Acute and chronic renal graft rejection influences the results of kidney transplantation and thus remains the major obstacle in the success of renal transplantation. In addition to human leukocyte antigen loci, several genes encoding cytokines and their receptors have been recently suggested to play an important role in the process of rejection (1–3).
As the pathogenesis of rejection is complex, there are many candidate genes that could potentially influence individual susceptibility. Tumor necrosis factor-alpha (TNF-α), transforming growth factor-beta (TGF-β), interferon gamma (IFN-γ), regulated upon activation normal T-cell expressed and secreted (RANTES), monocyte chemotactic protein-1 (MCP-1), CCR2, and CCR5 receptors have been all shown to be involved in pathogenesis of renal allograft rejection in experimental setting.
Animal models of acute and chronic kidney allograft rejection showed an increased intragraft expression of TGF-β (4), TNF-α, IFN-γ, RANTES, and MCP-1 (5) compared to control animals without rejection. In a mouse renal interstitial fibrosis model, Ccr2 and Ccr5 knockout animals exhibited reduced fibrosis comparable to that of their wild-type littermates (6, 7). Furthermore, overexpression of TGF-β in rats leads to mesangial expansion, accumulation of glomerular matrix proteins, and interstitial fibrosis leading to progressive glomerulosclerosis (8).
Although it is well accepted that cytokine-mediated inflammation promotes the development of cellular injury, interstitial fibrosis, and tubular atrophy with subsequent allograft dysfunction, there is less evidence that genetic variability in the cytokines determines the susceptibility for kidney graft rejection.
Recently, numerous polymorphisms in the promoter regions of TNF-α, MCP-1, RANTES and in the translated regions of CCR2, CCR5, IFN-γ, and TGF-β genes have been identified. In vitro studies showed that some of them influence either the expression of the respective gene or the function of the respective protein. Accordingly, the TNF-α −308G/A and the MCP-1 promoter polymorphisms −2518A/G enhance the production of TNF-α (9) and MCP-1, respectively, by monocytes (10). Haplotype [−403A; −109C; −28G] in the promoter of RANTES increase transcription of the gene (11). The polymorphism CCR2 +190G/A, corresponding to the amino acid exchange Val64Ile in the second exon of the gene, increases the transcription of CCR2 (12), whereas deletion of 32-base pairs in the coding region of CCR5 results in a frameshift and premature termination of translation at codon 185 (13). Polymorphism IFN-γ +874A/T is located within a binding site for the transcription factor nuclear factor-κB and causes a significant increase of IFN-γ transcription (14). Finally, two polymorphisms +869T/C and +915G/C in the coding region of TGF-β, corresponding to the amino acid exchange Leu10Pro and Arg25Pro, impeded the production of TGF-β by lymphocytes (15).
Consequently, the role of these polymorphisms in the susceptibility to the allograft nephropathy in humans has been studied and several groups reported an association with acute and chronic kidney graft rejection (16–22). Indeed, these findings could be utilized in the identification of kidney allograft recipients predisposed to allograft rejection and thus potentially benefiting from tailored immunosuppression. However, in view of the small numbers of individuals involved in the studies resulting in inadequate statistical power, and the fact that none of the reported associations has been replicated in an independent sample of patients, we sought to investigate the frequency of these polymorphisms in a large cohort of well-characterized patients who underwent renal transplantation.
PATIENTS AND METHODS
We consecutively included 436 white kidney transplant recipients who had undergone kidney transplantation at the Institute for Clinical and Experimental Medicine, Prague, the Czech Republic, from 1999 to 2004. In the posttransplant course, all patients received either cyclosporine or tacrolimus along with mycophenolate mofetil and steroids, recipients with panel reactive antibodies >50% received prophylaxis by muromonab-OKT3 or anti-thymocyte globulin. All acute rejection episodes (AR) were determined according to Banff 97 criteria (23) and were biopsy proven; borderline changes were included.
Clinical and laboratory data were collected on the date that the protocol 12-month biopsy was performed. A complete physical examination was undertaken in all patients. Laboratory evaluation included creatinine, total cholesterol, and total triglycerides. The glomerular filtration rate was estimated by the Cockcroft-Gault formula (24).
