Immunobiology 220 (2015) 1012–1024
Contents lists available at ScienceDirect
Immunobiology journal homepage: www.elsevier.com/locate/imbio
Association of PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with rheumatoid arthritis: A meta-analysis update Rami Elshazli a,∗ , Ahmad Settin b a b
Department of Biochemistry, Faculty of Science, Tanta University, Tanta, Egypt Genetics Unit, Children Hospital, Mansoura University, Mansoura, Egypt
a r t i c l e
i n f o
Article history: Received 14 March 2015 Accepted 20 April 2015 Available online 28 April 2015 Keywords: Meta-analysis Rheumatoid arthritis Polymorphism PTPN22 STAT4 RF Anti-CCP
a b s t r a c t Background: Rheumatoid arthritis (RA) is a common autoimmune disease with a complex genetic background. The genes encoding protein tyrosine phosphatase non-receptor type 22 (PTPN22) and signal transducer and activator of transcription 4 (STAT4) have been reported to be associated with RA in several ethnic populations. Objectives: This work aims to assess the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with RA susceptibility through an updated meta-analysis of available case–control studies. Methods: A literature search of all relevant studies published from January 2007 up to December 2014 was conducted using Pubmed and Science Direct databases. The observed studies that were related to an association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with RA susceptibility were identified. Meta-analyis of the pooled and stratified data was done and assessed using varied genetic models. Results: Thirty-seven case–control studies with a total of 47 comparisons (29 for PTPN22 rs2476601 polymorphism and 18 for STAT4 rs7574865 polymorphism) met our inclusion criteria. The meta-analysis showed an association between PTPN22 T allele, CT+TT and TT genotypes with RA susceptibility. Furthermore, The meta-analysis showed an association between STAT4 T allele, GT+TT and TT genotypes with RA susceptibility. Stratification of RA patients according to ethnic groups showed that PTPN22 T allele, CT+TT genotypes, STAT4 T allele and STAT4 GT+TT were significantly associated with RA in European, Asian, African subjects, while PTPN22 TT genotype was significantly associated with RA in European but not in Asian and African subjects and STAT4 TT genotype was significantly associated with RA in European and Asian but not in African subject. A subgroup analysis according to the presence or absence of rheumatoid factor (RF) and anti-cyclic citrullinated peptide (anti-CCP) antibodies revealed that the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with RA susceptibility may not be dependent on RF and anti-CCP antibodies. Conclusions: Our meta-analysis demonstrated that PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms confers susceptibility to RA in total subjects and in major ethnic groups. The association may not be dependent on RF and anti-CCP antibodies. © 2015 Elsevier GmbH. All rights reserved.
Introduction Rheumatoid arthritis (RA) is a chronic inflammatory disorder approximately affecting 1% of the adult population worldwide. The disease is characterized by an inflammation of the synovial tissue of multiple joints leading to pain, deformities and a reduced quality of life (Oliver et al., 2006; Firestein, 2003). Although the etiology
∗ Corresponding author. Tel.: +20 1064620110. E-mail address: [email protected] (R. Elshazli). http://dx.doi.org/10.1016/j.imbio.2015.04.003 0171-2985/© 2015 Elsevier GmbH. All rights reserved.
of RA is unknown, both genetic and environmental factors have been shown to play a role in its development. Genetic factors were thought to be responsible for up to 50–60% of the predisposition to RA. The genetic background of RA is complex and is likely involving multiple genes which encode proteins with significant functions in the regulation of immune response (Ikari et al., 2006; Goëb et al., 2008). Human leukocyte antigen (HLA) class II genes were portrayed as an important factors linked to RA, but they account for only one-third of genetic susceptibility, therefore non-HLA genes should be also considered (Urayama et al., 2013; Deighton et al., 1989).
