Expression of Human PTPN22 Alleles

ORIGINAL ARTICLE Expression of human PTPN22 alleles

C Nielsen1, T Barington1, S Husby2 and ST Lillevang1 1Department of Clinical Immunology, Odense University Hospital, Odense, Denmark and 2Department of Pediatrics, Odense University Hospital, Odense, Denmark

Considering the female predominance in most of the autoimmune disorders that associate with the PTPN22 Trp620 variant and the complexity by which this variant influences immunologic tolerance, the objective of this study was to ascertain if the allele- specific expression of the disease-associated Arg620Trp polymorphism is affected by cis-acting or sex-specific trans-acting factor/s (e.g. sex-hormones). The use of the allele-specific transcript quantification of the Arg620Trp encoding 1858T polymorphism revealed no difference in the expression of the 1858C- and T-alleles in non-stimulated peripheral blood mononuclear cells (PBMCs) from non-pregnant female subjects, male subjects or pregnant female subjects in first or third trimester (P ¼ 0.70), respectively. While the transcription of PTPN22 in anti-CD3/anti-CD28 stimulated PBMCs increased fourfold (Po0.0001) and 13-fold (Po0.0001) after 48 and 72 h of activation, respectively, the expression of PTPN22 1858C- and T-alleles increased to the same extent (P ¼ 0.64). The present result essentially excludes such phenomena as a partial explanation for the female predominance in most of the autoimmune disorders that associate with the PTPN22 Trp620 variant. Genes and Immunity (2007) 8, 131–137. doi:10.1038/sj.gene.6364369; published online 18 January 2007

Keywords: PTPN22; sex-specific; allele-specific expression; female predominance

Introduction enriched tyrosine phosphatase) on a C57/Bl6 back- ground show a lowering of the threshold for T-cell The lymphoid-specific phosphatase (LYP) encoded by receptor signaling and an enhanced expansion of effector PTPN22 (protein tyrosine phosphatase non-receptor type and memory T-cell population and increased serum 22; 1p13.3-p13.1) plays a very central role in maintaining immunoglobulin levels.10 The PTPN22 1858T changes immunologic tolerance and preventing autoimmunity. codon 620 from arginine into tryptophan and the effect of Since the first published association between a PTPN22 this amino acid change on LYP function has not been Trp620 variant (PTPN22 1858C to T changes codon 620 entirely defined but one consequence is to disrupt the from arginine into tryptophan) and type 1 diabetes,1 it interaction with an intracellular C-terminal Src tyrosine has become apparent that the PTPN22 Trp620 variant kinase (Csk),1,3 known to be an important suppressor is the strongest common genetic risk factor for human of kinases that mediate T-cell activation. The disease- autoimmunity outside the major histocompatibility predisposing PTPN22 Trp620 is located in the N- complex, with association with a wide variety of terminal-SH3 binding domain of the protein, and the autoimmune diseases that have a prominent humoral mutant protein fails to bind the SH3 domain of Csk.1 component (systemic lupus erythematosus: e.g. Kyogoku Unexpectively, the human PTPN22 Trp620 disease- et al.;2 rheumatoid arthritis: e.g. Begovich et al.;3 Graves’ associated variant has recently been found to be a gain disease: e.g. Velaga et al.;4 Addison’s disease: e.g. Velaga of function variant, with increased catalytic activity et al.;4 Hashimoto’s thyroiditis5). It is intriguing that the compared to the nonassociated variant encoded by PTPN22 Trp620 variant apparently is a gain-of-function PTPN22 Arg620.6 It currently remains unclear how this variant,6 but the mechanism by which the PTPN22 potential increase in the inhibition of TCR activation Trp620 variant exerts the disease-promoting effect has might lead to autoimmunity. A current hypothesis yet to be established. postulates that the threshold for TCR signaling during LYP is specifically expressed in lymphocytes3,7 and by thymic development is elevated in individuals expres- dephosphorylating and inactivating TCR-associated sing the mutant LYP protein, resulting in defective kinases Lck, Fyn, and ZAP-70, LYP is one of the most negative selection and survival of autoreactive T cells powerful inhibitors of T-cell activation.8,9 Mice knockout that would normally have undergone apoptosis.11 of the murine homolog of LYP called PEP (PEST domain- Further, an altered LYP function in peripheral CD4 þ CD25 þ T regulatory cells, making them less potent in suppressing immune responses against autoantigens, Correspondence: C Nielsen, Department of Clinical Immunology, has been suggested.11 Odense University Hospital, Sdr. Boulevard 29, 5000 Odense C, Owing to this apparent complexity, we aimed to Denmark. E-mail: [email protected] investigate whether cis-acting factors existed that inter- Received 18 October 2006; revised 27 November 2006; accepted 28 fered with the expression of the PTPN22 Trp620- November 2006; published online 18 January 2007 encoding variant and thus could explain the association Expression of human PTPN22 alleles C Nielsen et al 132 of this allele with autoimmune diseases. Moreover, we wanted to see if sex related trans factors could influence transcription of this allele in an allele-specific way considering the female predominance in most of the autoimmune disorders that associate with the PTPN22 Trp620 variant and autoimmune diseases in general.12 Several recent studies have indicated that a large proportion of human non-imprinted autosomal genes are subject to variable cis-acting influences resulting in differences in the allele-specific expression of the genes.13–16 Quantitative differences in allelic gene expression between individuals may contribute to the genetic basis for phenotypic differences and may play an important role in the development of complex diseases. We applied a quantitative method (fluorescent PCR- RFLP analysis) of allele discrimination to DNA and cDNA samples purified from peripheral blood mononuclear cells (PBMCs) obtained from healthy male subjects, female subjects, female subjects in the first trimester of pregnancy and female subjects in the third trimester of pregnancy, heterozygous for C1858T. Addi- tional variations in allelic expression may manifest in a cell type or state-specific manner,13 and as T cells play a role at least in the initiation phase of the autoimmune disorders that associate with the PTPN22 R620W polymorphism, we further investigated the allele-specific expression over time in activated T cells from heterozygous individuals.

