1998 Oxford University Press
Human Molecular Genetics, 1998, Vol. 7, No. 7
1143–1148
A mutation in human keratin K6b produces a phenocopy of the K17 disorder pachyonychia congenita type 2 Frances J. D. Smith1, Marcel F. Jonkman2, Harry van Goor3, Carrie M. Coleman1, Seana P. Covello1, Jouni Uitto1 and W. H. Irwin McLean1,* 1Epithelial
Genetics Group, Department of Dermatology and Cutaneous Biology, Jefferson Medical College, 233 South 10th Street, Philadelphia, PA 19107, USA, 2Department of Dermatology and 3Department of Pathology, University Hospital, Hanzeplein 1, NL-9700 Groningen, The Netherlands Received February 17, 1998; Revised and Accepted March 12, 1998
Type I and type II keratins form the heteropolymeric intermediate filament cytoskeleton, which is the main stress-bearing structure within epithelial cells. Pachyonychia congenita (PC) is a group of autosomal dominant disorders whose most prominent phenotype is hypertrophic nail dystrophy accompanied by other features of ectodermal dysplasia. It has been shown previously that mutations in either K16 or K6a, which form a keratin expression pair, produce the PC-1 variant (MIM 184510). Mutations in K17 alone, an unpaired accessory keratin, result in the PC-2 phenotype (MIM 184500). Here, we describe a family with PC-2 in which the K17 locus on 17q was excluded and linkage to the type II keratin locus on 12q was obtained (Zmax 3.31 at θ = 0). Mutation analysis of candidate keratins revealed the first reported missense mutation in K6b, implying that this keratin is the previously unknown expression partner of K17, analogous to the K6a/K16 pair. Co-expression of these genes was confirmed by in situ hybridization and immunohistochemical staining. These results reveal the hitherto unknown role of the K6b isoform in epithelial biology, as well as genetic heterogeneity in PC-2. INTRODUCTION Keratins are a multigene family of heteropolymeric intermediate filament proteins which are expressed in characteristic type I/type II expression pairs in specific epithelial tissues (1,2). Mutations which either (i) affect the assembly of or (ii) cause complete absence of keratin filaments result in diseases characterized by fragility of the particular epithelial cells expressing the mutant protein (3,4). Defects in 16 keratins are now associated with human diseases, the most recent additions being defects in corneal keratins K3 and K12 causing Meesmann’s corneal dystrophy (5), trichocyte (hair) keratins hHb6 or hHb1 in the hair
disorder monilethrix (6,7) and the association of a mutation in simple epithelial keratin K18 with cryptogenic cirrhosis (8). Pachyonychia congenita (PC) is a group of largely autosomal dominant genodermatoses characterized by hypertrophic nail dystrophy accompanied by varying features of ectodermal dysplasia (9). There are two main clinical subtypes. The Jadassohn–Lewandowsky form (PC-1), where nail dystrophy is accompanied by focal palmoplantar keratoderma and oral leukokeratosis, is caused by mutations in keratin K16 (10) or its expression partner K6a (11). In the Jackson–Lawler form (PC-2) there is minimal oral involvement, milder keratoderma and multiple steatocystomas are a major clinical feature. Steatocystoma, also known as eruptive vellus hair cyst, is a cystic hamartoma lined by sebaceous ductal epithelium (12). We have previously shown that PC-2 can be caused by mutations in K17 (10,13) and that some families carrying K17 mutations have steatocystomas with little or no nail changes (13,14), a phenotype known as steatocystoma multiplex. Similarly, focal keratoderma without nail changes can result from K16 mutations (15) and, by implication, K6a mutations. K17 is a type I keratin expressed in a number of epidermal appendages, such as nail bed, hair follicle, sebaceous gland and other structures (16,17). No specific type II expression partner for K17 has been reported and, to date, we have found K17 mutations in 11 unrelated PC-2 or steatocystoma families (10,13,14). Here, we report the first PC-2 kindred where the causative mutation resides in the type II keratin K6b, revealing genetic heterogeneity in PC-2 and suggesting that this type II keratin is the expression partner of K17. Co-expression of these genes in differentiated epithelia was confirmed experimentally. RESULTS Clinical findings and linkage analysis A pachyonychia family was identified with the hallmarks of autosomal dominant inheritance (Fig. 1) and the constellation of clinical features typical of the PC-2 variant (Fig. 2). In particular, affected individuals had multiple steatocystomas with the histo-
*To whom correspondence should be addressed. Tel: +1 215 503 3241; Fax: +1 215 923 9354; Email:
[email protected]
1144 Human Molecular Genetics, 1998, Vol. 7, No. 7
Figure 1. Pedigree of the pachyonychia congenita type 2 family studied, exhibiting male-to-male transmission, consistent with autosomal dominant inheritance. Asterisks indicate persons from whom DNA was obtained, arrow indicates proband.
