Despite recent advances in immunosuppressive therapy, acute allograft rejection remains a major problem after kidney transplantation, and has been shown to be a crucial determinant of chronic rejection and long-term graft function (1, 2). Identifying risk factors that influence the incidence and severity of acute rejection remains a priority of transplant biologists, as these may be used to develop algorithms to estimate individual patient risk, and ultimately may permit the development of personalized immunosuppressive regimens.
T lymphocytes are the major effector cells involved in acute allograft rejection, and they are capable of exhibiting a wide spectrum of immune responses after an encounter with antigen (3). The local environment in which an antigen-lymphocyte interaction takes place plays a major role in influencing the nature, severity and duration of the subsequent immune response, and cytokines are a major determinant of this milieu. These are small, short-acting proteins that amplify and direct the immune response, and are produced by a wide variety of cells, including lymphocytes, antigen-presenting cells, and parenchymal cells. Thus control of cytokine production is an important step in regulating the immune response.
Evidence is accumulating that the outcome of an immune response can be influenced by individual variation in cytokine receptor genes (4, 5, 6). However, cytokines do not act in isolation, but form a complex network of interacting proteins; thus the net cytokine milieu may be the product of variation in many polymorphic cytokine genes. In transplantation, the situation is further obscured by the fact that cytokines produced within the graft may be derived from both the recipient and the donor. As recipient lymphocytes play a central role in mediating the acute rejection response, previous studies have focused on the impact of recipient cytokine polymorphisms on outcome after transplantation (7–9), but cytokines produced by donor-derived graft tissue may also contribute to the milieu, and may therefore influence the rejection response. Our purpose was to explore the impact of donor cytokine and cytokine receptor gene polymorphisms on the incidence and severity of acute rejection after renal transplantation.
MATERIALS AND METHODS
Patient groups and selection criteria.
This study was approved by the Central Oxford Research Ethics Committee. Cytokine genotyping was performed on 145 caucasoid cadaveric organ donors who were selected on the basis of the presence or absence of acute rejection of the kidney after transplantation (Table 1). Acute rejection was defined using clinical, biochemical, and histological criteria including, in all cases, a rise in creatinine >15% above baseline and characteristic transplant biopsy features. Steroid-responsive acute rejection was said to have occurred if an individual experienced a single episode of acute rejection within 30 days of transplantation with a serum creatinine that returned to baseline following a course of methylprednisolone (usually 500 mg daily for 3 days). If the creatinine failed to fall or return to baseline, patients underwent a further biopsy to confirm the diagnosis of continuing acute rejection, and were classified as having steroid-resistant rejection. These individuals were then treated with antilymphocyte agents. Patients included in the “no acute rejection” group were also selected by specific criteria, namely the absence of a rise in creatinine in the first 30 days after transplantation, and no evidence of tubulitis, vasculitis, or a lymphocytic infiltrate on protocol biopsy (day 7 and 30). Individuals who failed to meet these criteria for the presence or absence of rejection were excluded from the study, and thus this group does not comprise a consecutive cohort.
A total of 56 individuals donated two kidneys to local recipients, and were classified by the rejection status of only one of the recipients, selected at random. Rejection status was classified by a single author (AMcL), and all genotyping was performed without knowledge of the rejection status of each individual. All recipients were treated with triple immunosuppressive therapy, namely cyclosporin, azathioprine, and prednisolone, and none received induction therapy with anti-lymphocyte agents. None were highly sensitized (panel reactive antibodies>85%) at the time of transplantation.
IL-6 genotype was also determined on 209 caucasoid renal allograft recipients, who were selected according to the same criteria as the donors. (Full cytokine and cytokine receptor genotyping of these individuals is the focus of an additional manuscript (10). IL-6 genotype was available for both donor and recipient in 126 pairs. A donor-recipient pair was considered to be mismatched for a polymorphism if the donor genotype included an allele not identified in the recipient.
All donors were genotyped for 20 polymorphisms in 11 cytokine and cytokine receptor genes, namely: IL-1 α [−889t/c (11), IL-1 β (+3962 t/c (12)], IL-1 receptor [Pst I 970c/t (13)], IL-1 receptor antagonist [Msp I 11100t/c (14)], IL-4 [−590t/c (15)], IL-4 receptor (1902 g/a (16)], IL-6 [−174 g/c (17)], IL-10 [−1082, −819, −592 (18)], tumor necrosis factor [−308, −238, +488 (19)], lymphotoxin [+249, +365, +720 (19)], and TGF-β [−880 g/a, −509c/t, aa10L/P (20)].
