A MUTATION IN THE GENE PTPN22 AS AN INDICATOR OF RISK FOR IDIOPATHIC THROMBOCYTOPENIC PURPURA By Matthew B. Zamora A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Clinical Laboratory Sciences 2010 ABSTRACT A MUTATION IN THE GENE PTPN22 AS AN INDICATOROF RISK IN IDIOPATHIC THROMBOCYTOPENIC PURPURA By Matthew B. Zamora Immune-mediated thrombocytopenic purpura (ITP) is an autoimmune disorder characterized by low platelet count and antibody mediated platelet destruction. ITP may be due to a failure of Tcell tolerance. Protein tyrosine phosphatase non receptor type 22 (PTPN22) is an important negative regulator of signal transduction from the T cell receptor. A single nucleotide polymorphism 1858C>T in the PTPN22 gene has been associated with various autoimmune disorders with an autoantibody component. We hypothesize that the PTPN22 mutation 1858C>T would be present in a higher proportion of ITP patients due to the humoral immunological component of the disease. We genotyped 45 patients with ITP by a Polymerase Chain Reaction (PCR) based restriction length polymorphism(RFLP) to identify the mutation 1858C>T in the gene PTPN22. Ten patients (22.2%) were heterozygous for the PTPN22 mutation and 2 were homozygous for the mutation (4.4%). The allele frequency of T allele in this ITP group is 15.5%. This compares with a T allele frequency of 8.6% observed in a published population study of 926 controls. Our results indicate that PTPN22 mutation allele is increased in ITP patients when compared to this control group with a P value of 0.02. ACKNOWLEDGMENTS I would like to express my thanks to Dr. John Gerlach for his support, patience and faith in me to finish. I would also like to say thanks to Dr. Kenneth Schwartz, Dr. Douglas Estry for being committee members. Special thanks to Dr. Karl D’Silva for his hard work in recruiting patients for the study his efforts are greatly appreciated. Lastly, I would like to thank my wife and kids for their constant encouragement and support. I couldn’t have done it without any of you. iii TABLE OF CONTENTS LIST OF TABLES .................................................................................................v LIST OF FIGURES ...............................................................................................vi INTRODUCTION .................................................................................................1 Idiopathic Thrombocytopenic Purpura ........................................................2 Basic Immunology......................................................................................4 T-Cell Development ...................................................................................6 Positive and Negative Selection ..................................................................9 Protein Tyrosine Kinases ............................................................................10 The regulation of T-cell receptor signaling .................................................14 Autoimmunity and the PTPN22 mutation ...................................................18 Comparison Studies ....................................................................................19 Possible relationship between PTPN22 and ITP ..........................................20 MATERIALS AND METHODS ............................................................................21 RESULTS ..............................................................................................................26 Statistics and Calculations ..........................................................................26 DISCUSSION ........................................................................................................34 CONCLUSIONS....................................................................................................38 RECOMMENDATIONS .......................................................................................39 FUTURE STUDIES OF PTPN22 ...........................................................................40 REFERENCES ......................................................................................................42 iv LIST OF TABLES Table 1: Summary of enzymes associated with TCR signaling and activation ........12 Table 2: Summary of Molecules associated with TCR signaling and activation .....13 Table 3: Summary of enzymes associated with inhibiting TCR activation and Signaling .................................................................................................15 Table 4: Frequency of genotypes for the PTPN22 mutation in patient groups analyzed ..................................................................................................27 Table 5: ITP genotype vs. Non-ITP control genotype 2x2 contingency table .........31 Table 6: ITP compared to published control group ................................................32 Table 7: Chi square calculation of ITP T allele compared to published control group T allele.....................................................................................................33 Table 8: ITP Patients genotypic frequencies compared with controls and other autoimmune disorders ..............................................................................36 Table 9: ITP Patients Allele frequencies compared with controls and other autoimmune disorders ..............................................................................37 v LIST OF FIGURES Figure 1: Components of the T-cell Receptor Complex .........................................8 Figure 2: MHC II interacting with the TCR causing activation of ITAMS by various protein kinases .......................................................................................11 Figure 3: Representation of the conformational change in the Src-family protein kinases .......................................................................................16 Figure 4: Graphic representation of TCR and enzymes involved with initiation and regulation of TCR signaling ...................................................................17 Figure 5: A listing of PCR sequence data, primers and recognition sites for XcmI .22 Figure 6: 2% agarose gel data ................................................................................25 2 Figure 7: χ statistic calculation..............................................................................28 Figure 8: Allele frequency calculation for the T-allele in ITP patients ...................29 Figure 9: Allele frequency calculation for the T-allele in thrombocytopenic non-ITP controls ..................................................................................................30 vi INTRODUCTION The term “horror autotoxicus” was phrased by the immunologist Paul Erlich. The biological mechanisms that result in horror autotoxicus are important to understand as they are at the core of concepts of self versus non-self in immunology [1]. Autoimmunity can be defined as an immune response toward one’s self. This response can be directed towards one specific organ as in diabetes or systemic as in the disease Lupus Erythematous. A major problem when studying autoimmunity is that the diseases are often complex, resulting from a set of complex pathophysiologic mechanisms. There is no “magic bullet” allowing detection of these diseases before onset. When autoimmune diseases present they are often recognized because of secondary outcomes associated with the disease such as antibody production. Although, mutations in certain genes can be looked at to determine the probability of developing a disease, this type of analysis is often an estimate, more specifically the likelihood for developing disease. It is not guaranteed that if you have a particular mutation or carry a certain gene you will develop disease. An approach is to test genes for association [1]. In this approach a candidate gene is identified and determined to be associated or not associated with a disease of interest [1]. Approximately 3 percent of the human population has some autoimmune disorder. With these autoimmune disorders there can be a cellular component or humoral component contributing to disease. Non-diseased individuals can have auto antibodies, the presence of which could be a risk factor for future autoimmune diseases [1]. The two main immune systems are the innate and the adaptive system. The innate system is thought of as primitive and nonspecific. This system responds to immediate threats to the body such as invasion by pathogens. The adaptive immune response is antigen specific and has memory [1]. 