Учебники / Genetic Hearing Loss Willems 2004

to be significantly higher in the region of the inner ridge cells, the main source of type II collagen for the tectorial membrane. Hybridization was not evident in the acellular tectorial membrane. Weaker Col11a2 RNA expression was detectable in the crista ampullaris and maculae of the saccule and utricle.

A transgenic mouse model for Col11a2 was developed through homologous recombination in which an inverted neomycin-resistance gene was inserted between restriction sites in exons 27 and 28 (12). Transcription of shortened mRNAs was shown, but translation did not occur owing to the introduction of premature termination codons. The heterozygous mice are normal in appearance suggesting that a single copy of the Col11a2 product is adequate to prevent detectable abnormalities. However, the homozygous Col11a2 mice exhibit smaller snouts with a shortened midface, are smaller in size, and have abnormal hind-limb placement.

Click-evoked auditory brainstem response (ABR) testing was performed on homozygous and heterozygous Col11a2-deficient mice as well as their wild-type littermates (5). Animals were tested at 2 months, 6 months, and 12 months of age. Heterozygous (+/-) and wild-type (+/+) mice had similar hearing thresholds. The homozygous mice (-/-) were found to have an elevated hearing threshold that averaged 43 dB when compared to wild-type and heterozygous littermates. The degree of hearing impairment was not influenced by age at testing.

Histological analysis was performed on the temporal bones of 10 animals (4 homozygotes, 2 heterozygotes, and 4 wild-type mice) (5). At the light microscopic level, the tectorial membranes of the Col11a2-deficient mice were noted to be larger and less compact when compared to the tectorial membrane of heterozygous and wild-type mice (Fig. 4A,B). No other cochlear abnormalities were identifiable. By electron microscopy, tectorial membrane changes were more pronounced (Fig. 5A,B). Collagen fibrils in the homozygous mutants were irregular placed with widened interfibrillar distances (5), but in both the heterozygous and wild-type mice, the fibrils were in a parallel array and closely approximated.

VI. SUMMARY

To date, 16 di erent mutations have been described in the COL11A2 gene and cause a diverse spectrum of phenotypic abnormalities, all with associated sensorineural hearing loss. Two mutations cause the subtlest phenotype— nonsyndromic prelingual sensorineural deafness. Three mutations have been reported to cause Stickler syndrome type 3—an exon-skipping mutation, a 27-bp in-frame deletion, and Gly955Glu. The remaining 11 mutations cause OSMED syndrome—all e ectively producing a null allele (13). Consanguin-

ity was observed in five of these cases. All parents of the OSMED patients were phenotypically normal, implying that haploinsu ency alone does not cause a clinical e ect.

The midfrequency hearing impairment associated with the DFNA13 phenotype is an unusual pattern of inherited hearing impairment. The ‘‘cookie-bite’’ audiometric pattern of hearing loss has classically been ascribed to genetic factors. However, of all the described nonsyndromic hearing loss loci, only the DFNA8/12, DFNB21 (TECTA), and DFNA21 audiometric phenotypes share the predominant midfrequency loss seen in DFNA13. The association of TECTA and COL11A2 with the tectorial membrane implies that the micromechanical properties of this structure are important for e ective mechanosensory transduction in the cochlea.

REFERENCES

1.Van Camp G, Smith RJH. Hereditary Hearing Loss Homepage. World Wide Web URL:http://dnalab-www.uia.ac.be/dnalab/hhh/.

2.Brown MR, et al. A novel locus for autosomal dominant nonsyndromic hearing loss, DFNA13, maps to chromosome 6p. Am J Hum Genet 1997; 6: 924–927.

3.Vikkula M, et al. Autosomal dominant and recessive osteochondrodysplasias associated with the COL11A2 locus. Cell 1995; 80:431–437.

4.Sirko-Osadsa DA, et al. Stickler syndrome without eye involvement is caused by mutations in COL11A2, the gene encoding the alpha-2 (XI) chain of type XI collagen. J Pediatr 1998; 132:368–371.

5.McGuirt WT, et el. Mutations in COL11A2 cause non-syndromic hearing loss (DFNA13). Nat Genet 1999; 23:413–419.

6.Kivirikko KI. Collagens and their abnormalities in a wide spectrum of diseases. Ann Med 1993; 25:113–126.

7.Ala-Kokko L, Prockop DJ. In: Kelly’s Textbook of Rheumatology, 6th ed.

8.Vuoristo MM, et al. Complete structure of the human COL11A2 gene: the exon sizes and other features indicate the gene has not evolved with genes for other fibriller collagens. Ann NY Acad Sci 1996; 785:343–344.

