Lung transplantation (LTx) is the last line of treatment for patients suffering from end-stage lung diseases including cystic fibrosis (CF), chronic obstructive pulmonary disease (COPD), and interstitial lung disease (ILD). LTx is the least successful solid organ transplantation due to the high incidence of chronic lung allograft dysfunction (CLAD). CLAD is clinically defined as obstructive (bronchiolitis obliterans syndrome, BOS), the most observed form of CLAD, or restrictive (restrictive allograft syndrome, RAS) (1). Latest numbers show that 5 year BOS-free survival is 50%, whereas only 25% of the transplanted patients remain free from BOS within the first 10 years after LTx (2).

BOS is a heterogeneous complication in which multiple parts of the immune system appear to play a role. Evidence exist for the involvement of humoral immunity, including complement and antibody formation, and features of autoimmunity (3). Also, cellular mediated processes, in which T cell reactivity seems to play a pivotal role, have been observed in patients diagnosed with BOS (4). A remarkable example of antibody formation is the association between autoantibodies directed against K-alpha 1 tubulin and collagen V and the development of BOS, also in the absence of donor specific anti-HLA antibodies (5). Furthermore, auto reactive CD4+ T cells directed against collagen V are observed in LTx patients, indicating that both B cell and T cell mediated autoimmunity plays a pivotal role in BOS development (6).

Upon T cell receptor-ligand interaction, T cell reactivity is controlled, among others, by intracellular proteases and phosphatases. One important protein tyrosine phosphatase (PTP) involved in T cell regulation is encoded by the PTPN22 gene, which is located on chromosome 1. The PTPN22 gene is considered to be one of the most important autoimmunity risk genes. Within the PTPN22 gene different single nucleotide polymorphisms (SNPs) are present, of which rs2476601 is the most striking clinically relevant example. Autoimmunity-related associations are, among others, rheumatoid arthritis (RA) (7), systemic lupus erythematosus (SLE) (8), vitiligo (9), and progressive systemic sclerosis (10, 11). The SNP rs2476601 is a missense mutation leading to an arginine tryptophan substitution within the first proline-rich sequence of the C-terminal domain, resulting in a disrupted interaction with CSK and enhanced enzyme activity (12). Currently, different functional models for the PTPN22–R620W substitution are described in literature (13).

Besides investigations on associations in autoimmunity and functional implications on T cell receptor signaling, PTPN22 SNPs are also studied in the field of transplantation. In kidney transplantation, PTPN22 gene polymorphisms were studied in two independent cohorts, but no associations with kidney function could be observed after transplantation (14, 15). Interestingly, Dullin et al. investigated selected SNPs within the PTPN22 gene and observed that rs2476601 is associated with multiple episodes of acute rejection after liver transplantation (16).

Taking into consideration the results obtained in liver transplantation, the associations of PTPN22 with autoimmunity and the observed features of autoimmunity in LTx, we hypothesized that SNPs in the PTPN22 gene could be associated to outcomes after LTx. For this end, we selected 6 SNPs within the PTPN22 gene and used both genotype and haplotype analysis on patient and donor samples and assessed correlations with transplantation outcome.

The PTPN22 gene is considered to be the most important non-HLA autoimmunity gene, of which the genetic variant rs2476601, 1858 C>T leading to an arginine–tryptophan substitution, is associated with multiple indications of autoimmunity including RA, SLE, and systemic sclerosis. Given the reported associations between autoimmunity and outcome after LTx we assessed the relation of the PTPN22 risk variant in both patients and donors and the incidence of chronic rejection after LTx.

Features of autoimmunity are often observed in or associated with chronic rejection after LTx (32). Type V collagen and K-α 1 tubulin are antigens for autoantibodies, presumably due to tissue remodeling and the activity of matrix metalloproteases that are induced by ischemia reperfusion injury and cleave collagen, thereby releasing antigenic fragments. Binding of autoantibodies to their respective targets on airway epithelial cells results in increases of profibrotic growth factors and proinflammatory cytokines (5, 33). Linked-recognition requires antigen recognition by T cells, and indeed auto reactive T cells directed at Type V collagen are present in LTx patients (6).

The non-receptor PTP encoded by PTPN22 consists of different domains, including the N-terminal catalytic active domain involved in dephosphorylation and a C-terminal domain which consists of proline-rich regions that facilitate binding of intracellular adapter proteins (29). PTPN22 is exclusively expressed by hematopoietic cells (34), and distinctive roles have been described in the regulation of immune cell signaling (13, 35). In T cell signaling, PTPN22 inhibits T cell activation via inhibiting downstream T-cell receptor signaling. In more detail, T cell activation after T cell receptor binding requires a cascade of tyrosine phosphorylation steps involving Src- (LCK, Fyn), and Syk family (ZAP-70) member kinases, that phosphorylate immunoreceptor tyrosine-based activation motifs located on the intracellular CD3-ζ chain (36). These all function as substrates for PTPN22, resulting in the dephosphorylation of their respective activation, and inhibition of T cell signaling (37). Furthermore, in T cells, the adapter protein binding domains on PTPN22 mediate the binding of tyrosine-protein kinase CSK, which also inhibits intracellular T cell receptor signaling (38, 39).