R. Elshazli, A. Settin / Immunobiology 220 (2015) 1012–1024
Two major genes have been contributed of the susceptibility with RA, these genes include the protein tyrosine phosphatase nonreceptor type 22 (PTPN22) and the signal transducer and activator of transcription 4 (STAT4). PTPN22 gene, is located on chromosome 1p13, and encodes a lymphoid specific phosphatase (LYP), which is important in the negative control of T cell activation and development (Cohen et al., 1999; Siminovitch, 2004; Totaro et al., 2011). The PTPN22 +1858 C>T SNP changes the amino acid at position 620 from an arginine (R) to a tryptophan (W) and disrupts the interaction between LYP and Csk and thus inhibits complex formation and suppresses T-cell activation, hence, is thought to predispose to multiple autoimmune diseases (Begovich et al., 2004; Bottini et al., 2004). On the other hand, STAT4 gene transmits signals induced by interleukin-12 (IL-12), interleukin-23 (IL-23) and interferon-␥ (IFN-␥), which are key cytokines and play important roles in the development of autoimmune diseases (Frucht et al., 2000; Mathur et al., 2007; Watford et al., 2004). The STAT4 gene maps to chromosome 2q33 and encodes a transcription factor, which plays pivotal roles in the differentiation and proliferation of both T helper 1 (Th1) and T helper 17 (Th17) cells (Watford et al., 2004). Since Th1 and Th17 lineages are crucial effectors in chronic inflammatory disorders, STAT4 gene may play an important role in the pathogenesis of autoimmune diseases in different ethnic populations, such as RA (Zhao et al., 2013; Tong et al., 2013; Mohamed et al., 2012; Settin et al., 2014). Recently, a large number of studies have explored the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with the susceptibility to RA in different ethnic populations (Song et al., 2013; Salama et al., 2014; Torres-Carrillo et al., 2012; Farago et al., 2009; Gu et al., 2014; Ben Hamad et al., 2011; Tong et al., 2013; Shen et al., 2013; Settin et al., 2014). However, some results from the previous studies were inconsistent which might be due to the sample size used, ethnic differences in allele frequencies, or publication bias (Song et al., 2013; TorresCarrillo et al., 2012; Gu et al., 2014; Settin et al., 2014). In the present study, we have done a meta-analysis to check for the contribution of PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms to RA susceptibility in different populations. We also examined their association with RA subtypes in terms of a positive or negative anti-cyclic citrullinated peptide (anti-CCP) or rheumatoid factor (RF). Methods Study identification and selection This meta-analysis followed the preferred reporting items for systematic reviews and meta-analyses (PRISMA) criteria (Moher et al., 2010). A literature search was conducted using Pubmed and Science Direct citation databases to identify articles published from January 2007 up to December 2014 that examined the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with the susceptibility to RA. Combinations of keywords such as: “Rheumatoid arthritis”, “PTPN22”, “STAT4”, “polymorphism” and “RA” were entered as both Medical Subject Headings (MeSHs) and text words without any restrictions on language or country. Inclusion and exclusion criteria Data were collected from the full-published paper, excluding any meeting or conference abstract. Inclusion criteria were defined as follows: (a) articles evaluating the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with RA susceptibility, (b) the design is a case–controlled study based on unrelated individuals, and (c) sufficient data (genotype distributions for cases
1013
and controls) available to estimate an odds ratio (OR) with its 95% confidence interval (CI). Studies were excluded if one of the following existed: (a) studies containing overlapping data; (b) studies in which the genotype frequencies or numbers could not be ascertained; (c) studies in which family members were studied, because their analysis is based on linkage considerations; (d) reviews and abstracts; and (e) studies in which the genotype distribution in controls was not consistent with Hardy–Weinberg equilibrium (HWE), because deviation from HWE among controls might be due to a biased control-selection or genotyping errors. If more than one article were published by the same authors using the same sample series, studies with the largest size of samples or the recently published ones were included. Data extraction The following information was extracted: (1) name of the first author; (2) year of publication; (3) country of origin; (4) ethnicity of the studied population; (5) the numbers of cases and controls of polymorphisms of PTPN22 and STAT4; (6) genotyping method; (7) Hardy–Weinberg equilibrium (HWE) of controls and (8) antibodies status (AS). Statistical analysis The Hardy–Weinberg equilibrium (HWE) was examined in control groups by Fisher’s exact test. If the study was found not to be in HWE with P value less than 0.05, it was considered to be disequilibrium. Allele frequencies of the PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms in each of the studies were determined using the allele counting method. Allelic effect contrast was examined for the minor allele vs. the common allele. The genetic models evaluated for pooled ORs of these two polymorphisms have included allelic contrast, recessive models, dominant models, overdominant models, homozygote contrast and heterozygote contrast. Furthermore meta-analyses were performed once using the total data set, and separately for the subgroups of different ethnicities, RF and anti-CCP status (Tables 2, 4 and 5). To evaluate the strength of association, the pooled odd ratios (ORs) and their 95% confidence intervals (CIs) were determined for each study, and within- and between-study heterogeneity were assessed using Cochran’s Q statistic. The heterogeneity test was used to assess the probability of the null hypothesis that all studies were evaluating the same effect. The random-effects model was used for meta-analysis when a significant Q statistic (P < 0.10) indicated heterogeneity across studies, while the fixed-effect model was used when heterogeneity was not indicated. The fixed-effect model assumes that genetic factors have similar effects on disease susceptibility across all studies and that observed variations between studies are caused by chance alone (Davey Smith and Egger, 1997). The random-effect model assumes that studies show substantial diversity, and assesses both within study sampling errors and between study variances (DerSimonian and Laird, 1986). When study groups are homogenous, the two models are similar, but if this is not the case, the random effects model usually provides wider CIs than the fixed effects model. The random effects model is best used in the presence of significant between study heterogeneity (DerSimonian and Laird, 1986). We quantified the effect of heterogeneity by using the recently developed I2 measure, where I2 = 100% x (Q − df)/Q (Higgins and Thompson, 2002). The I2 measure ranges between 0% and 100%, and it represents the proportion of inter-study variability attributable to heterogeneity rather than chance. I2 values of 25%, 50%, and 75% were defined as low, moderate, and high estimates, respectively. Statistical manipulations were performed using the comprehensive meta-analysis computer program (Biosta, Englewood, NJ, USA).
1014
R. Elshazli, A. Settin / Immunobiology 220 (2015) 1012–1024
Fig. 1. Flow chart of studies of (A) PTPN22 rs2476601 and (B) STAT4 rs7574865 polymorphisms in the meta-analysis.
Evaluation of publication bias Due to the limitations of funnel plotting, which requires a range of studies of varying sizes involving subjective judgments, we evaluated publication bias through Egger’s linear regression test (Egger et al., 1997; Duval and Tweedie, 2000), which measures funnel plot asymmetry using a natural logarithm scale for odds ratios (ORs).
