Genes and Immunity (2001) 2, 222–228  2001 Nature Publishing Group All rights reserved 1466-4879/01 $15.00 www.nature.com/gene IL-1B and IL-1Ra gene polymorphisms and disease severity in rheumatoid arthritis: interaction with their plasma levels

N Buchs1, FS di Giovine2, T Silvestri2, E Vannier3, GW Duff2 and P Miossec1 1Departments of Immunology and Rheumatology, Hoˆpital Edouard Herriot, Lyon, France; 2Division of Genomic Medicine, University of Sheffield, UK; 3Department of Medicine, Tufts University and New England Medical Center, Boston, MA, USA

The balance between -1 (IL-1) and its competitive antagonist IL-1 receptor antagonist (IL-1Ra) may contribute to the pathogenesis of rheumatoid arthritis (RA). We analysed the frequency of different alleles in the IL-1B gene (at −511 and at +3954) as well as in the IL-1Ra gene (at +2018) in an association study involving 297 RA patients and 112 healthy controls from the same geographic area. We tested associations with RA susceptibility or severity, and with circulating levels of IL-1Ra and IL-1␤. Carriage of the rare IL-1B (+3954) allele 2 was increased in destructive arthritis (DRA) as compared to non-destructive arthritis (NDRA) (OR 1.7, 95% CI 1.1–2.8, 49.0% vs 35.9%) and controls (OR 1.7, 95% CI 1.1–2.8, 35.8%). Patients carrying this allele had a more destructive (Larsen wrist radiological index: mean ± s.e.m., 2.1 ± 0.2 vs 1.6 ± 0.1, P = 0.005; Steinbrocker functional index: 2.4 ± 0.1 vs 1.9 ± 0.1, P = 0.002) and active disease (Ritchie articular index: 8.1 ± 0.8 vs 5.3 ± 0.6, P = 0.002; erythrocyte sedimentation rate (ESR): 36.6 ± 2.9 mm/h vs 25.3 ± 1.8 mm/h, P = 0.002). This contribution was independent from that of HLA DR4/DR1 to severity. IL-1Ra plasma levels adjusted to ESR values were significantly lower in IL-1B2 (+3954) positive than negative RA patients (1.0 ± 0.1 vs 1.2 ± 0.1 ng/ml, P = 0.01). This IL-1B (+3954) gene polymorphism may be an important marker for the severity of joint destruction in RA and is associated with an imbalance in IL-1Ra production. As this genetic association was independent and additive to the risk of HLA DR4/DR1 status, it could be a useful addition to HLA-DR4/1 as a genetic prognostic marker early in the course of the disease. Genes and Immunity (2001) 2, 222–228.

Keywords: interleukin 1; interleukin 1 receptor antagonist; rheumatoid arthritis; gene polymorphism; destruction