logically typical crenulated lining and some lesions containing vellus hairs in the lumen, a key diagnostic feature of PC-2 (data not shown). Exon 1 of the functional K17 gene (KRT17A) was amplified using conditions we have previously reported which do not amplify the two K17 pseudogenes (10). Direct sequencing of this fragment produced normal sequence identical to that published for K17 (17). To circumvent pseudogene contamination, we used cDNA derived from epidermal keratinocytes cultured from the skin of the proband to PCR amplify the entire K17 coding sequence. Full-length sequencing again revealed no mutations. DNA was obtained from additional family members as shown in Figure 1 and linkage analysis with informative marker D17S800 in the type I keratin cluster on 17q excluded this as the disease locus. However, linkage was obtained to marker D12S1651 in the type II keratin cluster on 12q, with a lod score of Zmax 3.31 at θ = 0. Candidate gene screening No type II keratin was known to have the same pattern of expression as K17; however, a clue was taken from the recent report of cloning six or more human genes encoding isoforms of keratin K6 (18). In this report, the close similarity in the coding sequences of these isoforms precluded detailed analysis of the expression patterns of these keratins. However, the representation in cDNA libraries of the two major expressed forms (K6a and K6b) was found to differ between different tissues of origin. Specifically, the K6b isoform was more common in scalp epidermis, relative to epidermal sites, where there are fewer hair follicles (18). K17 is most strongly expressed in the sebaceous gland and duct which accompany hair follicles (16,17), thus explaining the hyperplastic steatocystomas seen in PC-2 (10, 13). Therefore, we postulated that K6b might be the partner of K17 and harbor the mutation in family D. Primers and PCR conditions were developed to specifically amplify the full-length K6b isoform from keratinocyte cDNA, taking advantage of the fact that the 3′-untranslated region (UTR) has a number of fairly specific sequence sites (18). Sequencing of this fragment derived from the proband keratinocyte cDNA revealed a heterozygous missense mutation 1459G→A, abolish-
ing a BseRI restriction site and predicting the amino acid change E472K in the conserved helix termination motif of K6b (Fig. 3). A specific genomic PCR was developed for this region of the K6b gene and by BseRI digestion of this fragment, the mutation was found to co-segregate with the disease in family D and was excluded from 50 unrelated normal individuals. This mutation conforms to the CpG deamination model (19) and the analogous mutation has been seen in a number of other type II keratins where this sequence is conserved (4–6). Co-expression of K6b and K17 The close similarity in phenotype between PC-2 patients carrying K17 mutations and the family seen here with a K6b mutation strongly implies co-expression of these keratins in differentiated epithelia and so we set out to prove this hypothesis. No antibodies are known which can distinguish the K6a and K6b isoforms; however, we identified a 373 bp sequence in the 3′-UTR of the K6b mRNA which is poorly conserved in K6a (45% identity). This fragment of K6b was found to only hybridize with itself and not the equivalent K6a UTR sequence on Southern blots, as shown in Figure 4, demonstrating its specificity. In situ hybridization using the specific K6b UTR probe showed that K6b is expressed in the luminal cell layers of sebaceous ducts (Fig. 5A), where it co-localized with K17 protein, visualized by immunohistochemical staining (Fig. 5B). K6b was not expressed in epidermis, in sebaceous acini or in follicular epithelium. Similarly, K17 was not expressed in epidermis or in sebaceous acini, but, in contrast to K6b, was present in deeper parts of the follicular epithelium at the level of the istmus and the stem (data not shown). DISCUSSION Here, we have identified for the first time a second gene for keratin PC-2, K6b, and have shown that a heterozygous missense mutation E472K co-segregates with the disease in a large Dutch kindred. Thus, this disorder is genetically heterogeneous and can be caused by mutations in either protein of a keratin pair (4). The mutation was detected in the helix termination motif of the K6b