Results Meta-analysis of the association between the PTPN22 rs2476601 and RA Fifty-two relevant articles, which investigated the association between the PTPN22 rs2476601 polymorphism and RA susceptibility, were identified (Fig. 1A). Twenty-eight were excluded due to; previous meta-analysis studies (Tang et al., 2014; Song et al., 2013; Lee et al., 2012; Nong et al., 2011); data missing (Li et al., 2013; Ramirez et al., 2012; Viatte et al., 2012; Huang et al., 2012; Martín et al., 2011; Palomino-Morales et al., 2010; Briggs et al., 2010; Sfar et al., 2009; Gagnon et al., 2007); data overlap (Plant et al., 2010; Lee et al., 2009; Costenbader et al., 2008; Hinks et al., 2007; Lie et al., 2007); not a case–control study (Hopkins et al., 2014; Bayley et al., 2014; Davis et al., 2014; Harrison et al., 2012; Lamana et al., 2012; Salliot et al., 2011; Goëb et al., 2008; Feitsma et al., 2007); genotype distribution in controls was not consistent with HWE (Zhebrun et al., 2011) and the use of the same control sample (Majorczyk et al., 2007). Thus, twenty-four studies met the inclusion criteria (Fodil et al., 2014; Ferreiro-Iglesias et al., 2014; Salama et al., 2014; Salesi et al., 2014; Hashemi et al., 2013; Pradhan et al., 2012; Torres-Carrillo et al., 2012; Ates et al., 2011; El-Gabalawy et al., 2011; Eliopoulos et al., 2011; Mihailova et al., 2011; Rodríguez-Rodríguez et al., 2011; Totaro et al., 2011; Majorczyk et al., 2010; Chabchoub et al., 2009; Farago et al., 2009; Sahin et al., 2009; Stark et al., 2009; Eike et al., 2008; Naseem et al., 2008; Kokkonen et al., 2007; Viken et al., 2007; Mastana et al., 2007; Wesoly et al., 2007). We treated data of each study as a separate study, however, one of these eligible studies contained data on six different groups (Rodríguez-Rodríguez et al., 2011), and these groups were analyzed independently. These 24 articles have included 29 case–control studies involving 14,725 RA patients and 14,444 controls (Table 1). These encompassed various populations including: eighteen European, three African, two Asian, two Iranian, one Mexican, two Turkish and one Canadian.
However, because the sample populations were relatively inadequate in these populations, ethnicity-specific meta-analysis was conducted as European, Asian and African populations. Different genotyping methods were identified including RFLP, ARMS, TaqMan, Mass Array and allelic specific kinetic. Characteristics of PTPN22 rs2476601 studies included in the meta-analysis are listed in Table 1. Meta-analysis showed an association between RA and the PTPN22 T allele in all pooled subjects (OR = 1.745, 95% CI = 1.576–1.932, P < 0.001) (Table 2, Fig. 2). Regarding genotypes, there was a positive association in the dominant mode (CT+TT vs. CC, OR = 1.794, 95% CI = 1.604–2.006, P < 0.001); recessive model (TT vs. CC+CT; OR = 2.607, 95% CI = 2.079–3.270, P < 0.001); overdominant model (CT vs. CC+TT; OR = 1.566, 95% CI = 1.470–1.669, P < 0.001); heterozygote contrast (CT vs. CC; OR = 1.602, 95% CI = 1.504–1.707, P < 0.001) and homozygote contrast (TT vs. CC; OR = 2.924, 95% CI = 2.331–3.669, P < 0.001) (Table 2). After stratification by ethnicity, the PTPN22 T allele was significantly associated with RA in European (OR = 1.646, 95% CI = 1.554–1.743, P < 0.001); African (OR = 3.685, 95% CI = 1.020–13.312, P = 0.047) and Asian (OR = 3.573, 95% CI = 1.534–8.323, P = 0.003). Also, the PTPN22 CT+TT genotypes were significantly associated with RA in European (OR = 1.683, 95% CI = 1.579–1.793, P < 0.001); African (OR = 4.124, 95% CI = 1.008–16.873, P = 0.049) and Asian (OR = 3.587, 95% CI = 1.521–8.461, P = 0.004). Nonetheless, the PTPN22 TT genotype was significantly associated with RA in European (OR = 2.595, 95% CI = 2.061–3.266, P < 0.001), but not in African and Asian (OR = 3.918, 95% CI = 0.611–25.108, P = 0.150 and OR = 3.350, 95% CI = 0.135–82.969, P = 0.460, respectively) (Table 2).