1 The immune system is complex and involves various types of cells, molecular mechanisms, and cell communication. Therefore, it seems possible there could be potentially several causes for autoimmunity. For example: A loss of T-cell tolerance - peripheral and central, genetic defects resulting in a loss of central tolerance, and genetic defects in T regulation cells [1]. A loss of peripheral and central tolerance is due to the ability of self-reactive T-cells to survive negative selection in the thymus. This can happen to T-cells and T regulation cells [1]. Molecular mimicry and cross reactivity can be another cause of autoimmunity. This is due to an infectious agent having similar protein epitopes to the host therefore causing an immune response to the host proteins during an infection or after the infection has resolved [2]. Autoimmune diseases are different pathologically but are often classified as systemic or organ specific [1]. Even, based on this classification there are similarities among the diseases genetically and mechanistically [1]. An example of this is Lupus Erythematosus (SLE). SLE can affect every organ in the body. The disease can have various auto-antibodies regulated by Human Leukocyte Antigen (HLA) alleles thus resulting in various outcomes of the disease [1]. But it would appear that a disruption with the interferon pathway is central to SLE. Therefore, it can be hard to determine whether the disease is due to genetic defects, multiple autoimmune disorders, or an outcome of a disease [1]. Idiopathic thrombocytopenic purpura (ITP) is an autoimmune disease that can have multiple causes which include: molecular mimicry, loss of peripheral and central tolerance, and loss of cell communication [2]. Idiopathic Thrombocytopenic Purpura ITP is a disease that affects platelets. Also, known as primary immune thrombocytopenia [3] the outcome is a low platelet count with the exclusion of most causes of 2 thrombocytopenia [3]. Other causes of thrombocytopenia include, but not limited to, disseminated intravascular coagulation, vitamin deficiency, infection, thrombotic thrombocytopenic purpura, hemolytic-uremic syndrome, and other bone marrow disorders [4]. The disease affects both children and adults. ITP can be categorized by patient age and is either acute or chronic [3]. In children ITP is acute and most often results after a viral infection or immunization [3]. The disease usually resolves on its own [3]. In adults the disease onset is sudden without warning and may be chronic [3]. ITP presents with a wide range of symptoms from asymptomatic to mucocutaneous bleeding and bruising at any part of the body [3, 4]. Most cases of ITP are found after a complete blood cell count shows an abnormally low platelet count [3]. Reports vary on the incidence of the disease from 5.8-6.6 per 100,000 per year [4] to 1.6-3.2 cases per 100,000 per year [3]. These statistics are based on the criteria used to differentiate a low platelet count from a normal platelet count with normal being any result greater 150 x109[3]. Therefore, the differences in the criteria to define a low platelet count may explain the difference in incidence rate in the diagnosis of ITP. There are several types of antibodies associated with ITP. ITP antibodies can be demonstrated when healthy non-ITP individuals are injected with blood from ITP patients resulting in decreased platelet counts [4]. The prevalent antibody found in ITP is Immunoglobulin type G (IgG) directed towards various glycoproteins found on platelets. The most common being IIb/IIIa. Other antibodies directed towards other glycoproteins are anti: Ib/IX, Ia/IIa, IV, and V [4]. These antibodies are formed from B-cells which are stimulated by autoreactive T-cells reacting to platelet antigens [4]. The concentration of IgG antibody on the platelet surface is helpful in confirming the clinical diagnosis of ITP [5, 6]. Reports in 3 monozygotic twins and family studies suggest that some patients may have a genetic predisposition to auto-antibody development [7, 8]. Basic Immunology Components of the immune system may be classified into the innate immune system and adaptive immune system [9].The cellular component involves cells such as Macrophages, Dendritic cells, B-cells, T-cells, and Neutrophils. Each cell has a role in the innate and adaptive immune system and may depend on cells in the innate system to activate the adaptive system [9]. Macrophages are phagocytic and present antigen to T-cells. One function of Dendritic cells is to present antigen. B-cells function to produce antibody. T-cells function to release cytokines to activate macrophages and induce B-cells to make antibody. Neutrophils function to kill extra cellular pathogens. During T-cell development, T-cell precursors in the thymus undergo genetic rearrangement of genes that code for the T-cell receptor (TCR). After genetic rearrangement these precursor T-cells interact with thymic epithelium cells using the T-cell receptor (TCR) displaying antigen via the major histocompatibility complex (MHC). In order for T-cells to survive or die they need to process this interaction of antigen associated with MHC. If the self MHC and antigen is recognized weakly then positive selection occurs and T-cells move forward in the immune response. However, if the interaction of self MHC and antigen is very strong, negative selection occurs and these T-cells are eliminated from the immune repertoire. Therefore, depending on how strong this interaction is the T cells are either positively or negatively selected. Positively selected T-cells leave the thymus and circulate to lymphoid tissues to encounter foreign antigen based on how the T-cells were selected. If these cells recognize foreign antigen with co-signals from antigen presenting cells they proliferate to 4 generate an immune response becoming an effector T-cell. Effector T-cells, characterized by CD4 (Th1 Th2) and CD8, function in the adaptive immune response. Th1 cells produce IL-2, INF-Gamma, and TNF-beta [10]. These Th1 cells are involved in cell mediated responses [10]. Th2 provides help to B-cells to express high affinity antibodies to foreign antigens in an adaptive immune response and are involved with antibody mediated responses [10]. B-cells are dependent on T-cell interaction in order to progress toward antibody production by a process called linked recognition. This is a process by which a CD4 T-cell recognizes an antigen and the B-cell recognizes the same antigen although not necessarily the same epitope [11p.383]. B-cells display this antigen via MHC class II. MHC class II molecules display antigen that are produced by pathogens which are manufactured from intracellular vesicles within macrophages, phagocytic cells, and B-cells that contain the pathogen [11 p. 32-33]. T-cells specific for this antigen recognize this via the TCR and release cytokines IL-4 and IL-5 activating the B-cell and resulting in antibody production [10]. MHC class I molecules present antigens that are manufactured in the cytosol of cells [11 p. 32]. After recognizing cell surface MHC I and antigen with the appropriate signals CD8 cells are cytotoxic [11 p. 350-352]. Apoptosis is most often a result of cytotoxicity [11 p. 364]. Signaling of T-cells is regulated by intracellular biochemical messaging pathways. This is achieved through the MHC and co-receptors with various protein kinases such as: Lymphocyte Specific Protein Tyrosine Kinase (Lck), Proto-oncogene tyrosine-protein kinase encoded by the FYN gene (Fyn), and Zeta Chain Associated Protein Kinase (Zap-70). These kinases function to phosphorylate the Immunoreceptor Tyrosine Activation Motif (ITAM) thus activating the cell and inducing signaling pathways in the cell by a process called signal transduction. 