1145 Human Genetics, 1998, 7, No. NucleicMolecular Acids Research, 1994, Vol. Vol. 22, No. 1 7
Figure 2. Clinical features of pachyonychia congenita type 2 observed in the proband. (a) Focal palmoplantar keratoderma was observed on the pressure points of the feet. Similar focal palmar callosities were seen in some affected persons in the family. (b) Hypertrophic nail dystrophy (pachyonychia) was seen affecting all fingernails and toenails. (c) Steatocystomas, observed here on trunk skin, were widespread in post-pubescent affected individuals, typical of PC-2. (d) Mild lingual hyperkeratosis was observed on the margins of the tongue. Oral leukokeratosis is prominent in PC-1, since K16 and K6a are strongly expressed in these epithelia, but is not commonly observed in PC-2. These observations indicate that expression of K6b, like K17, is topographically restricted in oral and lingual epithelia.
1145
antibodies, so that this pairing had not been identified until now. In general, there is a tendency for mutations to occur more often in the type I protein of a keratin pair (4). This is reflected here by the fact that in 12 unrelated PC-2 families we have studied we detected 11 K17 mutations before finding the complementary K6b defect. The presence of a conserved CpG dinucleotide in the 1A domain of type I keratins only partly explains this skewing of mutations. The fact that this mutation in K6b produces a phenocopy of the K17 mutant phenotype implies that these keratins form a hitherto unknown expression pair. Here, we have confirmed this by a combination of in situ hybrization with a probe specific for K6b and immunohistochemical staining for K17 (Fig. 5). The expression of K6b and K17 in the sebaceous duct epithelium matches with Ackerman’s hypothesis that steatocystomas are cystic hamartomas of the sebaceous duct epithelium (12). From the PC-2 phenotype seen here, we expect that K6b is also expressed in the nail bed and in the epidermis of the palms and soles. Interestingly, expression of K6b and K17 was found to differ in deeper regions of the follicular epithelium (data not shown). This may be related to the occurrence of pili torti (twisted hair), a minor phenotype seen in PC-2. Although we have observed pili torti in several families with PC-2 due to mutations in K17 (10,13,14), this was not observed in the family studied here carrying the K6b mutation, consistent with the reduced follicular expression of K6b. In the follicular epithelium K17 may pair with a keratin other than K6b, perhaps another K6 isoform whose tissue distribution and function has yet to be elucidated (18). In view of this, the tissue distributions of the K6 isoforms and K17 require further study. It remains to be seen if the properties of K6a/K17 filaments differ significantly from K6b/K17 or K6b/K16 filaments. Among the seven amino acid substitutions between these K6 isoforms, four occur in less well-conserved regions of helix 1 and the H1 domain and are predicted to have a modest effect on polymerization (18). The remaining three substitutions occur in the V1 and V2 domains. The function of keratin variable domains is currently unclear, although two types of mutations in the V1 domain point to either a subtle role in filament assembly or in mediating interactions with non-keratin proteins (23,24). Future studies employing the yeast two-hybrid system, substitution of keratins in epithelial tissues by transgenics and in vitro biochemical assays are required to define the specific properties of the many differentially expressed keratin filament systems.