Introduction inflammatory synovium and appears to play a predomi- nant role in the destructive pattern associated with RA.4 The aetiology of rheumatoid arthritis (RA) remains It is a known activator of collagenase and stromelysin unknown, although susceptibility and severity to RA and inhibits the synthesis of collagen and proteoglycan. have been associated with several alleles of the HLA- Enhanced levels of IL-1␣ and IL-1␤ have been measured DRB1, which can account for over 30% of the genetic in RA synovial fluid.5 Circulating levels of IL-1␤ have 1 component of the disease. Despite this association, there been shown to correlate with disease activity6 and with are indications that other genes may be involved in sus- radiographic progression.7 2 ceptibility and severity of RA. from both normal individuals and RA Three known genes have been described in the IL1 patients have been shown to vary considerably in the gene family, clustered in a 430-kb region on the long arm amount of IL-1 released in response to lipopolysaccha- 3 ␣ ␤ of human 2 (2q13). IL1 and IL1 bind to ride stimulation.8 This raises the possibility of a genetic the IL1 receptor type I, eliciting signal transduction. The component associated to a higher or lower IL-1 pro- product of the third gene, IL1 receptor antagonist (IL- duction in response to a given stimulus. Physiologic 1Ra), also binds to the receptor but does not initiate intra- regulation of IL-1 occurs in part through the activity of cellular signalling, acting as a competitive inhibitor. The IL-1Ra. RA outcome may be determined by the balance pro-inflammatory IL-1 is produced by the existing between IL-1 and IL-1Ra and various studies have shown deficient production of the IL-1Ra by RA synovium.9,10 Correspondence: Pr P Miossec, Clinical Immunology Unit, Departments Polymorphisms for the IL-1B gene have been described of Immunology and Rheumatology, Hoˆpital Edouard Herriot, 69437 Lyon within the promoter region (at −511) and in exon V (at Cedex 03, France. E-mail: miossecȰlaennec.univ-lyon1.fr +3954),11,12 and related to IL-1␤ production12,13 (Figure 1). These studies have been supported in part by grants from the Hos- For the IL-1Ra gene, polymorphisms include a VNTR in pices Civils de Lyon, the European Union (Biomed-2 program con- intron 214 and a single base variation (C/T) in exon 2 at tract BMH4-CT96-1698), the Association de Recherche sur la Poly- position +2018.15 The importance of IL-1:IL-1Ra ratio has arthrite (ARP) and the Arthritis Research Campaign for UK (ARC). Nicolas Buchs was supported by a fellowship from the Swiss been underlined not only in RA but also in inflammatory National Science Foundation. bowel diseases and polymorphisms for the respective Received 30 January 2001; revised and accepted 19 April 2001 genes might regulate production and bioactivity of their IL-1B and IL-1Ra gene polymorphisms in RA N Buchs et al 223 (DRA). Similarly, frequencies of HLA-DR4 and RF posi- tivity, recognised markers of disease severity,21,22 were significantly higher in DRA patients in comparison with NDRA patients. Conversely, a higher frequency of HLA- DR3 was observed in the later, indicating a possible pro- tective effect (Table 1). Allelic frequencies of the IL-1 gene cluster are indicated in Table 2, panels B (DRA vs NDRA) and C (DRA vs controls). The IL-1B2 (+3954) allele carriage rate was higher in DRA patients than in NDRA patients (P = 0.03, OR: 1.7, 95% CI 1.1–2.8, 49.0 vs 35.9%) and in controls Figure 1 Allele description of the three IL-1 family gene polymor- (P = 0.03, OR: 1.7, 95% CI 1.1–2.8, 35.8%). There was no phisms. Refer to the method section for details. significant difference in allelic distribution in the other two markers tested (IL-1B −511 and IL1RN +2018). There was a significant association between IL-1B2 (+3954) posi- ␤ products. An association between alleles of IL-1 and IL- tive and IL-1RN2 (+2018) negative status (P = 0.02, OR = 1Ra polymorphisms has been described in inflammatory 2.2, 95% CI = 1.1–4.3), as it could be expected because of 16–18 bowel diseases and associated with severity. the linkage disequilibrium in this region.15 We analysed the frequency of these different alleles of To further assess the relationship between IL-1B ␤ IL-1 and IL-1Ra gene polymorphisms and their associ- (+3954) gene polymorphism and severity, we analysed its ation looking for a possible impact on RA susceptibility relationship with indices of disease activity and joint and severity. We also looked at possible relationship destruction. Erythrocyte sedimentation rate (ESR) values ␤ between IL-1 and IL-1Ra plasma levels and these poly- were higher in RA patients positive for IL-1B2 (ESR: morphisms. mean ± s.e.m.: 36.6 ± 2.9 vs 25.3 ± 1.8 mm/h, P = 0.002). Furthermore, the carriage of the rare allele 2 was associa- Results ted with a higher index of joint destruction (Larsen index: 2.1 ± 0.2 vs 1.6 ± 0.1, P = 0.005), clinical activity (Ritchie ± ± = IL-1B (+3954), IL-1B (−511) and IL-1RN (+2018) allelic articular index: 8.1 0.8 vs 5.3 0.6, P 0.002), and dis- ± ± frequencies in controls and patients ability (Steinbrocker functional index: 2.4 0.1 vs 1.9 = Characteristics of the patient population are shown in 0.1, P 0.002) (Figure 2). It should be noticed that there Table 1 according to the degree of joint destruction. The was no difference in disease duration between the group different alleles of IL-1␤ and IL-1Ra gene polymorphisms of patients positive and those negative for IL-1B2. are described in Figure 1 and their frequencies in the patient population and controls are reported in Table 2, Relationship between IL-1B(+3954) gene panel A. No association was found between the presence polymorphism and HLA-DR4/DR1 of a gene marker and that of RA. To look for a possible We looked at the occurrence of HLA-DR4/DR1/IL-1 association with disease severity, the patient population composite genotypes in these populations (Table 3). As was then divided into two groups, according to joint expected, there was an association between the DR4/DR1 destruction as defined by the wrist Larsen index.19 Indeed markers and disease severity. Fifty-eight percent of HLA- radiologic evaluation remains the gold standard to meas- DR4/DR1 positive patients vs 42% of those negative had ure the rate of disease progression.20 As expected and DRA (OR 2.8, 95% CI 1.7–4.6, P = 0.0001). Of the DRA shown in Table 1, patients with non-destructive RA patients, 37% were positive for both HLA-DR4/DR1 and (NDRA) had indices of disease activity and joint destruc- IL1-B2 (+3954), compared to 17% in the NDRA subgroup tion significantly lower than those with destructive RA (OR 2.9, 95% CI 1.6–5.3, P = 0.001). The analysis of

