letter - Weizmann Institute of Science

© 2001 Nature Publishing Group http://genetics.nature.com

letter


The UDP-N-acetylglucosamine 2-epimerase/Nacetylmannosamine kinase gene is mutated in recessive hereditary inclusion body myopathy Iris Eisenberg1, Nili Avidan3*, Tamara Potikha1*, Hagit Hochner1, Miriam Chen3, Tsviya Olender3, Mark Barash1, Moshe Shemesh1, Menachem Sadeh4, Gil Grabov-Nardini1, Inna Shmilevich1, Adam Friedmann1, George Karpati5, Walter G. Bradley6, Lisa Baumbach7, Doron Lancet3, Edna Ben Asher3, Jacques S. Beckmann3, Zohar Argov2 & Stella Mitrani-Rosenbaum1 * These authors contributed equally to this work.

Published online: 27 August 2001, DOI: 10.1038/ng718 Hereditary inclusion body myopathy (HIBM; OMIM 600737) is a unique group of neuromuscular disorders characterized by adult onset, slowly progressive distal and proximal weakness and a typical muscle pathology including rimmed vacuoles and filamentous inclusions1. The autosomal recessive form described in Jews of Persian descent2 is the HIBM prototype. This myopathy affects mainly leg muscles, but with an unusual distribution that spares the quadriceps3. This particular pattern of weakness distribution, termed quadriceps-sparing myopathy (QSM), was later found in Jews originating from other Middle Eastern countries as well as in non-Jews4. We previously localized the gene causing HIBM in Middle Eastern Jews on chromosome 9p12–13 (ref. 5) within a genomic interval of about 700 kb (ref. 6). Haplotype analysis around the HIBM gene region of 104 affected people from 47 Middle Eastern families indicates one unique ancestral founder chromosome in this community. By contrast, single non-Jewish families from India, Georgia (USA) and the Bahamas, with QSM

Fig. 1 Physical map of the HIBM locus and genomic organization of GNE. a, Relative positions of landmark microsatellite markers are indicated along the 700-kb HIBM candidate interval limited by markers 327GT4 and D9S1859 (red arrows), spanned by a contig consisting of six BAC clones (named by their accession numbers) sequenced at the Sanger Center. The genes and transcripts identified in the region at different stages of the positional cloning procedure are shown. OR2S2, an olfactory receptor, and CCIN (calicin) which is expressed only in testis, were not considered functional candidate genes. b, Expanded view of a 50-kb segment showing the genomic organization of GNE and the position of the genomic segments encoding the epimerase (pale gray) and kinase (dark gray) domains of the protein. Filled boxes, coding exons; open boxes, non-coding exons. A TATA box motif has been recognized 10 kb upstream from exon 1.

and linkage to the same 9p12–13 region, show three distinct haplotypes. After excluding other potential candidate genes, we eventually identified mutations in the UDP-N-acetylglucosamine2-epimerase/N-acetylmannosamine kinase (GNE) gene in the HIBM families: all patients from Middle Eastern descent shared a single homozygous missense mutation, whereas distinct compound heterozygotes were identified in affected individuals of families of other ethnic origins. Our findings indicate that GNE is the gene responsible for recessive HIBM.

We previously established a high-resolution physical and transcriptional map of the HIBM chromosomal region on chromosome 9p12–13 and defined the restricted HIBM critical region between markers 327GT4 and D9S1859, an interval of about 700 kb (Fig. 1; ref. 6). Based on their expression in skeletal muscle as well as other tissues, five genes located in this interval were potential candidate genes (Fig. 1)6. LOC64148 and FLJ21343 are newly defined mRNAs of unknown function; clathrin light chain

a

b

1Unit for Molecular Biology and 2Department of Neurology, Hadassah Hospital, The Hebrew University-Hadassah Medical School, Jerusalem, Israel. 3Department of Molecular Genetics and Crown Genome Center, The Weizmann Institute of Science, Rehovot, Israel. 4Department of Neurology, Wolfson Hospital, Holon, Israel. 5Department of Neurology and Neurosurgery, Montreal Neurological Institute, Montreal, Canada. 6Department of Neurology and 7Department of Pediatrics, University of Miami School of Medicine, Miami, Florida, USA. Correspondence should be addressed to S.M.-R.