Figure 1 Fragment length analysis by capillary electrophoresis of PTPN22 PCR products. TET- and FAM-coupled fragments are Results shown in black and blue, respectively. Only FAM-coupled fragments at 218 and 176 bp appear in the shown length interval and size Fragment length analysis by capillary electrophoresis markers are not visualized for clarity reasons. (a) Non-cleaved FAM- Following fragment length analysis by capillary electro- labelled and NED-labelled PCR products, (b) cleavage reaction phoresis, samples from individuals heterozygous for products of a C1858C homozygous sample, (c) cleavage reaction C1858T showed three FAM-coupled fragments of 218, products of a T1858T homozygous sample and (d) cleavage reaction 176 and 42 bp, whereas those homozygous for CC products of a C1858T heterozygous sample. Absence of a NED- coupled 148-bp fragment and presence of a NED-coupled 104 bp showed two FAM-coupled fragments of 176 and 42 bp fragment demonstrates complete cutting by the RsaI endonuclease. (Figure 1). Visualization of both FAM- and NED-coupled fragments was allowed, and the absence of a NED- coupled 148-bp fragment and presence of a NED- way Repeated Measures ANOVA, Figure 2), indicating coupled 104 bp fragment demonstrated complete cutting the absence of sex-specific trans-acting factor/s (e.g. sex- by the RsaI endonuclease throughout the study. hormones) affecting the expression of the PTPN22 Arg620Trp-encoding allele relative to the wild-type allele Sex-specific relative allelic expression in non-stimulated PBMCs. For every ANOVA, P-values Genomic DNA samples did not deviate from the were Bonferroni adjusted for the number of comparisons. expected equimolar ratio of alleles, indicating that the Repeat assays showed good reproducibility of indivi- present assay amplify and visualize the two alleles with dual cDNA allelic ratios with an average s.d. of 0.03. the same area, and rejecting that putative gene dosage effects or other biases influenced the measured allelic Regulation of PTPN22 mRNA and relative allelic expression ratio in cDNA samples (P ¼ 0.21; Two-way Repeated in stimulated PBMCs over time Measures analysis of variance (ANOVA), Figure 2). The mRNA Induction index of PTPN22 in PBMCs Therefore, no correction of cDNA peak height in each remained relatively unchanged until 48 h of anti-CD3 individual sample was necessary. and anti-CD28 stimulation, where the mRNA expression There was no overall effect of DNA type (genomic of PTPN22 was upregulated approximately four times DNA and cDNA) on the relative expression of the T- compared to non-stimulated cells (Po0.0001; One-way allele in PBMCs (P ¼ 0.08; Two-way Repeated Measures Repeated Measures ANOVA, Figure 3a). Following 72 h ANOVA, Figure 2), suggesting that the PTPN22 locus is of stimulation, the mRNA expression of PTPN22 was unlikely to contain common polymorphisms or epige- further upregulated compared to 48 h of stimulation netic modification that alter PTPN22 mRNA levels in (Po0.0001) and reached an approximately 13 times non-stimulated PBMCs, assayed by the level of allele- higher level compared to non-stimulated PBMCs specific expression of the PTPN22 Arg620Trp poly- (Po0.0001; One-way Repeated Measures ANOVA, morphism. Figure 3a). The level of DNA contamination was low as Nor was there any effect of sex and pregnancy status no-reverse transcription (NRT) controls always resulted