Table 1 Clinical and biological parameters of RA patients according to disease severity

Parameters Destructive RA Non-destructive RA P (n = 150) (n = 147)

Sex (F:M) (n = 297) 113:37 118:29 NS Age (yr) (n = 296) 56.1 ± 1.3 51.2 ± 1.1 0.003 Disease duration (yr) (n = 294) 9.0 ± 0.5 7.6 ± 0.5 0.05 Ritchie articular index (n = 264) 8.7 ± 0.7 3.8 ± 0.5 0.0001 Steinbrocker functional index (n = 230) 2.7 ± 0.1 1.5 ± 0.1 0.0001 Right Larsen wrist index (n = 297) 3.2 ± 0.1 0.2 ± 0.03 0.0001 ESR (mm/h) (n = 282) 39.8 ± 2.3 19.6 ± 1.6 0.0001 Pts with RF (%) (n = 266) 67.2 37.8 0.0001 Pts HLA-DR4+ or DR1+ (%) (n = 249) 72.5 48.8 0.0002 Pts HLA-DR3+ (%) (n = 210) 13.8 26.7 0.02 IL-1␤ (pg/ml) (n = 115) 2.9 ± 0.5 1.8 ± 0.2 0.02 IL-1Ra (ng/ml) (n = 276) 1.3 ± 0.2 0.9 ± 0.1 0.03 ESR adjusted IL-1Ra (ng/ml) (n = 271) 1.1 ± 0.2 1.0 ± 0.1 NS

ESR, erythrocyte sedimentation rate; Pts, patients; RF, rheumatoid factor; NS, not significant. Results are expressed as mean ± s.e.m. The number of patients in which data were available is indicated in brackets.

Genes and Immunity IL-1B and IL-1Ra gene polymorphisms in RA N Buchs et al 224 Table 2 Allelic distribution in controls and in patients with destructive (DRA) and non-destructive arthritis (NDRA)

IL-1B (−511) IL-1B (3954) IL-1RN (2018)

1,1 1,2 2,2 1,1 1,2 2,2 1,1 1,2 2,2

A RA 119 130 23 156 102 15 158 109 15 Controls 46 53 11 70 34 5 55 45 12 cont. table: 2 × 3 cont. table: 2 × 3 cont. table: 2 × 3 chi = 0.3 P = NS chi = 1.6 P = NS chi = 4.2 P = NS

B DRA 55 76 9 74 65 6 78 57 8 NDRA 64 54 14 82 37 9 80 52 7 cont. table: 2 × 3 cont. table: 2 × 2(2+ vs 1,1) cont. table: 2 × 3 chi = 5.3 P = NS chi = 4.7 P = 0.03 chi = 0.3 P = NS OR = 1.17, 95% CI = 1.1–2.8

C DRA 55 76 9 74 65 6 78 57 8 Controls 46 53 11 70 34 5 55 45 12 cont. table: 2 × 3 cont. table: 2 × 2(2+ vs 1,1) cont. table: 2 × 3 chi = 1.5 P = NS chi = 4.4 P = 0.03 chi = 2.5 P = NS OR = 1.7 95% CI = 1.0–2.9