9.Spranger J. The type XI collagenopathies. Pediatr Radiol 1998; 28:745–750.

10.De Leenheer EM, Kunst HH, McGuirt WT, Prasad SD, Brown MR, Huygen PL, Smith RJ, Cremers CW. Autosomal dominant inherited hearing impairment caused by a missense mutation in COL11A2 (DFNA13). Arch Otolaryngol Head Neck Surg 2001; 127(1):13–17.

11.Kunst H, Huybrechts C, Marres H, Huygen P, Van Camp G, Cremers C. The phenotype of DFNA13/COL11A2: nonsyndromic autosomal dominant midfrequency and high-frequency sensorineural hearing impairment. Am J Otol 2000; 21(2):181–187.

12.Li SW, Takanosu M, Arita M, Bao Y, Ren ZX, Maier A, Prockop DJ, Mayne R. Targeted disruption of Col11a2 produces a mild cartilage phenotype in transgenic mice: comparison with the human disorder otospondylomegaepiphyseal dysplasia (OSMED). Dev Dyn 2001; 222(2):141–152.

13.Melkoniemi M, et al. Autosomal recessive disorder otospondylomegaepiphyseal dysplasia is associated with loss-of-function mutations in the COL11A2 gene. Am J Hum Genet 2000; 66(2):368–377.

either induce or repress transcription of other genes, thus controlling the expression of other proteins.

Gene regulation in eukaryotes occurs at the level of transcriptional regulation, as well as RNA processing, mRNA stability, e ciency of translation of the mRNA into a polypeptide, and other posttranscriptional mechanisms. Transcriptional regulation involves binding of a general transcription machinery to the basal promoter that lies adjacent to the transcription start site, as well as binding of di erent transcription factors to specific DNA sequences that are at variable distances from the basal promoter (Fig. 1). The general transcription machinery consists of various proteins, among them transcription Factor IID, a multiprotein complex that binds to a DNA stretch containing an AT-rich region, named the TATA box. The specific transcription factors can bind to distant locations and probably draw the chromatin into a loop to establish protein-protein interactions with the basal promoter. The unique combination of transcription factors bound to the regulatory

Figure 1 Eukaryotic gene expression depends on both general and specific transcription factors. The general transcription factors, which are the same for all genes, are involved in the positioning of the RNA polymerase II at the promoter; separation of the two DNA strands so that transcription can begin; and release of the RNA polymerase once transcription has begun. Specific transcription factors, such as the POU-domain transcription factors, can bind to regulatory sequences at a variable distance from the promoter, even thousands of base pairs away. Regulation of spatiotemporal gene expression is mediated by the combination of specific transcription factors bound the regulatory sequences of a gene and their interaction with the proteins at its promoter. (Adapted with permission from Ref. 67.)

regions of a gene determines whether its transcription will be turned on or o in a particular cell at any given time.

Transcription factors can be classified according to the structure of their DNA-binding domain (e.g., helix-turn-helix, helix-loop-helix, zinc fingers, and leucine zipper), their function in transcriptional regulation (e.g., activation, coactivation, repression and corepression, architectural factors, chromatin remodeling factors, and transcription elongation factors), as well as their spatiotemporal expression and involvement in specific roles in the cell cycle, cell identity and fate, developmental processes, hormone response, and more. The transcription factors themselves are regulated by modifications induced by other genes, such as kinases or phosphatases, to name a few. Most important, di erent transcription factors from the same family can have similar but distinct functions.

Most transcription factors regulate the expression of a myriad of genes. It is therefore not surprising that mutations in transcription factors cause a multitude of diseases. Many of these mutations a ect more than one organ system, resulting in a syndrome (reviewed in Ref. 5). As key regulators of developmental processes, any of their identified targets can illuminate the pathways of cell commitment and maturation as well as represent valid candidate genes for similar or more restricted diseases (e.g., Ref. 6). A wellcharacterized example is that of the SOX10 pathway. Mutations in SOX10, a transcription modulator, underlie Shah-Waardenburg syndrome (OMIM #277580, http://www.ncbi.nlm.nih.gov/Omim), a neurocristopathy that associates intestinal aganglionosis, pigmentation defects, and sensorineural deafness (7). GJB1 (gap junction beta 1) encodes the human connexin 32 protein and is a downstream target of SOX10. Both mutations in this gene, and a recently reported mutation in the promoter of this gene, underlie an autosomal dominant demyelinating disease, an X-linked form of Charcot-Marie- Tooth disease (OMIM #302800) (8). Other targets of SOX10, PAX3 and MITF, cause di erent subtypes of Waardenburg syndrome (WS type I, OMIM #193500; WS type IIA, OMIM #193510), a disease with auditorypigmentary abnormalities (9,10).

