Mutations in MITF (microphthalmia transcription factor) cause Waardenburg syndrome type 2 (WS2A) in humans, an autosomal dominant disorder consisting of deafness and hypopigmentation. Phenotypes vary significantly within WS2 pedigrees, and there is generally no correlation between the predicted biochemical properties of mutant MITF proteins and disease severity. We have identified a nonsense mutation in the Mitf gene of the anophthalmic white (Wh) Syrian hamster that destabilizes its mRNA and prevents the encoded basic helix-loop-helix leucine zipper (bHLHzip) protein from dimerizing or binding DNA target sites. Although the resulting polypeptide does not act as a dominant-negative species in vitro, the Wh mutation is inherited as a semi-dominant trait. It thus more closely resembles WS2 than comparable Mitf alleles in laboratory mice and rats, which are expressed as purely recessive traits.
Waardenburg syndrome (WS) in humans is a classical autosomal dominant disorder marked by sensorineural hearing loss and pigment abnormalities (white forelock and heterochromia irides). Clinically, and more recently genetically, WS has been divided into two categories, type 1 (WS1, MIM 19350) and type 2 (WS2, MIM 19351) (1). These syndromes are collectively responsible for ~2% of patients with profound congenital deafness. WS1 is associated with mutations in the PAX3 gene on chromosome 2q (refs 1-3) and is distinguished from WS2 by the presence of dystopia canthorum (lateral displacement of the medial eye folds) (4). Within WS pedigrees, there is often significant variation in the extent of pigment loss and hearing defects (5,6).
The MITF locus encodes a basic helix-loop-helix leucine zipper (bHLHzip) transcription factor (7) that is required for proper development of all melanin-pigmented cells, including neural crest-derived melanocytes and the retinal pigment epithelium (RPE) (8,9). Mutations in MITF on chromosome 3p have been found in several human WS2 pedigrees (designated WS2A) (1,10-12). The WS2 phenotype is transmitted as a dominant trait so that affected individuals are heterozygous for mutant alleles. Laboratory mice carrying mutations at the orthologous microphthalmia locus (mi) also exhibit deafness and pigment abnormalities and, in addition, aberrant eye development (13,14). Many mi alleles are known, and the majority of them are inherited as recessive traits. Heterozygous mice usually show no phenotype unless a dominant-negative form of the protein is encoded, and only one such mutation (Miwh) causes a hearing deficiency in heterozygotes (8). Four dominant-negative proteins have been characterized; each has an alteration in the basic region that permits dimerization but prevents DNA binding (8,15,16). There is also a striking correlation between the predicted biochemical properties of mutant proteins and phenotypic severity of recessive Mitf alleles in adult homozygous mice. A truncating mutation in the mib (microphthalmie-blanc) rat is also inherited as a recessive trait (ref. 17, K. Opdecamp, unpublished data). In humans, only one MITF mutation, del(R217), is expected to produce a dominant-negative protein, and it is associated with extreme clinical findings (18). The majority of dominant phenotypes observed in WS2 thus arise from haploinsufficiency and cannot be explained by specific biochemical properties of the mutant proteins.
The anophthalmic white mutation (Wh) in Syrian hamsters causes pigmentation abnormalities and variable hearing loss in heterozygotes, and absent pigmentation, profound deafness and severe eye reduction in homozygotes (19). The Wh/+ phenotype is significantly enhanced by mutations at other loci, such as extension (20). In this report, we show that Wh encodes the hamster ortholog of MITF. We describe an Mitf mutation that co-segregates with the Wh phenotype and truncates the protein between helices 1 and 2 of the DNA-binding and dimerization (bHLHzip) domain, but does not act through a dominant-negative mechanism. The dominance characteristics of this allele are similar to those of human WS2, but unlike comparable mouse and rat Mitf mutations. These findings suggest that the Syrian hamster is a superior animal model for studying the pathogenesis of WS2.