5 T-Cell Development The activation of T-cells involves interaction of the antigen presenting cell (APC) and the T-cell. Part of this is accomplished by TCR and MHC [12]. The TCR is diverse due to genetic rearrangements during development. This allows the TCR to interact with self and MHC. The TCR and MHC complex have many receptors, co-receptors, and molecules that initiate T-cell signaling. The outcome of activation of the T-cell thru the T-cell receptor, results in gene transcription which leads to the proliferation of these cells [13]. A T-cell and its TCRs are specific for a single protein. Each T-cell is diverse among other T-cells [12]. The surface has as many as 30,000 clonally identical TCR [11 p. 123] molecules. The TCR recognizes peptides bound in MHC and by this recognition there is a determination made if the immune response should proceed forward with T-cell activation and proliferation [12]. The function of the MHC is to present antigen to the T-cell receptor. The MHC are differentiated dependent on the type of protein to be presented to the TCR. MHC class I molecules present antigens that are generated in the cell cytoplasm. MHC class II molecules present antigens that are considered exogenous to the cell [12]. As mentioned, the T-cell receptor functions to receive processed antigen presented by the MHC to the T-cell specific for the antigen presented. After antigen is recognized it needs to be processed forward to activate the T-cell. Since antigen recognition by the TCR alone cannot activate the T-cell, the TCR has other co-signaling molecules and co-receptors that initiate the signaling pathways within the cell. This process is known as signal transduction i.e. transforming one type of signal into another type of signal or an extracellular signal into a response [11p. 830& 14p. g:32]. One of the molecules associated with this complex are ITAMs these molecules contain tyrosine in their cytoplasmic domains that are phosphorylated by protein 6 tyrosine kinases known as sarcoma proto-oncogenic tyrosine kinases (Src) [11 p. 228]. When antigen is presented to the T-cell thru the immune complex of MHC and TCR the tyrosine in the ITAMS are phosphorylated by the Src kinases. The alpha-beta (αβ) TCR is composed of two heterodimers. They function to recognize antigen/MHC ligand. This is where the recognition antigen begins. A complex is associated with the TCR. This is called CD3. The CD3 molecule complex is also involved with the activation of T-cells. CD3 molecules are composed of four chains of molecules that are collectively known as the CD3 complex: CD3 gamma (γ), CD3 delta (δ), and two CD3 epsilon (ε) molecules [11 p. 228]. This complex provides ITAMS. These ITAMS are involved with the initial signal transduction of the T-cell. Also, associated with the TCR are the ITAMS of the zeta(δ) chain, which is a homodimer [11 p.228]. Together these molecules make up the signaling complex of the TCR (See Figure 1). 7 TCR ε δ α β γ cell surface cell surface ITAMS ε δ δ To T-cell Signaling Pathway Figure 1: Components of the T-cell Receptor Complex The structural components of TCR complex of a T-cell: TCR, CD3, and ITAMS with zeta (δ) chains. CD3 molecules are composed of four chains of molecules that are collectively known as the CD3 complex: CD3 gamma (γ), CD3 delta (δ), and two CD3 epsilon (ε) molecules. Diagram modified from [11 p. 229]. TCR ligation with cognate MHC and protein leads to further signal transduction. 8 Positive and Negative Selection Hematopoetic stem cells that migrate to the thymus give rise to T-cells [15]. The development of these cells is dependent on interaction with thymic epithelial cells and mesenchymal fibroblasts which are found in specific areas of the thymus [15]. There are three developmental stages classified for these cells. They are: double negative, double positive, and single positive [16]. This classification refers to the expression of CD4 protein molecules and CD8 protein molecules. Double negative refers to the absence of CD4 or CD8 on the cell. Double positive refers to the presence of CD4 and CD8. Single positive refers to the presence of either CD4 or CD8 but not both [16]. Specific protein cell surface markers, CD44 and CD25 [15, 16] can be used to determine the developmental stage of the pre-T-cell further. When these pre-T-cells enter the thymus, they are classified as double negative because they lack CD4 and CD8 expression [15]. As mentioned the double negative (DN) stage can further be subdivided into stages based on the expression of CD44 and CD25 [16]. These stages are defined as: DN1, DN2, DN3, and DN4. These DN cells enter the thymus at the corticomedullary junction and move through the cortex and mature becoming CD44+ /CD25-(DN1), CD44+ /CD25+ (DN2), CD44- /CD25+ (DN3) CD44-/CD25- (DN4) [15]. The DN3 stage can be classified when these cells are CD44- /CD25+ on the cell surface, this is when gene rearrangements occur for the β chain of the TCR [15,16] ultimately producing a TCR during the transition from the DN3 stage to the DN4 stage. The DN4 stage of T-cell development can be marked by α chain gene rearrangement [15,16]. The next stage of development is the double positive (DP) stage. The DP stage is defined by the presence of both CD4 and CD8 and an αβ TCR [11p. 279]. 9 It is accepted that positive selection and negative selection occurs when the TCR binds to self-peptide displayed by MHC [15, 17]. Spleen Tyrosine Kinase (Syk) and Src kinases mediate this process by signal transduction, which produces a signal to the cell that survival or positive selection should occur [15]. During positive selection the ability of the DP T-cells that are able to recognize self-peptide: self MHC with low affinity are positively selected [11 p. 279, 17]. Cells are negatively selected when the TCR recognizes self-peptide: self MHC at a high affinity resulting in these cells being eliminated by apoptosis [11 p. 258, 15, 17]. Protein Tyrosine Kinases Protein Tyrosine Kinases (PTK) are responsible for phosphorylating the ITAMS associated with the TCR signaling complex. The PTK enzymes involved with T-cell receptor signaling are: Lck, and Fyn of the Src family of kinases and Zap-70 of the Syk family of kinases [18] (see table 1 and table 2). Once antigen bound MHC is presented by the antigen presenting cell to the TCR of the T-cell and is recognized, the T-cell signaling pathway is activated. Lck and Fyn phosphorylate the ITAMS of the TCR causing diphosphorylation of the ITAMS of the CD3 and δ molecules which provides a docking site for Zap-70. Activated ZAP-70 amplifies these signals downstream, ultimately activating the effector function of the T-cell [18]. It is believed that ITAM phosphoylation provides docking sites for ZAP-70 Src Homology 2 (SH2) domains [13]. After ZAP-70 binds to the ITAMS it is phosphorylated by Lck. These bound ZAP-70 molecules autophosphorylate to create docking sites for other signaling proteins [13]. The substrate of ZAP-70 is linker for activation of T-cells (LAT) [13]. This adapter protein LAT connects the signaling event to other proteins downstream to initiate T-cell activation (See figure 2) [11 p. 233, 9]. 