Not all these polymorphisms have clear physiological significance, but this did not preclude their analysis, as they may serve as useful indicators for the involvement of other unidentified linked allelic variants in disease pathogenesis. All genotyping was conducted with PCR-SSP assays (polymerase chain reaction with sequence specific primers) that use identical amplification and detection conditions, enabling rapid and cost-efficient analysis of multiple polymorphisms. Assays for tumor necrosis factor, lymphotoxin, and IL-10 use both forward and reverse allele-specific primers, so that cis combinations of alleles can be directly amplified. Previous studies have demonstrated that, using this method, haplotypes can be inferred using a limited number of primer combinations due to linkage disequilibrium (18, 19). The IL-6–174 polymorphism was detected using the following primers: IL-6–174G sense primer 5′ TCGTGCATGACTTCAGCTTTA (3.4 μM) and antisense primer 5′ AATGTGACGTCCTTTAGCATC (3.4 μM); IL-6–174c sense primer 5′ TCGTGCATGACTTCAGCTTTA (3.4 μM) and antisense primer 5′ AATGTGACGTCCTTTAGCATG (3.4 μM). All other assays have been previously described (21, 19, 22, 23), and all use previously reported protocols for DNA extraction, PCR amplification, and gel electrophoresis (24). Full primer sequences and reaction conditions are available from the authors (see Appendix 1).
Study design and statistical analysis.
Phenotype, genotype, and allele frequencies were measured for all polymorphisms. Phenotype frequencies were obtained by counting the number of individuals in a population positive for an allele; allele frequencies were obtained by directly counting the number of chromosomes bearing an allele. Associations were assessed using 2×2 or 2×n contingency table analysis and the χ2 test, with Yates’ correction or Fisher’s exact tests where appropriate. Odds ratio (OR) and the OR 95% confidence interval (CI) were also calculated for significant associations.
To correct probability values for multiple comparisons, we applied a “two set” approach (25). Genotype, allele, and phenotype frequencies were first analyzed in a predetermined, randomly selected first set of 82 donors. Genotyping was then repeated in the second set of 63 donors. Only those associations found to be associated with rejection (P <0.05) in both cohorts were considered to be significant, and the data was merged.
Results of the analysis of 20 cytokine and cytokine receptor polymorphisms are shown in Table 2. This revealed an association of a polymorphism in the promoter of the donor IL-6 gene with acute rejection. The association of donor IL-6–174c/c genotype with acute rejection was demonstrated in both the first set of 82 donors, and the second set of 63 donors, and when the two cohorts were merged (overall P value=0.0002, OR 8.67, Table 3). No other polymorphism was associated with acute rejection according to the study criteria. Further analysis demonstrated that donor IL-6 genotype was also associated with severity of acute rejection, as defined by response to methylprednisolone (P no rejection versus steroid-resistant rejection=0.000007, OR 15.96;Table 4A).
A major genetic determinant of acute rejection after kidney transplantation is donor-recipient mismatching at HLA-DR (26), and as expected, this was also true in the selected patients studied here. After stratification for HLA-DR matching, the association of donor IL-6 genotype with rejection remained significant, both when the no rejection group was compared with the all rejection group, and when the steroid-resistant subgroup was analyzed separately (Table 5).
As recipient cells may also contribute to local IL-6 production we examined IL-6 genotype in 209 recipients of cadaveric renal transplants selected according to the same criteria as the donors. No influence of recipient IL-6 genotype on acute rejection was detected (Table 4B (10). Furthermore, no influence of donor-recipient IL-6 genotype mismatching on acute rejection incidence or severity was detected in 126 donor-recipient pairs (data not shown).
A total of 56 renal transplant donors had donated both kidneys to recipients at our transplant center. All had been classified according to the outcome of only one of the recipient pair, chosen at random. To determine whether donor IL-6–174c/c genotype could predict recipient outcome, we analyzed the positive predictive value of this polymorphism for acute rejection in the paired recipients who had not been used to classify the donors. Despite a relatively small sample size, this confirmed that donor IL-6 genotype was strongly predictive of recipient acute rejection (P =0.02, positive predictive value 78%, Table 6).