on the relative expression of the T-allele (P ¼ 0.70; Two- in CT values higher than the 40 cycles detected in the

Genes and Immunity Expression of human PTPN22 alleles C Nielsen et al 133

Figure 3 (a) Changes in the mRNA Induction index of PTPN22 in PBMCs stimulated with anti-CD3 and anti-CD28 over time (n ¼ 10). Data are shown as mean7s.d. and values with shared letters are not significantly different (P40.05); (b) allele-specific expression assayed at the PTPN22 1858C/T SNP in the same anti-CD3 and anti-CD28 over time. Data are expressed (n ¼ 10) as ratio of T:C alleles. The data are the peak area averages of two measurements for each individual sample.

Discussion Considering the female predominance in most of the autoimmune disorders that associate with the PTPN22 Trp620 variant and the apparent and not well under- stood complexity by which this variant influences the immunologic tolerance, we found it relevant to investigate the existence of cis-acting or sex-specific trans-acting factor/s (e.g. sex-hormones) affecting the allele-specific expression of the PTPN22 Arg620Trp polymorphism. The expression of genes conferring susceptibility to autoimmune disease may be regulated by sex hormones, and this sex-specific regulation could affect transcript synthesis or stability at the allele-specific level. The molecular basis for transcriptional regulation of gene expression is the binding of trans-acting proteins (transcription factors) to cis-acting sequences (binding sites). Polymorphic cis-acting sequences affect RNA Figure 2 Comparison between genomic DNA and cDNA ratios transcript synthesis or stability, or the translational assayed at the PTPN22 1858C/T SNP. Data are expressed (n ¼ 10– 12) as ratio of T:C alleles. The data are the peak area averages of two efficiency and protein stability in an allele-specific measurements for each individual sample. manner and are situated close to the gene(s) that they regulate.17 Several recent studies have indicated that a large proportion of human non-imprinted autosomal reverse transcription-polymerase chain reaction (RT- genes are subject to variable cis-acting influences on their PCR). expression, resulting in differences in the allele-specific There was no difference in the relative allelic expres- expression of the genes.13–16 Further, alleles with altered sion of the C1858T single nucleotide polymorphism expression have been associated to the polymorphism (SNP) between PBMCs stimulated with anti-CD3 and originally linked to human disease,15,18,19 and we anti-CD28 for 0, 6, 24, 48 and 72 h, respectively (P ¼ 0.64; speculated whether the PTPN22 gene also is subject to One-way Repeated Measures ANOVA; Figure 3b). cis-acting influence on its expression.