In comparison with wild-type golden hamsters, Wh/+ heterozygotes have white belly fur and a generalized coat color dilution (Fig. 1a and b), and Wh/Wh homozygotes are severely microphthalmic and unpigmented (Fig. 1c). When the Wh/+ genotype is combined with homozygous mutations at the extension locus (e/e), a black-eyed white phenotype is produced (Fig. 1d), which is more severe than either the Wh/+ or e/e (cream) pigmentation phenotype alone. Wh/Wh animals are profoundly deaf, with no measurable auditory brainstem evoked response (ABR) (21). Wh/+ animals also have reduced ABRs, with increased mean thresholds (>25 dB) in the low frequency (<4 kHz) range compared with wild-type hamsters (ref. 21, L.-B. Swenson and J. H. Asher, unpublished data). However, the expressivity of this hearing loss varies among Wh/+ heterozygotes from normal to severely impaired. This variability is similar to that found within human WS pedigrees (6).
Using antisera raised against bacterially expressed mouse MITF polypeptide (9), we detected hamster MITF protein in the retinal pigment epithelium (RPE) of E12.5 wild-type and Wh/+ embryos (Fig. 2). No immunoreactive MITF was detected in Wh/Wh RPE at the same developmental stage. To assess Mitf transcription, we tested poly(A)+ RNA from adult heart tissue, since melanocytes are absent from the skin of Wh/Wh hamsters. Northern blot analysis revealed an ~10-fold lower level of Mitf expression in anophthalmic white compared with wild-type cardiac tissue (Fig. 3). However, the transcripts in both samples were the same size (5.5 kb), indicating that the Wh mutation does not represent a deletion or gross rearrangement of the Mitf transcription unit.
Southern blot analysis of wild-type and anophthalmic white genomic DNA samples also revealed no gross alteration in the Mitf gene (data not shown). To screen for a point mutation, we amplified portions of the hamster Mitf gene and cDNA by PCR from mutant and wild-type tissues and compared these by sequence and single-strand conformational (SSC) analysis. The wild-type hamster (accession no. AF020900) and mouse Mitf cDNAs exhibit a high degree of sequence homology within the coding region (amino acids 43-419, ref. 13), at both the DNA (94.4% identity) and amino acid level (99.5% identity). In addition, the hamster transcipt is alternatively spliced in a pattern identical to mouse Mitf at exon 6 (data not shown) (15). We identified an SSC polymorphism between wild-type and mutant hamsters in a segment of the Mitf gene corresponding to human exon 8 (ref. 6). In an intercross between Wh/+ heterozygotes, this polymorphism co-segregated completely with the Wh genotype (Fig. 4a), indicating that Wh and Mitf are genetically linked (c2 = 23.92 for n = 8 and n = 13, P < 0.005). The Wh pigmentation phenotypes and decreased Mitf expression were therefore likely to result from a direct alteration in Mitf and not in an upstream regulatory gene in the same developmental pathway. Upon sequencing genomic PCR clones, we found a TGG -> TAG mutation in the Wh allele (Fig. 4b). The mutation causes the SSC polymorphism and affects a codon corresponding to residue 241 of the mouse MITF protein (13). It truncates MITF in the loop region, producing a polypeptide that lacks helix 2 and the leucine zipper (Fig. 4c). The presence of the mutation was confirmed at the RNA level by sequencing an RT-PCR product from adult Wh/Wh heart tissue. This alteration was not found in any wild-type genomic or RT-PCR product. Finally, in vitro translation products derived from the Wh allele terminate prematurely as expected (Fig. 5a, lane 2), whereas wild-type hamster and mouse MITF proteins are approximately the same size (Fig. 5a, lanes 1 and 3).cab
The mutant protein is predicted to be incapable of dimerization and, therefore, of DNA binding. To test this prediction and determine whether the Wh-encodedprotein acts in a dominant-negative manner, we performed electrophoretic mobility shift assays (EMSAs). Wild-type mouse and hamster MITF bound as a homodimer to a DNA probe containing an E-box motif, while the Wh-encodedprotein was unable to bind DNA (Fig. 5b, lanes 1-3). Similar results were obtained using an M-box (16) probe (data not shown). The Wh-encodedprotein also did not form DNA-binding heterodimers with mouse TFE3 (ref. 22), whereas wild-type hamster and mouse MITF proteins did heterodimerize with TFE3, resulting in intermediate mobility shifts (Fig. 5b, lanes 8-10). However, the Wh-encoded mutant protein did not obviously interfere with DNA binding of TFE3 or wild-type MITF homodimers. It differs in this respect from dominant-negative mouse MITF proteins, e.g. Miwh (ref. 16).