The protocol 12-month biopsy was performed in 273 patients with functioning kidney graft out of 436 patients who gave an inform consent with the protocol biopsy. Biopsies was performed using Tru-cut needle (Uni-Cut Nadeln, Angiomed, Germany) guided by ultrasound (Toshiba, Power Vision 6000, Japan). Samples were routinely stained with hematoxylin and eosin, periodic acid-Schiff, aldehyde-fuchsin orange G, Sirius red with elastic stain, and periodic acid silver-methenamine. Biopsy tissues were scored on the basis of the Banff 97 working classification (23). Subclinical rejection (SR) was defined as histological findings of AR including borderline changes at 12-month protocol biopsy and stable kidney graft function.
The study was approved by the institutional review board and written informed consent was obtained from all subjects.
Genotyping of Polymorphisms
The genomic DNA was isolated from whole blood samples using a commercial kit (Whole blood DNA purification kit; Fermentas, Canada). Single nucleotide polymorphisms were determined by polymerase chain reaction (PCR) followed by restriction fragment length polymorphism analysis: RANTES, MCP-1, CCR2 (19, 25–27) or by sequence specific priming (PCR-SSP): TNF-α, IFN-γ and TGF-β (28–30). The insertion-deletion 32bp polymorphism in CCR5 was determined by a simple PCR method (20).
To minimize genotyping errors, blank controls wells were left on the PCR plates and assays were wholly retyped in the call rate was under 80%. Three operators (I.B., P.H., K.H.) independently performed genotype assignment. Genotyping was duplicated in case of discrepancy between operators. After testing for Hardy-Weinberg equilibrium, allele frequencies were checked for consistency with data from Utah residents with Northern and Western European ancestry (CEU) population of European ancestry from the HapMap database (31).
We calculated the sample size required to detect the effect of an polymorphism on the risk of outcome using the DSTPLAN software (http://linkage.rockefeller.edu/soft). When the odds ratio (OR) of a polymorphism was assumed to be 2, the required sample size was 110 cases and 110 controls for the polymorphism with the variant allele frequency of 0.5 (TGF-β +915G/C). When the OR was assumed to be 4, the same sample size was sufficient to detect a true effect of a polymorphism with the allele frequency of 0.04 (RANTES −28C/G). In a model with 190 patients and 190 controls, the minimal detectable OR values for the same allele frequencies were 1.8 and 3.1, respectively. The calculations have been performed at a 5% level of significance for 80% statistical power.
Hardy-Weinberg equilibrium of alleles at individual loci was evaluated using the chi-square test. Linkage disequilibrium coefficients D’=D/Dmin or max and r2 were calculated using the standard formulas (32). The coefficients D’ and r2, ranging from 0 to 1, denote the strength of the LD. The value of D’=1 indicates a complete LD if one haplotype is not observed and r2=1 indicates a perfect LD in case that two haplotypes are not observed.
To evaluate the effect of continuous variables on the respective clinical outcomes, one-way analysis of variance was used. Single-locus association analyses with calculation of OR and 95% confidence intervals (CI) were performed by univariate logistic regression analysis using SPSS version 14.0 (SPSS Inc., Chicago, IL). Subsequently, multivariate regression adjusted for total number of mismatches, mismatches in DR locus, and immunosuppression regimen (either cyclosporine or tacrolimus-based) was performed. For haplotype analysis, univariate logistic regression was used. To test the effect of the polymorphisms on the graft survival, the Kaplan-Meier analysis was used. The level of significance was set at P<0.05. All P values were two-sided.
The age median of transplanted patients was 49.2 (range 19.0–76.0) years; of them, 287 (65.8%) were male. The mean number of mismatches between donor and recipient in human leukocyte antigen (HLA)-A, -B, and -DR loci was 3.1±1.2 and mean percentage of panel reactive antibodies was 19.7±27.3. The donor age median was 46.0 (range 2.0–76.0) years.
Upon the 12-month protocol kidney graft biopsy, chronic allograft nephropathy (CAN) was found in 122 (44.5%) patients (CAN grade I in 82 patients, grade II in 27 patients and grade III in 13 patients). Subclinical rejection was present in 38 patients (Table 1). Patients with CAN had higher creatinine level and proteinuria with correspondingly lower glomerular filtration rate values. Moreover, 35% of patients with CAN presented with AR during the first posttransplant year, compared to 21% of patients without CAN. One-hundred and ninety patients with AR had significantly increased creatinine level and proteinuria, compared to those without AR (Table 2). Apart from that, no clinically meaningful differences between the groups were observed.