patients which were double positive for both HLA- mation (hence the correlation with ESR.23 Accordingly, DR4/DR1 and IL-1B2 (+3954) vs those that were negative we have corrected the individual levels of IL-1Ra data by showed a clear association with increased joint destruc- normalising for the ESR levels measured at the same tion (Larsen index: 2.4 ± 0.2 vs 0.8 ± 0.2, P = 0.0001), ESR time. This was performed to take into account the contri- levels (40.7 ± 4.0 vs 17.4 ± 2.3, P = 0.0001), Ritchie articular bution to IL-1Ra production given by the degree of index (8.4 ± 1.0 vs 4.8 ± 1.3, P = 0.0009) and Steinbrocker inflammation. Corrected IL-1Ra plasma levels were lower functional index (2.4 ± 0.2 vs 1.8 ± 0.1, P = 0.003). in RA patients positive for IL-1B2 (+3954) than in those Importantly, the IL-1B2 (+3954) positive status was asso- negative for this allele (1.0 ± 0.1 vs 1.2 ± 0.1 ng/ml ciated with a more severe disease independently of the respectively, P = 0.01, Figure 3). HLA-DR4/DR1 status. Indeed patients either HLA- DR4/DR1 negative or positive but IL-1B2 (+3954) posi- tive had higher joint destruction (Larsen index: 1.6 ± 0.3 Discussion vs 0.8 ± 0.2 and 2.4 ± 0.2 vs 1.9 ± 0.2, P = 0.03 and P = 0.03, respectively), Ritchie articular index (8.2 ± 1.6 vs 4.8 Taken together, these results support the concept that IL- ± 1.3 and 8.4 ± 1.0 vs 5.3 ± 0.7, P = 0.01 and P = 0.02, 1 gene variants should be considered disease modifiers respectively) and Steinbrocker index (2.4 ± 0.2 vs 1.8 ± 0.1 in RA. We report that the rare allele of the IL-1B (+3954) and 2.4 ± 0.2 vs 2.0 ± 0.1, P = 0.01 and 0.03, respectively) polymorphism is associated with the destructive pattern than those IL-1B2 (+3954) negative. In fact, the associ- characteristic of chronic RA. The rare IL-1B (+3954) allele ations of destructive disease with IL-1B (+3954) geno- was over- represented in DRA patients, compared to types appeared very similar in both the HLA-DR4/DR1 NDRA individuals, rather than in the RA patients vs the positive and negative subgroups. This indicated an inde- control population. This suggests an implication of IL-1B pendent and additive effect over that of HLA-DR4/DR1. gene variants in severity rather than susceptibility to RA. A recent study showed a similar increase in the IL-1B Link to plasma levels of IL-1 and IL-1Ra (+3954) rare allele in erosive as compared to non-erosive In order to investigate the possible consequences of such RA that was however not significant, possibly due to a polymorphisms on IL-1 production and function, we smaller number of subjects analysed.24 A previous genome- looked at circulating plasma levels of IL-1Ra and IL-1␤. wide scan2 did not report evidence for linkage in this Both levels were positively correlated to ESR (r = 0.19, P region (2q13), but this is not surprising because the = 0.001, n = 262, for IL-1Ra; and r = 0.40, P = 0.0001, n = phenotype considered was susceptibility to RA, rather 105, for IL-1␤, respectively). Furthermore, IL1Ra plasma than risk of erosive disease. levels were positively correlated to those of IL1␤ (r = 0.58, We also studied, within the limits of the patient popu- P = 0.0001, n = 112). lation available, if the IL-1B (+3954) status could be con- We tested if IL-1␤ and IL-1Ra levels were associated sidered a risk factor on its own merit, similarly to HLA- with any particular genotype. Neither IL-1␤ nor IL-1Ra DR4/1 and other genetic factors implicated in RA. We levels differed in patients grouped according to the IL- could show that patients who were both HLA DR-4/DR1 1B (−511), IL-1B (+3954) or IL-1RN genotypes (data not and IL1-B2 (+3954) positive, were mostly in the destruc- shown). IL-1Ra has been described as an acute phase pro- tive erosive subgroup. Furthermore, within patients who tein, and like ESR its levels raise in the course of inflam- were HLA DR4/DR1 negative, those carrying IL-1B

Genes and Immunity IL-1B and IL-1Ra gene polymorphisms in RA N Buchs et al 225

Figure 3 IL-1Ra circulating levels in RA patients concentration (ng/ml) according to the IL-1B (+3954) allele 2 carriage. IL-1Ra plasma levels were measured by RIA and were corrected for the ESR value. Results are expressed as mean ± s.e.m. and compared using the Mann-Whitney U test.