(e-mail: [email protected]).

nature genetics • volume 29 • september 2001

83

letter


A (CLTA; ref. 7) is a regulatory element in clathrin gene function, known to be involved in several pathways of lysosomal proteolysis; reversion-inducing cysteine-rich protein with Kazal motifs (RECK; ref. 8) is a membrane-anchored glycoprotein with transformation suppressor activity; and UDP-N-acetylglucosamine 2epimerase/N-acetylmannosamine kinase (GNE) is a bifunctional enzyme known to regulate and initiate biosynthesis of N-acetyl-


a

b

c

d

84

neuraminic acid (NeuAc; ref. 9). Mutations in GNE have been related to the monogenic-dominant disease sialuria, a rare inborn error of metabolism characterized by cytoplasmic accumulation and increased urinary excretion of free sialic acid10. Systematic sequencing of the coding regions of LOC64148, FLJ21343, CLTA and RECK, which could not be excluded as possible functional candidate genes in HIBM, revealed no diseasecausing alleles in any of those four genes in HIBM families. By thorough screening of all 13 exons and cDNA of GNE in HIBM patients, we have identified 7 different mutations. All 104 patients from the 47 Middle Eastern families analyzed shared a single T→C transition in exon 12, converting methionine to threonine at codon 712 (Fig. 2). None of the 306 unaffected non–Middle Eastern control chromosomes were found to bear this mutation, whereas 5 of 73 unaffected, unrelated Persian Jewish individuals were found to be heterozygous carriers of the mutation. The latter figure agrees with the expected estimated frequency of heterozygotes derived from the prevalence of the disease in this population (1:1,500)4. The different haplotypes found in the ethnically unrelated HIBM patients, who were all heterozygotes with respect to the HIBM interval, suggest the existence of different disease-causing mutations. Indeed, each one of the families from India, Georgia and the Bahamas show two different mutations in various exons of GNE (Table 1 and Fig. 3). None of these mutations were found in any of the 132 control chromosomes analyzed. The identification of seven different mutations in GNE in HIBM patients contrasts with the paucity of polymorphic variants encountered by fully sequencing the coding region of eight healthy unrelated individuals: not a single polymorphism causing an amino acid change was recognized, nor was any found by an extensive search of the SNP database (http://www. ncbi.nlm.nih.gov/SNP). This observation strongly incriminates GNE as the gene that, when mutated, causes HIBM. UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase is a bifunctional enzyme that catalyzes the first two steps in the biosynthesis of N-acetylneuraminic acid (sialic acid)11. Site-directed mutagenesis has shown that the two functions of the protein are mediated by separate domains—the epimerase activity by the amino-terminus and the kinase activity by the carboxy-terminus—and can be Fig. 2 GNE mutation in Middle Eastern Jewish HIBM families. a, Pedigree of a Persian HIBM prototype family, showing the segregation of the ancestral founder chromosome with the disease. Diseased chromosomes are in gray boxes, healthy chromosomes in empty boxes. n, not done. b, Sequence chromatograms from a healthy (WT) and a diseased (M) genotype for the Middle Eastern HIBM mutation, a T→C transition in exon 12, converting methionine to threonine at codon 712. c,d, Segregation of the Middle Eastern mutation, as assayed by NlaIII restriction analysis. NlaIII digests the normal 677-bp PCR product generated between the F and R primers into fragments of 341, 156, 128 and 52 bp. The mutation results in the loss of the NlaIII site at position 284, and therefore restriction digestion produces a 208-bp fragment instead of the 156-bp and 52-bp fragments. In healthy individuals (+/+) III:1 and III:6, NlaIII cleavage generates a 156-bp fragment; in heterozygous carriers (+/–) II:3, II:4, II:5, II:6, III:2 and III:3, both 156-bp and 208bp fragments; and in affected individuals (–/–) II:1, II:2, III:4, III:5 and III:7, only the 208-bp fragment is generated. (The 128-bp fragment is present in all individuals; the 341-bp fragment, also common to all individuals, is not shown in the figure.)