All alleles at individual loci were in Hardy-Weinberg equilibrium with nonsignificant χ2 values in all groups. The genotype frequencies in the control groups for all polymorphisms were in concordance with the reference HapMap database (31). There was a significant linkage disequilibrium (LD) between the TGF-β +869T/C and TGF-β +915G/C loci, with 69% of the inferred haplotypes [TGF-β +869; TGF-β +915] consisting of either T-G or C-C. The linkage between these two loci calculated in the CAN-negative control group was almost complete (D’=0.99), but not perfect (r2=0.11).
Single Locus Analysis
Neither univariate analysis nor multivariate analysis adjusted for number of mismatches, DR-locus haplotype and immunosuppression regimen showed significant difference in the distribution of the genotype frequencies between patients with and without AR (Table 3), and between patients with CAN or SR and individuals with normal 12-month protocol biopsy (Table 4). Furthermore, no influence of any polymorphism on the graft survival was observed (Fig. 1).
Haplotype (TGF-β +869G; TGF-β +915C) seemed to be associated with the presence of SR (OR: 3.45, 95% confidence interval: 1.19–9.99, P=0.023), but the association was nonsignificant due to the insufficient power (data not shown).
In the white population, promoter polymorphisms TNF-α −308G/A, MCP-1 −2518A/G, RANTES −403G/A, −109T/C and −28C/G, and exonal polymorphisms CCR2+ 190G/A, CCR5 del32bp, IFN-γ +874A/T, TGF-β +869T/C and +915G/C have been associated with an increased risk of systemic lupus erythematodes (33), psoriatic arthritis (34), human immunodeficiency virus resistance (35, 36), tuberculosis (14), and progression of chronic renal insufficiency (37), supporting their role in the pathogenesis of human inflammatory disorders. In line with these reports, several groups found an association of the polymorphisms with acute or chronic kidney allograft rejection, with odds ratio for acute graft rejection ranging from 1.8 (38) to 10 (16) and for chronic allograft nephropathy ranging from 1.8 (39) to 3.1 (22).
In our study, we included only those cytokines whose role in the pathogenesis of renal allograft rejection was proved in animal models and those polymorphisms whose impact on the expression of the respective gene or the function on the respective protein was described in in vitro studies. However, in contrast to the published reports (16, 22, 40, 41), we found no association between any of the functionally relevant polymorphisms and the risk of acute, subclinical rejection as well as chronic allograft nephropathy. There are several potential explanations for the discrepancy between our results and those of previous studies. These explanations refer to the currently accepted prerequisites for the design of genotype/phenotype association studies (42, 43), to which the previous studies did not closely adhere.
At the outset, there should be a logical rationale for the chosen candidate genes and a coherent hypothesis based on the functional significance of the studied genetic variants. As regards the former, previous papers correctly studied cytokines involved in the Th1 and Th2-mediated pathogenesis of renal allograft rejection. However, in a study published by Sankaran et al. (16), individuals with hypersecretory phenotype of both TNF-α and interleukin (IL)-10 had an increased risk of acute allograft rejection, although IL-10 is an example of anti-inflammatory cytokine. In a transgenic mouse model, IL-10 inhibited production of TNF-α and neutrophil accumulation (44), which seems contradictory to the synergism of both cytokines postulated by Sankaran et al. (16).
Then, several functionally-related genes should be tested since this approach has a higher chance to detect genetic risk factors than the screening of single genetic variants. However, as an increasing number of comparisons increases the false-positive rate (45), an appropriate correction for multiple testing has to be implemented. Such correction would show that the association of the TNF-α −308A, TGF-β +915G and IL-10 −1082G alleles with acute allograft rejection reported by Alakulppi et al. (40) is a flawed interpretation of the data.