Animal models allow researchers to study the pathogenesis of human genetic disease in detail. The laboratory mouse, by virtue of the powerful genetic resources available and the ability to target specific loci, is a favored model species (23). However, mouse mutations do not always generate phenotypes equivalent to that observed in humans with orthologous gene defects (24,25), although the mechanisms by which the phenotypes arise might be identical. The Wh hamster mutation is biochemically similar to the mouse microphthalmia-cloudy eyes allele (mice) (8,15,26). Both truncate the encoded protein (Fig. 4c) and destabilize the mRNA, and neither acts as a dominant-negative species in vitro. Four human MITF mutations also truncate this protein (1,2,11); two terminate within the bHLHzip region (Fig. 4c) and are unable to bind DNA or transactivate an M-box-containing tyrosinase reporter construct (11). These two mutant proteins have also been shown not to interfere with DNA binding of wild-type MITF or mTFE3. Given the inability of any of truncation mutants studied to act as a dominant-negative species in vitro, the overall level of mRNA or mutant protein is unlikely to have a significant influence on the observed disease phenotype. The mice allele produces a purely recessive phenotype, whereas the human and hamster mutations produce dominant phenotypes. Surprisingly, hamsters and humans, but not mice or rats, with similar loss-of-function mutations show visible and physiological phenotypes due to haploinsufficiency of MITF protein.
Analysis of WS pathogenesis is complicated in mice, since Mitf homozygotes generally lack melanocytes, and affected heterozygotes express one of the dominant-negative proteins that are capable of altering the activities of other transcription factors. In contrast, comparison of cochlea structures between wild-type, heterozygous and homozygous hamsters may reveal the mechanism that leads to hearing deficiency. Ultrastructural analysis of the RPE from mutant hamsters suggests that the Wh mutation affects cell-cell interactions (32). Cultured melanocytes from Wh/+ hamsters may also have altered growth factor requirements. Our findings suggest it may be possible to produce deafness in mice by further lowering the functional level of MITF protein. This could be accomplished by combining mice and misp alleles, or miVGA-9 and misp alleles. The misp (microphthalmia-spotted) allele expresses only one MITF protein isoform, which is shorter and transactivates less efficiently than the longer form (33). The miVGA-9 allele expresses no mRNA. These compound heterozygotes may have aural phenotypes similar to WS2A patients or Wh/+ hamsters.
Intragenic MITF mutations do not account for all cases of human WS2. While some affected individuals may have MITF regulatory defects, it is likely that mutations in other genes are also capable of causing WS2. The endothelin-B receptor (EDNRB) and its ligand endothelin-3 (EDN3) are essential for the development of neural crest-derived melanocytes as well as myenteric ganglion neurons. Mutations in these genes cause Hirschsprung disease and WS4 (Shah-Waardenburg syndrome), respectively, in humans and white spotting and megacolon in mice (reviewed in ref. 27). An EDNRB missense mutation predisposes homozygous and heterozygous individuals to pigmentation and hearing abnormalities, and colonic agangliosis. Penetrance is incomplete, but higher among those with two mutant alleles. In contrast, mutations in the mouse Ednrb gene (piebald, s) are recessively inherited. Similarly, laboratory mice are less sensitive than humans to heterozygous mutations in Pax3 (splotch) (28) and the c-ret receptor tyrosine kinase (29), both of which affect neural crest derivatives. Thus, in humans, but not mice, haploinsufficiency at any one of several loci leads to WS.
