A Genetic Marker Within the CD44 Gene Confirms Linkage at 11P13 In

Home , Genetic marker

Genes and Immunity (2002) 3 (Suppl 1), S86–S88  2002 Nature Publishing Group All rights reserved 1466-4879/02 $25.00 www.nature.com/gene BRIEF COMMUNICATION A genetic marker within the CD44 gene conﬁrms linkage at 11p13 in African-American families with lupus stratiﬁed by thrombocytopenia, but genetic association with CD44 is not present

KM Kaufman1,3, J Rankin1, ITW Harley1, JA Kelly1, JB Harley1,2,3 and RH Scoﬁeld1,2,3 1Arthritis and Immunology Program, Oklahoma Medical Research Foundation, Oklahoma City, OK, USA; 2Department of Medicine, University of Oklahoma Health Sciences Center and the Medical Service, Oklahoma City, OK, USA; 3Department of Veterans Affairs Medical Center, Oklahoma City, OK, USA

Systemic lupus erythematosus (SLE) is complicated from both a clinical and genetic standpoint. We have stratified SLE families by the presence of thrombocytopenia, which is associated with increased mortality among SLE patients, and found genetic linkage at chromosome 11p13 in African-American families. In the present study we have evaluated CD44, a gene very close (0.5 cM) to the peak LOD score marker, as a candidate gene. Using a newly identified short DNA repeat within the CD44 gene, we find a LOD score of 2.7, which confirms linkage within this genetic interval. However, using a panel of four single nucleotide markers spanning the CD44 gene, we find no genetic association with SLE. Therefore, these data further suggest an SLE susceptibility gene at 11p13, but also imply that an ancestral mutation in the CD44 gene does not account for the linkage. Genes and Immunity (2002) 3 (Suppl 1), S86–S88. doi:10.1038/sj.gene.6363887