severe disease conferred by the IL-1 locus is additional and independent of the HLA-DR4/DR1 status. Interestingly, the rare allele of IL-1B (+3954) has pre- viously been associated with an IL-1␤ high production phenotype.12,13 This rare allele may therefore represent a risk factor for joint destruction by being a marker for chromosomal events related to IL-1␤ production, shifting the cytokine profile toward a proinflammatory pattern. Figure 2 Indices of joint destruction (Larsen wrist X-rays index, An impact of the rare allele of this polymorphism has Steinbrocker functional index) and joint inflammation (ESR, Ritchie already been shown in various diseases such as ulcerative articular index) according to IL-1B (+3954) allele 2 carriage in RA colitis, insulin-dependent diabetes mellitus, myasthenia, patients. Results are expressed as mean ± s.e.m. and compared periodontitis.12,29–31 Pro-inflammatory events at the IL-1 using the Mann-Whitney U test. locus would include modulation of production, secretion, or agonist bioactivity of IL-1␤, IL-1␣ or IL-1Ra. This is in Table 3 Association between IL-1B (+3954) and HLADR4/1 geno- line with the imbalance between IL-1␤ and IL-1Ra, with types in patients with destructive (DRA) and non-destructive a relative excess in IL-1␤ over IL-1Ra described in RA.9,10 arthritis (NDRA) In our study, the IL-1B (+3954) rare allele was not asso- ciated with an increased plasma level of IL-1␤ in contrast IL1B +3954 2/2 or 1/2 1/1 to other studies.12,13 Data on the production of single genotype in relation to single chromosomal markers in DRA NDRA DRA NDRA the IL-1 locus are highly controversial, and mostly = = = = n 58 n 35 n 57 n 77 plagued by experimental errors or shortcomings in a very complex system. In our case, we can only speculate that positive 43 19 42 37 HLA-DR4/1 circulating levels do not reflect the situation at the level negative 15 16 15 40 of the inflamed synovium, as we have shown in early studies.32 P = 0.001 for HLA-DR4/1 and IL1B2 (+3954) positive DRA vs NDRA patients. Mechanisms other than direct co-regulation need to be invoked to explain the association of IL-1B (+3954) with lower plasma levels of IL-1Ra in RA patients. It is poss- (+3954) had significantly higher Larsen, Steinbrocker and ible that this is related to a linkage disequilibrium with Ritchie indices. another IL-1 RN gene polymorphism (IL-1RN VNTR), This confirms that RA patients are more prone to which is in 100% linkage disequilibrium with IL-1RN develop destruction when additional genetic factors are (+2018), the one analysed in this study. This IL-1RN present, as first described for HLA-DR4,21 then for CTLA marker has been linked to an increase in IL-1Ra pro- 4 in some25 but not all populations.26 The risk of severe duction for its rare allele in stimulated circulating mono- arthritis associated with IL-1 locus genotypes had pre- cytes33 and in diabetic patients34 but to a decreased viously been reported in family27 and association stud- mucosal production in the bowel.35 The linkage disequi- ies.28 We clearly indicate in this study that the risk of librium between the rare allele of IL-1B (+3954) and IL-

