Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2010 Jul 23.
Published in final edited form as: J Mol Med (Berl). 2006 May 4;84(6):484–490. doi: 10.1007/s00109-005-0023-3

Localization of a novel autosomal recessive nonsyndromic hearing impairment locus DFNB65 to chromosome 20q13.2–q13.32

Aamira Tariq 1, Regie Lyn P Santos 2, Mohammad Nasim Khan 3, Kwanghyuk Lee 4, Muhammad Jawad Hassan 5, Wasim Ahmad 6, Suzanne M Leal 7
PMCID: PMC2909102  NIHMSID: NIHMS218864  PMID: 16596430

Abstract

Abstract Autosomal recessive nonsyndromic hearing impairment (ARNSHI) is the most frequent form of prelingual hereditary hearing loss in humans. Between 75 and 80% of all nonsyndromic deafness is inherited in an autosomal recessive pattern. Using linkage analysis, we have mapped a novel gene responsible for this form of nonsyndromic hearing impairment, DFNB65, in a consanguineous family from the Azad Jammu and Kashmir regions, which border Pakistan and India. A maximum multipoint LOD score of 3.3 was obtained at marker D20S840. The three-unit support interval is contained between markers D20S902 and D20S430, while the region of homozygosity is flanked by markers D20S480 and D20S430. The novel locus maps to a 10.5-cM region on chromosome 20q13.2–q13.32 and corresponds to a physical map distance of 4.3 Mb. DFNB65 represents the first ARNSHI locus to map to chromosome 20.

Keywords: Autosomal recessive nonsyndromic hearing impairment, DFNB65, Pakistan, 20q13.2–q13.32

Introduction

It is estimated that 1–2 per 1,000 newborns are affected with profound hearing impairment (HI) [1]. For congenital HI, roughly half of the cases are expected to be genetic. Of the genetic cases of HI, 30% are known to be syndromic, with the remaining cases displaying nonsyndromic hearing impairment (NSHI). For NSHI, 70–75% of the cases are due to autosomal recessive (AR) inheritance, 15–20% display autosomal dominant inheritance, 2–3% of cases show X-linked inheritance, and less than 1% of cases are due to mitochondrial inheritance [2]. Thus far, for ARNSHI more than 60 loci have been mapped and 21 genes identified [3]. This extreme heterogeneity reflects involvement of different molecular mechanisms within the auditory system that malfunction to cause HI. In the current study, we investigated a consanguineous family segregating ARNSHI and mapped the gene responsible for the defect to chromosome 20q13.2–q13.32.

Materials and methods

Family history

Before the start of the study, approval was obtained from the Quaid-I-Azam University and the Baylor College of Medicine Institutional Review Boards. Informed consent was obtained from all family members who agreed to participate in the study. Family 4020 (Fig. 1), which is from the city of Rawlakot in Azad Jammu and Kashmir region that borders Pakistan and India, provided convincing evidence of autosomal recessive mode of inheritance. All affected individuals have a history of prelingual profound HI and use sign language for communication. In addition, all available medical records of the affected individuals were scrutinized for evidence of childhood illness. Each HI individual underwent a physical and ophthalmologic examination. Defects in ear morphology, dysmorphic facial features, eye disorders including night blindness and tunnel vision, limb deformities, mental retardation, and other clinical features that could indicate that HI was syndromic were not identified. The clinical history and physical examination of affected individuals indicated no vestibular involvement.

Fig. 1.

Fig. 1

Open in a new tab

Pedigree drawing of family 4020. Black symbols represent individuals with hearing impairment. Clear symbols represent unaffected individuals. The gender of some of the family members was changed to protect the anonymity of the family. Haplotypes are shown below each individual for whom genotypes are available. For individuals in generation IV, maternal haplotypes are displayed on the left-hand side and paternal haplotypes on the right-hand side. Shaded areas indicate the region of homozygosity

Figure 2 exhibits the audiometric profile of hearing-impaired individuals IV-2 and IV-3, aged 17 and 15 respectively, who have profound HI at all frequencies in the two ears. The audiograms also display a flat configuration.

Fig. 2.