Meta-analysis of the association between the STAT4 rs7574865 and RA Another thirty-five relevant studies, which investigated the association between the STAT4 rs7574865 and RA susceptibility, were identified (Fig. 1B). Twenty-two were excluded due to; metaanalysis studies (Jiang et al., 2014; Gu et al., 2014; Zheng et al., 2013; Tong et al., 2013; Liang et al., 2012; Lee et al., 2010; Ji et al., 2010); data missing (Seddighzadeh et al., 2012; Viatte et al., 2012; Ben Hamad et al., 2011; El-Gabalawy et al., 2011; Plant et al., 2010; Kelley et al., 2010; Suarez-Gestal et al., 2009; Daha et al., 2009; Remmers et al., 2007); no case–control study (Davis et al., 2014; Zervou et al., 2013; Lamana et al., 2012; Rodríguez-Rodríguez et al.,
Table 1 Characteristics of the studies of PTPN22 rs2476601 polymorphism included in the meta-analysis. Year
Country
Ethnicity
Fodil Ferreiro-Iglesias Salama Salesi Hashemi Pradhan Torres-Carrillo Ates El-Gabalawy Eliopoulos Mihailova Rodriruez-Rodriruez-1 Rodriruez-Rodriruez-2 Rodriruez-Rodriruez-3 Rodriruez-Rodriruez-4 Rodriruez-Rodriruez-5 Rodriruez-Rodriruez-6 Totaro Majorczyk Chabchoub Farago Sahin Stark Eike Naseem Kokkonen Vilken Mastana Wesoly
2014 2014 2014 2014 2013 2012 2012 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2010 2009 2009 2009 2009 2008 2008 2007 2007 2007 2007
Algeria Spain Egypt Iran Iran India Mexico Turkey Canada Greece Latvia Spain New Zealand UK Norway Netherlands Germany Italy Poland Tunisia Hungary Turkey Slovakia Norway UK Sweden Norway South Asian Netherlands
African European African Iranian Iranian Asian Mexican Turkish Canadian European European European European European European European European European European African European Turkish European European European European European Asian European
Numbers RA
Controls
110 1743 112 100 120 130 309 323 490 378 94 1689 735 664 944 931 183 396 371 150 399 167 514 686 832 504 861 129 661
197 1650 122 100 120 100 347 426 333 430 238 1855 555 598 1100 834 279 477 543 236 107 177 302 952 412 970 557 143 284
Genotyping method
Antibodies status
TaqMan Single base extension PCR-RFLP PCR-RFLP T-ARMS-PCR PCR-RFLP PCR-RFLP PCR-RFLP TaqMan PCR-RFLP TaqMan TaqMan TaqMan TaqMan TaqMan TaqMan TaqMan PCR-RFLP PCR-RFLP PCR-RFLP PCR-RFLP PCR-RFLP TaqMan TaqMan Sequenom MassArray TaqMan TaqMan PCR-RFLP Allelic specific Kinetic-PCR
NA NA Anti-CCP/RF NA NA Anti-CCP/RF Anti-CCP Anti-CCP/RF NA NA NA NA NA NA NA NA NA NA RF NA Anti-CCP/RF NA NA NA Anti-CCP/RF Anti-CCP/RF NA NA RF
RA/controls (Genotype)
RA/controls (Allele)
CC
CT
TT
C
T
76/189 1383/1405 73/112 89/92 102/116 118/96 278/333 296/403 457/305 353/405 53/193 1364/1613 518/448 482/483 662/858 651/683 122/220 352/443 242/425 143/224 241/85 156/168 356/239 473/761 641/342 338/761 597/438 115/140 478/229
33/7 325/235 36/9 11/8 18/4 12/4 28/14 27/23 32/27 25/25 36/42 294/230 192/104 174/105 261/230 253/142 51/56 43/33 112/111 7/12 121/20 11/9 144/61 197/181 178/66 151/196 244/111 13/3 162/55
1/0 35/10 3/1 0/0 0/0 0/0 3/0 0/0 1/1 0/0 5/3 31/12 25/3 8/10 21/12 27/9 10/3 1/1 17/7 0/0 37/2 0/0 14/2 16/10 13/4 15/13 20/8 1/0 21/0
185/385 3091/3045 182/233 189/192 222/236 248/196 584/680 619/829 946/637 731/835 142/428 3022/3456 1228/1000 1138/1071 1585/1946 1555/1508 295/496 747/919 596/961 293/460 603/190 323/345 856/539 1143/1703 1460/750 827/1718 1438/987 243/283 1118/513
35/7 395/255 42/11 11/8 18/4 12/4 34/14 27/23 34/29 25/25 46/48 356/254 242/110 190/125 303/254 307/160 71/62 45/35 146/125 7/12 195/24 11/9 172/65 229/201 204/74 181/222 284/127 15/3 204/55
HWE (P)
0.79 0.96 0.11 0.68 0.85 0.84 0.70 0.56 0.63 0.53 0.68 0.23 0.24 0.13 0.43 0.59 0.79 0.64 0.93 0.69 0.52 0.73 0.37 0.83 0.68 0.92 0.75 0.89 0.07
R. Elshazli, A. Settin / Immunobiology 220 (2015) 1012–1024
First author [Ref]
PCR-RFLP: polymerase chain reaction-restriction fragment length polymorphism; T-ARMS-PCR: tetra amplification refractory mutation system-polymerase chain reaction; anti-CCP: anti-cyclic citrullinated peptide; RF: rheumatoid factor; HWE: Hardy–Weinberg equilibrium.