Furthermore, subjects with comparable baseline characteristics should be included in order to eliminate possible confounding (46). In particular, the influence of genetic variability on rejection could be masked by immunosuppression even when present. For instance, Asderakis et al. (22) reported an increased risk for acute allograft rejection for carriers of the allele *2 of the IFN-γ microsatellite polymorphism. The association was found only in the subgroup of 28 patients on cyclosporine monotherapy, but not in the whole cohort of 88 patients, of whom 60 received also steroids and azathioprine. It is likely that some of these individuals carrying the high producer IFN-γ *2 allele might have developed rejection if they had been only on cyclosporine. The treatment heterogeneity caused an underestimation of patients with acute rejection and subsequent failure to assign them into the appropriate outcome group.
In addition, the reliability of the genotyping assays should be assured and the test for Hardy-Weinberg equilibrium (HWE) should be performed. Deviation from the HWE points at systematic genotyping error, which hampers any interpretation of the results, as for example in the study of Dmitrienko et al. (41) in which the low P value for HWE at the TNF-α −308 locus in the control group of healthy subjects indicates a significant genotyping inaccuracy.
Finally, the most obvious explanation for the discrepancy between our findings and previous studies is the lack of statistical power in vast majority of them. The key determinant of quality in an association study is sample size (43), which should be determined by the power calculation in the study-designing phase. Results obtained from inadequately powered studies tend to have a decreased probability of detecting a true effect of a polymorphism due to the type II error (false negativity). Moreover, for any choice of significance level, the proportion of false-positive results among all positive results (type I error) is greatly increased as power decreases (47). Our calculation revealed that none of the studies reporting an association of polymorphisms in the proinflammatory genes with renal allograft rejection complied with the current demands for 80% power (16, 22, 38, 40, 41, 48). Moreover, none of the studies included an independent validation cohort, which should be implemented, particularly in small studies that are likely to overestimate the true effect size (49).
In conclusion, none of the functionally relevant polymorphisms in the TNF-α, MCP-1, RANTES, CCR2, CCR5, IFN-γ and TGF-β genes increased the risk of acute, chronic, or subclinical renal allograft rejection in our cohort. Although there seems a little doubt that Th1- and Th2-mediated immune reactions play a role in the pathogenesis of renal allograft rejection, hereditary susceptibility to acute or chronic renal allograft rejection in white subjects has not been unequivocally proven. It remains possible that genetic predisposition to renal allograft rejection is determined by multiple polymorphisms with a low individual contribution to the phenotype that cannot be assessed in allelic association studies. Alternatively, other as yet unidentified polymorphisms may significantly affect the risk of renal allograft rejection. Therefore, identification of such polymorphisms warrants future studies that will be fully compliant with the currently accepted standards for polymorphism-disease association studies.
We thank Alena Louzecka for excellent technical assistance, Michaela Prokopova for her help with clinical data collection, Věra Lánská for her help with statistical analysis, and Milan Jirsa for comments.
1. Waller JR, Nicholson ML. Molecular mechanisms of renal allograft fibrosis. Br J Surg
2001; 88: 1429.
2. Baan CC, Weimar W. Intragraft cytokine gene expression: Implications for clinical transplantation. Transpl Int
1998; 11: 169.
3. Khan F, Agrawal S, Agrawal S. Genetic predisposition and renal allograft failure: Implication of non-HLA genetic variants. Mol Diagn Ther
2006; 10: 205.
4. Paul LC, Saito K, Davidoff A, Benediktsson H. Growth factor transcripts in rat renal transplants. Am J Kidney Dis
1996; 28: 441.
5. Nadeau KC, Azuma H, Tilney NL. Sequential cytokine dynamics in chronic rejection of rat renal allografts: Roles for cytokines RANTES and MCP-1. Proc Natl Acad Sci U S A
1995; 92: 8729.
6. Kitagawa K, Wada T, Furuichi K, et al. Blockade of CCR2 ameliorates progressive fibrosis in kidney. Am J Pathol
2004; 165: 237.
7. Eis V, Luckow B, Vielhauer V, et al. Chemokine receptor CCR1 but not CCR5 mediates leukocyte recruitment and subsequent renal fibrosis after unilateral ureteral obstruction. J Am Soc Nephrol
2004; 15: 337.
8. Sanderson N, Factor V, Nagy P, et al. Hepatic expression of mature transforming growth factor beta 1 in transgenic mice results in multiple tissue lesions. Proc Natl Acad Sci U S A
1995; 92: 2572.