This study identifies a new factor associated with the development of acute rejection after renal transplantation. Analysis of polymorphisms in eleven cytokine and cytokine receptor genes demonstrated that donor IL-6 genotype was strongly associated with the incidence and severity of acute rejection. This was significant in a “two set” analysis, and furthermore analysis of paired kidneys demonstrated that donor IL-6–174c/c is strongly predictive of acute rejection. The effect of IL-6 genotype on both incidence and severity of acute rejection is independent of the other major genetic determinant of acute rejection, namely HLA-DR matching. Analysis of recipient pairs who had both received a kidney from the same donor provided an opportunity to determine the predictive value of donor IL-6 genotype for acute rejection, as both recipients were likely to have been equally subject to any temporal differences in patient management. Despite a relatively small patient group, donor IL-6–174c/c genotype was again shown to be strongly predictive of both incidence and severity of acute rejection.
This is the first demonstration that donor cytokine genotype impacts on the recipient immune response to a transplant. IL-6 is a pleotropic cytokine involved in many different aspects of the response to inflammation. It is produced by a wide range of cells of both hematopoetic and nonhematopoetic origin, including macrophages, lymphocytes, endothelial cells, mesangial cells, and vascular smooth muscle (27). IL-6 is an integral mediator of the acute phase response, and modulates both local and systemic immunity (28, 29). Although it was originally thought to have predominantly pro-inflammatory effects, IL-6 has now been recognised to have considerable anti-inflammatory properties as well (30, 31, 32).
Renal allograft rejection has been associated with increased intragraft IL-6 protein and mRNA production in a number of studies, and it may be produced by both renal tubular cells and graft-infiltrating mononuclear cells (33, 34). However, other studies have failed to demonstrate any correlation between IL-6 production and rejection (35, 36), but perhaps this inconsistency is not surprising, given the pleotropic effects of IL-6 in vivo. The source of IL-6 in rejecting allografts has not been addressed: although recipient mononuclear cells within the rejecting graft may express IL-6 (34), IL-6 expression is also increased on donor-derived parenchymal tissue (33) and may be expressed by donor tissue before implantation (37). This suggests that at least some of the IL-6 present within a graft is derived from donor tissue, and thus supports the hypothesis that donor IL-6 gene variants may influence the recipient immune response within the graft.
A number of hypotheses are suggested by the data that donor IL-6 genotype is associated with acute rejection. One possibility is that differential production of IL-6 in the vicinity of an immune response can significantly influence that immune response. However, we failed to demonstrate any association of recipient IL-6 genotype with acute rejection, suggesting either that the effect of donor IL-6 genotype within the graft overwhelms any contribution by recipient lymphocytes, or that before implantation donor IL-6 genotype determines events critical to acute rejection. Another explanation is that the IL-6 polymorphism directly elicits graft rejection by acting as a minor transplantation antigen. These polymorphic proteins are not involved in antigen presentation, but can nevertheless initiate rejection of a transplant if the donor and recipient express different allelic forms (38). However, no influence of mismatching of donor-recipient IL-6 genotype was identified.
Although the mechanism by which donor IL-6 genotype exerts its effect remains unknown, it is likely to be associated with differential production or activity of the IL-6 protein. This may be a direct effect of the single nucleotide substitution altering promoter activity, or indirectly through linkage with an additional, as yet unrecognized polymorphism. Indeed, the IL6–174 polymorphism has been associated with differential basal and induced cytokine levels (17), with the –174C allele being associated with decreased IL-6 production both in vivo and in vitro. Although at first glance this appears contradictory, a number of possible explanations may be postulated: defining the biological significance of individual cytokine promoter variants is notoriously difficult to discern (39) and the association between promoter alleles and levels of cytokine production may be complex. Furthermore, IL-6 has profound anti-inflammatory as well as pro-inflammatory effects, and increased IL-6 produced by donor-derived dendritic cells may result in the differentiation of recipient lymphocytes to a nonaggressive, IL-4-producing phenotype (30). Alternatively, donor IL-6 genotype may exert its effect not through any direct action, but through its influence on response to immunosuppressive therapy. The –174 polymorphism is found near two steroid response elements (17), and immunosuppressive therapies including steroids and cyclosporin have been shown to influence the transcripton of the IL-6 gene (40). This polymorphism may therefore confer differential responsiveness of the IL-6 gene to corticosteroids. As corticosteroids are a mainstay of the standard immunosuppressive regimen in our unit, and are used both for maintenance immunosuppression and as first-line treatment for graft rejection, the donor IL-6–174c/c genotype may act as a surrogate marker for individuals who are inadequately immunosuppressed using corticosteroid-based therapy.
In summary, this study identifies donor IL-6 genotype as a major genetic risk factor for the development of acute rejection after renal transplantation, independent of donor-recipient HLA-DR matching. This may enable the development of therapeutic algorithms to predict those individuals at particularly high risk of acute renal allograft rejection.Copyright © 2001 Wolters Kluwer Health, Inc. All rights reserved.