letter


a

b

d

f

c

e

g

Fig. 3 GNE mutations in non–Middle Eastern, non-Jewish HIBM families. a, Pedigrees of the three HIBM families linked to 9p12-13, presenting heterozygous haplotypes in the HIBM restricted interval. b,c, Sequence chromatograms for the two different mutations found in the India family: a nonsense mutation (C303X) in exon 5 causing premature termination of the protein and a missense transition (G→A) converting valine to methionine (V696M) in exon 12. d,e, Sequence chromatograms for the two heterozygous missense mutations found in the Georgia family: G576E, changing glycine to glutamic acid in exon 10, and A631T, converting alanine to threonine in exon 11. f,g, Sequence chromatograms for the two mutations identified in the Bahamas family, both in exon 4: D225N, converting aspartic acid to asparagine, and R246Q, converting arginine to glutamine.

carried out independently12. In sialuria, the three dominant missense mutations identified at codons 263 and 266 occur in the epimerase domain of the protein, leading to a loss of feedback inhibition of the enzyme. This, in turn, causes constitutive overproduction of free sialic acid in these patients10. By contrast, the unique phenotype of HIBM can be caused by mutations either in the regions encoding the epimerase domain, in the kinase domain, or in both. The missense homozygous or compound heterozygous mutations (which do not overlap with the sialuria mutations) do not, however, result in detectable overproduction of urinary sialic acid in HIBM patients. The broad distribution of mutations along the gene tends to exclude the impairment of a solitary aspect of the enzyme’s function as an underlying cause for HIBM. It suggests instead a more nonspecific disruption that results in a general loss of function. The high proportion of mis-

sense mutations, combined with the fact that all patients have at least one missense mutation (and in most cases two), suggests that loss of function is partial rather than complete. The fact that none of the patients of any of the various ethnic origins carries two truncating mutations raises the possibility that a homozygous complete loss of function of GNE might be lethal. To gain insight into the significance of the effect of HIBM mutations, we compared the human enzyme with its orthologs. GNE has orthologs in rat and mouse, both with 98% identity to the human protein, but none in Drosophila melanogaster, Caenorhabditis elegans or yeast. Indeed, GNE is one of 40 genes that are exclusively shared by vertebrates and bacteria and are examples of horizontal transfer from bacteria to vertebrates13. The two enzymatic activities of GNE are carried out by separate proteins in bacteria. We compared two sets

Table 1• GNE mutations in HIBM patients Family origin

Number of families

Genotypea

Exon

Nucleotide positionb

Amino acid change

Protein domain

Restriction site changec

Middle Eastern Jews

47

homozygous

12

2186T→C

Met→Thr (M712T)

kinase

− NlaIII

India

1

compound heterozygous

12 5

2137G→A 960T→A

Val→Met (V696M) Cys→stop (C303X)

kinase epimerase

− FokI + BbvCI

Georgia

1


10 11

1778G→A 1942G→A

Gly→Glu (G576E) Ala→Thr (A631T)

kinase kinase

− Sau96I − Fnu4HI

Bahamas

1


4 4

788G→A 724G→A

Arg→Gln (R246Q) Asp→Asn (D225N)

epimerase epimerase

− NlaIV − TspRI

aIn

each of the compound-heterozygous families, the first substitution described is the inherited maternal mutation and the second one is the inherited paternal mutation. bNucleotide position in GNE cDNA, accession no. XM_005500. c (–), mutation abolishes the restriction site; (+), mutation generates the restriction site. As controls for the Middle Eastern mutation, 352 chromosomes from healthy unrelated individuals were analyzed. For each of the heterozygous mutations found in the families from India, Georgia and Bahamas, 132 chromosomes from unrelated control individuals were analyzed.