Fig. 2

Open in a new tab

Audiograms of hearing-impaired individuals a IV-2 and b IV-3 of family 4020, demonstrating profound hearing impairment involving all frequencies in the two ears. The normal audiograms of family members c III-2, d IV-1 and e IV-5 are also shown. Individual IV-5 is aged 21 years, while IV-2 is 17 and IV-3 15 years of age. Circles and crosses represent air conduction for right and left ear, respectively. Triangles represent masked air conduction thresholds

Extraction of genomic DNA and genotyping

Peripheral blood samples were obtained from six family members, four of whom were hearing-impaired. Genomic DNA was extracted from whole blood following a standard protocol [4], quantified by spectrophotometric reading at optical density 260, and diluted to 40 ηg/μl for polymerase chain reaction (PCR) amplification. A genome scan was carried out on six DNA samples at the Center for Inherited Disease Research (CIDR). A total of 405 fluorescently labeled short tandem repeat markers were genotyped. These markers are spaced at approximately 10 cM apart and are located on 22 autosomes and X and Y chromosomes. After the completion of the genome scan, two additional unaffected (IV-1 and IV-5) family members were ascertained and their DNA samples were used for fine mapping of the DFNB65 locus.

For fine mapping, microsatellite markers were amplified by PCR according to standard procedure in a total volume of 25 μl with 40 ηg genomic DNA, 0.3 μl of primer (Invitrogen Corp., Carlsbad, CA, USA), 200 μM deoxyribonucleotide triphosphate and 1 unit of Taq DNA Polymerase (Fermentas Life Sciences, Burlington, ON, Canada) in a thermal cycler (Whatman Biometra, Goettingen, Germany). PCR products were resolved on 8% nondenaturing polyacrylamide gel.

Linkage analysis

The National Center for Biotechnology Information Build 34 sequence-based physical map was used to determine the order of the genome scan markers and fine mapping markers [5]. Genetic map distances according to the Rutgers combined linkage-physical map of the human genome [6] were used to carry out the multipoint linkage analysis for the fine map and genome scan markers. For those genome scan markers where no genetic map position was available, interpolation was performed to place these markers on the Rutgers combined linkage-physical map. PEDCHECK [7] was used to identify Mendelian inconsistencies, while the MERLIN [8] program was utilized to detect potential genotyping errors that did not produce a Mendelian inconsistency. Haplotypes were constructed using SIMWALK2 [9, 10]. Two-point linkage analysis was carried out using the MLINK program of the FASTLINK computer package [11], and multipoint linkage analysis was performed using ALLEGRO [12]. An autosomal recessive mode of inheritance with complete penetrance and a disease allele frequency of 0.001 were assumed. Marker allele frequencies were estimated from the founders and the reconstructed genotypes of founders from this family and 32 additional Pakistani families that underwent a genome scan at the same time at CIDR. Equal allele frequencies were used for fine mapping markers because it was not possible to estimate allele frequencies from the founders, because these markers were only genotyped in this family. Because false positive results can be obtained when analyzing the data using too low of an allele frequency for the allele segregating with the disease locus [13], a sensitivity analysis was carried out for the multipoint linkage analysis by varying the allele frequency of the allele that is segregating with the disease locus from 0.2 to 0.7 for the fine mapping markers.

Results

Two point linkage analysis of the genome scan markers generated maximum LOD scores at θ=0 of 1.2 and 1.9 with markers D20S480 and D20S451, respectively. A maximum multipoint LOD score of 1.4 was obtained at D20S451. Figure 3 is a genome-wide representation of the multipoint LOD scores for the 22 autosomes. Only on chromosome 20 was a LOD score greater than 1.0 achieved.

Fig. 3.