There is strong evidence that epistatic interactions and possibly stochastic mechanisms underlie the appearance of WS2 phenotypes. In mice, strain background can significantly affect phenotypic severity in general (25). In the case of Mitf, the miVGA-9 allele is generally recessive but can display a dominant spotting phenotype in some genetic backgrounds (H. Arnheiter, unpublished data). Among hamsters, black-eyed white (Wh/+, e/e) animals have lower hearing thresholds than Wh/+, E/- individuals, although both have moderate to severe hearing impairment (ref. 21, L.-B. Swenson and J. H. Asher, unpublished data). Likewise, cream hamsters (+/+, e/e) have a significantly reduced response latency compared with wild-type (+/+, E/-). Thus, the interactions between Wh and e are complex with respect to pigmentation (Fig. 1) and hearing loss. Comparable interactions may underlie the variable expressivity of WS2 in humans. In one WS2 pedigree, a missense mutation in the tyrosinase gene (TYR) was shown to significantly exacerbate phenotypes arising from an unlinked MITF mutation (12). Similarly, human WS1 and mouse Pax3 phenotypes are particularly sensitive to genetic background effects (5,30,31). Thus, although mutations at specific loci have a primary determining effect on the appearance of the Waardenburg phenotype, the degree of severity is determined in a complex manner. This property is underscored by our analysis of the Wh/+ hamster model.
Syrian hamsters (Mesocricetus auratus) were derived from a congenic inbred Wh/+ colony maintained at Michigan State University (F14+ generation). For immunohistochemistry, embryos were staged from the appearance of a copulation plug (designated day E0.5), harvested in phosphate-buffered saline (PBS), fixed overnight in 4% paraformaldehyde-PBS at 4°C, equilibrated sequentially in 10 and 20% sucrose-PBS and frozen in OCT medium (Tissue-Tek). Heterozygous embryos were obtained by crossing Wh/Wh and +/+ animals. Cryosections (14 mm) were stained with polyclonal rabbit antiserum raised against full-length histidine-tagged mouse MITF protein (9), followed by a horseradish peroxidase (HRP)-conjugated secondary antibody, and were developed using a diaminobenzidine reagent (Vector Laboratories).
RNA was prepared from adult hamster heart tissue by guanidinium isothiocyanate extraction. Poly(A)+ fractions were purified by oligo(dT)-cellulose chromatography, electrophoresed through an agarose-formaldehyde gel, and transferred to a Genescreen Plus nylon membrane (NEN). This northern blot was then hybridized sequentially with mouse Mitf and [beta]-actin cDNA probes as described (13). Mitf transcript levels were quantitated using a phosphorimager (Molecular Dynamics) and normalized to [beta]-actin.
Wild-type hamster Mitf cDNA sequence was obtained by directly sequencing a 1276 bp RT-PCR product obtained from Syrian hamster melanoma cell line CCL-49 (American Type Culture Collection). RT-PCR was performed using mouse primers 5'-GGCACCAGGTAAAGCAGTAC-3' (forward) and 5'-AGAAGCTTGGAAAATTATCAAGAAAACC-3' (reverse). For SSC analysis, genomic DNA was extracted from footpad blood, preserved on isocode filters (Schleicher and Schuell) or cheek pouch swab (epithelial cell) samples. A 222 bp DNA fragment encompassing most of exon 8 was amplified by PCR using primers 5'-TTCAAGGTCCTTTGCACGCTGCTG-3' (forward) and 5'-TGAGGGGTGAGCACCTAGTGAG-3' (reverse). Reaction conditions were 95°C for 5 min, followed by 30 cycles of 95°C for 30 s, 55°C for 30 s and 72°C for 1 min, with a final extension step of 72°C for 10 min. The forward primer was end-labeled with [[gamma]-32P]ATP and T4 polynucleotide kinase. Single-strand products were resolved by electrophoresis (6 V/cm) through a 0.4 mm thick MDE gel (FMC) in 0.5× TBE buffer at room temperature for 16 h. PCR products from Wh/Wh and +/+ homozygotes were subcloned in pCRII by A-T annealing (Invitrogen) and sequenced on both strands. For each genotype, three independent amplification reactions were performed and at least six clones were sequenced.