1015
1016
Table 2 Meta-analysis of the association between PTPN22 rs2476601 polymorphism and RA susceptibility. Comparison
Population
No. of studies
Sample size
Test of association
Test of heterogeneity
Publication bias
RA
Control
OR
95% CI
P-value
Model
Q test
P-value
I2 (%)
P-value (Egger’s)
Overall European African Asian
29 18 3 2
29,450 25,170 744 518
28,888 24,286 1110 486
1.745 1.646 3.685 3.573
1.576–1.932 1.554–1.743 1.020–13.312 1.534–8.323
Contents lists available at ScienceDirect
Immunobiology journal homepage: www.elsevier.com/locate/imbio
Association of PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with rheumatoid arthritis: A meta-analysis update Rami Elshazli a,∗ , Ahmad Settin b a b
Department of Biochemistry, Faculty of Science, Tanta University, Tanta, Egypt Genetics Unit, Children Hospital, Mansoura University, Mansoura, Egypt
a r t i c l e
i n f o
Article history: Received 14 March 2015 Accepted 20 April 2015 Available online 28 April 2015 Keywords: Meta-analysis Rheumatoid arthritis Polymorphism PTPN22 STAT4 RF Anti-CCP
a b s t r a c t Background: Rheumatoid arthritis (RA) is a common autoimmune disease with a complex genetic background. The genes encoding protein tyrosine phosphatase non-receptor type 22 (PTPN22) and signal transducer and activator of transcription 4 (STAT4) have been reported to be associated with RA in several ethnic populations. Objectives: This work aims to assess the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with RA susceptibility through an updated meta-analysis of available case–control studies. Methods: A literature search of all relevant studies published from January 2007 up to December 2014 was conducted using Pubmed and Science Direct databases. The observed studies that were related to an association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with RA susceptibility were identified. Meta-analyis of the pooled and stratified data was done and assessed using varied genetic models. Results: Thirty-seven case–control studies with a total of 47 comparisons (29 for PTPN22 rs2476601 polymorphism and 18 for STAT4 rs7574865 polymorphism) met our inclusion criteria. The meta-analysis showed an association between PTPN22 T allele, CT+TT and TT genotypes with RA susceptibility. Furthermore, The meta-analysis showed an association between STAT4 T allele, GT+TT and TT genotypes with RA susceptibility. Stratification of RA patients according to ethnic groups showed that PTPN22 T allele, CT+TT genotypes, STAT4 T allele and STAT4 GT+TT were significantly associated with RA in European, Asian, African subjects, while PTPN22 TT genotype was significantly associated with RA in European but not in Asian and African subjects and STAT4 TT genotype was significantly associated with RA in European and Asian but not in African subject. A subgroup analysis according to the presence or absence of rheumatoid factor (RF) and anti-cyclic citrullinated peptide (anti-CCP) antibodies revealed that the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with RA susceptibility may not be dependent on RF and anti-CCP antibodies. Conclusions: Our meta-analysis demonstrated that PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms confers susceptibility to RA in total subjects and in major ethnic groups. The association may not be dependent on RF and anti-CCP antibodies. © 2015 Elsevier GmbH. All rights reserved.
Introduction Rheumatoid arthritis (RA) is a chronic inflammatory disorder approximately affecting 1% of the adult population worldwide. The disease is characterized by an inflammation of the synovial tissue of multiple joints leading to pain, deformities and a reduced quality of life (Oliver et al., 2006; Firestein, 2003). Although the etiology
∗ Corresponding author. Tel.: +20 1064620110. E-mail address: [email protected] (R. Elshazli). http://dx.doi.org/10.1016/j.imbio.2015.04.003 0171-2985/© 2015 Elsevier GmbH. All rights reserved.
of RA is unknown, both genetic and environmental factors have been shown to play a role in its development. Genetic factors were thought to be responsible for up to 50–60% of the predisposition to RA. The genetic background of RA is complex and is likely involving multiple genes which encode proteins with significant functions in the regulation of immune response (Ikari et al., 2006; Goëb et al., 2008). Human leukocyte antigen (HLA) class II genes were portrayed as an important factors linked to RA, but they account for only one-third of genetic susceptibility, therefore non-HLA genes should be also considered (Urayama et al., 2013; Deighton et al., 1989).
R. Elshazli, A. Settin / Immunobiology 220 (2015) 1012–1024
Two major genes have been contributed of the susceptibility with RA, these genes include the protein tyrosine phosphatase nonreceptor type 22 (PTPN22) and the signal transducer and activator of transcription 4 (STAT4). PTPN22 gene, is located on chromosome 1p13, and encodes a lymphoid specific phosphatase (LYP), which is important in the negative control of T cell activation and development (Cohen et al., 1999; Siminovitch, 2004; Totaro et al., 2011). The PTPN22 +1858 C>T SNP changes the amino acid at position 620 from an arginine (R) to a tryptophan (W) and disrupts the interaction between LYP and Csk and thus inhibits complex formation and suppresses T-cell activation, hence, is thought to predispose to multiple autoimmune diseases (Begovich et al., 2004; Bottini et al., 2004). On the other hand, STAT4 gene transmits signals induced by interleukin-12 (IL-12), interleukin-23 (IL-23) and interferon-␥ (IFN-␥), which are key cytokines and play important roles in the development of autoimmune diseases (Frucht et al., 2000; Mathur et al., 2007; Watford et al., 2004). The STAT4 gene maps to chromosome 2q33 and encodes a transcription factor, which plays pivotal roles in the differentiation and proliferation of both T helper 1 (Th1) and T helper 17 (Th17) cells (Watford et al., 2004). Since Th1 and Th17 lineages are crucial effectors in chronic inflammatory disorders, STAT4 gene may play an important role in the pathogenesis of autoimmune diseases in different ethnic populations, such as RA (Zhao et al., 2013; Tong et al., 2013; Mohamed et al., 2012; Settin et al., 2014). Recently, a large number of studies have explored