Genes and Immunity Expression of human PTPN22 alleles C Nielsen et al 134 The predominance of autoimmune disease among Several recent studies inbred mouse strain,22 chicken,23 women has focused attention on the role of sex human fetuses24 suggest that allelic expression differ- hormones, mainly estrogen, progesterone and testoster- ences in non-imprinted autosomal genes occur at a one, as primary mediators of the sex differences.20 One tissue-specific level, suggesting that the transcription way estrogens exert their biological effect is through factors that bind to cis-acting elements may vary in nuclear estrogen receptors (ERs), which interact with abundance or in ability to function in different tissues. promoter elements to regulate gene transcription.21 Theoretically, sex-specific differences in the allele-specific Although no potential targets of ER in the promoter of regulation in the selection of thymocytes during thymic PTPN22 were found, the existence of putative sex- differentiation, could lead to higher risk of female specific trans-acting factor/s affecting the allele-specific subjects heterozygous for the PTPN22 C1858T SNP to expression of PTPN22, was investigated in heathy males, develop autoimmune diseases. non-pregnant female subjects, pregnant female subjects In conclusion, comparison between cDNA from non- in first and third trimester, respectively, during which the stimulated and anti-CD3 and anti-CD28 stimulated estrogen and progesterone levels increase greatly. PBMCs revealed that the expression of PTPN22 1858C- We developed a fragment length analysis by capillary and T alleles are greatly increased by stimulation but to electrophoresis using the natural RsaI reaction-restriction the same extent. Furthermore, the absence of any fragment length polymorphism at the PTPN22 C1858T significant sex-specific variability in relative mRNA SNP with complete cutting by the RsaI endonuclease allele ratio suggests that the PTPN22 locus is unlikely throughout the study. There was no deviation from the to contain common polymorphisms or epigenetic mod- expected 1:1 ratio of the two alleles in genomic DNA ification that alter PTPN22 mRNA levels in non- samples from heterozygous individuals, showing that no stimulated PBMCs and during T-cell activation, and inequality in allelic representation was introduced by the essentially exclude such phenomena as a partial expla- assay, and rejecting that putative gene dosage effects or nation for the female predominance in most of the other biases influenced the measured allelic ratio in autoimmune disorders that associate with the PTPN22 cDNA samples. There was no difference in the relative Trp620 variant and autoimmune diseases in general. expression of the PTPN22 1858T-allele between genomic DNA and cDNA samples from non-stimulated human PBMCs, suggesting that the PTPN22 locus is unlikely to Materials and methods contain common polymorphisms or epigenetic modifica- RFLP analysis (genotyping) tion that alter PTPN22 mRNA levels in non-stimulated Genomic DNA was extracted from peripheral blood PBMCs, assayed by the level of allele-specific expression leukocytes using DNA extraction with proteinase K as of the PTPN22 Arg620Trp polymorphism. Further, there described previously.25 The PTPN22 C1858T polymorph- was no effect of sex or pregnancy status on the relative ism creates a natural RsaI restriction fragment length expression of the PTPN22 1858T-allele, indicating the polymorphism (RFLP) with the C (GTAC) but not the T absence of sex-specific trans-acting factor/s (e.g. sex- (GTAT) allele cleaved by the restriction enzyme. Indivi- hormones) affecting the allele-specific transcription duals heterozygous for the PTPN22 C1858T SNP were efficiency or RNA stability of the PTPN22 Arg620Trp selected following genotyping by RFLP analysis as polymorphism in non-stimulated PBMCs. described previously.26 Samples showing one 218 bp This essentially excludes such phenomena as a partial band were typed as homozygous TT, samples displaying explanation for the female predominance in most of the both 176 and 42 bp bands were typed as homozygous autoimmune disorders that associate with the PTPN22 CC, and samples exhibiting 218, 176 and 42 bp bands Trp620 variant. It should be noted that cis-acting variants, were typed as heterozygous. besides affecting RNA transcript synthesis or stability, which is measured in the present study, may affect the Relative expression of 1858T- and C-allele by fluorescent translational efficiency and protein stability. Based on the PCR-RFLP analysis present results, it cannot be ruled out that the transla- To investigate whether a sex-specific differential expres- tional efficiency of the two PTPN22 Arg620Trp alleles sion of the PTPN22 C1858T alleles exists, a fluorescent and/or the stability of the variant and wild-type LYP PCR-RFLP analysis was established to discriminate proteins are affected by sex-specific factors. between the two allele-specific transcripts. Additional allelic variation in expression may manifest in a cell type or cell state-specific manner,13 and as T cells cDNA synthesis. Mononuclear cells from 1 ml of fresh play a role at least in the initiation phase of the auto- full blood from healthy males, non-pregnant female immune disorders that associate with the C1858T poly- subjects, and female subjects in the first and third morphism, we further investigated whether the PTPN22 trimester of pregnancy were isolated using Ficoll-Hypa- gene is subject to allele-specific regulation during T-cell que gradient centrifugation and thawed. mRNA extrac- activation. While the transcription of PTPN22 in anti- tion was performed using the MagAttract kit (Qiagen, CD3 and anti-CD28 stimulated PBMCs increased 4- and Hilden, Germany) and KingFisher mL (Thermo Labsys- 13-fold by 48 and 72 h post-activation, respectively, no tems, Helsinki, Finland). cDNA synthesis was performed allele-specific response to the stimulation was observed using an oligo-dT reverse primer (Applied Biosystems, at any sample time points during the experiment. Sex- Foster City, CA, USA) and the following temperature stratification was not possible due to the small sample profile: 421C for 30 min and 941Cfor5min.ANRTcontrol size and as the stimulated cells were harvested from was assayed alongside the RT reaction for each individual. healthy individuals, any putative disease-dependent differences in the relative allelic expression of the Polymerase chain reaction. For the fluorescent capillary C1858T SNP during T-cell activation was not investigated. electrophoresis, the PCR was performed using a forward