10 MHC II Antigen ε δ α β γ ε cell surface cell surface P Fyn Lck P P ZAP-70 δ P δ IL-2 DNA Intracellular P Other signaling enzymes and molecules Nucleus Lck IL-2 mRNA Figure 2: MHC II interacting with the TCR resulting in Lck phosphorylation of: ITAMS, Fyn, and ZAP-70. P Indicates phosphorylation Indicate further downstream signaling events resulting in Interleukin-2(IL-2) production by Deoxyribonucleic Acid (DNA) and Messenger Ribonucleic Acid (mRNA) Modified from[11 figure 6.18, 19] 11 Table 1: Summary of enzymes associated with TCR signaling and activation Enzyme Lck Lymphocyte Specific Protein Tyrosine Kinase Fyn Proto-oncogene tyrosineprotein kinaseencoded by the FYN gene Zap-70 Zeta Chain Associated Protein Kinase CD45 Protein tyrosine phosphatase, receptor type, C Function Phosphorylates ITAMS of TCR/CD3 complex which then provides a docking site for ZAP-70 which is also phosphorylated by Lck Same as Lck Enzyme Family Src Family Kinase Phosphorylates T-cell specific adapters such as LAT which leads to activation of other enzymes which leads to T-cell activation Promotes signaling by dephosphorylating inhibitory tyrosine on Lck and Fyn Syk Family Kinase 12 Src Family Kinase Protein tyrosine Phosphatase Table 2: Summary of Molecules associated with TCR signaling and activation Molecule Function ITAM Immunoreceptor tyrosine activation Motif Recruits Src family protein kinase to coordinate signal transduction CD3 T-Cell Co-Receptor Associated with TCR provides ITAMS TCR T-cell receptor Receives antigen from Major histocompatibility complex Co-receptors for signaling CD4 or CD8 Cluster around MHC peptide TCR complex to promote signaling 13 The regulation of T-cell receptor signaling As with the components involved to activate the T-cell, there are enzymes involved to regulate the amount of signaling or to shutoff the activation of the T-cell (see table 3). Failure to maintain normal function of T-cells can result in autoimmunity and various diseases [13]. As with the activation of the TCR several enzymes exist to regulate this process. C-terminal Src kinase (Csk) is a protein kinase that functions to phosphorylate tyrosine residues Tyr-505 in Lck and Tyr-528 in Fyn. After this phosphorylation occurs there is a conformational change in which the tail of the enzyme blocks the active site in the SH2 domain [13]. (See figure 3) CD45 a protein tyrosine phosphatase is a positive regulator of T-cell activation [13]. It functions to dephosphorylate Src-family protein kinases [13] Ptpn22 is an enzyme encoded by the gene PTPN22. Also, it is referred to as Lyp and PTP. This enzyme is reported to be a negative regulator of T-cell receptor (TCR) signaling due to the enzymes ability to inactivate the various protein tyrosine kinases involved with T-cell activation [18]. In T-cells, when there is interaction of the TCR and antigen bound in the MHC, there is an increase in the amount of phosphorylation by various protein tyrosine kinases. This phosphorylation induces cellular events such as: transcription, proliferation, and cytokine production and release [18]. (See figure 4) 14 Table 3: Summary of enzymes associated with inhibiting TCR activation and signaling Enzyme Function Enzyme Family Csk C-Terminal Src Kinase Inhibits signaling Associates with PTPN22 Src Family Kinase PTPN22 Protein Tyrosine Phosphatase receptor type 22 Inhibits T-cell receptor signaling associates with CSK Dephosphorylates activating enzymes Lck, Fyn, and Zap-70 Protein tyrosine Phosphatase 15 SH3 SH2 kinase phosphorylation at carboxyl terminal tyrosine Csk CD45 Figure 3: Representation of the conformational change in the Src-family protein This figure represents the activation and inhibition of Src Kinases by CD45 and Csk. In the inactive state the phosphorylated kinase interacts with the SH2 domain therefore blocking the active site of the enzyme. This is accomplished by Csk. CD45 however dephosphorylates this enzyme at the carboxyl terminal tyrosine therefore activating the kinase. Modified from [11 p. 232] 16 Antigen Presenting Cell MHC II Antigen CD3 T-cell surface TCR Complex CD3 δ ε α γ β δ ε δ Fyn P P P PTPN 22 Csk P Intracellular Csk Lck ZAP-70 P P PTPN22 P Other signaling enzymes and molecules IL-2 Nucleus IL-2 DNA mRNA Figure 4: Graphic representation of TCR and enzymes involved with initiation and regulation of TCR signaling. In this diagram an example of regulation of T-cell activation by the enzymes PTPN22 and Csk. CD3 molecules are composed of what is known as four chains of molecules that are collectively known as the CD3 complex: CD3 gamma (γ), CD3 delta (δ), and two CD3 epsilon (ε) molecules. Each enzyme for activation: Lck, ZAP-70, Fyn are shown. Figure modified from [11 p.235, 19] P Represents Phosphorylation P 17 Represents Dephosphorylation Autoimmunity and PTPN22 Mutation It is hypothesized that autoimmunity, in the presence of the mutation1858C>T in the PTPN22 gene, develops due to the altered signaling pathway allowing T-cells that are selfreactive to escape negative selection in the thymus [19]. In these cells activation isn’t as strong as auto reactive cells without the mutation. This is because an altered ptpn22 enzyme is more efficient at restricting T-cell signaling therefore auto-reactive T-cells survive which would have been deleted by negative selection in normal individuals [19] i.e. less signal to restrict T-cell signaling therefore allows positive selection of self- reactive T-cells with moderate affinity of self MHC. Another hypothesis is that mutation effects peripheral T-cell tolerance through Tregulatory cells by decreasing intracellular signaling which would result in less regulation by these cells thus leading to autoimmunity [1]. Ptpn22 enzyme is known to be expressed in other cells such as: B-cells, NK cells, Macrophages, Monocytes, and Dendritic cells [19]. Therefore, it is reasonable to assume there could be other mechanisms or other cell types involved in autoimmunity with this altered enzyme. A recent study demonstrated an association with the C/T mutation and a reduced amount of memory B-cell population [20]. Although, the exact role the C/T variant may play in this decrease in memory B-cell population is unclear, it was demonstrated that the memory Bcells stimulated at the B-cell receptor (BCR) by cross-linking with IgM had a decrease in the amount of intracellular calcium produced, suggesting a decreased signaling effect upon the memory B-cells due to this mutation [20]. 