the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with the susceptibility to RA in different ethnic populations (Song et al., 2013; Salama et al., 2014; Torres-Carrillo et al., 2012; Farago et al., 2009; Gu et al., 2014; Ben Hamad et al., 2011; Tong et al., 2013; Shen et al., 2013; Settin et al., 2014). However, some results from the previous studies were inconsistent which might be due to the sample size used, ethnic differences in allele frequencies, or publication bias (Song et al., 2013; TorresCarrillo et al., 2012; Gu et al., 2014; Settin et al., 2014). In the present study, we have done a meta-analysis to check for the contribution of PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms to RA susceptibility in different populations. We also examined their association with RA subtypes in terms of a positive or negative anti-cyclic citrullinated peptide (anti-CCP) or rheumatoid factor (RF). Methods Study identification and selection This meta-analysis followed the preferred reporting items for systematic reviews and meta-analyses (PRISMA) criteria (Moher et al., 2010). A literature search was conducted using Pubmed and Science Direct citation databases to identify articles published from January 2007 up to December 2014 that examined the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with the susceptibility to RA. Combinations of keywords such as: “Rheumatoid arthritis”, “PTPN22”, “STAT4”, “polymorphism” and “RA” were entered as both Medical Subject Headings (MeSHs) and text words without any restrictions on language or country. Inclusion and exclusion criteria Data were collected from the full-published paper, excluding any meeting or conference abstract. Inclusion criteria were defined as follows: (a) articles evaluating the association between PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms with RA susceptibility, (b) the design is a case–controlled study based on unrelated individuals, and (c) sufficient data (genotype distributions for cases
1013
and controls) available to estimate an odds ratio (OR) with its 95% confidence interval (CI). Studies were excluded if one of the following existed: (a) studies containing overlapping data; (b) studies in which the genotype frequencies or numbers could not be ascertained; (c) studies in which family members were studied, because their analysis is based on linkage considerations; (d) reviews and abstracts; and (e) studies in which the genotype distribution in controls was not consistent with Hardy–Weinberg equilibrium (HWE), because deviation from HWE among controls might be due to a biased control-selection or genotyping errors. If more than one article were published by the same authors using the same sample series, studies with the largest size of samples or the recently published ones were included. Data extraction The following information was extracted: (1) name of the first author; (2) year of publication; (3) country of origin; (4) ethnicity of the studied population; (5) the numbers of cases and controls of polymorphisms of PTPN22 and STAT4; (6) genotyping method; (7) Hardy–Weinberg equilibrium (HWE) of controls and (8) antibodies status (AS). Statistical analysis The Hardy–Weinberg equilibrium (HWE) was examined in control groups by Fisher’s exact test. If the study was found not to be in HWE with P value less than 0.05, it was considered to be disequilibrium. Allele frequencies of the PTPN22 rs2476601 and STAT4 rs7574865 polymorphisms in each of the studies were determined using the allele counting method. Allelic effect contrast was examined for the minor allele vs. the common allele. The genetic models evaluated for pooled ORs of these two polymorphisms have included allelic contrast, recessive models, dominant models, overdominant models, homozygote contrast and heterozygote contrast. Furthermore meta-analyses were performed once using the total data set, and separately for the subgroups of different ethnicities, RF and anti-CCP status (Tables 2, 4 and 5). To evaluate the strength of association, the pooled odd ratios (ORs) and their 95% confidence intervals (CIs) were determined for each study, and within- and between-study heterogeneity were assessed using Cochran’s Q statistic. The heterogeneity test was used to assess the probability of the null hypothesis that all studies were evaluating the same effect. The random-effects model was used for meta-analysis when a significant Q statistic (P < 0.10) indicated heterogeneity across studies, while the fixed-effect model was used when heterogeneity was not indicated. The fixed-effect model assumes that genetic factors have similar effects on disease susceptibility across all studies and that observed variations between studies are caused by chance alone (Davey Smith and Egger, 1997). The random-effect model assumes that studies show substantial diversity, and assesses both within study sampling errors and between study variances (DerSimonian and Laird, 1986). When study groups are homogenous, the two models are similar, but if this is not the case, the random effects model usually provides wider CIs than the fixed effects model. The random effects model is best used in the presence of significant between study heterogeneity (DerSimonian and Laird, 1986). We quantified the effect of heterogeneity by using the recently developed I2 measure, where I2 = 100% x (Q − df)/Q (Higgins and Thompson, 2002). The I2 measure ranges between 0% and 100%, and it represents the proportion of inter-study variability attributable to heterogeneity rather than chance. I2 values of 25%, 50%, and 75% were defined as low, moderate, and high estimates, respectively. Statistical manipulations were performed using the comprehensive meta-analysis computer program (Biosta, Englewood, NJ, USA).
1014
R. Elshazli, A. Settin / Immunobiology 220 (2015) 1012–1024
Fig. 1. Flow chart of studies of (A) PTPN22 rs2476601 and (B) STAT4 rs7574865 polymorphisms in the meta-analysis.
Evaluation of publication bias Due to the limitations of funnel plotting, which requires a range of studies of varying sizes involving subjective judgments, we evaluated publication bias through Egger’s linear regression test (Egger et al., 1997; Duval and Tweedie, 2000), which measures funnel plot asymmetry using a natural logarithm scale for odds ratios (ORs).