MATERIALS AND METHODS polypeptide, a short sequence which marks the end of the coiled coil rod domain. This motif has been implicated in molecular overlap interactions in the higher order polymerization of keratin heterodimers by virtue of its remarkable degree of evolutionary conservation (1), the results of in vitro mutagenesis studies (20) and chemical cross-linking analysis (21). The prevalence of pathogenic mutations in human keratin disorders at the helix boundary motifs further attests to the functional importance of these sites (4,22). The finding that K17 has an expression partner is surprising. The fact that the K6 isoforms only differ by seven isolated amino acid substitutions (18) has precluded their distinction using
Genotyping and linkage analysis Microsatellite markers were PCR amplified using 33P-labeled primers, analyzed on standard 6% sequencing gels and visualized by autoradiography. Marker D17S800 was used to exclude the type I keratin cluster on 17q; marker D12S1651 in the type II keratin cluster on 12q was fully informative and showed linkage to the disease. Two-point lod scores were computed using the MLINK algorithm of Linkage v.5.1, assuming a mutant allele frequency of 0.001 and 100% penetrance. Marker allele frequencies were assumed to be equal in the population and in the case of marker D12S1651 recalculation using a population
1146 Human Molecular Genetics, 1998, Vol. 7, No. 7
Figure 3. Detection and confirmation of a K6b mutation in affected members of family D. (a) Normal cDNA sequence in the region encoding the helix termination motif of the K6b polypeptide. (b) Analogous region of K6b cDNA sequence derived from the proband, showing heterozygous missense mutation 1459G→A (arrow), which disrupts a BseRI restriction enzyme site and is predicted to cause amino acid substitution E472K. (c) BseRI digestion of specific K6b genomic PCR fragments containing the mutation. Lanes 1–3, 5 and 6 represent affected persons from family D, which are heterozygous for the BseRI site. Fragments derived from unaffected family members (lanes 4 and 7) showed complete BseRI digestion, as did 50 normal unrelated controls (not shown).
Figure 4. Southern analysis of K6a and K6b 3′-UTR fragments. (A) Gel used for Southern analysis: lane 1, DNA standard type VI (Boehringer); lane 2, 373 bp PCR fragment derived from the 3′-UTR region of K6a; lane 3, corresponding PCR fragment derived from the 3′-UTR of K6b. (B) Southern blot of the gel shown in (A), probed with digoxygenin-labeled K6b fragment, showing specificity of this fragment for the K6b isoform.
frequency of 50% for the linked allele yielded the same lod score of Zmax = 3.31 at θ = 0. Mutation detection and confirmation Primary skin keratinocytes were cultured from punch biopsies in KGM medium (Gibco) by standard methods (25). Poly(A)+ mRNA was extracted from the cultured cells using a QuickPrep Micro mRNA Purification Kit (Pharmacia) and reverse transcribed using oligo(dT) and AMV reverse transcriptase (Promega). A 2013 bp fragment of the K6b cDNA was amplified using Boehringer Mannheim High Fidelity buffer containing 1 mM MgCl2, 4% dimethylsulfoxide and primers specific for K6b
