A missense single-nucleotide polymorphism (SNP) (rs2476601, C1858T) in PTPN22, which encodes a tyrosine phosphatase, has been associated with multiple autoimmune diseases including rheumatoid arthritis (RA) (1–3). This SNP results in the substitution of a conserved arginine with tryptophan at codon 620 (R620W) in the protein's SH3-binding domain. The 1858T risk allele, which disrupts the interaction between PTPN22 and the C-Src kinase CSK (3), potentially alters these proteins' function as negative regulators of T cell activation. In the present study, we investigated the association of the PTPN22 C1858T SNP with RA, undifferentiated arthritis (UA), and both quantitative (rate of joint destruction) and qualitative (autoantibody status and remission or progression) RA characteristics.

All RA cases (n = 416) were participants in the Leiden Early Arthritis Clinic (EAC), a population-based inception cohort of patients with recent onset arthritis described by van Aken et al (4), and fulfilled the American College of Rheumatology (formerly, the American Rheumatism Association) criteria for RA (5) at 1 year followup. EAC participants who could not be properly classified at 1 year (n = 265) were categorized as UA cases. All of the patients and unrelated control individuals (6) were Dutch Caucasians. The appropriate institutional review boards approved the protocol.

The C1858T SNP and HLA–DRB1 alleles were genotyped as previously described (1, 7), and genotyping accuracy was >99.8%. Individuals were considered positive for the shared epitope (SE) if they carried at least 1 copy of any of the following HLA–DRB1 alleles: 0101, 0102, 0401, 0404, 0405, 0408, 1001, or 1402. Radiography of hands and feet was performed at baseline, 6 months, 1 year, and annually thereafter. The chi-square test or Fisher's exact test, unconditional logistic regression, and tests for trend were used for statistical analysis. Reported P values are 2-tailed; values less than 0.05 were considered significant.

PTPN22 C1858T genotypes were generated for 416 RA cases, 265 UA cases, and 891 controls, and were in Hardy-Weinberg equilibrium in all groups. The frequency of the PTPN22 risk allele in our control group was similar to previously reported allele frequencies in US white controls (0.091 and 0.087, respectively) (1). We observed a higher frequency of the 1858T risk allele in RA cases compared with controls (0.119 versus 0.091; P = 0.0257) (Table 1). Genotypic analysis showed an increased risk of RA associated with the presence of 1 or 2 copies of the 1858T allele (odds ratio [OR] 1.37, P = 0.0336). Consistent with previous reports (1, 2), TT homozygotes (OR 1.95) appeared to be at greater RA risk than CT heterozygous individuals (OR 1.34) when compared with CC homozygotes.

Table 1. PTPN22 C1858T SNP case–control analysis*
Study groupTGenotypeCT, OR (95% CI)TT, OR (95% CI)PtrendCT + TT
FrequencyPCCCTTTOR (95% CI)P§
  • *

    SNP = single nucleotide polymorphism; 95% CI = 95% confidence interval; RA = rheumatoid arthritis; anti-CCP = anti–cyclic citrullinated peptide; UA = undifferentiated arthritis.

  • Significance of the difference in allele frequency compared with control group.

  • Test of linear trend for genotypic odds ratios (ORs).

  • §

    Significance of the genotypic OR, 1858T carriers versus noncarriers.

  • One-year diagnosis.

  • #

    Data were not available for all cases.