A selection of transcription factors play a role in pivotal steps of cell fate decision and di erentiation in the mammalian inner ear, as can be learned from both mouse mutants and human disease genes (reviewed in Ref. 11). For example, the basic helix-loop-helix (bHLH) genes neurogenin-1 (Ngn-1) and NeuroD are essential for the formation of the otic ganglion (12–14), while mutations in the zinc finger protein GATA3 cause hypomorphic development of the VIIIth cranial nerve and the whole inner ear (15). The bHLH transcription factor Math1 is required for hair cell fate specification in mice and ectopic expression of this protein in the rat results in the development of supernumerary hair cells (16), while the sensory epithelium of a mouse knock-

out model of this gene never develops hair cells (17). Gene-targeted mutagenesis of another group of bHLH proteins, the hairy/enhancer of split (Hes) transcriptional repressors, leads to supernumerary inner and outer hair cells in Hes1 and Hes5 mutants, respectively (18). Last but not least, mutations in the Pou3f4 and Pou4f3 genes cause deafness and distinct inner ear phenotypes, as will be further discussed.

III.THE POU FAMILY OF TRANSCRIPTION FACTORS

The POU-domain family of transcription factors was identified on the basis of amino acid sequence homology in the DNA-binding domain of the transcription factors PIT1/GHF1, OCT1 and OCT2, and UNC86 (19). PIT1/ GHF1 is involved in controlling transcription of the growth hormone, prolactin, and other pituitary-specific genes. OCT1 activates transcription of histone H2B genes and OCT2 activates immunoglobulin genes transcription in B lymphocytes. UNC86 is a protein that determines neuroblast fate in the nematode Caenorhabditis elegans (20,21). Since the discovery of the first POU-domain proteins, many additional proteins have been identified in this family, sharing high sequence homology within the POU domain and di ering substantially elsewhere (2). The POU domain consists of 147–156 amino acids and is comprised of two distinct DNA binding domains: a 69–78- amino-acid POU-specific domain located amino terminal to a 60-amino-acid POU homeodomain (22) (Fig. 2). The two POU domains are separated by a variable linker, a flexible stretch of amino acids that increases the repertoire of the specific sequences to which these proteins can bind to and improves the kinetics of the binding (reviewed in Ref. 23).

The three-dimensional structure of the Oct-1 POU-domain protein was the first to be resolved and shed some light on the protein-DNA interactions of this bipartite DNA-binding domain family (24,25). The POU-specific domain contains four alpha-helices surrounding a hydrophobic core, with the second and third comprising a helix-turn-helix motif. The POU-domain homeodomain motif is a homeodomain DNA-binding domain consisting of three alpha helices, creating a helix-turn-helix tertiary structure. This domain is part of the homeodomain family of transcription factors, DNA-binding motifs. Nevertheless, unlike the other transcription factors within this group (e.g., Hox and engrailed), both the POU-specific domain and POU homeodomain are required for high-a nity sequence-specific DNA binding (26). This is done via the third helix of the POU homeodomain and the third helix of the POU-specific domain, both binding to the major groove of the DNA alpha-helix structure.

Figure 2 POU-domain transcription factors contain a bipartite DNA-binding domain, comprised of a POU-specific domain and a POU homeodomain, separated by a variable linker. A model of the position of the POU domains on DNA. The POUspecific domain contacts the DNA opposite from the POU homeodomain in the adjacent major groove. (Adapted with permission from Ref. 68.)

A.The POU-Domain Class III Transcription Factors

The class III of POU-domain genes was identified after the initial definition of this transcription factor family. The genes of this class were identified using the polymerase chain reaction (PCR) and degenerate oligonucleotides representing codons of the nine conserved amino acids in the original POU genes. Initially, three POU-domain class III genes were isolated from cDNA derived from human brain and rat brain and testes, namely, Pou3f1 (also named Tst- 1, Oct-6, SCIP, or Otf-6), Pou3f2 (also named Brn-2, N-Oct3, N-Oct5, or Otf- 7), and Pou3f3 (also named Brn-1 or Otf-8) (27). The fourth member of this group, Pou3f4 (also named Brn-4, RHS2, N-Oct4, or Otf-9), was the thirteenth POU domain protein to be identified in mammals, this time via screening of a rat hypothalamic cDNA library using the Brn-2 POU domain as a probe (28). POU domain class III transcription factors have also been identified in Drosophila melanogaster (29), C. elegans (30), Xenopus laevis (31),

Danio rerio (zebrafish) (32), and metazoans (33). All mammalian class III