Keywords: CD44; SLE; African-American

Systemic lupus erythematosus (SLE) is a clinically com- African-American SLE families with thrombocytopenia at plicated disease of unknown origin. As evidenced by the chromosome 11p13 with a LOD score of 5.72 (Scofield classification criteria, the manifestations of SLE vary RH, Bruner GR, Kelly JA, Kilpatrick J, Bacino D, Nath widely among individuals with the disease.1,2 Epidemio- SK, Harley JB. Thrombocytopenia identifies a severe logical studies of families and twins have indicated a familial phenotype of systemic lupus erythematosus substantial genetic predisposition to SLE.3,4 Recently, and reveals genetic linkages at 1q21 and 11p13, manu- genome-wide scans have found over a dozen chromo- script submitted). somal intervals linked to the SLE phenotype in families In the present study, we have investigated CD44 as a in which more than one individual has SLE.5–9 candidate gene. The CD44 gene lies only 0.5 cM from the We have hypothesized that the use of the clinical peak LOD score found at marker D11s1392. CD44 is a variation found among patients with SLE can be used in cell surface glycoprotein that is a major receptor for hyal- studies of the genetics of the disease, and this approach uronan. The molecule is involved in cell adhesion and has proved useful so far.10 Thrombocytopenia is a mani- is expressed on immune cells such as lymphocytes and festation of SLE that is of particular clinical interest leukocytes as well as a wide variety of other cells.12 CD44 because of its strong association with mortality.11 Low is implicated in the pathogenesis of several inflammatory platelet count is part of the diagnostic criteria for SLE1 illnesses and its interaction with hyaluronan is con- and occurs in 10–20% of SLE patients. We have stratified sidered a potential therapeutic target in inflammatory families multiplex for SLE on the basis of the presence of diseases.12 The CD44 protein is polymorphic on the basis thrombocytopenia in at least one SLE patient in an effort of isoforms resulting from alternative splicing. Some iso- to achieve a more homogeneous set of families for genetic forms are expressed in a tissue or cell specific manner.13 study. This study demonstrates that these families have Thus, we hypothesized that CD44 was an important can- a severe form of SLE, even among SLE patients without didate within the linkage interval defined on chromo- thrombocytopenia in these families, regardless of race. some 11p13 among African-American SLE families with Furthermore, we identified a genetic linkage effect in thrombocytopenia. Families were recruited, screened and interviewed as previously described.5 All patients met the American Col- Correspondence: H Scofield, 825 NE 13th Street, Oklahoma City, OK lege of Rheumatology (ACR) criteria for the classification 1,2 73104, USA. E-mail: hal-scofieldȰomrf.ouhsc.edu of systemic lupus. Typing of short repeat microsatellite Received 14 December 2001; accepted 20 March 2002 markers was performed as previously described.5 Throm- CD44 gene and African-American families with lupus KM Kaufman et al S87 bocytopenia was determined by review of all available Table 2 A short DNA repeat marker (number 6 in Table 1) within medical records, patients interview and questionnaire as CD44 was analyzed for linkage to the SLE phenotype in 13 African- well as interview of treating physicians. A platelet count American families where at least one member had SLE manifested with thrombocytopenia of less than 100 000/mm3 was considered significant. This is the definition used by the ACR criteria.1 Marker LOD score Model Basepairs Frequency Six short tandem repeating DNA segments within the position CD44 gene were identified (Table 1). Three of these markers were found to be polymorphic. There were 13 CD44-rpt6 2.721 Dom mix African-American families useful for linkage in which at Allele least one SLE patient had thrombocytopenia. These famil- 1 286 0.08 ies were previously studied to establish linkage at 11p13 2 282 0.45 (Scofield et al, manuscript submitted). A total of 78 indi- 3 278 0.20 viduals were included in these pedigrees, all of whom 4 274 0.18 5 270 0.06 were studied with the three polymorphic short tandem 6 266 0.02 repeat markers. We found a maximum two-point LOD 7 294 0.006 score of 2.721 for one of the markers in these 13 families with a dominant inheritance pattern (Table 2). In this The 78 participants DNA samples were genotyped by polymerase model, penetrance values varied between men and chain reaction and separated the PCR products by size as described women (49% and 92%, respectively). This LOD score, elsewhere5,10 at the three polymorphic tandem repeat markers. One then, confirms