Genes and Immunity Expression of human PTPN22 alleles C Nielsen et al 135 primer (50-ACTGATAATGTTGCTTCAACGG-30) labeled absence of an RT step. As an additional control of the at its 50 end with the fluorescent dye FAM (6-carboxy- complete cutting by the RsaI endonuclease in the fluorescein) and a reverse primer: 50-TCAC cleavage reactions, a separate cleavage reaction of a CAGCTTCCTCAACCAC-3026 designed to amplify a FAM-labelled PCR product using cDNA from a PTPN22 218 bp FAM-labeled fragment of the coding region of 1858CC homozygous healthy individual was included in PTPN22 containing the C1858T SNP. PCR amplification every PCR-RFLP analysis. was performed in a total reaction volume of 25 ml containing the following reagents: For genomic DNA: Fragment length analysis by capillary electrophoresis 1 ml (50 ng) of genomic DNA, 1.25 ml (5.0 pmol/ml) of Two microliters of the fluorescently labelled cleavage each primer, 2 ml dNTP (2.5 mM), 2.5 ml10Â PCR buffer reaction products were mixed well with a mixture

(Applied Biosystems) containing 15 mM MgCl2, and containing 0.5 ml of the GeneScan ROX-labelled G-350 0.25 ml AmpliTaq Gold polymerase (Applied Biosys- size standard (Applied Biosystems) and 13.5 ml of Hi-Di tems). For cDNA:10ml of cDNA, 1.0 ml (5.0 pmol/ml) of Formamide (Applied Biosystems). each primer, 2.5 ml10Â PCR buffer (Applied Biosystems) The mixture was denatured for 5 min at 951C and without MgCl2, 0.5 ml MgCl2 (25 mM) and 0.125 mlTaq placed on ice for 5 min, and to resolve the RFLP profiles, polymerase (Applied Biosystems). the fragments were size-separated using a four-color- Amplification was carried out using the following laser-induced 16-capillary electrophoresis system (ABI specific temperature cycling profile: an initial hold at Prism 3100 Genetic Analyzer), polymer POP 7 and 1 Â 941C for 12 min (genomic DNA) or 2 min (cDNA); 30 Genetic Analyzer buffer with EDTA (PE Applied cycles of 941C for 30 s, 601C for 30 s and 721C for 30 s; and Biosystems). The results were collected and the magni- a final extension step of 721C for 7 min. tudes of peak heights for all samples were determined using the ABI Genescan and Genemapper software Cleavage control. A 284-bp fragment of the coding (Applied Biosystem, Foster City, CA, USA). region of PTPN22 containing the C1858T SNP was Absence of a NED-coupled fluorescent 148-bp frag- amplified (forward primer: 50-CCTCCTGGGTTTGTA ment and presence of a 104 bp fragment was used as an CCTTA-30; reverse primer: 50-TCACCAGCTTCCTCA internal control demonstrating complete cutting by the ACCAC-30) using cDNA from a CC homozygous healthy RsaI endonuclease. Throughout the study failure of the individual. The PCR amplification conditions were the restriction enzyme was never seen. same as above with 30 PCR cycles. The PCR fragments were ligated into the pCR2.1-TOPO (Invitrogen, Taastr- Relative allelic expression analysis. For each cDNA sam- up, Denmark) cloning vector (using the manufacturer’s ple from heterozygous individuals, the relative expres- instructions) and transfected into competent bacteria that sion of the PTPN22 1858T- and 1858C-allele was were spread on carbenicillin (Sigma, Vallensbæk, Den- calculated as the ratio of the peak areas of the 218-bp mark) and X-gal (Qbiogen via KemEnTec, Copenhagen, fragment and the 176-bp fragment. The same PCR and Denmark) containing agar plates for selection. Positive analytic conditions were used for gDNA and cDNA to clones were grown overnight in carbenicillin supple- allow, if necessary, the use of the ratios observed from mented LB-media and plasmids were purified using gDNA to correct the allelic ratios obtained from cDNA Wizard SV 9600 Plasmid Purification System (Promega analyses for any inequalities in allelic representation via Ramcon, Birkerød, Denmark). introduced by fluorescent dye bias. Using purified plasmids (1:50) as template, a 148-bp fragment was amplified (forward primer: 50- Relative expression of PTPN22 mRNA and relative allelic GAATTTCCTTTGGATTGTTC-30; reverse primer: 50- expression in stimulated T cells over time TCACCAGCTTCCTCAACCAC-30) with the forward Using Ficoll-Hypaque gradient centrifugation, PBMCs primer being labelled at its 50-end with the fluorescent were isolated from full blood from 10 healthy individuals dye NED (2,70,80-benzo-50-fluoro-20,4,7-trichloro-5-car- heterozygous for the PTPN22 C1858T polymorphism. boxyfluorescein). PCR amplification using the same PBMCs were cultured in complete medium (1 Â106 condition as above and the following specific tempera- cells/ml) consisting of RPMI 1640 supplemented with ture cycling profile: an initial hold at 941C for 2 min; 35 10% heat-inactivated AB-serum, 200 mML-glutamine, cycles of 941C for 30 s, 601C for 30 s, and 721C for 30 s; 100 U/ml penicillin G and 100 mg/ml streptomycin and a final extension step of 721C for 10 min. sulphate (Invitrogen Corporation). To assess the effects The NED-labelled 148-bp fragment PCR product was of T-cell activation, cells were stimulated with 1 mg/ml used as an internal cleavage control to demonstrate anti-CD3mAb (clone OKT-3; Ortho Biotech, Bridgewater, complete cutting by the RsaI endonuclease in each NJ, USA) and 0.05 mg/ml anti-CD28 mAb (clone CD28.1; cleavage reaction, performed as described below. DAKO A/S, Glostrup, Denmark) in 96-wells (Nunc) microtiter plates incubated at 371C in an atmosphere

Cleavage reaction. Initially labelled PCR products with 5% CO2. Anti-CD3 and anti-CD28 mimic the were digested by RsaI. The reaction conditions were as binding of the peptide:MHC complex to the T-cell follows: 8 ml of FAM-labelled PCR product, 1 ml of NED- receptor and the binding of the co-stimulatory molecule labelled PCR product and 1000 U/ml RsaI incubated at B7 to CD28, respectively. A fixed number (100 000) of 371C overnight. Each sample was tested in duplicate and living cells were harvested following 0, 6, 24, 48 and 72 h in two dilutions (undiluted and 1:10) to check for of stimulation, respectively, lysed, and stored at 801C reproducibility. To check for any significant level of until mRNA extraction. The extraction of mRNA and genomic DNA amplification, NRT controls were run in cDNA synthesis was performed as described above. duplicates for each sample. Throughout the study, RNA Real-time PCR product accumulation was monitored samples did not yield detectable levels of product in the in the ABI Prism 7300 Sequence Detector (Applied