Genes and Immunity IL-1B and IL-1Ra gene polymorphisms in RA N Buchs et al 226 1RN (+2018), and the possibility of tissue-specific gene DNA analysis regulation could in part explain our findings. For each individual enrolled in the study, a 7-ml sample The biological significance of the IL-1 locus genetic of venous blood was collected in EDTA. Plasma was association with erosive RA remains unclear. The associ- freshly separated and stored in 0.5 ml aliquots at −20°C ation of IL-1B2 (+3954) with greater disease severity may before cytokine determination. DNA was extracted from suggest that this is a genetic marker for an IL-1 haplotype uncoagulated blood by a modification of the salt-out which, by multiple pathways, determines different activi- technique (Nucleon, Scotlab, UK) and stored at a final ties on the IL-1 receptor. The occurrence of specific IL-1 concentration of 200 ␮g/ml until used for genotyping. locus haplotypes may lead to a variation of the IL-1␤/IL- The cytokine gene polymorphisms are presented in Fig- 1Ra ratio in specific tissues, or making more IL-1␣ ure 1. bioavailable intercellularly and leading for example to destructive RA. IL-1B (−511): A single base variation (C/T) has been The functional correlates of IL-1 locus genetic variants described in the promoter region of IL-1B gene at pos- are likely to play an important role in inflammatory dis- ition −511(11). Allele 1 is C and allele 2 is T. Primer eases, but simple experimental approaches have to date sequences and PCR conditions were: Forward primer: failed to explain mechanisms of altered gene regulation. 5ЈTGG CAT TGA TCT GGT TCA TC-3Ј; Reverse primer: Simple functional hypotheses have generated contrasting 5ЈGTT TAG GAA TCT TCC CAC TT-3Ј; PCR cycles: studies and new experimental approaches are needed to [95°C, 1 min] × 1; [95°C, 1 min; 53°C, 1 min; 72°C, 1 min] unravel the functional significance of these genetic associ- × 35; [72°C, 5 min] × 1. ations of the IL-1 gene variants. It is confirmed however The resulting products contained an Ava I restriction that IL-1 gene markers can be very useful in predicting site in the frequent allele (‘1’) and a Bsu 361 site in the disease severity in RA, as risk factors which are inde- rarer ‘2’ allele. Products were analysed by PAGE, and pendent and additional to that of DR4/DR1. allele 1 (C), Ava I digestion of PCR products yielded 190 and 114 bp fragments whereas for allele 2 (T), a single 304 Patients and methods bp product was observed. In contrast, Bsu 361 digestion produced 190 and 114 bp for allele 2, while allele 1 was Patients and controls uncut (304 bp). A total of 378 patients with chronic polyarthritis were included. All were followed by the same investigators IL-1B (+3954): This polymorphism was described as Taq (NB, PM) at the same hospital in Lyon. Approval for this I RFLP of IL-1B (12). We sequenced the most likely region study was obtained from the local ethical committee. All implicated and a C/T single base variation was found at patients fulfilled the commonly used criteria for RA +3954 in Exon V, explaining the RFLP. Allele 1 is C and (American College of Rheumatology (ACR)) criteria.36 All allele 2 is T. PCR primers were designed to engineer a patients had a disease duration of at least 2 years. control Taq I site in the PCR product. Primer sequences Arthritis patients who had recent wrist X-rays (n = 297) and genotyping conditions were: Forward primer: 5ЈCTC were classified into two groups: those with destructive AGG TGT CCT CGA AGA AAT CAA A-3Ј; Reverse arthritis (DRA), ie with a mean Larsen wrist X-rays index primer: 5ЈGCT TTT TTG CTG TGA GTC CCG-3Ј; PCR Ͼ2 (3.21 ± 0.07, mean ± s.e.m., n = 150) and those without cycles: [95°C, 2 min] × 1; [95°C, 1 min; 67.5°C, 1 min; (NDRA), ie with Larsen wrist X-rays index Ͻ2 (0.18 ± 72°C, 1 min] × 35; [72°C, 5 min] × 1. 0.03, n = 147). The control population was composed of PCR products were size fractionated by PAGE. Taq I 108 anonymous blood donors of the Lyon blood bank, yielded a constant band of 12 bp and either two further France, from the same geographic area as the patients. bands of 85 and 97 bp (allele 1), or a single band of 182 Biological and genetic studies were performed blindly by bp (allele 2). scientists unaware of the clinical status. Clinical indices of disease activity and joint destruction IL-1-RN (+2018): A single base variation (C/T) has been were evaluated by a single observer. The clinical data described in Exon 2 at position +2018 (14). Allele 1 is C included: age, sex, disease duration, Ritchie articular and allele 2 is T. Primer sequences and genotyping con- index, Steinbrocker functional index and right Larsen ditions were: Forward primer: 5ЈCTA TCT GAG GAA wrist X-rays index. Biological data included: ESR, rheu- CAA CCA ACT AGT AGC-3Ј; Reverse primer: 5ЈTAG matoid factor (RF) tested with Rose-Waaler assay. HLA GAC ATT GCA CCT AGG GTT TGT-3Ј; PCR cycles: typing for HLA-DR4 and DR1 was performed by a sero- [96°C, 1 min] × 1; [94°C, 1 min; 57°C, 1 min; 70°C, 2 min] logic method as described.37 × 35; [70°C, 5 min] × 1. Clinical and biological parameters of RA patients PCR primers were designed to engineer two enzyme according to the disease severity are presented in Table cutting sites on the two alleles. Alu I digestion of PCR 1. As expected, patients with non-destructive arthritis products yielded 126 and 28 bp fragments for allele 1, (mean Larsen wrist X-rays index Ͻ2) had indices of dis- whereas for allele 2, a single 154 bp product was ease activity and joint destruction significantly lower observed. Msp I digestion produced 125 and 29 bp for than patients with destructive arthritis (mean Larsen allele 2, while allele 1 was uncut (154 bp). wrist X-rays index Ͼ2). Plasma levels of IL-1Ra were assessed with a specific Statistical analysis radioimmunoassay (RIA) exactly as previously The genotype distributions of the three IL-1B and IL1RN described.38 Levels of IL-1␤ were measured with the polymorphisms were compared between RA patients and human IL-1 beta Quantikine High Sensitivity ELISA from controls, and between patients with destructive or non- R&D (Abington, UK) according to the manufacturer’s destructive arthritis (Table 1). Heterozygous and homo- instructions. zygous odds ratios (ORhet and ORhom) were calculated