Controls (n = 891)0.091 7361487     
RA cases (n = 416)0.1190.02573238761.34 (1.00–1.80)1.95 (0.65–5.86)0.08381.37 (1.03–1.82)0.0336
Rheumatoid factor          
 Positive (n = 249)0.1310.00911895551.45 (1.02–2.05)2.78 (0.87–8.86)0.03191.51 (1.08–2.11)0.0173
 Negative (n = 167)0.1020.52881343211.19 (0.78–1.82)0.79 (0.10–6.43)0.70581.17 (0.77–1.78)0.4636
 Positive (n = 197)0.1320.01321494441.47 (1.00–2.15)2.82 (0.82–9.76)0.04471.53 (1.06–2.21)0.0237
 Negative (n = 153)0.0920.97341252801.11 (0.71–1.74) 0.89361.06 (0.68–1.66)0.7858
UA (n = 265)0.1260.01632016131.51 (1.08–2.11)1.57 (0.40–6.12)0.04921.51 (1.09–2.10)0.0141

Next, we investigated this SNP in the patients stratified according to autoantibody status (Table 1). Compared with the allele frequency in the controls, the 1858T allele frequency was elevated in both rheumatoid factor (RF)–positive (P = 0.0091) and cyclic citrullinated peptide (CCP)–positive (P = 0.0132) RA patients, but not in patients who were negative for these autoantibodies. Genotypic analysis revealed that carriers of the 1858T allele were at an increased risk for RF-positive RA (P = 0.0173) and anti-CCP–positive RA (P = 0.0237). This was not the case for autoantibody-negative patients, which suggests that there is an association between the PTPN22 1858T risk allele and autoantibody production in RA. Consistent with findings in previous studies (1, 2), PTPN22 C1858T genotype frequencies were similar in HLA–DRB1 SE–positive and SE-negative cases (0.125 versus 0.105; P = 0.4022). This suggests that the PTPN22 risk allele acts independently of HLA–DRB1 susceptiblility alleles to influence RA risk.

Allele frequencies of C1858T were similar in 45 RA patients whose disease entered remission (defined as absence of arthritis without use of disease-modifying antirheumatic drugs or nonsteroidal antiinflammatory drugs) versus 319 patients with persistent inflammation (0.10 versus 0.118; P = 0.707) (8). Mean baseline and yearly Sharp-van der Heijde scores (9) of radiographs of the hands and feet of RA patients with different PTPN22 genotypes (Figure 1) were also similar in 1858T carriers and noncarriers. These data suggest that there is no association of the PTPN22 risk allele with the rate of joint destruction.

thumbnail image

Figure 1. Sharp-van der Heijde scores over time among rheumatoid arthritis patients, by PTPN22 1858T carrier status (CT + TT versus CC). Radiographs of the hands and feet of 220 CC and 68 CT + TT patients were available at baseline. Data were analyzed for 254 CC and 72 CT + TT patients at year 1, 178 and 56 at year 2, 117 and 36 at year 3, and 101 and 27 at year 4, respectively. Values are the mean ± SEM.

Download figure to PowerPoint

The 1858T allele frequency was also elevated in UA cases compared controls (0.126 versus 0.091; P = 0.0163) (Table 1). Genotypic analysis indicated that carriers of the 1858T allele were at significantly higher risk of UA (OR 1.51).

These results support the role of the PTPN22 C1858T SNP as a common, HLA-independent, genetic risk factor for RA. Moreover, this study also replicated the dosage effect that occurs when comparing CT and TT risk genotypes (1–3). We confirm the predominant PTPN22 association with RF-positive RA (1, 2) and provide the first evidence that this SNP is also associated with anti-CCP–positive RA. These findings support the hypothesis that the PTPN22-1858T variant may predispose individuals to autoimmunity by facilitating the production of certain disease-associated autoantibodies (1, 2). Although no association between the PTPN22 C1858T SNP and rate of RA joint destruction, remission, and progression was observed in this study, the sample size is too small to rule out the possibility of a Type II error. Our results also indicate that the 1858T allele is a risk factor for UA. In conclusion, our data suggest that the PTPN22 1858T variant acts as a susceptibility allele for autoantibody-positive RA, but does not appear to influence RA severity.


  1. Top of page
  2. Acknowledgements

Drs. Begovich and Chokkalingam have stock options in Applied Biosystems and Celera Genomics. Drs. Wesoly and Chokkalingam contributed equally to this work.