–Marie–Tooth disease impairs the P2 promoter activity of human connexin-32 gene. Brain Res. Mol. Brain Res., 78, 146–153.


49. Bergmann, C., Schroder, J.M., Rudnik-Schoneborn, S., Zerres, K. and Senderek, J. (2001) A point mutation in the human connexin32 promoter P2 does not correlate with X-linked dominant Charcot–Marie–Tooth neuropathy in Germany. Brain Res. Mol. Brain Res., 88, 183–185. 50. Ionasescu, V.V., Searby, C., Ionasescu, R., Neuhaus, I.M. and Werner, R. (1996) Mutations of the noncoding region of the connexin32 gene in X-linked dominant Charcot–Marie–Tooth neuropathy. Neurology, 47, 541–544. 51. Nagarajan, R., Svaren, J., Le, N., Araki, T., Watson, M. and Milbrandt, J. (2001) EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron, 30, 355–368. 52. Musso, M., Balestra, P., Bellone, E., Cassandrini, D., Di Maria, E., Lamba Doria, L., Grandis, M., Mancardi, G., Schenone, A., Levi, G. et al. (2001) The D355V mutation decreases EGR2 binding to an element within the Cx32 promoter. Neurobiol. Dis., 8, 700–706. 53. Pareyson, D., Taroni, F., Botti, S., Morbin, M., Baratta, S., Lauria, G., Ciano, C. and Sghirlanzoni, A. (2000) Cranial nerve involvement in CMT disease type 1 due to early growth response 2 gene mutation. Neurology, 54, 1696–1698. 54. Stojkovic, T., Latour, P., Vandenberghe, A., Hurtevent, J.F. and Vermersch, P. (1999) Sensorineural deafness in X-linked Charcot-Marie-Tooth disease with connexin 32 mutation (R142Q). Neurology, 52, 1010–1014. 55. Boerkoel, C.F., Takashima, H., Stankiewicz, P., Garcia, C.A., Leber, S.M., Rhee-Morris, L. and Lupski, J.R. (2001) Periaxin mutations

56.

57.

58.

59.

60.

cause recessive Dejerine–Sottas neuropathy. Am. J. Hum. Genet., 68, 325–333. Guilbot, A., Williams, A., Ravise, N., Verny, C., Brice, A., Sherman, D.L., Brophy, P.J., LeGuern, E., Delague, V., Bareil, C. et al. (2001) A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot–Marie–Tooth disease. Hum. Mol. Genet., 10, 415–421. Ellis, L., Clauser, E., Morgan, D.O., Edery, M., Roth, R.A. and Rutter, W.J. (1986) Replacement of insulin receptor tyrosine residues 1162 and 1163 compromises insulin-stimulated kinase activity and uptake of 2-deoxyglucose. Cell, 45, 721–732. Sock, E., Enderich, J., Rosenfeld, M.G. and Wegner, M. (1996) Identification of the nuclear localization signal of the POU domain protein Tst-1/Oct6. J. Biol. Chem., 271, 17512–17518. Timmerman, V., De Jonghe, P., Ceuterick, C., De Vriendt, E., Lofgren, A., Nelis, E., Warner, L.E., Lupski, J.R., Martin, J.J. and Van Broeckhoven, C. (1999) Novel missense mutation in the early growth response 2 gene associated with Dejerine–Sottas syndrome phenotype. Neurology, 52, 1827–1832. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Press, Cold Spring Harbor, NY, pp. 16.66–16.67.