Genes and Immunity IL-1B and IL-1Ra gene polymorphisms in RA N Buchs et al 227 for each locus. These indicate the increased risk to an correlates with IL-1 beta secretion in vitro. Eur J Clin Invest 1992; individual heterozygote or homozygote carrier for a spe- 22: 396–402. cific allele compared with homozygous non-carriers, and 13 Di Giovine FS, Cork MJ, Crane A, Mee JB, Duff GW. Novel gen- + are estimates of the genotypic relative risks (GRR).36 On etic association of an IL-1B gene variation at 3953 with IL-1 inspection of these, and comparison of observed vs beta protein production and psoriasis. Cytokine 1995; 7: 606 (Abstr. A665). expected values, data were either analysed in a 2 × 3 × 14 Demeter J, Messer G, Ramisch S et al. Polymorphism within the genotype table or a 2 2 table for the carriage of a specific second intron of the IL-1 receptor antagonist gene in patients allele (as indicated in the table). These contingency tables with hematopoietic malignancies. Cytokines Mol Ther 1996; 2: were tested using chi-squared analysis, and standard 239–242. odds ratios with 95% confidence intervals were calcu- 15 Clay FE, Tarlow JK, Cork MJ, Cox A, Nicklin MJ, Duff GW. lated. Novel interleukin-1 receptor antagonist exon polymorphisms Nominal variables (eg HLADR4/DR1 positivity) were and their use in allele-specific mRNA assessment. Hum Genet compared in different groups using the chi-square test, 1996; 97: 723–726. and continuous variables (like extent of erosions) by the 16 Bidwell J, Keen L, Gallagher G et al. Cytokine gene polymor- unpaired Student’s t-test or the Mann-Whitney U test. In phism in human disease: on-line databases. Genes Immun 1999; 1 order to show a possible correlation between the tested : 3–19. 17 Heresbach D, Alizadeh M, Dabadie A et al. Significance of polymorphisms and IL-1Ra plasma levels, we have taken interleukin-1beta and interleukin-1 receptor antagonist genetic into account the positive impact of inflammation on IL- polymorphism in inflammatory bowel diseases. Am J Gastro- 23 1Ra plasma levels. Accordingly, IL-1Ra plasma concen- enterol 1997; 92: 1164–1169. trations were corrected for the level of ESR, measured at 18 Bioque G, Crusius JB, Koutroubakis I et al. Allelic polymorphism the same time. Indeed a linear correlation was established in IL-1 beta and IL-1 receptor antagonist (IL-1Ra) genes in in the patient population between IL-1Ra and ESR levels inflammatory bowel disease. Clin Exp Immunol 1995; 102: 379– where IL-1Ra level = aESR level + b. Single IL-1Ra levels 383. were then corrected for the level of inflammation for each 19 Larsen A, Dale K, Eek M. Radiographic evaluation of rheuma- sample, as measured by ESR, using the following for- toid arthritis and related conditions by standard reference films. mula: corrected IL-1Ra level = measured IL-1Ra level + Acta Radiol Diagn (Stockh) 1977; 18: 481–491. 20 van der Heijde DM. Radiographic imaging: the ‘gold standard’ mean IL-1Ra level in the population − (aESR level+b). for assessment of disease progression in rheumatoid arthritis. Rheumatology (Oxford) 2000; 39 (Suppl 1): 9–16. 21 van Zeben D, Hazes JM, Zwinderman AH et al. Association of References HLA-DR4 with a more progressive disease course in patients with rheumatoid arthritis. Results of a follow up study. Arthritis 1 Deighton CM, Walker DJ, Griffiths ID, Roberts DF. The contri- Rheum 1991; 34: 822–830. bution of HLA to rheumatoid arthritis. Clin Genet 1989; 36: 22 Corbett M, Dalton S, Young A, Silman A, Shipley M. Factors 178–182. predicting death, survival and functional outcome in a prospec- 2 Cornelis F, Faure S, Martinez M et al. New susceptibility locus tive study of early rheumatoid disease over fifteen years. Br J for rheumatoid arthritis suggested by a genome-wide linkage Rheumatol 1993; 32: 717–723. study. Proc Natl Acad Sci USA 1998; 95: 10746–10750. 23 Gabay C, Smith MF, Eidlen D, Arend WP. Interleukin 1 receptor 3 Nicklin MJ, Weith A, Duff GW. A physical map of the region antagonist (IL-1Ra) is an acute-phase protein. J Clin Invest 1997; encompassing the human interleukin-1 alpha, interleukin-1 99 beta, and interleukin-1 receptor antagonist genes. Genomics 1994; : 2930–2940. 19: 382–384. 24 Cantagrel A, Navaux F, Loubet-Lescoulie P et al. Interleukin- 4 Arend WP, Dayer JM. Inhibition of the production and effects 1beta, interleukin-1 receptor antagonist, interleukin-4, and of Interleukin-1 and ␣ in rheumatoid interleukin-10 gene polymorphisms: relationship to occurrence arthritis. Arthritis Rheum 1995; 38: 151–160. and severity of rheumatoid arthritis. Arthritis Rheum 1999; 42: 5 Hopkins SJ, Humphreys M, Jayson MI. Cytokines in synovial 1093–1100. fluid. I. The presence of biologically active and immunoreactive 25 Seidl C, Donner H, Fischer B et al. CTLA4 codon 17 dimorphism IL-1. Clin Exp Immunol 1988; 72: 422–427. in patients with rheumatoid arthritis. Tissue Antigens 1998; 51: 6 Eastgate JA, Symons JA, Wood NC, Grinlinton FM, di Giovine 62–66. FS, Duff GW. Correlation of plasma interleukin 1 levels with 26 Barton A, Myerscough A, John S, Gonzalez-Gay M, Ollier W, disease activity in rheumatoid arthritis. Lancet 1988; 2: 706–709. Worthington J. A single polymorphism in exon 1 of 7 North J, Situnayake RD, Tikly M et al. Interleukin 1 beta, hand cytotoxic T-lymphocyte-associated-4 (CTLA-4) is not associated and foot bone mineral content and the development of joint ero- with rheumatoid arthritis. Rheumatology (Oxford) 2000; 39: 63–66. sions in rheumatoid arthritis. Ann Rheum Dis 1994; 53: 543–546. 27 Cox A, Camp NJ, Cannings C et al. Combined sib-TDT and TDT 8 Danis VA, Kulesz AJ, Nelson DS, Brooks PM. The effect of gold provide evidence for linkage of the interleukin-1 gene cluster to sodium thiomalate and auranofin on lipopolysaccharide- erosive rheumatoid arthritis. Hum Mol Genet 1999; 8: 1707–1713. induced interleukin-1 production by blood monocytes in vitro: 28 Jouvenne P, Chaudhary A, Buchs N, di Giovine FS, Duff GW, ␣ variation in healthy subjects and patients with arthritis. Clin Exp Miossec P. Possible genetic association between IL-1 gene poly- Immunol 1990; 79: 335–340. morphism and the severity of arthritis. Eur Cyt Netw 1999; 10: 9 Chomarat P, Vannier E, Dechanet J et al. The balance of IL-1 33–36. receptor antagonist/IL-1␤ in rheumatoid synovium and its 29 Stokkers PC, van Aken BE, Basoski N, Reitsma PH, Tytgat GN, regulation by IL-4 and IL-10. J Immunol 1995; 154: 1432–1439. van Deventer SJ. Five genetic markers in the interleukin 1 family 10 Firestein GS, Boyle DL, Yu C et al. Synovial interleukin-1 recep- in relation to inflammatory bowel disease. Gut 1998; 43: 33–39. tor antagonist and interleukin-1 balance in rheumatoid arthritis. 30 Kornman KS, Crane A, Wang HY et al. The interleukin-1 geno- Arthritis Rheum 1994; 37: 644–652. type as a severity factor in adult periodontal disease. J Clin Per- 11 di Giovine FS, Takhsh E, Blakemore AI, Duff GW. Single base iodontol 1997; 24: 72–77. polymorphism at -511 in the human interleukin-1 beta gene (IL1 31 Huang D, Pirskanen R, Hjelmstrom P, Lefvert AK. Polymor- beta). Hum Mol Genet 1992; 1: 450. phisms in IL-1beta and IL-1 receptor antagonist genes are asso- 12 Pociot F, Molvig J, Wogensen L, Worsaae H, Nerup J. A TaqI ciated with myasthenia gravis. J Neuroimmunol 1998; 81: 76–81. polymorphism in the human interleukin-1 beta (IL-1 beta) gene 32 Saxne T, Di Giovine FS, Heinegard D, Duff GW, Wollheim FA.