linkage for this chromosomal region to of the tandem repeat markers located near the middle of the gene SLE in African-American families where SLE is mani- showed linkage to lupus with a LOD score of 2.721 and is shown above. For the polymorphic markers (1, 5 and 6 in Table 1), primers fested by thrombocytopenia since the result exceeds the to unique sequences were designed that flanked tandem repeats. level required for conformation (P Ͻ 0.01). The other two PCR was performed as previously described,5 using the following tandem repeats within the CD44 gene did not have sig- primers: repeat 1: Forward 5′ CTCTTTCTACTAGTGGC nificant LOD scores, but the informativity of these mark- TTTGTA 3′/reverse 5′ GAGCATGAGAGGCTTCCTTT 3′, repeat 5: Forward 5′ GGGCTGCAAGGGATTTTCAA 3′/reverse 5′ GGTT ers was low. ′ ′ In order to assess the possibility that CD44 was in fact ACTAGAGACCTTGAAG 3 ,repeat6:Forward5GATAGAGACAA TTAAGAGATAGAG 3′/reverse 5′ TGGATGCATCCTGTGAGTTC 3′. the gene responsible for the observed linkage, we studied Each of the forward primers also contained an M13–29 primer tail the linkage and association of single nucleotide poly- (5′ CACGACGTTGTAAAACGAC 3′). Maximum-likelihood model- morphisms (SNPs) lying within the CD44 gene to SLE in based linkage analysis was performed as previously described this population. We found four informative SNPs within using a screening set of six models. The maximum LOD score was the CD44 gene (Table 2). There was no linkage to the SLE obtained under an autosomal dominant model with a mixed pene- phenotype for any of these SNPs in the 13 African- trance (dom mix) in which the penetrance in men is 49% and that of women is 92%.5,14 American families with thrombocytopenia when studied by maximum-likelihood model-based linkage analysis, or by affected sibling pair analysis. In addition, when per- found to be in linkage disequilibrium (data not forming the transmission disequilibrium test, we found presented). no transmission distortion of any allele within the four We have stratified SLE families by the presence of SNPs (Table 3). CD44 haplotypes were generated with thrombocytopenia in at least one SLE patient. This mani- these four SNPs from GENEHUNTER-PLUS. Expected festation of SLE is strongly associated with an increased and observed counts of transmitted and non-transmitted mortality.10,12 In our families with SLE, the presence of alleles were evaluated using: (1) trios and, (2) one discor- thrombocytopenia in one member of a family is associa- dant sibpair per pedigree. Statistical significance was ted with severe disease in other members of the family, measured by the McNemar ␹2 test. Only informative despite the fact that these other SLE affected members do haplotypes were used for the analyses. No haplotype was not tend to have thrombocytopenia (Scofield et al, submitted) Thus, thrombocytopenia identifies not just individual patients, but entire families with severe SLE. Table 1 The position, sequence and number of alleles for the six Therefore, our effort in which SLE families are stratified microsatellite markers found within the CD44 gene by thrombocytopenia is likely to identify genes associated with severe disease. Identification of such genes will a Repeat Sequence Position Genbank Alleles undoubtably prove useful to understand the disease itself no. accession no. and in developing new therapeutic strategies. There are, of course, potential problems associated 1 CA 34538 to 34569 AL356215 16 with our approach in which a small portion of the famil- 2 TATT 65629 to 65660 AL356215 1 3 AC 140171 to 140214 AL133330 1 ies are examined. This may lead to false positive results. 4 AT 170075 to 170220 AL133330 1 However, we have now confirmed our initial results with 5 TG 182422 to 182461 AL133330 5 the microsatellite marker within the CD44 gene. In 6 GATA 5074 to 5145 AL356215 7 addition, we have identified a number of other microsat- ellites in this region but not within CD44 that also con- aPosition of refers to the position in the Genbank entry. CD44 was firm linkage (not shown). Finally, we have performed gene downloaded from Genbank in Fasta format. The sequence was simulation studies by randomly selecting 13 African- processed using the software program Tandem Repeat Finder Ver American pedigrees from the 63 total families of sub- 2.02.15 Six tandem repeats that had a period size of 2, 3 or 4, a copy number greater than 10 and a percent match greater than 90% were Saharan origin. These simulations indicate that the results found either within or near the CD44 gene with three of these obtained are highly statistically significant (unpublished regions polymorphic (see above). data, manuscript in preparation).

Genes and Immunity CD44 gene and African-American families with lupus KM Kaufman et al S88 Table 3 Association and linkage were tested using four single ing genes within the linkage interval, such as CD59. In nucleotide polymorphisms found within the CD44 gene. Sibpair the near future we hope to be in the position to confirm analysis and the TDT analysis also showed non-significant results or refute linkage in a new cohort of African-American for each of the five SNPs tested families with SLE manifested by thrombocytopenia. We have a new cohort collection assembled that is equal in Positiona LOD score Polymorphism Allele frequency size to the population previously studied. When typing at a set of genome-wide markers and clinical assessment 5132 0.19 T/C 0.753/0.247 5226 −0.003 C/T 0.927/0.073 is complete, we can begin analyses. If confirmed, we will 5252 0.002 C/A 0.167/0.833 then be confident that a gene, likely associated with sev- 39549 0.661 A/G 0.0473/0.9527 ere lupus, but not CD44, is present within the linkage interval at chromosome 11p13. aPosition refers to Genbank accession number AL133330. The CD44 gene is ෂ800 000 bp downstream of the D11S1392 marker. The CD44 gene is contained in the Genbank contig NTF024299. Examination References of the approximately 30 kbp that encode the CD44 gene revealed the presence of 192 single nucleotide polymorphisms (SNP). Primers 1 Tan EM, Cohen AS, Fries JF et al. The 1982 revised criteria for were designed to amplify four regions at either the 5′, middle or 3′ the classification of systemic lupus erythematosus. Arthritis end of the gene. Each region contained at least two SNPs. These four Rheum 1982; 25: 1271–1277. primer pairs amplified a total of 10 different putative SNPs. Seventy- 2 Hochberg MC. Updating the American College of Rheuma- eight DNA samples from thrombocytopenic families multiplex for tology revised criteria for the classification of systemic lupus lupus and showing linkage at D11S1392 were amplified and erythematosus. Arthritis Rheum 1997; 40: 1725. sequenced. Four of the 10 SNPs were found to actually be polymor- 3 Hochberg MC. The application of genetic epidemiology to sys- phic and informative, and are shown above. The oligonucleotide temic lupus erythematosus. J Rheumatol 1987; 14: 867–869. primers were as follows—for dbSNP 932633 through 932635 4 Deapen D, Escalante A, Weinrib L et al. A revised estimate of ′ ′ forward primer 5 GCCTGTGAGGAGCATATAGC 3 /reverse twin concordance in systemic lupus erythematosus. Arthritis primer 5′ GTCATCTCCTTCATCCTGT 3′. For dbSNP 1467558 for- ′ ′ Rheum 1992; 35: 311–318. ward primer 5 GTGGCTGAGCATCGTTATTC 3 /reverse primer 5 Moser KL, Neas BR, Salmon JE et al. Genome scan of human 5′ TTTCCTCTTGGCCAGATGTG 3′. Each forward primer con- systemic lupus erythematosus: Evidence for linkage on chromo- tained an M13 −20 tail (5′ TTGTAAAACGACGGCCAGT 3′) and some 1q in African-American pedigrees. Proc Natl Acad Sci USA each reverse primer contained an M13 reverse primer tail (5′ GGATAACAATTTCACACAGG 3′). PCR reactions were perfor- 1998; 95: 14869–14874. med under standard conditions as described.5 PCR products were 6 Gaffney PM, Kearns GM, Shark KB et al. A genome-wide search directly sequenced in order to determine the allele in each of the for susceptibility genes in human systemic lupus erythematosus 78 individuals in the 13 African-American families with SLE mani- sib-pair families. Proc Natl Acad Sci USA 1998; 95: 14875–14879. fested by thrombocytopenia. 7 Lindqvist AKB, Steinsson K, Johanneson B et al. A susceptibility locus for human systemic lupus erythematosus (hSLE1) on chromosome 2q. J Autoimmun 2000; 14: 169–178. CD44 is within 1 cM of the peak LOD score previously 8 Gaffney PM, Ortmann WA, Selby SA et al. Genome screening found (Scofield et al, submitted) Further, this gene is of in human systemic lupus erythematosus: results from a second interest because of its multiple functions, including its Minnesota cohort and combined analyses of 187 sib-pair famil- apparent role in inflammation. The CD44 protein is poly- ies. Am J Hum Genet 2000; 66: 547–556. morphic based on extensive alternative splicing that pro- 9 Shai R, Quismorio FP Jr, Li L et al. Genome-wide screen for duces tissue specific alternative proteins. The CD44 gene systemic lupus erythematosus susceptibility genes in multiplex is large. Our data with the SNPs within CD44 strongly families. Hum Mol Genet 1999; 8: 639–644. suggest that an ancient, ancestral CD44 allele is not the 10 Nath SK, Kelly JA, Namjou B et al. Evidence for a susceptibility gene, SLEV1, on chromosome 17p13 in families with vitiligo- actual gene producing the observed linkage. However, related systemic lupus erythematosus. Am J Hum Genet 2001; 69: association operates over a short genetic interval, so these 1401–1406. data do not eliminate the possibility that CD44 is 11 Reveille JD, Bartolucci A, Alarcoń GS. Prognosis in systemic important to SLE linkage in this chromosomal region. lupus erythematosus. Arthritis Rheum 1990; 33:37–48. The small number of families studied and the informativ- 12 Pure E, Cuff CA. A crucial role for CD44 in inflammation. Trends ity of the SNP markers limit the power of the TDT test Mol Med 2001; 7: 213–221. to detect association. A series of SNPs each in linkage 13 Bajorath J. Molecular organization, structural features, and disequilibrium with its closest neighbors that do not ligand binding characteristics of CD44, a highly variable cell show genetic association and can be studied as extended surface glycoprotein with multiple functions. Proteins 2000; 39: 103–111. haplotypes will be needed to eliminate CD44 as a candi- 14 Bias WB, Reveille JD, Beaty TH, Meyers DA, Arnett FC. Evi- date gene. dence that autoimmunity in man is a Mendelian dominant trait. Nonetheless, these data suggest that CD44 is not the Am J Hum Genet 1986; 39: 584–602. gene responsible for the linkage observed. This result 15 Benson G. Tandem repeats finder: a program to analyze DNA means that our attention will be drawn to other interest- sequences. Nucleic Acids Res 1999; 27: 573–580.

Genes and Immunity