brief communications nature genetics • volume 26 • september 2000 15 Bardet-Biedl syndrome (BBS) is anautosomal recessive disorder with locus heterogeneity5–9. None of the ‘responsible’ genes have previously been identified. Some BBS cases (approximately 10%) remain unassigned to the five previ- ously mapped loci10. McKusick-Kaufma syndrome (MKS) includes hydrometrocol- pos, postaxial polydactyly and congenital heart disease, and is also inherited in an autosomal recessive manner11,12. We ascer- tained 34 unrelated probands with classic features of BBS including retinitis pigmen- tosa (RP), obesity and polydactyly. The probands were from families unsuitable for linkage because of family size. We found MKKS mutations in four typical BBS probands (Table 1). The first is a 13-year- old Hispanic girl with severe RP, PAP, men- tal retardation and obesity (BMI >40). She was a compound heterozygote for a mis- sense (1042G→A, G52D) and a nonsense (1679T→A, Y264stop) mutation in exon 3. Cloning and sequencing of the separate alleles confirmed that the mutations were present in trans. A second BBS proband (from Newfoundland), born to consan- guineous parents, was homozygous for two deletions (1316delC and 1324-1326del- GTA) in exon 3, predicting a frameshift. An affected brother was also homozygous for the deletions, whereas an unaffected sibling had two normal copies of MKKS. Both the proband and her affected brother had RP, PAP, mild mental retardation, morbid obe- sity (BMI >50 and 37, respectively), lobu- lated kidneys with prominent calyces and diabetes mellitus (diagnosed at ages 33 and 30, respectively). A deceased sister (DNA unavailable) had similar phenotypic fea- tures (RP with blindness by age 13, BMI >45, abnormal glucose tolerance test and IQ=64, vaginal atresia and syndactyly of both feet). Both parents and the maternal grandfather were heterozygous for the deletions. Genotyping with markers from the MKKS region12 confirmed homozygos- ity at 20p12 in both affected individuals. A third family included a 4-year-old male proband (the offspring of consan- guineous parents) with reduced visual acu- ity, PAP, obesity and cystic kidneys, and a sibling with hypospadias, PAP, obesity and lobular cystic kidneys who died at age 18 months (eye examination and DNA were unavailable). A fourth BBS family consisted of a female proband diagnosed at age five years because of severe RP, PAP, morbid obesity and diabetes mellitus, and a male sibling with RP, PAP, obesity, lobulated cys- tic kidneys and diabetes mellitus. Although not known to be related, both families are from the same region of Newfoundland. Sequencing revealed that the three affected individuals in these two families from whom DNA is available are homozygous for a frameshift mutation (1167 delT). Both parents of family 3 and the mother of family 4 were heterozygous for this muta- tion. No DNA was available from the father of family 4. Families 3 and 4 share a haplo- type of 5 SNPs within MKKS and 5 STRPs flanking the gene, indicating that they probably inherited the 1167delT mutation from a common ancestor. The mutations reported here were not found in 102 Northern European control chromo- somes13 or, in the case of the mutations in family 1, in 102 chromosomes from His- panic controls. MKKS is expressed in tis- sues affected by BBS, including retina, pancreas, brain and kidney (Fig. 1). MKKS represents a sixth locus for BBS and is the first gene shown to be mutated in patients with BBS. Mutations in MKKS cause MKS in the Amish4, in whom retinal degeneration, obesity and learning disabil- ity have not been reported12,14. We hypoth- esize that MKS is a distinct phenotype caused by hypomorphic alleles of MKKS, whereas BBS is caused by null alleles. Although the numbers are small, it is of interest that all known MKS patients have at least one missense mutation, whereas homozygous frameshift mutations are pre- sent in three BBS families. Alternatively, amino-acid polymorphisms in MKKS (ref. 4) or variations in the promoter region may influence the phenotype. The frequency of MKKS mutations in BBS patients (4/34 =11.8%; 95% CI 1–24%) is consistent with the percentage of cases unassigned to the five previously known loci. Identification of MKKS as a BBS gene may assist in the identification of other BBS genes. We hypothesize that the clini- cal features of BBS may be caused by the inability of the MKKS putative chaper- onin to maintain protein integrity in the retina, brain, pancreas and other organs. Our results suggest that genes encoding chaperonins and their substrates are can- didates for other BBS loci, RP, diabetes, obesity and mental retardation. Acknowledgements We thank A. McClain, S. Naylor, T. Young and D. Hefferton for patient ascertainment and collection Mutations in MKKS cause Bardet- Biedl syndrome Fig. 1 Human multiple tissue northern blots analysed with a probe from exons 3-6 of the full- length MKKS cDNA showing a 2.4-kb transcript in tissues affected by BBS. The numbering of muta- tions is based on the 5´ end of the MKKS cDNA with exon 1B (Genbank AF221993). Table 1 • Diagnosis of BBS