cDNA: K6b.P1 (sense), 5′-CTC CAG CCT CTC ACA CTC TCC TA-3′; K6b.P2 (antisense), 5′-CTT CTC AGA ATT ATG GCA GAC TCA G-3′. The reactions were subjected to a ‘hot start’ with 1 U High Fidelity Enzyme Mix. PCR conditions were 94C for 5 min; 35 cycles of 94C for 30 s, 55C for 1 min, 72C for 2 min; a final extension of 72C for 5 min. PCR products were purified using a QIAquick PCR purification kit (Qiagen) and directly sequenced with both primers on an ABI 377 automated sequencer using the ABI PRISM fluorescent dye terminator system (Perkin Elmer). Genomic DNA was extracted from whole blood by standard methods and the mutation was confirmed by direct sequencing of K6b-specific genomic PCR products. A fragment of ∼1200 bp was amplified using primers K6b.P5 (sense, 5′-TTT CTC TTC CTT TGC CCT CCT-3′) and K6b.P2 (antisense, 5′-CTT CTC AGA ATT ATG GCA GAC TCA-G-3′) as above, except that the following conditions were used: 94C for 5 min; 35 cycles of 94C for 30 s, 55C for 1 min, 72C for 2 min; a final extension of 72C for 5 min. PCR products were purified and sequenced as above. Mutation E472K abolishes a recognition site for the restriction enzyme BseRI. Nested PCR was carried out using primers K6b.P2 and K6b.P5 as above, followed by reamplification using primers K6b.P9 (sense, 5′-GCT GGA AGG GCT GGA GGA TG-3′) and K6b.p10 (antisense, 5′-TGC TCA GTG CCA GAA CCT TGA A-3′). The resultant 210 bp fragment, which contains a single BseRI site, was exhaustively digested with BseRI. An uncut band was observed only in digests derived from affected members of family D. Cloning of a specific K6b probe and Southern analysis A 373 bp fragment from the 3′-UTR of the K6b mRNA was amplified from normal human keratinocyte cDNA using primers K6B.U1 (+ strand, 5′-GCC CTC ACT TTT CTT CTC ATC
1147 Human Genetics, 1998, 7, No. NucleicMolecular Acids Research, 1994, Vol. Vol. 22, No. 1 7
1147
Figure 5. (A) K6b mRNA visualized by non-isotopic in situ hybridization co-localizes with (B) K17 protein in the luminal cell layers of sebaceous glands in the epidermis. The latter was visualized by subsequent immunohistochemical staining of the same tissue section.
AA-3′) and K6B.U2 (– strand, 5′-CAA AGA GAG CAG AGA AAG CAG TG-3′). Similarly, the corresponding fragment of the K6a 3′-UTR was amplified using primers K6A.U1(+ strand, 5′-CTC ACT TCT TCT CTC TCT CTC TAT ACC AT-3′) and K6A.U2 (– strand, 5′-CAG TGA AGA GCA CAG AAA TCA TCA-3′). These fragments were cloned into the pCR2.1 vector (TA Cloning kit; Invitrogen) and fully sequenced. Clones in both forward and reverse orientations were selected for synthesis of sense and antisense RNA probes for in situ hybridization. PCR fragments generated from the K6a and K6b clones were Southern blotted and probed with digoxigenin-labeled K6b fragment. Blots were washed at 45C and the signal was detected by antidigoxigenin Fab fragments conjugated to alkaline phosphatase, with the color substrate nitroblue tetrazolium/5-bromo-4chloro-3-indolyl phosphate (NBT/BCIP; Boehringer Mannheim GmbH). In situ hybridization and immunohistochemistry A skin biopsy was obtained from the face of a normal individual. The specimen was fixed in 10% formalin overnight at 37C and embedded in paraffin. To produce digoxigenin-labeled RNA, sense and antisense K6b cDNA was transcribed in vitro with T7 RNA polymerase in the presence of digoxigenin-UTP according to the manufacturer’s protocol (Boehringer Mannheim GmbH). After overnight hybridization at 55C the sections were treated with RNase (Sigma R4875). The hybridized digoxigenin-labeled probes were immunodetected with anti-digoxigenin Fab fragments conjugated to alkaline phosphatase and the bound conjugate was visualized with NBT/BCIP (Boehringer Mannheim GmbH). Digital micrographs were taken with a high resolution full color CCD camera (Lumina