Results Meta-analysis of the association between the PTPN22 rs2476601 and RA Fifty-two relevant articles, which investigated the association between the PTPN22 rs2476601 polymorphism and RA susceptibility, were identified (Fig. 1A). Twenty-eight were excluded due to; previous meta-analysis studies (Tang et al., 2014; Song et al., 2013; Lee et al., 2012; Nong et al., 2011); data missing (Li et al., 2013; Ramirez et al., 2012; Viatte et al., 2012; Huang et al., 2012; Martín et al., 2011; Palomino-Morales et al., 2010; Briggs et al., 2010; Sfar et al., 2009; Gagnon et al., 2007); data overlap (Plant et al., 2010; Lee et al., 2009; Costenbader et al., 2008; Hinks et al., 2007; Lie et al., 2007); not a case–control study (Hopkins et al., 2014; Bayley et al., 2014; Davis et al., 2014; Harrison et al., 2012; Lamana et al., 2012; Salliot et al., 2011; Goëb et al., 2008; Feitsma et al., 2007); genotype distribution in controls was not consistent with HWE (Zhebrun et al., 2011) and the use of the same control sample (Majorczyk et al., 2007). Thus, twenty-four studies met the inclusion criteria (Fodil et al., 2014; Ferreiro-Iglesias et al., 2014; Salama et al., 2014; Salesi et al., 2014; Hashemi et al., 2013; Pradhan et al., 2012; Torres-Carrillo et al., 2012; Ates et al., 2011; El-Gabalawy et al., 2011; Eliopoulos et al., 2011; Mihailova et al., 2011; Rodríguez-Rodríguez et al., 2011; Totaro et al., 2011; Majorczyk et al., 2010; Chabchoub et al., 2009; Farago et al., 2009; Sahin et al., 2009; Stark et al., 2009; Eike et al., 2008; Naseem et al., 2008; Kokkonen et al., 2007; Viken et al., 2007; Mastana et al., 2007; Wesoly et al., 2007). We treated data of each study as a separate study, however, one of these eligible studies contained data on six different groups (Rodríguez-Rodríguez et al., 2011), and these groups were analyzed independently. These 24 articles have included 29 case–control studies involving 14,725 RA patients and 14,444 controls (Table 1). These encompassed various populations including: eighteen European, three African, two Asian, two Iranian, one Mexican, two Turkish and one Canadian.
However, because the sample populations were relatively inadequate in these populations, ethnicity-specific meta-analysis was conducted as European, Asian and African populations. Different genotyping methods were identified including RFLP, ARMS, TaqMan, Mass Array and allelic specific kinetic. Characteristics of PTPN22 rs2476601 studies included in the meta-analysis are listed in Table 1. Meta-analysis showed an association between RA and the PTPN22 T allele in all pooled subjects (OR = 1.745, 95% CI = 1.576–1.932, P < 0.001) (Table 2, Fig. 2). Regarding genotypes, there was a positive association in the dominant mode (CT+TT vs. CC, OR = 1.794, 95% CI = 1.604–2.006, P < 0.001); recessive model (TT vs. CC+CT; OR = 2.607, 95% CI = 2.079–3.270, P < 0.001); overdominant model (CT vs. CC+TT; OR = 1.566, 95% CI = 1.470–1.669, P < 0.001); heterozygote contrast (CT vs. CC; OR = 1.602, 95% CI = 1.504–1.707, P < 0.001) and homozygote contrast (TT vs. CC; OR = 2.924, 95% CI = 2.331–3.669, P < 0.001) (Table 2). After stratification by ethnicity, the PTPN22 T allele was significantly associated with RA in European (OR = 1.646, 95% CI = 1.554–1.743, P < 0.001); African (OR = 3.685, 95% CI = 1.020–13.312, P = 0.047) and Asian (OR = 3.573, 95% CI = 1.534–8.323, P = 0.003). Also, the PTPN22 CT+TT genotypes were significantly associated with RA in European (OR = 1.683, 95% CI = 1.579–1.793, P < 0.001); African (OR = 4.124, 95% CI = 1.008–16.873, P = 0.049) and Asian (OR = 3.587, 95% CI = 1.521–8.461, P = 0.004). Nonetheless, the PTPN22 TT genotype was significantly associated with RA in European (OR = 2.595, 95% CI = 2.061–3.266, P < 0.001), but not in African and Asian (OR = 3.918, 95% CI = 0.611–25.108, P = 0.150 and OR = 3.350, 95% CI = 0.135–82.969, P = 0.460, respectively) (Table 2).