Genes and Immunity Expression of human PTPN22 alleles C Nielsen et al 136 Biosystems, Foster City, CA, USA) using the intercalating 3 Begovich AB, Carlton VE, Honigberg LA, Schrodi SJ, dye SYBR Green I. The reaction was carried out with the Chokkalingam AP, Alexander HC et al. A missense single- 2 Â SYBR Green PCR Master Mix, 3 ml of cDNA, and nucleotide polymorphism in a gene encoding a protein 5 pmol of each specific primer (forward primer: 50- tyrosine phosphatase (PTPN22) is associated with rheumatoid ACTGATAATGTTGCTTCAACGG-30; reverse primer: arthritis. Am J Hum Genet 2004; 75: 330–337. 50-TCACCAGCTTCCTCAACCAC-30) in a total volume 4 Velaga MR, Wilson V, Jennings CE, Owen CJ, Herington S, of 15 ml. Thermocycling program was 40 cycles of 951C Donaldson PT et al. The codon 620 tryptophan allele of the lymphoid tyrosine phosphatase (LYP) gene is a major for 10 min, 951C for 15 s, and 601C for 1 min with an 1 determinant of Graves’ disease. J Clin Endocrinol Metab 2004; initial cycle of 50 C for 2 min. All amplifications and 89: 5862–5865. detections were carried out in capped 96-well optical 5 Criswell LA, Pfeiffer KA, Lum RF, Gonzales B, Novitzke J, plates (Applied Biosystems). Kern M et al. Analysis of families in the multiple autoimmune Each sample was tested in duplicates to check for disease genetics consortium (MADGC) collection: the PTPN22 reproducibility. To control for genomic DNA amplifica- 620W allele associates with multiple autoimmune phenotypes. tion, NRT controls were run in duplicates for each Am J Hum Genet 2005; 76: 561–571. sample. 6 Vang T, Congia M, Macis MD, Musumeci L, Orru V, Zavattari The ‘mRNA Induction index’ of PTPN22 was calcu- P et al. Autoimmune-associated lymphoid tyrosine phos- ÀDCT phatase is a gain-of-function variant. Nat Genet 2005; 37: lated using the formula 2 , where DCT is the CT value 1317–1319. at the stimulation time in question minus the CT value at 7 Cohen S, Dadi H, Shaoul E, Sharfe N, Roifman CM. Cloning 0 h (non-stimulated cells). and characterization of a lymphoid-specific, inducible Using the same experimental conditions as in the human protein tyrosine phosphatase, Lyp. Blood 1999; 93: present study, previous experiments in our laboratory 2013–2024. have revealed that the mRNA expressions of eight 8 Cloutier JF, Veillette A. Cooperative inhibition of T-cell antigen frequently used reference genes including GAPDH and receptor signaling by a complex between a kinase and a b-actin, fluctuate during stimulation of PBMCs (unpub- phosphatase. J Exp Med 1999; 189: 111–121. lished data). The expression level of reference genes is 9 Hill RJ, Zozulya S, Lu YL, Ward K, Gishizky M, Jallal B. The therefore unreliable for normalizing the quantitative RT- lymphoid protein tyrosine phosphatase Lyp interacts with the adaptor molecule Grb2 and functions as a negative regulator PCR assay in the present study and hence we normalized of T-cell activation. Exp Hematol 2002; 30: 237–244. the expression of PTPN22 in cells stimulated for different 10 Hasegawa K, Martin F, Huang G, Tumas D, Diehl L, Chan AC 27 time periods to the input cell number. et al. PEST domain-enriched tyrosine phosphatase (PEP) The fluorescent PCR-RFLP analysis was performed as regulation of effector/memory T cells. Science 2004; 303: described above with 32 cycles in the PCR. 685–689. 