Genes and Immunity IL-1B and IL-1Ra gene polymorphisms in RA N Buchs et al 228 Synovial fluid concentrations of interleukin-1 beta and proteog- 36 Schaid DJ, Sommer SS. Genotype relative risks: methods for lycans are inversely related. J Autoimmun 1988; 1: 373–380. design and analysis of candidate-gene association studies. Am J 33 Danis VA, Millington M, Hyland VJ, Grennan D. Cytokine pro- Hum Genet 1993; 53: 1114–1126. duction by normal human monocytes: inter-subject variation 37 Freidel AC, Betuel H, Gebuhrer L, Farre A, Lambert J. Cellular and relationship to an IL-1 receptor antagonist (IL-1Ra) gene and serological subtypes of HLA-DR2. Tissue Antigens 1987; 29: polymorphism. Clin Exp Immunol 1995; 99: 303–310. 129–140. 34 Mandrup-Poulsen T, Pociot F, Molvig J et al. antagon- 38 Jouvenne P, Vannier E, Dinarello CA, Miossec P. Elevated levels ism is reduced in patients with IDDM. Diabetes 1994; 43: of soluble interleukin-1 receptor type II and interleukin-1 recep- 1242–1247. tor antagonist in patients with chronic arthritis: correlations 35 Dionne S, D’Agata ID, Hiscott J, Vanounou T, Seidman EG. with markers of inflammation and joint destruction. Arthritis Colonic explant production of IL-1 and its receptor antagonist Rheum 1998; 41: 1083–1089. is imbalanced in inflammatory bowel disease (IBD). Clin Exp Immunol 1998; 112: 435–442.

Genes and Immunity