85

letter


b


a

c

Fig. 4 Blocks of multiple alignments of GNE with epimerases (a,b) and sugar kinases (c); bacterial orthologs are shown around each amino acid substitution. The amino acid position in the protein is indicated to the right of each block. Mutation positions are indicated by the top arrows and the amino acid substitutions are marked under each block. The sialuria pathogenic mutation block is included in b. Percent identities of the different proteins are shown in parenthesis.

of multiple alignments with several bacterial orthologs (Fig. 4). Three of the HIBM mutations are located at positions that are highly conserved in bacteria, possibly implying that the modified proteins might be severely altered functionally. The other three missense mutations occur at non-conserved positions; their functional significance has yet to be determined. The fact that pathogenic mutations that cause sialuria also occur at non-conserved positions (Fig. 4b), however, suggests that comparison to evolutionarily remote bacteria should not be viewed as a sole criterion for functionality in higher organisms. UDP-N-acetylglucosamine-2-epimerase/N-acetylmannosamine kinase has been shown to be the rate-limiting enzyme in the sialic acid biosynthetic pathway14. Sialic acid modification of glycoproteins and glycolipids expressed at the cell surface is crucial for their function in many biological processes, including cell adhesion and signal transduction14. Measuring sialic acid and sialylation of proteins in normal and affected muscle could clarify some of the issues related to the pathophysiology of HIBM. GNE thus joins a growing list of single genes that are associated with several syndromes or diseases (http://www.ncbi.nlm.nih. gov/OMIM). It is a molecule of ubiquitous function, encoded by a single gene, that is the template for a protein with multiple functional attributes. Such complexity is concordant with the recent estimate of a relatively small number of genes in the human genome15,16. The identification of the gene causing HIBM and of the unique mutation occurring in the Jewish community of Middle Eastern descent will contribute to genetic screening and counseling. We suggest further study of the involvement of GNE in Nonaka’s distal myopathy with rimmed vacuoles, described in Japanese patients17, and which colocalizes with HIBM18, as well as in worldwide sporadic inclusion body myositis, the most common myopathy in individuals aged 50 and older19. In addition, elucidation of the mechanism by which UDP-N-acetylglucosamine-2epimerase/N-acetylmannosamine kinase eventually leads to hereditary inclusion-body myopathy may provide insight into related myopathologies20. 86

Methods Families. Fifty HIBM families were available for this study, 47 of Jewish Middle Eastern descent (37 Persian, 5 Afghani, 3 Egyptian and 2 Iraqi) and the remaining 3 of non-Jewish origin (1 from India, 1 from Georgia, USA and 1 from the Bahamas). All patients had a similar phenotype of slowly progressive chronic myopathy. All of the affected individuals in the three non-Jewish families had onset at the third decade, showing distal leg-muscle weakness. Quadriceps were unaffected in two of the four patients of the Bahamas family, mildly affected in the remaining two patients, unaffected in the patients from Georgia, and mildly affected, compared to other hip muscles, in the two patients of Indian origin. The shoulder girdle became involved later and facial muscles remained unaffected in all these patients. DNA was extracted from whole blood or from Epstein-Barr virus–transformed lymphocytes, as described21. These studies were approved by the institutional review board of Hadassah Hospital, Jerusalem. Genotyping. We genotyped all individuals at polymorphic markers (Fig. 2a) as previously described6. We carried out additional genotyping with three newly identified markers 84TAGA (in BAC84P7, AL161792), AAAT15 (in BAC117L14, AL354935) and ATT45 (in BAC117L21, AL354932). Sequences of oligonucleotides and PCR conditions are available upon request. Sialic acid detection. We measured sialic acid in urine samples by thin-layer chromatography followed by resorcinol staining, as described22. Mutation identification. We sequenced all 13 amplified GNE exons and cDNA fragments using Big-Dye Terminator (Perkin-Elmer) on an ABI PRISM 377 sequencer (Perkin-Elmer; see Web Table A for primer sequences). We performed sequence comparisons using DNASTAR package, STADEN package and Sequencher software (Gene Codes). Mutation analysis. We analyzed all mutations by restriction enzyme digestion (Table 1). In addition, we also analyzed mutations M712T and D225N by amplification-refractory mutation system (ARMS)23. All restriction enzymes used were from NEB. ARMS primers are available upon request. Multiple alignment. We carried out the selection of GNE orthologs using a PSI-BLAST (ref. 24) search of GNE against the nonredundant database, based on both biological function and protein similarity. We conducted multiple alignments with ClustalW (ref. 25) using the default pairwise gap nature genetics • volume 29 • september 2001