18 Comparison Studies In 2004 the first report of PTPN22 variant associated with type I diabetes (T1D) was published [21]. Today, there have been numerous diseases that have been reported demonstrating an association. This mutation 1858C>T in the PTPN22 gene which codes for the enzyme ptpn22 has been associated with various autoimmune disorders [21-29]. The genotype frequency of this mutation appears to be approximately 16-17% for the C/T genotype in the normal population [22, 23] and has been found to be higher in populations associated with autoimmune diseases. Some of these autoimmune diseases with C/T genotype frequency include: Rheumatoid Arthritis (RA) 25% [22], T1D 30% [21], SLE approximately 20% [23] and Graves disease (GD) 25% [24]. All of which have the generation of autoantibodies as an outcome of the disease [30]. For example, RA IgM antibodies, known as rheumatoid factor, are produced against self IgG antibodies [31]. Another autoantibody that is associated with RA is anti-cyclic citrullinated peptides (CCP) [31]. Autoantibodies associated with T1D are autoantibodies against glutamic acid decarboxylase-65, and autoantibodies directed against the islet of the pancreas [31]. In SLE autoantibodies are produced against DNA, RNA, and chromatin [11 p. 611 & 831]. In GD antibodies are produced against the Thyroid Stimulating Hormone (TSH) receptor in the Thyroid gland [11 p. 611] causing the thyroid to release thyroid hormone resulting in hyperthyroidism [11 p. 620]. An interesting aspect of the mutation is a strong association with autoantibodies. In one study of RA, the presence of CCP antibodies in combination with the PTPN22 C/T mutation was associated with an increased risk for the disease RA [32]. This study involved RA patients who were identified to have donated blood samples before any RA symptoms were noted [22]. These 19 patients were analyzed statistically and shown to have 100% specificity for RA when the PTPN22 C/T mutation was present in combination with anti-CCP [32]. The disease Alopecia areata is believed to be autoimmune with the presence of autoantibodies to hair follicle proteins and the presence of T-cell infiltrates at and around the hair follicle [33]. In a study of this disease it was found that there was no difference between mild cases of Alopecia areata in individuals with PTPN22 mutation and healthy controls. However, when severe forms of Alopecia areata were compared to a healthy control group there was a statistical association with the disease and the PTPN22 C/T mutation [33]. Of interest, increased representation of this polymorphism is not observed in Th1 cell mediated autoimmune disorders: multiple sclerosis [35-37], inflammatory bowel diseases (Crohn’s Disease and ulcerative colitis) [34, 38-41] and rheumatoid factor negative rheumatoid arthritis that are not associated with circulating autoantibodies [35]. Possible relationship between PTPN22 and ITP Due to the association of this mutation with antibody mediated autoimmune diseases, it is reasonable to hypothesize that there is an association with ITP, an autoantibody mediated disease. The objective of this project is to determine if the association of PTPN22 mutation and ITP exists. The hypothesis being the inability of a mutation in PTPN22 to regulate signaling pathways in T-cells through the T-cell receptor in affected individuals. This predisposes these individuals to ITP by the loss of tolerance to self-antigens during thymic development. 20 Materials and Methods A case control study was designed to compare a control group of thrombocytopenic individuals due to chemotherapy to those with ITP. The investigative protocol was approved by the institutional review board of Michigan State University. After informed consent a venous blood specimen was collected from patients diagnosed with ITP and non-ITP thrombocytopenic controls. Additional inclusion criteria for ITP patients were: platelet counts below 100,000, bone marrow morphology, when obtained, compatible with a diagnosis of ITP, no concomitant autoimmune disorders, no medications known to cause thrombocytopenia, no associated viral disorders and no current diagnosis of malignancy. Blood samples were obtained from 45 unrelated consecutive patients diagnosed with ITP. These patients had a positive direct platelet antibody of at least 500 molecules IgG/platelet [42]. A radiometric technique was used to quantify autologous platelet surface IgG [5]. Controls were obtained from 38 individuals being treated with chemotherapy with no history of ITP or other autoimmune disorders. After collection specimens were centrifuged for 10 minutes at 3000 rpm. 200 microliters (uL) of the buffy coat was harvested for isolation of the genomic DNA. Isolation was performed using the QIAamp blood mini kit (QIAGEN Inc., Valencia, CA). Isolation procedure was performed following manufacturer specifications. The PTPN22 1858C/T genotypes were determined by polymerase chain reaction (PCR) Restriction Fragment Length Polymorphism (RFLP) using a restriction enzyme from Xanthomonascampestris (XcmI). Oligonucleotide primers were synthesized by Michigan State University Macromolecular Synthesis Facility (see figure 5). 21 The forward and reverse primers used: 5’-TCACCAGCTTCCTCAACCACA-3’ Forward primer 5’-GATAATGTTGCTTCAACGGAATTTA-3’ Reverse primer When mutation is present XcmI cuts at g producing 2 fragments of 46 bp and 169 bp R= A or G The recognition sequence for XcmI enzyme: 5’-CCANNNNNNNNNTGG-3’ 3’-GGTNNNNNNNNNACC-5’ 5’- tactcaccagcttcctcaaccacaataaatgattcaggtgtccRtacaggaagtggaggg 061gggatttcat catctatccttggagcagttgctatccaaaatgtcaaaaatattgtaaca 121attgttaatt agaacaatccaaaggaaattcttatattctaatattaaatataaatttac 181cataatttat atttaaattccgttgaagcaacattatcag t -3’ Figure 5: A listing of PCR sequence data, primers and recognition sites for XcmI Single Nucleotide Polymorphism Reference sequence 2476601 22 PCR amplification was performed in 50 uL reactions containing: 1 uL of genomic DNA, 2.5 units of Taq polymerase (Roche Applied Science, Mannheim Germany), 0.1mM of each 2’deoxyadensone 5’-triphosphate, 2’-deoxyguanosine 5’-triphosphate, 2’-deoxycytosine 5’triphosphate, 2’-deoxythymidine 5’-triphosphate (Invitrogen Carlsbad, California), 0.44uL of each primer, 1.5mM MgCl2, 10mM Tris-HCL, 50mM KCL, and 47 uL PCR grade water. This reaction was performed using the GeneAmp PCR system 9700 (Applied Biosystems, Carlsbad, California). The cycling parameters were as follows, modified from Kemp et.al [33]: Denaturation step-94 degrees Celsius (°C) for 1 minute, annealing step 57°C for 1 minute, extension/elongation step 72°C for 1 minute. This cycle was repeated 35 times. After 35 cycles there was a final extension of 10 minutes at 72°C. To confirm fidelity of the PCR product, 10 uL of the 50 uL amplified sample was resolved along with a 50 base pair (bp) DNA marker (Invitrogen, Carlsbad, California). This was completed in a 2 percent (% ) (w/v) agarose gel for 30 minutes at 100 volts in Tris borate buffer (TBE) (0.089 M Tris base, 0.089 M Boric acid .002M EDTA). The gel was stained with ethidium bromide 0.5 micrograms per milliliter (ug/ml) for 7 minutes and destained with double distilled water for 3 minutes. The gel was visualized with an ultra violet transilluminator 254 nanometers (nm) (UVP, Inc. San Gabriel California) A picture was taken to document results using DS-34 camera (Polaroid Corporation, Cambridge, Massachusetts) with 667 Polaroid film (Polaroid Corporation, Cambridge, Massachusetts) (See figure 6). In order to test for the polymorphism by RFLP, the amplified product was digested with the enzyme XcmI (New England Biolabs Inc., Ipswich, Massachusetts). This reaction was performed by adding 1ul of enzyme and 2ul of NE2 buffer (New England Biolabs Inc., Ipswich, MA): (50mM NaCl, 10mM Tris HCL (pH 7.9), 10mM MgCl2, 1mM dithiothreitol) provided 23 with the enzyme to 17ul amplified product. The amplified product mixed with digestion enzyme was incubated at 37°C overnight. The enzyme was inactivated by warming the same mixture in a 65°C water bath for 10 minutes. Following digestion 20ul of the post digested product was loaded into a 2% agarose gel and the DNA bands were separated with a current of 100 volts for 1.5 hours in TAE buffer. After separation the gel was stained with ethidium bromide 0.5 ug/ml and visualized with by UV transilluminator 254 nm. A picture was taken for documentation. All DNA bands seen were compared to a 50 base pair (bp) marker. If no mutation was present a 215 bp band of DNA would be observed. A heterozygous mutation was observed by producing a 215 bp fragment, 169 bp fragment, and a 46 bp fragment. A homozygous mutation was noted with a 169 bp fragment and a 46 bp fragment (See figure 6). 