Meta-analysis of the association between the STAT4 rs7574865 and RA Another thirty-five relevant studies, which investigated the association between the STAT4 rs7574865 and RA susceptibility, were identified (Fig. 1B). Twenty-two were excluded due to; metaanalysis studies (Jiang et al., 2014; Gu et al., 2014; Zheng et al., 2013; Tong et al., 2013; Liang et al., 2012; Lee et al., 2010; Ji et al., 2010); data missing (Seddighzadeh et al., 2012; Viatte et al., 2012; Ben Hamad et al., 2011; El-Gabalawy et al., 2011; Plant et al., 2010; Kelley et al., 2010; Suarez-Gestal et al., 2009; Daha et al., 2009; Remmers et al., 2007); no case–control study (Davis et al., 2014; Zervou et al., 2013; Lamana et al., 2012; Rodríguez-Rodríguez et al.,
Table 1 Characteristics of the studies of PTPN22 rs2476601 polymorphism included in the meta-analysis. Year
Country
Ethnicity
Fodil Ferreiro-Iglesias Salama Salesi Hashemi Pradhan Torres-Carrillo Ates El-Gabalawy Eliopoulos Mihailova Rodriruez-Rodriruez-1 Rodriruez-Rodriruez-2 Rodriruez-Rodriruez-3 Rodriruez-Rodriruez-4 Rodriruez-Rodriruez-5 Rodriruez-Rodriruez-6 Totaro Majorczyk Chabchoub Farago Sahin Stark Eike Naseem Kokkonen Vilken Mastana Wesoly
2014 2014 2014 2014 2013 2012 2012 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2010 2009 2009 2009 2009 2008 2008 2007 2007 2007 2007
Algeria Spain Egypt Iran Iran India Mexico Turkey Canada Greece Latvia Spain New Zealand UK Norway Netherlands Germany Italy Poland Tunisia Hungary Turkey Slovakia Norway UK Sweden Norway South Asian Netherlands
African European African Iranian Iranian Asian Mexican Turkish Canadian European European European European European European European European European European African European Turkish European European European European European Asian European
Numbers RA
Controls
110 1743 112 100 120 130 309 323 490 378 94 1689 735 664 944 931 183 396 371 150 399 167 514 686 832 504 861 129 661
197 1650 122 100 120 100 347 426 333 430 238 1855 555 598 1100 834 279 477 543 236 107 177 302 952 412 970 557 143 284
Genotyping method
Antibodies status
TaqMan Single base extension PCR-RFLP PCR-RFLP T-ARMS-PCR PCR-RFLP PCR-RFLP PCR-RFLP TaqMan PCR-RFLP TaqMan TaqMan TaqMan TaqMan TaqMan TaqMan TaqMan PCR-RFLP PCR-RFLP PCR-RFLP PCR-RFLP PCR-RFLP TaqMan TaqMan Sequenom MassArray TaqMan TaqMan PCR-RFLP Allelic specific Kinetic-PCR
NA NA Anti-CCP/RF NA NA Anti-CCP/RF Anti-CCP Anti-CCP/RF NA NA NA NA NA NA NA NA NA NA RF NA Anti-CCP/RF NA NA NA Anti-CCP/RF Anti-CCP/RF NA NA RF
RA/controls (Genotype)
RA/controls (Allele)
CC
CT
TT
C
T
76/189 1383/1405 73/112 89/92 102/116 118/96 278/333 296/403 457/305 353/405 53/193 1364/1613 518/448 482/483 662/858 651/683 122/220 352/443 242/425 143/224 241/85 156/168 356/239 473/761 641/342 338/761 597/438 115/140 478/229
33/7 325/235 36/9 11/8 18/4 12/4 28/14 27/23 32/27 25/25 36/42 294/230 192/104 174/105 261/230 253/142 51/56 43/33 112/111 7/12 121/20 11/9 144/61 197/181 178/66 151/196 244/111 13/3 162/55
1/0 35/10 3/1 0/0 0/0 0/0 3/0 0/0 1/1 0/0 5/3 31/12 25/3 8/10 21/12 27/9 10/3 1/1 17/7 0/0 37/2 0/0 14/2 16/10 13/4 15/13 20/8 1/0 21/0
185/385 3091/3045 182/233 189/192 222/236 248/196 584/680 619/829 946/637 731/835 142/428 3022/3456 1228/1000 1138/1071 1585/1946 1555/1508 295/496 747/919 596/961 293/460 603/190 323/345 856/539 1143/1703 1460/750 827/1718 1438/987 243/283 1118/513
35/7 395/255 42/11 11/8 18/4 12/4 34/14 27/23 34/29 25/25 46/48 356/254 242/110 190/125 303/254 307/160 71/62 45/35 146/125 7/12 195/24 11/9 172/65 229/201 204/74 181/222 284/127 15/3 204/55
HWE (P)
0.79 0.96 0.11 0.68 0.85 0.84 0.70 0.56 0.63 0.53 0.68 0.23 0.24 0.13 0.43 0.59 0.79 0.64 0.93 0.69 0.52 0.73 0.37 0.83 0.68 0.92 0.75 0.89 0.07
R. Elshazli, A. Settin / Immunobiology 220 (2015) 1012–1024
First author [Ref]
PCR-RFLP: polymerase chain reaction-restriction fragment length polymorphism; T-ARMS-PCR: tetra amplification refractory mutation system-polymerase chain reaction; anti-CCP: anti-cyclic citrullinated peptide; RF: rheumatoid factor; HWE: Hardy–Weinberg equilibrium.
1015
1016
Table 2 Meta-analysis of the association between PTPN22 rs2476601 polymorphism and RA susceptibility. Comparison
Population
No. of studies
Sample size
Test of association
Test of heterogeneity
Publication bias
RA
Control
OR
95% CI
P-value
Model
Q test
P-value
I2 (%)
P-value (Egger’s)
Overall European African Asian
29 18 3 2
29,450 25,170 744 518
28,888 24,286 1110 486
1.745 1.646 3.685 3.573
1.576–1.932 1.554–1.743 1.020–13.312 1.534–8.323