opening penalty of 10 and gap extension penalty of 0.2 for the epimerase domain. For the kinase domain we used the same parameters, except for “delay divergent sequences,” which was set to 20%. Accession numbers. GNE cDNA for exons 2–13, XM_005500; cDNA clone used for GNE exon 1 definition, AU117634; BAC clone RP11-421H8 used for alignment of genomic and cDNA sequences, AL158830. GenInfo Identifiers (GI) of the epimerase orthologs are: 6688602 (Legionella pneumophila), 11095589 (Campylobacter jejuni), 12004284 (Campylobacter jejuni), 1073208 (Neisseria meningitidis), 11095585 (Campylobacter jejuni), 11095582 (Campylobacter coli), 5823221 (Streptococcus agalactiae), 6685689 (E. coli). GI of the kinase orthologs are: 12060297 (Streptomyces griseus), 2052193 (Renibacterium salmoninarum), 8843910 (Corynebacterium glutamicum), 7229539 (Streptomyces lividans), 10173411 (Bacillus halodurans), 12725071 (Lactococcus lactis), 12055569 (Streptococcus pyogenes), 4982035 (Thermotoga maritima).

3. 4.

5. 6

7. 8.

9.

10.

Note: Supplementary information is available on the Nature Genetics web site (http://genetics.nature.com/supplementary_info/). Acknowledgments

We are grateful to all of the family members who made these studies possible, and we extend our appreciation to S. Nazarian for her extensive efforts and our special thanks to M. Banayan for his endless and warm support. We also thank M. Korner, A.Tuvy, M. Mordechaishvili and all the staff from The Laboratory of DNA Analysis for skillful assistance in sequencing, A. Sanilevich for oligonucleotide synthesis, M.E. Ahearn for technical help, M. Zeigler for sialic acid measurements, D. Darvish and H. Raz for their involvement in blood collection, and T. Levi for her continuous support. This study was supported by Hadasit (Medical Research Services Development Co., a subsidiary for R&D of Hadassah Medical Organization; S.M.-R., Z.A.); by a special donation from Hadassah Southern California– Persian Group Council, Haifa Metro Group, Malka Group, Haifa San Diego Group, Vanguard II, Healing Spirit and the ARM organization (S.M.-R., Z.A.); by a special donation in memory of N. Hollo-Bencze (Z.A.); by an Israel Ministry of Science grant to the National Laboratory for Genome Infrastructure, The Crown Human Genome Center (D.L.); by the Krupp foundation (D.L.); and by the Weizmann Institute Glasberg, Levy, N. Brunschwig and Levine funds (D.L.). T.P. is a recipient of a Kamea fellowship from the Israeli Ministry of Science and Ministry of Absorption.

11.

12.

13. 14. 15. 16. 17.

18. 19.

20. 21. 22.

23.

Received 27 April; accepted 25 July 2001. 1. 2.

Griggs, R.C. et al. Inclusion body myositis and myopathies. Ann. Neurol. 38, 705–713 (1995). Sadeh, M. & Argov, Z. Hereditary inclusion body myopathy. in Inclusion Body


24. 25.