Dimerization and DNA-binding experiments were performed as described by Hemesath et al. (16), using partial proteins translated in vitro. DNA templates for transcription and protein synthesis were generated by PCR using Pwo polymerase, with a T3 promoter and an optimal translation initiation sequence appended to the 5' primer. Products were sequenced to confirm that the template coded for the correct protein. Coupled in vitro transcription and translation (IVTT) reactions were then performed using wheat germ extract (TNT, Promega) according to the manufacturer's instructions. MITF proteins were detected by including [35S]methionine in the IVTT reactions, followed by SDS-PAGE and autoradiography. Parallel IVTT reactions were performed using a full-length mouse TFE3 cDNA as template (22). To improve resolution of TFE3 heterodimers in the EMSA, the MITF proteins contained N-terminal deletions. These polypeptides correspond to amino acids 112-419 of the mouse protein (13) and encompass the bHLHzip motif. The E-box DNA probe was generated by annealing oligonucleotide 5'-CGCCTCGAGCTAGAACCCGGTCACGTGGCCTACTGCAGTCGACCAG-3' and its complement, radiolabeled by kinase reaction.
We thank Thomas Friedman for his assistance, which allowed the completion of this study following the untimely death of Dr James H. Asher, Jr. The murine TFE3 cDNA was generously provided by Kathryn Calame. We are grateful to David Dolan and Lisa Gillikan for sharing results of hearing tests and Marilyn Gandy for assistance in preparing this manuscript. T.G. is an assistant investigator of the Howard Hughes Medical Institute.
Human Molecular Genetics
Pages
Introduction
Results
Pigmentation and hearing defects in Wh mutants
Reduced Mitf expression in Wh/Wh hamsters
Mitf point mutation
Wh is not a dominant-negative mutation
Discussion
Materials And Methods
Animals
Northern analysis
PCR and mutation detection
Protein and electrophoretic mobility shift assays (EMSA)
Acknowledgements
References
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 14 Mar 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. L. Arduini, K. M. Bosse, and P. D. Henion Genetic ablation of neural crest cell diversification Development, June 15, 2009; 136(12): 1987 - 1994. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
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]
|
|
|
|
|
|
K. Bharti, W. Liu, T. Csermely, S. Bertuzzi, and H. Arnheiter Alternative promoter use in eye development: the complex role and regulation of the transcription factor MITF Development, March 15, 2008; 135(6): 1169 - 1178. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. A. Leegwater, M. A. van Hagen, and B. A. van Oost Localization of White Spotting Locus in Boxer Dogs on CFA20 by Genome-Wide Linkage Analysis with 1500 SNPs J. Hered., August 3, 2007; (2007) esm022v2. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Baumer, T. Marquardt, A. Stoykova, D. Spieler, D. Treichel, R. Ashery-Padan, and P. Gruss Retinal pigmented epithelium determination requires the redundant activities of Pax2 and Pax6 Development, July 1, 2003; 130(13): 2903 - 2915. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Graw, W. Pretsch, and J. Loster Mutation in Intron 6 of the Hamster Mitf Gene Leads to Skipping of the Subsequent Exon and Creates a Novel Animal Model for the Human Waardenburg Syndrome Type II Genetics, July 1, 2003; 164(3): 1035 - 1041. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Steingrimsson, L. Tessarollo, B. Pathak, L. Hou, H. Arnheiter, N. G. Copeland, and N. A. Jenkins Mitf and Tfe3, two members of the Mitf-Tfe family of bHLH-Zip transcription factors, have important but functionally redundant roles in osteoclast development PNAS, April 2, 2002; 99(7): 4477 - 4482. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. D Smith, P. M Kelley, J. B Kenyon, and D. Hoover Tietz syndrome (hypopigmentation/deafness) caused by mutation of MITF J. Med. Genet., June 1, 2000; 37(6): 446 - 448. [Abstract] [Full Text]
|
|
|
|
|
|
J. Lister, C. Robertson, T Lepage, S. Johnson, and D. Raible nacre encodes a zebrafish microphthalmia-related protein that regulates neural-crest-derived pigment cell fate Development, January 9, 1999; 126(17): 3757 - 3767. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||