Fig. 3

Open in a new tab

Genome-wide plot of multipoint LOD scores from the genome scan genotypes of family 4020. The LOD score is plotted against the genetic map distance in centimorgans (cM) per chromosome. Only on chromosome 20 at marker D20S451 was a multipoint LOD score of more than 1.0 (LOD=1.4) derived

To fine-map this region on chromosome 20, 33 additional markers located in the vicinity of D20S480 and D20S451 were selected from the Marshfield genetic map [14]. Sixteen of the markers were uninformative, although all markers that were selected had a reported heterozygosity of >0.7. Of the 17 fine mapping markers that are informative for linkage, eight markers (including D20S428, D20S877, D20S606, and D20S902) lie between genome scan markers D20S481 and D20S480, while seven markers (D20S840, D20S876, D20S1085, D20S832, D20S100, D20S102, and D20S430) are between D20S480 and genome scan marker D20S451. In addition, D20S164 and D20S171, two of the fine mapping markers, are distal to the genome scan marker D20S451. Analysis of the genome scan and fine mapping marker genotypes within this region with PEDCHECK and MERLIN did not elucidate any genotyping errors. Table 1 summarizes the two-point LOD scores obtained after fine mapping. The maximum two-point LOD score of 2.4 was obtained with five markers (D20S840, D20S1085, D20S832, D20S100 and D20S102) at θ=0. Multipoint linkage analysis resulted in a maximum LOD score of 3.3 at marker D20S840 (Fig. 4). The three-unit support interval was between markers D20S902 and D20S430, spanning 11.7 cM according to the Rutgers combined linkage-physical map of the human genome [6] and corresponding to a physical map distance of 4.4 Mb (Table 1).

Table 1.

Two-point LOD score results between the DFNB65 locus and chromosome 20 markers after fine mapping

Markera Genetic map positionb Physical map positionc LOD score at θ=
0.0 0.01 0.02 0.04 0.05 0.1 0.2 0.3
D20S428 80.34 51,588,497 – ∞ –1.39 –0.83 –0.32 –0.17 0.20 0.34 0.24
D20S877 80.49 51,799,558 0.85 0.83 0.81 0.78 0.76 0.67 0.48 0.29
D20S606 81.41 52,247,558 GO 0.31 0.56 0.77 0.81 0.88 0.70 0.41
D20S902 81.41 52,436,690 1.15 1.13 1.10 1.05 1.03 0.91 0.66 0.41
D20S480 d 82.62 52,542,629 1.31 1.28 1.26 1.20 1.17 1.03 0.75 0.46
D20S840 82.80 52,832,756 2.35 2.30 2.25 2.15 2.09 1.83 1.29 0.75
D20S876 83.13 53,033,103 1.15 1.13 1.10 1.05 1.03 0.91 0.66 0.41
D20S1085 84.67 53,366,902 2.35 2.30 2.25 2.15 2.09 1.83 1.29 0.75
D20S832 86.80 54,580,765 2.35 2.30 2.25 2.15 2.09 1.83 1.29 0.75
D20S100 88.19 54,999,475 2.35 2.30 2.25 2.15 2.09 1.83 1.29 0.75
D20S102 89.21 55,676,797 2.35 2.30 2.25 2.15 2.09 1.83 1.29 0.75
D20S430 93.11 56,835,225 – ∞ 0.61 0.86 1.06 1.10 1.15 0.93 0.57
D20S451 d 94.79 57,349,950 – ∞ 0.81 1.06 1.25 1.29 1.31 1.04 0.63
D20S164 95.96 57,738,776 –1.17 –0.49 –0.28 –0.08 –0.02 0.07 0.02 –0.03
D20S171 98.00 58,493,299 – ∞ –2.19 –1.66 –1.15 –1.00 –0.56 –0.23 –0.09
Open in a new tab
a

Markers in bold type flank the haplotype. Genome scan markers are shown in italics

b

Cumulative sex-averaged Kosambi cM genetic map distances from the Rutgers combined linkage-physical map of the human genome [6]

c

Sequence-based physical map distances in bases according to Build 34 of the human reference sequence [5]

d

The two unaffected individuals who were not included in the genome scan were also genotyped for these genome scan markers.

Therefore, the results reported in this table for the genome scan markers are different from those obtained from the initial genome scan

Fig. 4.