Systems, Westborough, MA) mounted on a light microscope (Carl Zeiss Jena GmbH). The same section used for K6b in situ hybridization was washed with acetone to remove the NBT/BCIP precipitate and processed further for K17 detection. For the detection of K17 we used mouse monoclonal antibody E3 (16) as primary antibody, followed by rabbit anti-mouse IgG conjugated to horse radish peroxidase as second antibody (Dako A/S, Denmark). The bound conjugate was visualized with 3-amino-9-ethyl-carbazole in the peroxidase reaction. ACKNOWLEDGEMENTS We wish to thank Jane den Dunnen-Brigss, Jettie van der Wijk, Klaas Heeres and Miranda Nijenhuis (Gröningen) for their
technical assistance and Hans-Jürg Alder and his staff (Nucleic Acid Facility, Kimmel Cancer Center, Jefferson Medical College, Philadelphia, PA) for DNA synthesis and sequencing. Antibody E3 against K17 was a generous gift of Dr S. Trojanovsky (Moscow). This work was supported by grants from the Jan Kornelis de Cock-Stichting (to M.F.J.), the Dystrophic Epidermolysis Bullosa Research Associations (DEBRA) of the UK and USA (to W.H.I.M.) and by the US Public Health Service, National Institutes of Health (grant PO1-AR38923 to J.U.). NOTE ADDED IN PROOF Since submission of this manuscript, we have learned that another group has identified a PC-2 family which showed linkage to the type II keratin cluster. These data were presented at the 5th Joint Clinical Genetics Meeting, Los Angeles, CA, February 1998, by Drs D. Ross McLeod and Gail E. Graham, Alberta Children’s Hospital, Calgary, Alberta (G.E. Graham, personal communication). REFERENCES 1. Quinlan, R.A., Hutchison, C.J. and Lane, E.B. (1994) Intermediate filaments. In Sheterline, P. (ed.), Protein Profiles. Academic Press, London. 2. Fuchs, E. and Weber, K. (1994) Intermediate filaments: structure, dynamics, function and disease. Annu. Rev. Biochem., 63, 345–382. 3. McLean, W.H.I. and Lane, E.B. (1995) Intermediate filaments in disease. Curr. Opin. Cell Biol., 7, 118–125. 4. Corden, L.D. and McLean, W.H.I. (1996) Human keratin diseases: hereditary fragility of specific epithelial tissues. Exp. Dermatol., 5, 297–307. 5. Irvine, A.D., Corden, L.D., Swensson, O., Swensson, B., Moore, J.E., Frazer, D.G., Smith, F.J.D., Knowlton, R.G., Christophers, E. et al. (1997) Mutations in cornea-specific keratins K3 or K12 cause Meesmann’s corneal dystrophy. Nature Genet., 16, 184–187. 6. Winter, H., Rogers, M.A., Langbein, L., Stevens, H.P., Leigh, I.M., Labreze, C., Roul, S., Taieb, A., Kreig, T. and Schweizer, J. (1997) Mutations in the hair cortex keratin hHb6 cause the inherited hair disease monilethrix. Nature Genet., 16, 372–374. 7. Winter, H., Rogers, M.A., Gebhardt, M., Wollina, U., Boxall, L., Chitayat, D., Babul-Hirji, R., Stevens, H.P., Zlotogorski, A. and Schweizer, J. (1997) A new mutation in the type II hair cortex keratin hHb1 involved in the inherited hair disorder monilethrix. Hum. Genet., 101, 165–169. 8. Ku, N.O., Wright, T.L., Terrault, N.A., Gish, R. and Omary, M.B. (1997) Mutation of human keratin 18 in association with cryptogenic cirrhosis. J. Clin. Invest., 99, 19–23. 9. Stevens, H.P., Kelsell, D.P., Bryant, S.P., Bishop, D.T., Spurr, N.K., Weissensbach, J., Marger, D., Marger, R.S. and Leigh, I.M. (1996) Linkage of an American pedigree with palmoplantar keratoderma and malignancy (palmoplantar ectodermal dysplasia type III) to 17q24. Literature survey and proposed updated classification of the keratodermas. Arch. Dermatol., 132, 640–651.