24 1 2 3 4 Bl M A 215 bp 50 bp B 215 bp 169 bp 46 bp 50 bp Figure 6 2% agarose gel data: DNA bands 215 bp, 169 bp, 46 bp. Compared to 50 bp marker A Post amplification gel electrophoresis to check fidelity of amplified product. Wells 1, 2, 3, 4 amplified product. B is the blank well i.e. all PCR reagents without DNA template. M is the track it 50 bp marker. B Post digestion electrophoresis to check for mutation. Wells 1&2 are double mutant. Wells 3&4 are heterozygous for mutation. Followed by a blank and a 50 bp marker. 25 Results Statistics and Calculations Allele frequencies were calculated for the T allele for ITP and the thrombocytopenic nonITP control group. The genotype data (see table 4) was analyzed by Chi square (χ2) to determine if there is an association with the PTPN22 mutation and ITP (see figure 7). The T allele frequencies for ITP and the non-ITP thrombocytopenic control group were calculated from genotypic frequencies (see figures 8 and 9). Using 2 by 2 contingency tables the data was analyzed by Chi square analysis as followed: ITP patients genotype frequencies compared to the non-ITP thrombocytopenic control group genotype frequencies (see table 5), ITP patients genotype frequencies compared to the published control group genotype frequencies (see table 6) and ITP patients allele frequencies compared to published control group allele frequencies (see table 7). The T allele frequencies for ITP and the published controls were calculated from the percent positive allele frequencies. The allele frequency calculated by Begovich [22] was 8.7% in 926 controls with 1852 total alleles. This information was used to derive the number of alleles that were possible for either a TC or TT combined total in this control group. The same was done for the allele frequency calculated for the 45 ITP patients , total alleles 90, with an allele frequency of 15.5%. 26 Table 4: Frequency of genotypes for the PTPN22 mutation in patient groups analyzed PATIENT TYPE HETEROZYGOUS DOUBLE WILDTYPE MUTANTS CT CC 10 2 33 (22.2%) ITP TT (4.4%) (73.3%) 7 0 31 Non-ITP (thrombocytopenic controls) (18.4%) TOTALS 17 (81.6%) 2 27 64 2 χ =∑ 2 Figure 7: χ (chi square) statistic calculation Chi square statistic calculation used for all comparisons for categorical data 2 χ = chi square test statistic O=an observed frequency E= an expected frequency 2 χ critical value at 95% confidence is 3.84 2 χ critical value at 90% confidence is 2.70 Degrees of freedom (df) used for all comparisons is 1 28 T= T allele frequency TT= number of TT genotypes CT= number of CT genotypes Number of possible alleles from entire ITP group 45 x 2=90 Using data from genotype frequency table: CC=33 CT=10 TT=2 T= T=15.5% Figure 8: Allele frequency calculation for the T allele in ITP patients 29 T= T allele frequency TT= number of TT genotypes CT= number of CT genotypes Number of possible alleles from entire control group 38 x 2=76 Using data from genotype frequency table: CC=31 CT=7 TT=0 T= T=9.2% Figure 9: Allele frequency calculation for the T allele in non-ITP thrombocytopenic control group 30 Table 5: ITP genotype vs. Non ITP control genotype 2x2 contingency table PTPN22 mutation observed No Yes ITP 33 12 CONTROL 31 7 Hypothesis: H0: The variables are independent there is no association between ITP and the PTPN22 mutation. H1: The variables are not independent, there is an association between ITP and the PTPN22 mutation. 2 2 χ calc=Σ[(O-E) /E] where O=observed frequency and E= expected frequency 2 χ calc=0.794 with a degree of freedom of 1 and a probability of 0.373 2 χ crit=3.84 2 2 χ calc <χ crit therefore accept H0 Accept H0:The variables are independent there is no association between ITP and the PTPN22 mutation 31 Table 6: ITP genotype compared to published control group genotype ITP Published control PTPN22 mutation observed No 33 774 Yes 12 152 Hypothesis: H0: The variables are independent there is no association between ITP and the PTPN22 mutation H1: The variables are not independent-There is an association between ITP and PTPN22 mutation 2 2 χ calc=Σ[(O-E) /E] Degrees of freedom 1 2 χ calc=3.21 with a probability 0.073 2 χ crit=3.84 2 2 χ calc <χ crit therefore accept H0 H0:The variables are independent there is no association between ITP and the PTPN22 mutation 2 2 χ calc > χ critical value of 2.70 at 90% confidence therefore reject H0 and accept H1 H1: The variables are not independent-There is an association between ITP and PTPN22 mutation 32 Table 7: Chi square calculation of ITP T allele compared to published control group T allele ITP Published control PTPN22 mutation observed CC 76 1691 CT or TT combined 14 161 Hypothesis: H0: The variables are independent there is no association between ITP and the PTPN22 T allele H1: The variables are not independent-There is an association between ITP and PTPN22 T allele 2 2 χ calc=Σ[(O-E) /E] Degrees of freedom 1 2 χ calc=4.128 with a probability 0.0211 2 χ crit=3.84 2 2 χ calc >χ crit therefore reject H0 and accept H1: The variables are not independent-There is an association between ITP and PTPN22 T allele 33 Discussion The reported incidence for the allelic frequency of the PTPN22 mutation in European populations demonstrates a gradient with a low of 2% to 3% in Italy and Sardinia to 7-8% for the United Kingdom, increasing to more than 10% for Scandinavians and over 15% for Finland [43]. In most European-American populations genotypic frequencies are in the range of 16-17% [22, 23] and allele frequencies 8-9% [25]. This is likely due to this population having predominately western European and British ancestors. In African and Asian populations the C>T mutation is absent [25]. PTPN22 mutation has been associated with several autoimmune disorders. This study statistically fails to reject the null hypothesis by Chi square analysis. This is likely due to a lack of power with the sample size used in the control group. To determine if a true association exists between ITP and the PTPN22 mutation a larger sample size would be needed for both ITP and control subjects. Since, the genotypic frequencies are increased in the ITP group it seems reasonable to compare the observed genotypic frequencies in this study with data in a control population used by Begovich [22]. Based on the data there is an increase in the allele frequency in the ITP group compared to the control group. The genotype and allele frequencies presented here for the ITP group correlate well with genotype and allele frequencies presented in the published study by Begovich [22]. But it should be noted that when comparing the ITP patients from this study to the 2 published study of controls by χ analysis the outcome fails to reject the null hypothesis : The variables are independent there is no association between ITP and the PTPN22 mutation. Although, at the 0.05 significance level statistically the study fails to reject the null hypothesis it is suggestive that an association exists. Specifically the p value for this is 0.073 34 which is close to 0.05 showing significance. Interestingly, if the significance level is changed 2 from .05 to 0.10 an association can be established when comparing the χ calculation to the 2 χ critical value. Therefore, concluding that an association between ITP and PTPN22 mutation exists. However, by increasing the significance level the possibility exists for increased error, the likelihood of rejecting the null hypothesis when in fact it should be retained. It would seem that an association of ITP and PTPN22 mutation would be plausible based on the etiology of ITP. The allele and genotype frequencies are consistent with other autoimmune disease (see tables 8 and 9). Besides determining that the allele frequency for the T-allele is higher in ITP patients when compared to the published control data, the disease is: autoimmune, appears to have a T-cell tolerance failure, has an autoantibody component, has no defined mechanism for disease. Also, two ITP patients were homozygous for the mutation. Homozygous alleles for this mutation are rare. Autoimmune diseases are complex due to a genetic defects and responses to environmental factors. As with many autoimmune diseases determining an association would give some insight to how these diseases develop. 