Fig. 4

Open in a new tab

The plot of the multipoint LOD scores including the fine mapping markers on chromosome 20 (broken line) is superimposed on the LOD score plot from the genome scan markers (solid line)

Haplotypes were constructed to determine the critical recombination events in the family (Fig. 1). All hearing-impaired individuals were homozygous for markers D20S902 and D20S480, but unaffected individual IV-1 is also homozygous for the same markers. Thus, the centromeric boundary of the interval is assigned between markers D20S480 and D20S840. The telomeric boundary of the interval corresponds to a recombination event between markers D20S102 and D20S430, which occurred in unaffected individual IV-1. The region of homozygosity that is flanked by markers D20S480 and D20S430 contains 4.3 Mb and is 10.5-cM long according to the Rutgers combined linkage-physical map of the human genome (Table 1). The region of homozygosity is therefore narrower than the three-unit support interval and most likely includes the gene for DFNB65.

When the marker allele frequencies for the alleles segregating with HI were varied for the fine mapping markers between 0.2 and 0.4, the maximum multipoint LOD score remained to be 3.3 at marker D20S840. When the allele frequencies were increased, the maximum LOD score still occurred at marker D20S840 but decreased to 3.2 and 3.1 at allele frequencies of 0.6 and 0.7, respectively.

Discussion

The linkage data presented here suggest that a novel gene for ARNSHI is located on chromosome 20q13.2–q13.32. This is the first HI locus identified on human chromosome 20. Many of the known nonsyndromic HI loci have been identified in single large families in which linkage can be established independently. However, due to the limited number of meiosis within each family, the genetic interval for the HI locus is often large, which makes gene identification difficult.

The DFNB65 interval spans 4.3 Mb and contains 17 known genes (Table 2). Of these, bone morphogenetic protein 7(BMP7, MIM 112267) has extensive expression in the otic placode of embryonic chick, which eventually becomes restricted to cochlear supporting cells over time [18]. This makes the BMP7 gene the best candidate of the known genes in the DFNB65 locus. However, sequencing of DNA for the BMP7 gene from one unaffected and two hearing-impaired individuals from family 4020 resulted in no functional sequence variants within the promoter and exonic regions.

Table 2.

DFNB65 candidate genes

Gene Protein MIM Physical map position Cytogenetic band Description
ZNF217 Zinc finger protein 217 602967 52,869,034 20q13.2 Putative oncogene
BCAS1 Breast carcinoma amplified sequence 1 602968 53,245,939 20q13.2 Putative oncogene
CYP24A1 a Cytochrome P450, family 24 precursor 126065 53,455,410 20q13.2 Has role in calcium metabolism and blood cell differentiation
PFDN4 Prefoldin 4 604898 53,509,924 20q13.2 Molecular chaperone; putative oncogene
DOK5 DOK5 protein isoform b 608334 53,777,679 10q13.2 Docking protein
CBLN4 Cerebellin 4 precursor – 55,257,918 20q13.31 Neuromodulatory glycoprotein in adrenal medulla
MC3R b Melanocortin 3 receptor 155540 55,509,211 20q13.31 Associated with obesity
STK6 Serine/threonine protein kinase 6 602687 55,629,867 20q13.31 Has role in spindle formation and chromosomal segregation
CSTF1 Cleavage stimulation factor sub-unit 1 600369 55,653,041 20q13.31 Ensures correct polyadenylation of mRNA
TFAP2C c Transcription factor AP-2 gamma 601602 55,889,780 20q13.31 Regulator of gene expression for ectodermal tissue development
BMP7 Bone morphogenetic protein 7 precursor 112267 56,430,977 20q13.31 Member of transforming growth factor-beta superfamily
SPO11 Meiotic recombination protein SPO11 isoform a 605114 56,590,253 20q13.31–q13.32 Required for meiotic double strand break formation
RAE1 RAE1 (RNA export 1, S.pombe) homolog 603343 56,611,567 20q13.32 May function in nucleocyto-plasmic transport and attaching cytoplasmic mRNA-binding protein to cytoskeleton
RNPC1 RNA-binding region containing protein 1 isoform – 56,651,885 20q13.32 Probable RNA-binding protein
HMG1L1 High-mobility group (nonhistone chromosomal) – 56,748,868 20q13.32 May be involved in tumorigenesis
CTCFL CCCTC-binding factor-like protein 607022 56,757,645 20q13.32 Putative oncogene
PCK1 Cytosolic phosphoenolpyruvate carboxykinase 1 261680 56,821,558 20q13.32 Main target for regulation of gluconeogenesis
Open in a new tab