9. Wilson AG, Symons JA, McDowell TL, et al. Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation. Proc Natl Acad Sci U S A
1997; 94: 3195.
10. Rovin BH, Lu L, Saxena R. A novel polymorphism in the MCP-1 gene regulatory region that influences MCP-1 expression. Biochem Biophys Res Commun
1999; 259: 344.
11. Guignard F, Combadiere C, Tiffany HL, Murphy PM. Gene organization and promoter function for CC chemokine receptor 5 (CCR5). J Immunol
1998; 160: 985.
12. Nakayama EE, Tanaka Y, Nagai Y, et al. A CCR2–V64I polymorphism affects stability of CCR2A isoform. AIDS
2004; 18: 729.
13. Samson M, Libert F, Doranz BJ, et al. Resistance to HIV-1 infection in caucasian individuals bearing mutant alleles of the CCR-5 chemokine receptor gene. Nature
1996; 382: 722.
14. Rossouw M, Nel HJ, Cooke GS, et al. Association between tuberculosis and a polymorphic NFkappaB binding site in the interferon gamma gene. Lancet
2003; 361: 1871.
15. Awad MR, El-Gamel A, Hasleton P, et al. Genotypic variation in the transforming growth factor-beta1 gene: Association with transforming growth factor-beta1 production, fibrotic lung disease, and graft fibrosis after lung transplantation. Transplantation
1998; 66: 1014.
16. Sankaran D, Asderakis A, Ashraf S, et al. Cytokine gene polymorphisms predict acute graft rejection following renal transplantation. Kidney Int
1999; 56: 281.
17. Pelletier R, Pravica V, Perrey C, et al. Evidence for a genetic predisposition towards acute rejection after kidney and simultaneous kidney-pancreas transplantation. Transplantation
2000; 70: 674.
18. Kruger B, Schroppel B, Ashkan R, et al. A Monocyte chemoattractant protein-1 (MCP-1) polymorphism and outcome after renal transplantation. J Am Soc Nephrol
2002; 13: 2585.
19. Abdi R, Tran TB, Sahagun-Ruiz A, et al. Chemokine receptor polymorphism and risk of acute rejection in human renal transplantation. J Am Soc Nephrol
2002; 13: 754.
20. Fischereder M, Luckow B, Hocher B, et al. CC chemokine receptor 5 and renal-transplant survival. Lancet
2001; 357: 1758.
21. McDaniel DO, Barber WH, Nguyan C, et al. Combined analysis of cytokine genotype polymorphism and the level of expression with allograft function in African-American renal transplant patients. Transpl Immunol
2003; 11: 107.
22. Asderakis A, Sankaran D, Dyer P, et al. Association of polymorphisms in the human interferon-gamma and interleukin-10 gene with acute and chronic kidney transplant outcome: The cytokine effect on transplantation. Transplantation
2001; 71: 674.
23. Racusen LC, Solez K, Colvin RB, et al. The Banff 97 working classification of renal allograft pathology. Kidney Int
1999; 55: 713.
24. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron
1976; 16: 31.
25. Szalai C, Duba J, Prohaszka Z, et al. Involvement of polymorphisms in the chemokine system in the susceptibility for coronary artery disease (CAD). Coincidence of elevated Lp(a) and MCP-1-2518 G/G genotype in CAD patients. Atherosclerosis
2001; 158: 233.
26. Azzawi M, Pravica V, Hasleton PS, Hutchinson IV. A single nucleotide polymorphism at position −109 of the RANTES proximal promoter. Eur J Immunogenet
2001; 28: 95.
27. al Sharif F, Ollier WE, Hajeer AH. A rare polymorphism at position −28 in the human RANTES promoter. Eur J Immunogenet
1999; 26: 373.
28. Verjans GM, Brinkman BM, Van Doornik CE, et al. Polymorphism of tumour necrosis factor-alpha (TNF-alpha) at position −308 in relation to ankylosing spondylitis. Clin Exp Immunol
1994; 97: 45.
29. Pravica V, Perrey C, Stevens A, et al. A single nucleotide polymorphism in the first intron of the human IFN-gamma gene: Absolute correlation with a polymorphic CA microsatellite marker of high IFN-gamma production. