Interethnic studies of TNF polymorphisms confirm the likely ... - Nature

Genes and Immunity (2000) 1, 418–422  2000 Macmillan Publishers Ltd All rights reserved 1466-4879/00 $15.00 www.nature.com/gene

Interethnic studies of TNF polymorphisms confirm the likely presence of a second MHC susceptibility locus in ankylosing spondylitis A Milicic1, F Lindheimer2, S Laval3, M Rudwaleit4, H Ackerman3, P Wordsworth1, T Hohler2 and MA Brown1 1

Nuffield Orthopaedic Centre, Windmill Rd, Headington, OX3 7LD, UK; 2Johannes Gutenberg Universita¨t Mainz, Mainz, Germany; 3Wellcome Trust Centre for Human Genetics, Oxon, OX3 7BN, UK; 4University Hospital Benjamin Franklin, Berlin, Germany

The objective of this study was to investigate TNF promoter region polymorphisms for association with susceptibility to ankylosing spondylitis (AS). The TNF −238 and −308 polymorphisms were genotyped in 306 English AS cases and 204 ethnically matched healthy B27-positive controls, and 96 southern German AS cases, 58 B27-positive and 251 B27negative ethnically matched controls. Additionally, the TNF −376 polymorphism was genotyped in the southern German cases and controls. In the southern German AS patients a significant reduction in TNF −308.2 alleles was seen, compared with B27 positive controls (odds ratio 0.4, P = 0.03, 95% confidence interval 0.2–0.9), but no difference in allele frequencies was observed at TNF −238. Significant association between AS and both TNF −238 and TNF −308 was excluded in the English cases. These results confirm previous observations in the southern German population of association between TNF promoter region polymorphisms and AS, but the lack of association in the English population suggests that these polymorphisms themselves are unlikely to be directly involved. More likely, a second, non-HLA-B, MHC locus is involved in susceptibility to AS in these two populations. Genes and Immunity (2000) 1, 418–422. Keywords: TNF; ankylosing spondylitis; association

Introduction Ankylosing spondylitis (AS) is a common inflammatory rheumatic disease with a strong familial predisposition. Susceptibility to AS is largely determined by genetic factors, with heritability in twins estimated to be ⬎90%.1 Although ⬎90% of subjects with AS carry HLA-B27 (B27),2,3 only a small proportion (1–6%) of B27-positive individuals develop the condition.4,5 Evidence from twin and family studies suggest that genes not linked to the major histocompatibility complex (MHC) are the primary determinants of whether B27-positive individuals develop AS.1,6 Nonetheless, there is suggestive evidence that MHC genes other than B27 and HLA-B60 may influence susceptibility to AS, or clinical manifestations of the disease such as the occurrence of iritis, peripheral arthritis, disease severity and age of onset (reviewed in Brown and Wordsworth).7 In particular, association between an B27/DR1 haplotype and AS has been reported in case-control,8 twin1 and family studies,9 and between the DR1 subtype HLA-DRB1*0103 and B27positive spondyloarthritis complicating inflammatory bowel disease.10 Correspondence: Dr Matthew Brown, Spondyloarthritis Research Group, Wellcome Trust Centre for Human Genetics, Roosevelt Drive, Headington, OX3 7BN, UK. E-mail: mbrown얀well.ox.ac.uk MAB was funded by the Arthritis Research Campaign (UK). Received 20 May 2000; revised and accepted 21 June 2000

Tumour necrosis factor (TNF) is a potent immunomodulator with an essential role in the inflammatory response, and raised concentrations of TNF mRNA have been detected in the synovium of sacroiliac joints of AS patients.11 A role for the TNF gene in susceptibility to AS has been suggested by recent reports of association of allelic variants of the gene, with AS.12,13 Although several polymorphisms of the TNF gene exist, two variants at position −238 and −308 have been most widely studied. These two variants are both located in the putative transcription regulatory region, and involve G to A transitions. Several studies have suggested that the −308 polymorphism influences secretion of TNF, the −308.2 allele (A at −308) being associated with greater TNF secretion.14–18 Whether this polymorphism influences gene transcription remains uncertain, with some studies suggesting increased transcription associated with the −308.2 allele,17,19 and others demonstrating no difference.20,21 No effect of the −238 polymorphism on TNF secretion,22 or transcriptional activation,23 has been demonstrated. The TNF-376 (G/A) polymorphism is associated with increased susceptibility to cerebral malaria and in some studies is associated with altered TNF secretion by monocytes,24 although this also has not been consistently reproduced.23 Several other promoter region polymorphisms (including at positions −163, −244, −857, −863 and −1031), have been reported but their functional significances are unknown. Whether TNF promoter region polymorphisms influ-

Interethnic studies of TNF polymorphisms A Milicic et al

ence susceptibility to AS has been uncertain, as not all published studies have demonstrated any association. However, the case-control studies reporting no association have had adequate power to detect only large genetic effects,23,25,26 making interpretation of their findings difficult. To clarify this important question we have studied the role of the TNF −376, −308 and −238 polymorphisms in AS in southern German AS cases and a large cohort of English Caucasian AS cases, and healthy, ethnically-matched, controls.

and a non-significant trend was observed in southern German healthy controls (D/Dmax = 0.19, P = 0.4). A similar trend has previously been reported in this population,13 although even pooling the data with the current study this remains statistically non-significant (D/Dmax = 0.16, P = 0.2). Between B27 and TNF −238, strong linkage disequilibrium was observed in the southern German population (D/Dmax = 0.81, P = 0.03), but none was present in the English population.

419

Discussion

Results The genotype and allele frequencies of TNF −238, −308 and −376 in English and southern German patients and controls are given in Table 1. Amongst English cases and controls, no significant genotypic or allelic differences were noted between cases and B27-positive controls at either TNF −238 or −308. This study had ⭓90% power to detect a 2% difference in allele frequencies at TNF −238 and a 3% difference in allele frequencies at TNF −308, at a P-value of 0.05. In the southern German AS patients no difference in allele frequencies were observed at TNF −238, but a significant reduction in TNF −308.2 alleles was seen in patients compared with B27 positive controls (odds ratio 0.4, P = 0.03, 95% confidence interval 0.2–0.9). There were no signficant differences in either TNF −238 or −308 genotype frequencies. This study had ⭓80% power to detect a 4% difference in allele frequencies at TNF −238 at a Pvalue of 0.05. Positive linkage disequilibrium was observed between B27 and TNF −308.1 in English (D/Dmax = 0.19, P = 0.02),

There is increasingly strong evidence of the presence of further genes within the MHC influencing disease susceptibility in AS, but identifying the genes involved is complicated by the strong linkage disequilibrium present across the MHC, and its varying patterns in different populations. Several studies have now reported association between HLA-DR alleles and AS, in particular with HLA-DR1 in primary AS,1,9,29 and the HLA-DRB1*0103 allele in AS complicating inflammatory bowel disease (IBD).10,30 The strength of these findings and the fact that they have been replicated in different patient cohorts indicate that it is highly likely that one or more further genes lying on a B27-DRB1*0103 haplotype increase susceptibility to AS, particularly if complicated by IBD. Studies of the TAP and LMP loci have produced contradictory results in ethnically similar populations and therefore no firm conclusion can be made about the role of these genes in AS at this point. TNF −238.2 and TNF −308.2 alleles have been reported to be under-represented in AS patients in the southern German13 and western Scottish populations,12 but studies

Table 1 TNF genotype and allele frequencies in patients and controls. Previously reported figures were used for general population frequencies for English controls.27 Power calculations were performed by standard methods.28 All P-values quoted are two-tailed. No correction is required for the number of loci examined, as this is a confirmation study of a previously reported finding.12,13 1,1 n (%) TNF −238 southern German AS patients B27 positive controls B27 negative controls English AS patients B27 positive controls General controls TNF −308 southern German AS patients B27 positive controls B27 negative controls English AS patients B27 positive controls General controls TNF −376 southern German AS patients B27 positive controls B27 negative controls

1,2 n (%)

2,2 n (%)

Allele 1 (G) n (%)

Allele 2 (A) n (%)

87 (96) 52 (98) 130 (88)

4 (4) 1 (2) 15 (10)

0 0 2 (1)

178 (98) 105 (99) 275 (94)

4 (2) 1 (1) 19 (6)

269 (88) 189 (92)

36 (12) 15 (7)

1 (0) 1 (0)

574 (94) 393 (96) 181 (95)

38 (6) 17 (4) 9 (5)

87 (91) 47 (81) 116 (76)

9 (9) 9 (16) 27 (18)

0 (0) 2 (3) 9 (6)

183 (95 103 (89) 259 (85)

9 (5)* 13 (11)* 45 (15)

235 (77) 167 (82)

70 (23) 36 (18)

1 (0) 1 (0)

540 (88) 370 (91) 161 (84)

72 (12) 38 (9) 31 (16)

87 (100) 53 (100) 128 (98)

0 0 2 (2)

174 (100) 106 (100) 258 (99)

0 0 2 (1)

0 0 0

*P ⬍ 0.05. Genes and Immunity


420

from Holland23 and Spain25 have not reported any association. The current study confirms the previously reported under-representation of TNF −308.2 alleles in southern German AS patients. Combining these results with those previously reported for southern German cases and controls, the strength of B27-independent association is considerable (P = 0.001, odds ratio 0.4, 95% confidence interval 0.2–0.7). No significant difference in TNF −238.2 allele frequencies was noted between southern German AS patients and controls, although the power to detect a difference in this study was only moderate. Combining the results of the current study with that previously reported from southern Germany,13 the results remain non-significant (P = 0.1), suggesting that any association at this locus in southern Germans must be weak. In the English Caucasian population no significant differences in either TNF −238 or TNF −308 allele or genotype frequencies were noted, although the study had high power to detect even small effects. This indicates that it is unlikely that either TNF −238 or −308 influence susceptibility to AS themselves. One can also conclude from this finding that the polymorphism TNF −1031, which is in near complete linkage disequilibrium with TNF −238, is unlikely to be involved.31 Furthermore, contradictory findings have been reported on whether there is an HLA-DR-independent effect of the TNF polymorphisms on susceptibility to systemic lupus erythematosus (SLE).27,32 A more likely explanation is that other genes in linkage disequilibrium with these loci may play a role. Linkage disequilibrium across the MHC varies in different populations, and there are significant differences in the observed linkage disequilibrium found in English and southern Germans. In this study significant linkage disequilibrium is observed between TNF −238 and B27 in southern Germans (D/Dmax = 0.8), but no linkage disequilibrium was observed in English. There are also major differences in MHC haplotype frequencies between the two populations. In English healthy B27-positive individuals, the DR1 allele frequency is 0.2,29 whereas in southern Germans it is 0.09 (data not presented). Clearly B27-bearing MHC haplotypes have different distributions in these two related but different ethnic groups. Although further studies will be required to prove it, these findings suggests that any further MHC gene influencing susceptibility to AS lies centromeric to TNF towards the HLA-DRB1 locus. Some other studies also support the existence of an independent RA susceptibility locus in the MHC class III region.33,34 The gene/s involved must lie on different haplotypes in the southern German and English populations. In a southern German population the allele would most likely lie on a TNF −308.1 bearing haplotype, whereas in the English population it would lie on a DR1 haplotype, with linkage disequilibrium breaking down around TNF. Further studies of haplotypes across the MHC addressing particularly the area centromeric of TNF will be required to determine the precise location of the probably further susceptibility gene/genes. No significant difference was noted in TNF −376 polymorphism frequencies between cases and controls in the southern German population. The −376A allele frequency was so low that the study had little power to detect even moderate sized association of this allele with AS, but the low allele frequency excludes it as a significant contributor to the linkage disequilibrium observed at TNF −238.

Genes and Immunity

In summary, this study confirms the reported association between TNF −308.1 and AS in the southern German population. At both this locus and the TNF −238 locus, previously also reported to be associated with AS in southern Germany, no association was found in the English population. This suggests that these two loci are unlikely to contribute themselves to susceptibility to AS, but strongly supports the presence of genes other than B27 influencing susceptibility to AS.

Materials and methods Cases and controls DNA was obtained by standard methods from peripheral venous blood samples from 306 English AS cases (Nuffield Orthopaedic Centre, Oxford, UK) and 204 ethnically-matched healthy B27-positive controls (Oxford Regional Transfusion Centre), and 96 southern German AS cases (Johannes Gutenberg Universita¨t Mainz, Mainz, Germany), 58 B27-positive and 251 B27-negative ethnically matched controls (Mainz regional blood transfusion service). All cases and controls were unrelated, and no results of TNF promoter polymorphisms in either the cases or controls have previously been published. AS was defined according to the modified New York criteria.35 Ethical approval for the study was obtained from the Central Oxford Research Ethics Committee. Methods All cases and controls were typed for HLA-B27 by PCR using sequence-specific primers.3 Genotyping of the TNF −308 and −238 polymorphisms was performed by PCRARMS (PCR-amplification refractory mutation system) method which enables simultaneous typing for both variant alleles (G and A) in a single amplification reaction, using four PCR primers (see Figure 1). All subjects were also typed for the TNF −376 polymorphism by PCR/ARMS using the following allele specific primers: −376G 5⬘-CCT GCA TCC TGT CTG GAA G and −376A 5’- TCC TGC ATC CTG TCT GGA AA, and the constant primer −376R 5⬘-GGC TGG GTG TGC CAA CAA C which produce fragments of 396 or 397 bp respectively. In each reaction, a 796 bp fragment was amplified from a conserved region of DRB1 using the following primers: 5⬘- TGC CAA GTG GAG CAC CCA A and 5⬘-GCA TCT TGC TCT GTG CAG AT. Thermocycling conditions were 96°C for 1 min; followed by 5 cycles of 96°C for 35 s, 68°C for 45 s, and 72°C for 35 s; followed by 21 cycles of 96°C for 25 s, 64°C for 50 s, and 72°C for 40 s; followed by six cycles of 96°C for 35 s, 56°C for 60 s, and 72°C for 90 s. Statistical analysis Genotype relative risk was calculated by standard methods.36 For comparison of allelic frequencies, Pvalues were calculated by the ␹2 statistic, with odds ratios calculated using the cross-products ratio. Linkage disequilibrium was assessed using the standardised D⬘ statistic,37 which measures linkage disequilibrium on a scale from 1 to −1.

Acknowledgements AM and FL contributed equally to this project.


10

11

12

13

14

15 Figure 1 Genotyping of TNF −238 and −308 polymorphisms. Lanes 1, 5 and 9—DNA Marker VIII. Lane 2 238.1/238.1, lane 3 238.1/238.2, lane 4 238.2/238.2, lane 6 308.1/308.1, lane 7 308.1/308.2, lane 8 308.2/308.2. The method employed is similar to that previously reported,27 with minor modifications of the primers selective for the TNF −308.1 and 308.2 alleles to improve amplification characteristics. The primers for the TNF −308.1 allele were: C5: 5⬘-GCC CCT CCC AGT TCT AGT TCT ATC-3⬘ and 308Gw: 5⬘ACC CTG GAG GCT GAA CCC CGA CCC-3⬘ (139 bp product), while the rare TNF −308.2 allele (A at −308) was amplified using: C3: 5⬘-AAG CGG TAG TGG GCC CTG CAC CTT-3⬘ and 308Aw: 5⬘-GGC AAT AGG TTT TGA GGG GCT TGA-3⬘ (216 bp product). Primers C5 and C3 also result in a constant 354 bp band which spans the mutation site and serves as a control band for the amplification. Thermocycling conditions were 32 cycles of 94°C, 63°C and 72°C (45 s each).

16

17

18

19

20

References 1 Brown MA, Kennedy LG, MacGregor AJ et al. Susceptibility to ankylosing spondylitis in twins: the role of genes, HLA, and the environment. Arthritis Rheum 1997; 40: 1823–1828. 2 Brewerton DA, Hart FD, Nicholls A, Caffrey M, James DC, Sturrock RD. Ankylosing spondylitis and HL-A 27. Lancet 1973; 1: 904–907. 3 Brown MA, Pile KD, Kennedy LG et al. HLA class I associations of ankylosing spondylitis in the white population in the United Kingdom. Ann Rheum Dis 1996; 55: 268–270. 4 van der Linden SM, Valkenburg HA, de Jongh BM, Cats A. The risk of developing ankylosing spondylitis in HLA-B27 positive individuals. A comparison of relatives of spondylitis patients with the general population. Arthritis Rheum 1984; 27: 241–249. 5 Braun J, Bollow M, Remlinger G et al. Prevalence of spondylarthropathies in HLA-B27 positive and negative blood donors. Arthritis Rheum 1998; 41: 58–67. 6 Brown MA, Pile KD, Kennedy LG et al. A genome-wide screen for susceptibility loci in ankylosing spondylitis. Arthritis Rheum 1998; 41: 588–595. 7 Brown MA, Wordsworth BP. Predisposing factors to spondyloarthropathies. Curr Opin Rheumatol 1997; 9: 308–314. 8 Brown MA, Pile KD, Klinke J et al. HLA Class II associations of ankylosing spondylitis. Br J Rheumatol 1995; 34 (Abstracts Supplement 1): 146. 9 Nahal R, Gautreau C, Amor B et al. Genetic investigation in fam-

21

22

23

24

25

26

27

ilial spondyloarthropathy (SpA) with focus on HLA region using transmission desequilibrium test (TDT). Arthritis Rheum 1996; 39: S122. Laval SH, Bradbury L, Darke C, Brophy S, Calin A, Brown M. The role of HLA-DR genes in ankylosing spondylitis complicating inflammatory bowel disease. Rheumatology 2000; 39 (Suppl. 1): 64. Braun J, Bollow M, Neure L et al. Use of immunohistologic and in situ hybridization techniques in the examination of sacroiliac joint biopsy specimens from patients with ankylosing spondylitis. Arthritis Rheum 1995; 38: 499–505. McGarry F, Walker R, Sturrock R, Field M. The −308.1 polymorphism in the promoter region of the tumor necrosis factor gene is associated with ankylosing spondylitis independent of HLAB27. J Rheumatol 1999; 26: 1110–1116. Hohler T, Schaper T, Schneider PM, Meyer zum Buschenfelde KH, Marker-Hermann E. Association of different tumor necrosis factor alpha promoter allele frequencies with ankylosing spondylitis in HLA-B27 positive individuals. Arthritis Rheum 1998; 41: 1489–1492. Whichelow CE, Hitman GA, Raafat I, Bottazzo GF, Sachs JA. The effect of TNF*B gene polymorphism on TNF-alpha and -beta secretion levels in patients with insulin-dependent diabetes mellitus and healthy controls. Eur J Immunogenet 1996; 23: 425–435. Bouma G, Crusius JB, Oudkerk Pool M et al. Secretion of tumour necrosis factor alpha and lymphotoxin alpha in relation to polymorphisms in the TNF genes and HLA-DR alleles. Relevance for inflammatory bowel disease. Scand J Immunol 1996; 43: 456–463. Galbraith GM, Steed RB, Sanders JJ, Pandey JP. Tumor necrosis factor alpha production by oral leukocytes: influence of tumor necrosis factor genotype. J Periodontol 1998; 69: 428–433. Kroeger KM, Carville KS, Abraham LJ. The -308 tumor necrosis factor-alpha promoter polymorphism effects transcription. Mol Immunol 1997; 34: 391–399. Louis E, Franchimont D, Piron A et al. Tumour necrosis factor (TNF) gene polymorphism influences TNF-alpha production in lipopolysaccharide (LPS)-stimulated whole blood cell culture in healthy humans. Clin Exp Immunol 1998; 113: 401–406. Wilson AG, Symons JA, McDowell TL, McDevitt HO, Duff GW. Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation. Proc Natl Acad Sci USA 1997; 94: 3195–3199. Stuber F, Udalova IA, Book M et al. −308 tumor necrosis factor (TNF) polymorphism is not associated with survival in severe sepsis and is unrelated to lipopolysaccharide inducibility of the human TNF promoter. J Inflamm 1995; 46: 42–50. Brinkman BM, Zuijdeest D, Kaijzel EL, Breedveld FC, Verweij CL. Relevance of the tumor necrosis factor alpha (TNF alpha) −308 promoter polymorphism in TNF alpha gene regulation. J Inflamm 1995; 46: 32–41. Pociot F, D’Alfonso S, Compasso S, Scorza R, Richiardi PM. Functional analysis of a new polymorphism in the human TNF alpha gene promoter. Scand J Immunol 1995; 42: 501–504. Kaijzel EL, Brinkman BM, van Krugten MV et al. Polymorphism within the tumor necrosis factor alpha (TNF) promoter region in patients with ankylosing spondylitis. Hum Immunol 1999; 60: 140–144. Knight JC, Udalova I, Hill AV et al. A polymorphism that affects OCT-1 binding to the TNF promoter region is associated with severe malaria. Nat Genet 1999; 22: 145–150. Fraile A, Nieto A, Beraun Y, Vinasco J, Mataran L, Martin J. Tumor necrosis factor gene polymorphisms in ankylosing spondylitis. Tissue Antigens 1998; 51: 386–390. Verjans GM, Brinkman BM, Van-Doornik CE, Kijlstra A, Verweij CL. Polymorphism of tumour necrosis factor-alpha (TNF-alpha) at position −308 in relation to ankylosing spondylitis. Clin Exp Immunol 1994; 97: 45–47. Rudwaleit M, Tikly M, Khamashta M et al. Interethnic differences in the association of tumor necrosis factor promoter poly-

421

Genes and Immunity


422 28

29

30

31

32

morphisms with systemic lupus erythematosus. J Rheumatol 1996; 23: 1725–1728. Cohen J. 1988 Statistical Power Analysis for the Behavioural Sciences, 2nd edn. Lawrence Erlbaum Associates, Hillsdale, New Jersey, 1988. Brown MA, Kennedy LG, Darke C et al. The effect of HLA-DR genes on susceptibility to and severity of ankylosing spondylitis. Arthritis Rheum 1998; 41: 460–465. Roussomoustakaki M, Satsangi J, Welsh K et al. Genetic markers may predict disease behaviour in patients with ulcerative colitis. Gastroenterology 1997; 112: 1845–1853. Skoog T, van’t Hooft FM, Kallin B et al. A common functional polymorphism (C→A substitution at position −863) in the promoter region of the tumour necrosis factor-alpha (TNF-alpha) gene associated with reduced circulating levels of TNF-alpha. Hum Mol Genet 1999; 8: 1443–1449. Rood MJ, van Krugten MV, Zanelli E et al. TNF-308A and HLA-

Genes and Immunity

33

34

35

36 37

DR3 alleles contribute independently to susceptibility to systemic lupus erythematosus. Arthritis Rheum 2000; 43: 129–134. Bali D, Gourley S, Kostyu D et al. Genetic analysis of multiplex rheumatoid arthritis families. Genes Immun 1999; 1: 28–36. Mulcahy B, Waldron Lynch F, McDermott MF et al. Genetic variability in the tumor necrosis factor-lymphotoxin region influences susceptibility to rheumatoid arthritis. Am J Hum Genet 1996; 59: 676–683. van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984; 27: 361–368. Lathrop GM. Estimating genotype relative risks. Tissue Antigens 1983; 22: 160–166. Lewontin R. The interaction of selection and linkage. I. General considerations; heterotic models. Genetics 1964; 49: 49–67.