This list does not include five genes that encode hypothetical proteins: C20orf108 (LOC116151), C20orf32 (HEF-like protein), C20orf43 (LOC51507), C20orf106 (LOC200232) and LOC388799

a

The Cyp24a1 rat manifests with albuminuria and hyperlipidemia [15]

b

Mc3r–/– mice have increased fat mass [16]

c

The Tfap2c knockout is embryonic lethal in mice [17]

Other genes within the interval that are included in expression databases for the inner ear [19, 20] are breast carcinoma amplified sequence 1 (BCAS1, MIM 602967) and prefoldin 4 (PFDN4, MIM 604898). These two genes have been implicated in several soft tissue cancers (e.g., breast, prostate, colon), which may indicate the oncogenic potential of these genes, although their specific functions remain unknown.

Acknowledgements

We wish to thank the family members for their invaluable participation and cooperation. This research was funded by the National Institutes of Health—National Institute of Deafness and other Communication Disorders grant R01-DC03594-06A1 and the Higher Education Commission, Islamabad, Pakistan. Genotyping services were provided by the Center for Inherited Disease Research (CIDR). CIDR is fully funded through a federal contract from the National Institutes of Health to The Johns Hopkins University, Contract Number N01-HG-65403.

Abbreviations

HI

hearing impairment

NSHI

nonsyndromic HI

ARNSHI

autosomal recessive NSHI

Biography

graphic file with name nihms-218864-b0005.gif Aamira Tariq received her MSc and MPhil degrees in Biochemistry/Molecular Biology from Quaidi-Azam University, Islamabad, Pakistan. She is presently serving as a Research Associate at the COMSATS Institute of Biosciences, Islamabad, Pakistan. Her area of teaching and research is molecular genetics of human diseases. She has been teaching courses and supervising research of students working in the area of human genetic disorders.

graphic file with name nihms-218864-b0006.gifSuzanne M. Leal received her M.S. in biostatistics and Ph.D. in epidemiology from Columbia University in New York City, USA. She is currently an Associate Professor in the Department of Molecular and Human Genetics at Baylor College of Medicine. Her research interests include gene mapping of complex and Mendelian traits, the genetics of nonsyndromic hearing impairment, and the study of methodological problems in statistical genetics.

Contributor Information

Aamira Tariq, Department of Biological Sciences, Quaid-I-Azam University, Islamabad, Pakistan.

Regie Lyn P. Santos, Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, Alkek N1619.01, Houston, TX 77030, USA

Mohammad Nasim Khan, Department of Biological Sciences, Quaid-I-Azam University, Islamabad, Pakistan.

Kwanghyuk Lee, Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, Alkek N1619.01, Houston, TX 77030, USA.

Muhammad Jawad Hassan, Department of Biological Sciences, Quaid-I-Azam University, Islamabad, Pakistan.

Wasim Ahmad, Department of Biological Sciences, Quaid-I-Azam University, Islamabad, Pakistan.

Suzanne M. Leal, Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, Alkek N1619.01, Houston, TX 77030, USA