11 Bottini N, Vang T, Cucca F, Mustelin T. Role of PTPN22 in type Search for potential targets of estrogen receptor by identifying 1 diabetes and other autoimmune diseases. Semin Immunol estrogen response elements in the promoter of PTPN22. The 2006; 18: 207–213. web-based program PAGen@UIC28 was used to search 12 Whitacre CC. Sex differences in autoimmune disease. Nat Immunol 2 for transcription factor binding sites in the promoter of 2001; : 777–780. 13 Yan H, Yuan W, Velculescu VE, Vogelstein B, Kinzler KW. the PTPN22 gene. Allelic variation in human gene expression. Science 2002; 297: 1143. 14 Lo HS, Wang Z, Hu Y, Yang HH, Gere S, Buetow KH et al. Statistical analysis Allelic variation in gene expression is common in the human Two-way Repeated Measures ANOVA was used to genome. Genome Res 2003; 13: 1855–1862. compare the relative expression of the T-allele in gDNA 15 Bray NJ, Buckland PR, Owen MJ, O’Donovan MC. Cis-acting and cDNA samples between and within the different variation in the expression of a high proportion of genes in sample groups. One-way Repeated Measures ANOVA human brain. Hum Genet 2003; 113: 149–153. was used to compare the changes in the mRNA 16 Pastinen T, Sladek R, Gurd S, Sammak A, Ge B, Lepage P et al. A survey of genetic and epigenetic variation affecting human Induction index of PTPN22, and the relative expression gene expression. Physiol Genom 2004; 16: 184–193. of the T-allele in cDNA samples from anti-CD3 and anti- 17 Bray NJ, O’Donovan MC. Investigating cis-acting regulatory CD28 stimulated PBMCs over time. When required, data variation using assays of relative allelic expression. Psychiatr were log(x þ 1) transformed to satisfy the ANOVA Genet 2006; 16: 173–177. assumption of homogeneity of variances. Po0.05 were 18 Yan H, Dobbie Z, Gruber SB, Markowitz S, Romans K, considered statistically significant. For every ANOVA, Giardiello FM et al. Small changes in expression affect P-values were Bonferroni adjusted for the number of predisposition to tumorigenesis. Nat Genet 2002; 30: 25–26. comparisons made. 19 Yang WS, Tsou PL, Lee WJ, Tseng DL, Chen CL, Peng CC et al. Allele-specific differential expression of a common adiponec- tin gene polymorphism related to obesity. J Mol Med 2003; 81: 428–434. References 20 Whitacre CC, Reingold SC, O’Looney PA. A gender gap in autoimmunity. Science 1999; 283: 1277–1278. 1 Bottini N, Musumeci L, Alonso A, Rahmouni S, Nika K, 21 Ackerman LS. Sex hormones and the genesis of autoimmunity. Rostamkhani M et al. A functional variant of lymphoid Arch Dermatol 2006; 142: 371–376. tyrosine phosphatase is associated with type 1 diabetes. 22 Cowles CR, Hirschhorn JN, Altshuler D, Lander ES. Detection Nat Genet 2004; 36: 337–338. of regulatory variation in mouse genes. Nat Genet 2002; 32: 2 Kyogoku C, Langefeld CD, Ortmann WA, Lee A, Selby S, 432–437. Carlton VE et al. Genetic association of the R620W poly- 23 Zhen FS, Du HL, Xu HP, Luo QB, Zhang XQ. Tissue and morphism of protein tyrosine phosphatase PTPN22 with allelic-specific expression of hsp70 gene in chickens: basal and human SLE. Am J Hum Genet 2004; 75: 504–507. heat-stress-induced mRNA level quantified with real-time