35 Table 8: ITP Patients genotypic frequencies compared with controls and other autoimmune disorders Subjects Total N Number Positive for the PTPN22 allele CT or TT Percent Positive Reference Controls 926 152 16.4 22 ITP 45 12 27 RA 475 125 26 22 SLE 705 157 22 23 T1D 294 101 34 21 Graves D. 901 240 26.6 26 36 Table 9: ITP Patients allele frequencies compared with controls and other autoimmune disorders Subjects Total N Number Positive for the PTPN22 allele CT or TT Allele Frequency of T Controls 926 152 8.7% ITP 45 12 15.5% RA 475 125 13.8% 22 SLE 705 157 12.6% 23 T1D 294 101 19% 21 Graves D. 901 240 14.3% 26 37 Reference 22 Conclusions The PTPN22 gene encodes an enzyme ptpn22 that is a negative regulator of T-cell activation. The enzyme associates with Csk and regulates T-cell signaling after MHC and protein are displayed to the TCR. The C>T mutation in these gene causes a conformational change that no longer allows the enzyme to associate with Csk. This conformational change is believed to allow the ptpn22 enzyme to more efficiently dephosphorylate the ITAMS of the CD3 molecule and the activation enzymes Lck, Fyn, and Zap-70. The hypothesis is this causes Tcells that are auto-reactive to survive negative selection due to an attenuated signal. The PTPN22 mutation has been associated with several autoimmune diseases that have auto antibodies as an outcome of the disease such as: GD, SLE, T1D, and RA. It seems reasonable that this mutation would be associated with ITP due to its antibody mechanism for disease. The conclusion from this study is ITP is not associated with the PTPN22 mutation using Chi square test statistic at the 95% confidence level. However, the data suggests that a possible relationship could exist. The data is suggestive of this when you look at the allele and genotypic frequencies compared to other autoimmune diseases where an association exists using a published control group [42]. The T allele does have an increased representation in ITP patients with a P value of 0.02 [42]. However, this is relative to the small number of ITP patients studied, 45, and should be interpreted prudently. The data suggests that a larger sample size is needed to confirm an association. 38 Recommendations The data suggests for future studies that more specimens should be analyzed to determine if an association exists between ITP and the gene PTPN22. An association can be established at 0 .10 confidence level using Chi square with the data used. The Chi square calculated value is close to being greater than the Chi square critical value at the 0.05 level which is very suggestive. The data strongly suggest that hundreds of known cases of ITP and equal number of controls are used to increase the power of the study. Perhaps the data collected in this study could be used in meta-analysis to achieve a sample group number comparable to other studies where the mutation has an association. If more specimens are to be analyzed an efficient technique should be employed to detect the 1858 C>T mutation in the PTPN22 gene. This would decrease the time for determining the genotypes of samples being analyzed. Although, with the method used many specimens can be analyzed it would be cumbersome trying to perform restriction digest and resolve many specimens at one time. If one were to undertake such a task of analyzing many specimens, the use of real time PCR would be ideal. It would be very efficient since amplification and analysis of the mutation can be done at the same time. The RFLP method used in this study could be used as a quality control method on random samples to confirm amplification and analysis against the real time PCR method. The real time PCR method would eliminate the time involved with digesting the amplified product with the enzyme XcmI. It would also eliminate the need to resolve the digested product on an agarose gel. Therefore, many specimens could be analyzed and one could have results the same day. 39 Another benefit of using real time PCR is that it could reduce the amount of human error introduced when setting up the restriction digest and gel electrophoresis. For example, the process of setting this up was tedious even when analyzing only 7 specimens at a time. The restriction digest involves several pipetting steps. The XcmI enzyme needs to be added to the NE2 buffer supplied with enzyme. The amplified product needs to be pipetted with the correct amount of XcmI enzyme NE2 buffer. Therefore, it is plausible when analyzing many samples one could add the wrong amount enzyme or specimen. It is reasonable to argue that the possibility exists of adding two different specimens together. Future studies of PTPN22 Another project to consider is the effect of this mutation 1858 C>T in the PTPN22 gene on B-cells. So far, the investigations of this mutation are with different diseases with an autoantibody component and proposed mechanisms for autoimmunity. However, there seems to be a lack of investigation into the B-cells role with this mutation. The B-cell receptor is similar to that of the TCR. They both use ITAMS for signaling. Both are initiated by Src-family kinases. Specifically, it would be of interest to determine if B-cell signaling is affected in many of the autoimmune diseases associated with this mutation. The hypothesis of autoimmunity developing due to this mutation is: T-cells survive negative selection due to an attenuated signal caused by the efficiency of the ptpn22 enzyme. Another hypothesis is that peripheral tolerance is lost due to the PTPN22 mutation effecting Tregulatory signaling when it encounters an auto-reactive T-cell. T-regulatory cells are known to regulate autoimmunity by destroying cells in the periphery that are auto-reactive [1]. A future study would be to explore the effects of the PTPN22 mutation specifically in these T-regulatory cells to determine if the altered enzyme disrupts the normal function of these cells. 40 Autoimmunity is complex and has many causes. Genetic mutations are just one possible predisposition to disease. More research into factors that contribute to autoimmunity such as environment, genetic variants, molecular mimicry and loss of tolerance will explain autoimmune diseases more thoroughly. 41 REFERENCES REFERENCES [1]Gregersen, Peter K., and Timothy W. Behrens. Genetics of autoimmune diseases-disorders of immune homeostasis. Nature 7 (2006): 917-28. [2] Zhou Bin, Hui Zhao, Ren Chi Yang, Zhong Chao Han. Multi-dysfunctional pathophysiology in ITP. Critical Reviews in Oncology/Hematology 54(2005) 107-116. [3]Stasi, Roberto, Maria L. Evangelista, Elisa Stipa, Francesco Buccisano, AndrianoVenditti, and Sergio Amadori. Idiopathic thrombocytopenic purpura: Current Concepts in pathophysiology and management. Thrombosis and Haemostasis 99 (2008): 4-13. [4]Gernsheimer, Terry. Epidemiology and pathophysiology of immune thrombocytopenic purpura.European Journal of Haematology 80.Supplemental 69 (2008): 3-8. [5]Schwartz KA, Gauger JA, Davis JM. Precise Quantitation of PAIgG: A new radiometric microtechnique. American Journal of Hematology. 