letter

Myositis and Myopathies: Jews of Persian Origin: Clinical and Laboratory Data. (eds. Askanas, V., Engel, W.K. & Serratrice, G.) 191–199 (Cambridge Press, Cambridge, 1997). Argov, Z. & Yarom, R. “Rimmed vacuole myopathy” sparing the quadriceps. A unique disorder in Iranian Jews. J. Neurol. Sci. 64, 33–43 (1984). Argov, Z. & Mitrani-Rosenbaum, S. Hereditary inclusion body myopathy (H-IBM) with quadriceps sparing: epidemiology and genetics. in Inclusion Body Myositis and Myopathies; (eds. Askanas, V., Engel, W.K. & Serratrice, G.) 200–210 (Cambridge Press, Cambridge, 1997). Eisenberg, I. et al. Fine structure mapping of the hereditary inclusion body myopathy locus. Genomics 55, 43–48 (1999). Eisenberg, I. et al. Physical and transcriptional map of the hereditary inclusion body myopathy locus on chromosome 9p12-p13. Eur. J. Hum. Genet. 9, 501–509 (2001). Ponnambalam, S. et al. Chromosomal location and some structural features of human clathrin light-chain genes (CLTA and CLTB). Genomics 24, 440–444 (1994). Takahashi, C. et al. Regulation of matrix metalloproteinase-9 and inhibition of tumor invasion by the membrane-anchored glycoprotein RECK. Proc. Natl. Acad. Sci. USA 95, 13221–13226 (1998). Hinderlich, S., Stasche, R., Zeitler, R. & Reutter, W. A bifunctional enzyme catalyzes the first two steps in N-acetylneuraminic acid biosynthesis of rat liver. Purification and characterization of UDP-N-acetylglucosamine 2-epimerase/Nacetylmannosamine kinase. J. Biol. Chem. 272, 24313–24318 (1997). Seppala, R., Lehto, V.P. & Gahl, W.A. Mutations in the human UDP-Nacetylglucosamine 2-epimerase gene define the disease sialuria and the allosteric site of the enzyme. Am. J. Hum. Genet. 64, 1563–1569 (1999). Stasche, R. et al. A bifunctional enzyme catalyzes the first two steps in Nacetylneuraminic acid biosynthesis of rat liver. Molecular cloning and functional expression of UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase. J. Biol. Chem. 272, 24319–24324 (1997). Effertz, K., Hinderlich, S. & Reutter, W. Selective loss of either the epimerase or kinase activity of UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase due to site-directed mutagenesis based on sequence alignments. J. Biol. Chem. 274, 28771–28778 (1999). Salzberg, S.L., White, O., Peterson, J. & Eisen, J.A. Microbial genes in the human genome: lateral transfer or gene loss? Science 292, 1903–1906 (2001). Keppler, O.T. et al. UDP-GlcNAc 2-epimerase: a regulator of cell surface sialylation. Science 284, 1372–1376 (1999). International Human Genome Sequencing Consortium. Initial sequencing and analysis of the human genome. Nature 409, 860–921 (2001). Claverie, J.M. Gene number. What if there are only 30,000 human genes? Science 291, 1255–1257 (2001). Nonaka, I., Sunohara, N., Ishiura, S. & Satayoshi E. Familial distal myopathy with rimmed vacuole and lamellar (myeloid) body formation. J. Neurol. Sci. 51,141–155 (1981). Ikeuchi, T. et al. Gene locus for autosomal recessive distal myopathy with rimmed vacuoles maps to chromosome 9. Ann. Neurol. 41, 432–437 (1997). Askanas, V. & Engel, W.K. Sporadic inclusion-body myositis and hereditary inclusion-body myopathies: current concepts of diagnosis and pathogenesis. Curr. Opin. Rheumatol. 10, 530–542 (1998). Argov, Z., Eisenberg, I. & Mitrani-Rosenbaum, S. Genetics of inclusion body myopathies. Curr. Opin. Rheumatol. 10, 543–547 (1998). Mitrani-Rosenbaum, S. et al. Hereditary inclusion body myopathy maps to chromosome 9p-q1. Hum. Mol. Genet. 5, 159–163 (1996). Humbel, R. & Collart, M. Oligosaccharides in urine of patients with glycoprotein storage diseases. I. Rapid detection by thin-layer chromatography. Clin. Chim. Acta. 60, 143–145 (1975). Little, S. Amplification-refractory mutation system (ARMS) analysis of point mutations. in Current Protocols in Human Genetics (eds. Dracopoli, N.C. et al.) 9.8.1–9.8.2 (Wiley, New York, 1995). Altschul, S.F. et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402 (1997). Higgins, D.G., Thompson, J.D. & Gibson, T.J. Using CLUSTAL for multiple sequence alignments. Methods Enzymol. 266, 383–402 (1996).

87