Genes and Immunity Expression of human PTPN22 alleles C Nielsen et al 137 reverse transcriptase polymerase chain reaction. Br Poultry Sci 26 Zheng W, She JX. Genetic association between a lymphoid 2006; 47: 449–455. tyrosine phosphatase (PTPN22) and type 1 diabetes. Diabetes 24 Lo HS, Wang Z, Hu Y, Yang HH, Gere S, Buetow KH et al. 2005; 54: 906–908. Allelic variation in gene expression is common in the human 27 Nielsen C, Ohm-Laursen L, Barington T, Husby S, Lillevang genome. Genome Res 2003; 13: 1855–1862. ST. Alternative splice variants of the human PD-1 gene. Cell 25 Olerup O, Zetterquist H. HLA-DR typing by PCR amplifica- Immunol 2005; 235: 109–116. tion with sequence-specific primers (PCR-SSP) in 2 h: an 28 Kamalakaran S, Radhakrishnan SK, Beck WT. Identification of alternative to serological DR typing in clinical practice estrogen-responsive genes using a genome-wide analysis of including donor–recipient matching in cadaveric transplant- promoter elements for transcription factor binding sites. J Biol ation. Tissue Antigens 1992; 39: 225–235. Chem 2005; 280: 21491–21497.

Genes and Immunity