References

  • 1.Parving A. Epidemiology of hearing loss and aetiological diagnosis of hearing impairment in childhood. Int J Pediatr Otorhinolaryngol. 1983;7:29–38. doi: 10.1016/s0165-5876(83)80020-2. [DOI] [PubMed] [Google Scholar]
  • 2.Morton NE. Genetic epidemiology of hearing impairment. Ann NY Acad Sci. 1991;630:16–31. doi: 10.1111/j.1749-6632.1991.tb19572.x. [DOI] [PubMed] [Google Scholar]
  • 3.Van Camp G, Smith RJH. Hereditary hearing loss homepage. 2005 http://webhost.ua.ac.be/hhh/
  • 4.Grimberg J, Nawoschik S, Bellusico L, McKee R, Trucks A, Eisenberg A. A simple and efficient non-organic procedure for the isolation of genomic DNA from blood. Nucleic Acids Res. 1989;17:83–90. doi: 10.1093/nar/17.20.8390. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.International Human Genome Sequence Consortium Initial sequence and analysis of the human genome. Nature. 2001;409:860–921. doi: 10.1038/35057062. http://genome.ucsc.edu/cgi-bin/hgGateway. July 2003 reference sequence as viewed in: Genome Bioinformatics Group of UC Santa Cruz. UCSC Genome Browser.
  • 6.Kong X, Murphy K, Raj T, He C, White PS, Matise TC. A combined linkage-physical map of the human genome. Am J Hum Genet. 2004;75:1143–1148. doi: 10.1086/426405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Connell JR, Weeks DE. PedCheck: a program for identification of genotype incompatibilities in linkage analysis. Am J Hum Genet. 1998;63:259–266. doi: 10.1086/301904. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Abecasis GR, Cherny SS, Cookson WO, Cardon LR. Merlin—rapid analysis of dense genetic maps using sparse gene flow trees. Nat Genet. 2002;30:97–101. doi: 10.1038/ng786. [DOI] [PubMed] [Google Scholar]
  • 9.Weeks DE, Sobel E, O'Connell JR, Lange K. Computer programs for multilocus haplotyping of general pedigrees. Am J Hum Genet. 1995;56:1506–1507. [PMC free article] [PubMed] [Google Scholar]
  • 10.Sobel E, Lange K. Descent graphs in pedigree analysis: applications to haplotyping, location scores, and marker-sharing statistics. Am J Hum Genet. 1996;58:1323–1337. [PMC free article] [PubMed] [Google Scholar]
  • 11.Cottingham RW, Jr, Indury RM, Schaffer AA. Faster sequential genetic linkage computation. Am J Hum Genet. 1993;53:252–263. [PMC free article] [PubMed] [Google Scholar]
  • 12.Gudbjartsson DF, Jonasson K, Frigge ML, Kong A. Allegro, a new computer program for multipoint linkage analysis. Nat Genet. 2002;25:12–13. doi: 10.1038/75514. [DOI] [PubMed] [Google Scholar]
  • 13.Freimer NB, Sandkuijl LA, Blower S. Incorrect specification of marker allele frequencies: effects on linkage analysis. Am J Hum Genet. 1993;52:1102–1110. [PMC free article] [PubMed] [Google Scholar]
  • 14.Broman K, Murray JC, Scheffield VC, White RL, Weber JL. Comprehensive human genetics maps: individual and sex specific variation in recombination. Am J Hum Genet. 1998;63:861–869. doi: 10.1086/302011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kasuga H, Hosogane N, Matsuoka K, Mori I, Sakura Y, Shimakawa K, Shinki T, Suda T, Taketomi S. Characterization of transgenic rats constitutively expressing vitamin D-24-hydroxylase gene. Biochem Biophys Res Commun. 2002;297:1332–1338. doi: 10.1016/s0006-291x(02)02254-4. [DOI] [PubMed] [Google Scholar]
  • 16.Chen AS, Marsh DJ, Trumbauer ME, Frazier EG, Guan SM, Yu H, Rosenblum DI, Vongs A, Feng Y, Cao L, Metzger JM, Strack AM, Camacho RE, Mellin TN, Nunes CN, Min W, Fisher J, Gopal-Truter S, MacIntyre DE, Chen HY, Van der Ploeg LH. Inactivation of the mouse melanocortin-3 receptor results in increased fat mass and reduced lean body mass. Nat Genet. 2000;26:97–102. doi: 10.1038/79254. [DOI] [PubMed] [Google Scholar]
  • 17.Werling U, Schorle H. Transcription factor gene AP-2 gamma essential for early murine development. Mol Cell Biol. 2002;22:3149–3156. doi: 10.1128/MCB.22.9.3149-3156.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Oh SH, Johnson R, Wu DK. Differential expression of bone morphogenetic proteins in the developing vestibular and auditory sensory organs. J Neurosci. 1996;16:6463–6475. doi: 10.1523/JNEUROSCI.16-20-06463.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.The Hearing Research Group at Brigham & Women's Hospital Human cochlear cDNA library and EST database. 2002 http://hearing.bwh.harvard.edu/estinfo.HTM.
  • 20.Holme RH, Bussoli TJ, Steel KP. Table of gene expression in the developing ear. 2002 http://www.ihr.mrc.ac.uk/hereditary/genetable/search.shtml.

RESOURCES