PTPN22 rs2476601 is associated with the occurrence of autoantibodies in RA, including rheumatoid factor and anti-citrullinated peptide (7), presumably due to the contributive effect on the generation on autoreactive B cells. Dai et al. have demonstrated spontaneous autoimmunity with infiltrations in both lungs and livers in a C57BL/6x129 mouse model homologous to the human PTPN22 risk variant (40). Also, deregulation in the deletion of clonal B cells and escape of autoreactive B cells from deletion stimuli have been observed in studies concerning the PTPN22 risk variant (41). These observations suggest a contributing factor of rs2476601 in autoantibody production, associated with impaired clinical outcome after LTx.

T regulatory cells are known to induce tolerance to self-antigens and suppress Th-1 autoimmunity, which has also been shown in human LTx (6). An altered T cell differentiation model is introduced as explanation for the pathogenesis toward autoimmunity. Disbalance between T cells and T regulatory cells is observed in PTPN22 knockout mice, suggesting tight regulation between effector and regulatory T cells via PTPN22. This has been strengthened by research of Vang et al. showing that T cells from individuals homozygous for the PTPN22 risk variant depicted increased Th-1 induced IFN-γ secretion and a decreased suppression of Th-1 cells via regulatory T cells (42). Translating these observations to the field of LTx suggests that a decreased suppression of Th-1 immune responses in patients genotyped with the risk variant of PTPN22 could contribute to autoimmunity and the pathogenesis of BOS development.

Here, we analyzed six selected PTPN22 gene polymorphisms in a cohort of patients who underwent LTx. Also, coupled donors were analyzed for these specific SNP configurations. Our analysis shows that a single missense SNP in the PTPN22 gene in patients is associated with outcome after LTx and the incidence of obstructive chronic lung allograft dysfunction. From our Kaplan-Meier analysis we can conclude that this association is predominantly with late chronic allograft dysfunction, indicated by the difference in inclination after month 30 post-LTx. Of the analyzed haplotypes only one haplotype showed a significant association with BOS incidence. However, this haplotype was only present in a relatively small number of BOS patients (15%), and validation of this finding is imperative.

In this study we have not been able to include an auto-antibody screening for anti-collagen V or anti-K-alpha 1 tubulin antibodies. However, our group has previously identified autoantibodies against ETAR, AT1R and BPIFA1/SPLUNC1 in pre- and post-LTx sera of the investigated cohort. Anti-ETAR and anti-AT1R autoantibodies are correlated with kidney transplantation outcome and anti-BP1FA1/SPLUNC1 autoantibodies are present in patients with endstage CF (43, 44). We found no correlation between autoantibody titers or stratified rs2476601 SNP configuration (data not shown). Within our cohort, we observed differences between the age at the time of LTx procedure between BOS+ and BOS− patients. Patients developing BOS were significantly older compared to non-BOS patients. This differs from other studies, were the reverse was observed, explained by the fact that younger patients have a less active immune system, although this remains purely speculative (45, 46). After Cox regression modeling in our cohort, including both the risk allele for PTPN22 and age at time of LTx, SNP rs2476601 was still an independent risk factor for BOS development (p = 0.004, OR: 2.748, 95% CI: 1.369–5.516, data not shown).

Targeted SNP analyses and genetic studies focused on solid organ transplantation and especially LTx outcome have increased over the past few years (47). Our group previously reported a genetic promotor polymorphism in complement regulation to be associated with chronic rejection after LTx, with the rationale that after binding to the graft, antibodies are less functional (48). Furthermore, the LTx research group from Leuven recently published findings on a polymorphism in the IL-17 receptor that predisposes to primary graft dysfunction (49), and correlations between SNP genotypes and immunosuppressive therapy efficacy have been described (50). These results are encouraging and indicate the importance of this emerging field, despite the need of prospective multicenter and validation studies. The recently installed aforementioned iGeneTRAiN consortium could facilitate these needs (18, 19).

Studies concerning the influence of rs2476601 on transplantation outcome have been conducted on other solid organ transplantations, albeit with different results. In liver transplantation, associations between rs2476601 and multiple episodes of acute rejection were observed, although chronic rejection associations were not assessed (16). No associations with outcome after kidney transplantation and PTPN22 gene polymorphisms were found by two independent groups (14, 15). Inter-organ differences could be explained by variations in either humoral or cellular immunologic mechanisms predisposing chronic rejection pathology, or differences in immunosuppressive regimens (51).

To summarize, our results are the first to show a deleterious effect of the autoimmune-related risk factor of PTPN22 in patients on LTx outcome. These results could be used for the further optimization of risk stratification algorithms to identify LTx outcome prior to, or earlier after transplantation. More research is expedient to elucidate the specific pathogenic pathways affected by PTPN22 in the development of chronic rejection after LTx. Importantly, larger study cohorts are needed to replicate our findings and to incorporate other genetic polymorphisms, in both patients and donors, associated with the development of rejection after transplantation. Our results could serve to further optimize patient treatment and follow-up therapy, and increase the long-term efficacy of LTx treatment for selected end-stage lung disease patients.

KB, JvS, TK-H, and OvR performed the research. KB, JvS, EvdG, OvR, CH, and HO participated in data analysis. EvdG, JK, and E-JO contributed patient material. EvdG and HO participated in research design. KB, JvS, EvdG, CH, and HO wrote the paper. All authors provided final approval of the version to be published.