and MKS Primary (P) or secondary (S) diagnostic Major diagnostic Clinical feature criteria for BBS criteria for MKS Family 1 Family 2 Family 3 Family 4 pigmentary retinopathy P + + + + polydactyly P + + + + + obesity P + + + + learning disability P + + + – hypogonadism P – – – – renal anomalies P – + + + diabetes mellitus S – + – + hydrometrocolpos + – – – – congenital heart disease S + – – – – A diagnosis of BBS requires four primary features or three primary features plus two secondary features2. A diagnosis of MKS requires all three major criteria12. © 2000 Nature America Inc. • http://genetics.nature.com © 2 0 0 0 N a tu re A m e ri c a I n c . • h tt p :/ /g e n e ti c s .n a tu re .c o m brief communications 16 nature genetics • volume 26 • september 2000 of DNA samples; C. Searby and H. Naik for technical assistance; and R. Swiderski for collection of RNA and assistance with northern- blot analysis. This work was supported in part by NIH grant EY11298 and the Foundation Fighting Blindness (V.C.S. and E.M.S.). V.C.S. is an associate investigator of the Howard Hughes Medical Institute. Anne M. Slavotinek1, Edwin M. Stone2, Kirk Mykytyn3, John R. Heckenlively4, Jane S. Green5, Elise Heon6, Maria A. Musarella7, Patrick S. Parfrey5, Val C. Sheffield3 & Leslie G. Biesecker1 1Genetic Diseases Research Branch, National Human Genome Research Institute, NIH, Bethesda, Maryland, USA. 2Department of Ophthalmology, and 3Howard Hughes Medical Institute and Department of Pediatrics, University of Iowa, Iowa City, Iowa, USA. 4Department of Ophthalmology, Harbor-UCLA Medical Center, Torrance, California, USA. 5Faculty of Medicine, Memorial University, St. Johns, Newfoundland, Canada. 6Department of Ophthalmology and Vision Science Research Program, University of Toronto, Toronto, Canada. 7Long Island College Hospital, Brooklyn, New York, USA. Correspondence should be addressed to V.C.S. (e-mail: val-sheffield@uiowa.edu). 1. Green, J. et al. N. Engl. J. Med. 321, 1002–1009 (1989). 2. Beales, P., Elioglu, N., Woolf, A., Parker, D. & Flinter, F. J. Med. Genet. 36, 437–446 (1999). 3. David, A. et al. J. Med. Genet. 36, 599–603 (1999). 4. Stone, D. et al. Nature Genet. 25, 79–82 (2000). 5. Kwitek-Black, A.E. et al. Nature Genet. 5, 392–396 (1993). 6. Leppert, M. et al. Nature Genet. 7, 108–112 (1994). 7. Sheffield, V.C. et al. Hum. Mol. Genet. 3, 1331–1335 (1994). 8. Carmi, R. et al. Hum. Mol. Genet. 4, 9–13 (1995). 9. Young, T.L. et al. Am. J. Hum. Genet. 64, 900–904 (1999). 10. Bruford, E.A. et al. Genomics 41, 93–99 (1997). 11. McKusick, V.A., Bauer, R.L., Koop, C.E. & Scott, R.B. JAMA 189, 813–816 (1964). 12. Stone, D. et al. Hum. Mol. Genet. 7, 475–481 (1998). 13. Rosenberg, M.J. et al. Am. J. Hum. Genet. 66, 419–427 (2000). 14. McKusick, V.A. Am. J. Hum. Genet. 30, 105–122 (1978). Fig. 1 CDH1 inactivation in HDGC tumours. a, Pedigrees of families 4201 and CHG72. ‘Y’, ‘N’ and ‘U’ indicate CDH1 germline mutations as present, absent or unde- termined. b, Methylation-specific PCR (ref. 8) of the CDH1 promoter5,10. Following sodium bisulphite treatment of genomic DNA, the CDH1 promoter was amplified using PCR primers specific for methylated (M) or unmethylated (U) promoters. CDH1 methylation is demonstrated in HDGC tumours 4201-5, CHG72-II-3 and CHG72- II-4. ‘MCF7 methylated’ and ‘MCF7’ are positive and negative controls. c, HDGC tumour summary showing CDH1-LOH, E-cadherin expression, CDH1 promoter methy- lation (filled circle, methylated; open circle, unmethylated) and CDH1 sequence (open box, wild type; filled box, somatic mutation; diagonal hashed box, unknown). ‘X’ designates presumptive silencing of methylated promoters. d, Map of the methylation status of the CpG dinucleotides (circles) located in the CDH1 promoter. a d c b Aberrant promoter methylation andthe associated loss of gene expression is a common accompaniment of human cancers1. Nonetheless, it has been challeng- ing to demonstrate in any given tumour that methylation of a specific gene was causal and not consequent to malignant transformation. In this regard, our atten- tion was drawn to the genesis of gastric can- cers in individuals with hereditary diffuse gastric cancer (HDGC). These individuals harbour germline mutations in the gene encoding E-cadherin, CDH1 (refs 2–4), but their cancers have consistently demon- strated absence of loss of heterozygosity at the CDH1 locus2–4. These findings sug- gested the hypothesis that CDH1 promoter methylation might function as the ‘second genetic hit’ in the genesis of these cancers. Our study was carried out in two previ- ously identified HDGC kindreds bearing known germline CDH1 mutations4 (Fig. 1a). Immunohistochemistry showed absence of E-cadherin protein expression from five of six gastric cancers from these families (Fig. 2d–f), suggesting these Methylation of the CDH1 promoter as the second genetic hit in hereditary diffuse gastric cancer © 2000 Nature America Inc. • http://genetics.nature.com © 2 0 0 0 N a tu re A m e ri c a I n c . • h tt p :/ /g e n e ti c s .n a tu re .c o m Mutations in MKKS cause Bardet-Biedl syndrome Acknowledgements References