1990;33:167-176. [6]Estrada DT, Davis J, Schwartz KA. Radiometric Quantitation of PAIgG stratified Treatment response in ITP patients. Central Society for Clinical Research. 2004; Chicago [7]Laster AJ, Conley CL, Kickler T, Dorsch CA, Bias W. Chronic Immune Thrombocytopenic Purpura in Monozygotic Twins. New England Journal of Medicine. 1982;307:1495-1498. [8]Bizzaro N. Familial Association of autoimmune thrombocytopenia and hyperthyroidism. American Journal of Hematology. 1992;39:294-298. [9]Kim, Irene K., Damanpreet S. Bedi, XupengGe, and Stefan G. Tullius. Impact of Innate and Adaptive Immunity on Rejection and Tolerance. Transplantation 86 (7) (2008): 889-94. [10]Constant S, Bottomly K. Induction of Th1 and Th2CD4CTCellRespones: The Alternative Approaches. Annual Review of Immunology 1997; 15 297-322. [11]Murphy K, Travers P, Walport M. ImmunoBiology, 7thedn.Garland Science Publishing 2008. [12]Dustin M. T-cell activation through immunological synapse and kinapses. Immunological Reviews 2008; 221 77-89. [13]Mustelin T, Tasken K. Positive and negative regulation of T-cell activation through kinases and phosphatases. The Biochemical Journal 2003; 371 15-27. [14]Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P. Molecular Biology of the Cell, 4thedn. Garland Science Publishing 2002. 42 [15]Starr, Timothy K., Stephen C. Jameson, and Kristin A. Hogquist."Positive and Negative Selection of T Cells."Annual Review of Immunology 21 (2003): 139-76. [16]Abbey, Janice L., and Helen C. O'Neill. "Expression of T-cell receptor genes during early Tcell development. "Immunology and Cell Biology 86 (2008): 166-74. [17]Werlen, Guy, Barbara Hausmann, Dieter Naeher, and Ed Palmer. “Signaling Life and Death in the Thymus: Timing is Everything.” Science 299 (2003):1859-1863. [18]Cloutier J, Veillette A. Cooperative Inhibition of T-cell Antigen Receptor Signaling by a Complex between a Kinase and a Phosphatase. Journal of Experimental Medicine 1999; 189:111-121. [19]Bottini N, Vang T, Cucca F, Mustelin T. Role of PTPN22 in type 1 diabetes and other autoimmune diseases. Seminars in Immunology 2006; 18:207-213. [20]Rieck M, Arechiga A, Onengut-Gumuscu S, Greenbaum C, Concannon P, Buckner J. Genetic Variation in PTPN22 Corresponds to Altered Function of T and B lymphocytes. The Journal of Immunology 2007;179:4704-4710. [21]Bottini N, Musumeci L, Alonso A, Rahmouni S, Konstantina N, Rostankhani M, MacMurray J, Meloni GF, Paolo L, Pellecchia M, Eisenbarth GS, Comings D, Mustelin T. A functional variant of lymphoid tyrosine phosphatase is associated with type I diabetes. Nature Genetics 2004; 36:337-338. [22]Begovich AB, Carlton VE, Honigberg LA, Schrodi SJ, Chokkalingam AP, Alexander HC, et al. A Missense Single-Nucleotide Polymorphism in a Gene Encoding a Protein Phosphatase (PTPN22) Is Associated with Rheumatoid Arthritis. American Journal of Human Genetics 2004; 75:330-337. [23]Kyoguku C, Langefeld CD, Ortmann WA, Lee A, Selby S, Carlton VEH, et al. Genetic association of the R620W polymorphism of protein tyrosine phosphatase PTPN22 with human SLE. American Journal of Human Genetics 2004; 75:504-507. [24]Velaga, Wilson V, Jennings CE, Owen CJ, Herington S, Donaldson PT, Ball SG, James RA, Quinton R, Perros P, Pearce SHS. The codon 620 Tryptophan Allele of the Lymphoid Tyrosine Phosphatase (LYP) Gene is a major determinant of Graves’ Disease. The Journal of Clinical Endocrinology & Metabolism 2004; 89(11):5862-5865. [25]Gregersen P, Lee H-S, Batliwalla F, Begovich AB. PTPN22: Setting thresholds for autoimmunity. Seminars in Immunology 2006; 18:214-223. [26]Smyth D, Cooper JD, Collins JE, et al. Replication of an Association Between the Lymphoid Tyrosine Phosphatase Locus (LYP/PTPN22) with Type 1 Diabetes and Evidence for Its Role as a General Autoimmunity Locus. Diabetes 2004; 53 3020-3023. 43 [27]Criswell LA, Pfeiffer KA, Lum RF, et al. Analysis of families in the Multiple Autoimmune Disease Genetics Consortium (MADGC) Collection: The PTPN22 620W Allele Associates with Multiple Autoimmune Phenotypes. American Journal of Human Genetics 2005; 76:561-571. [28]Plenge RM, Padyukov L, Remmers E, et al. Replication of Putative Candidate-Gene Associations with Rheumatoid Arthritis in >4000 Samples from North America and Sweden: Association of Susceptibility with PTPN22, CTLA4 and PAD14. American Journal of Human Genetics 2005;77:1044-1060. [29]Anaya J-M, Gomez L, Castiblanco J. Is there a common genetic basis for autoimmune disease. Clinical and Developmental Immunology. 2006;13:185-195 [30]Gregersen P. Gaining insight into PTPN22 and autoimmunity. Nature Genetics 2005; 37:1300-1302. [31]Reveille JD. The genetic basis of autoantibody production. Autoimmunity reviews 2006; 5:389-398. [32]Johansson M, Arlestig L, Hallmans G, Rantapaa-Dalqvist S.PTPN22 polymorphism and anti-cyclic citrullinated peptide antibodies in combination strongly predicts future onset of rheumatoid arthritis and has a specificity of 100% for the disease. Arthritis Research & Therapy 2006, 8:R19 [33]Kemp E. H., Mcdonagh A., Wengraf D, Messenger A, Gawkrodger D, Cork M, Tazi-Ahnini. The Non-Synonymous C1858T Substitution in the PTPN22Gene is Associated with Susceptibility to the Severe Forms of Alopecia Areata. Human Immunology 2006; 67 535-539 [34] Barrett JC, Hansoul S, Nicolae DL, Cho JH, Duerr RH, et al. 2008. Genome-wide association defines more than 30 distinct susceptibility loci for Crohn’s disease. Nature Genetics.40:955–62 [35] Begovich AB, Caillier SJ, Alexander HC, Penko JM, Hauser SL, Barcellos LF, Oskenberg JR. The R620W Polymorhphism of the Protein Tyrosine Phosphatase PTPN22 is Not Associated with Multiple Sclerosis. American Journal of Human Genetics 2005; 76:184-187 [36] Matesanz F, Rueda B, Orozco G, Fernandez O, Leyva L, Alcina A, et al. Protein tyrosine phosphatase gene (PTPN22) polymorphism in multiple sclerosis. Journal of Neurology 2005; 252:994-995. [37] De Jager PL, Sawcer S, Waliszewska A, Farwell L, Wild G, Cohen A, et. al. Evaluating the role of the 620W allele of protein tyrosine phosphatase PTPN22 in Crohn’s disease and multiple sclerosis. European Journal of Human Genetics 2006; 14:317-321. 44 [38] van Oene M, Wintle RF, Liu X, Yazdanpanah M, Gu X, Newman B, et al. Association of the lymphoid tyrosine phosphatase R620W variant with rheumatoid arthritis, but not Crohn’s disease, in Canadian populations. Arthritis Rheum 2005;52:1993-1998. [39] Zhernakova A, Eerligh P, Wijmenga C, Barrera P, Roep BO, KoelemanBP. Differential association of the PTPN22 coding variant with autoimmunediseases in a Dutch population. Genes Immunity 2005;6:459–61. [40] Martin MC, Oliver J, Urcelay E, Orozco G, Gomez-Garcia M, Lopez- Nevot MA, et al. The functional genetic variation in the PTPN22 gene has a negligible effect on the susceptibility to develop inflammatory bowel disease. Tissue Antigens 2005;66:314–7. [41] Prescott NJ, Fisher SA, Onnie C, Pattni R, Steer S, Sanderson J, et al. A general autoimmunity gene (PTPN22) is not associated with inflammatory bowel disease in a British population. Tissue Antigens 2005;66:318–20. [42] D’Silva KJ, Zamora MB, Gerlach J, Schwartz KA. Increased representation of the PTPN22 mutation in patients with immune thrombocytopenia. Journal of Thrombosis and Haemostasis 2010; 8:2076-8. [43] Gregersen P, Olsson L. Recent advances in the Genetics of Autoimmune Disease. Annual Review Immunology 2009; 27:363-391 45