1148 Human Molecular Genetics, 1998, Vol. 7, No. 7 10. McLean, W.H.I., Rugg, E.L., Lunny, D.P., Morley, S.M., Lane, E.B., Swensson, O., Dopping-Hepenstal, P.J.C., Griffiths, W.A.D., Eady, R.A.J., Higgins, C. et al. (1995) Keratin 16 and keratin 17 mutations cause pachyonychia congenita. Nature Genet., 9, 273–278. 11. Bowden, P.E., Haley, J.L., Kansky, A., Rothnagel, J.A., Jones, D.O. and Turner, R.J. (1995) Mutation of a type II keratin gene (K6a) in pachyonychia congenita. Nature Genet., 10, 363–365. 12. Steffen, C. and Ackerman, A.B. (1994) Steatocystomas. In Steffen, C. and Ackerman, A.B. (eds), Neoplasms with Sebaceous Differentiation. Lea and Febiger, Philadelphia, PA, pp. 148–181. 13. Smith, F.J.D., Corden, L.D., Rugg, E.L., Ratnavel, R., Leigh, I.M., Moss, C., Tidman, M.J., Hohl, D., Huber, M., Kunkeler, L. et al. (1997) Missense mutations in keratin 17 cause either pachyonychia congenita type 2 or a phenotype resembling steatocystoma multiplex. J. Invest. Dermatol., 108, 220–223. 14. Covello, S.P., Smith, F.J.D., Sillevis Smitt, J.H., Paller, A., Munro, C.S., Jonkman, M.F., Uitto, J. and McLean, W.H.I. (1998) Keratin 17 mutations cause either steatocystoma multiplex or pachyonychia congenita type 2. Br. J. Dermatol., in press. 15. Shamsher, M.K., Navsaria, H.A., Stevens, H.P., Ratnavel, R.C., Purkis, P.E., Kelsell, D.P., McLean, W.H.I., Cook, L.J., Griffiths, W.A.D., Gschmeissner, S. et al. (1995) Novel mutations in keratin 16 gene underly focal non-epidermolytic palmoplantar keratoderma (NEPPK) in 2 families. Hum. Mol. Genet., 4, 1875–1881. 16. Troyanovsky, S.M., Guelstein, V.I., Tchipysheva, T.A., Krutovskikh, V.A. and Bannikov, G.A. (1989) Patterns of expression of K17 in human epithelia: dependency on cell position. J. Cell Sci., 93, 419–426. 17. Troyanovsky, S.M., Leube, R.E. and Franke, W.W. (1992) Characterization of the human gene encoding cytokeratin 17 and its expression pattern. Eur. J. Cell Biol., 59, 127–137.
18. Takahashi, K., Paladini, R.D. and Coulombe, P.A. (1995) Cloning and characterisation of multiple human genes and cDNAs encoding highly related type II keratin 6 isoforms. J. Biol. Chem., 270, 18581–18592. 19. Cooper, D.N. and Krawczak, M. (1993) Human Gene Mutation. Bios Scientific, Oxford. 20. Letai, A., Coulombe, P.A., and Fuchs, E. (1992) Do the ends justify the means? Proline mutations at the ends of the keratin coiled-coil rod segment are more disruptive than internal mutations. J. Cell Biol., 116, 1181–1195. 21. Steinert, P.M., Marekov, L.N., Fraser, R.D.B. and Parry, D.A.D. (1993) Keratin intermediate filament structure: crosslinking studies yield quantitative information on molecular dimensions and mechanism of assembly. J. Mol. Biol., 230, 436–452. 22. Steinert, P.M., Yang, J.M., Bale, S.J. and Compton, J.G. (1993) Concurrence between the molecular overlap regions in keratin intermediate filaments and the locations of keratin mutations in genodermatoses. Biochem. Biophys. Res. Commun., 197, 840–848. 23. Kimonis, V., DiGiovanna, J.J., Yang, J.-M., Doyle, S.Z., Bale, S.J. and Compton, J.G. (1994) A mutation in the V1 end domain of keratin 1 in non-epidermolytic palmar-plantar keratoderma. J. Invest. Dermatol., 103, 764–769. 24. Uttam, J., Hutton, E., Coulombe, P.A., Anton-Lamprecht, I., Yu, Q.-C., Gedde-Dahl, T., Fine, J.-D. and Fuchs, E. (1996) The genetic basis of epidermolysis bullosa simplex with mottled pigmentation. Proc. Natl Acad. Sci. USA, 93, 9079–9084. 25. Mitra, R. and Nickoloff, B. (1994) Cultivation of human epidermal keratinocytes in serum-free growth medium. In Leigh, I.M. and Watt, F.M. (eds), Keratinocyte Methods. Cambridge University Press, Cambridge, pp. 17–19.
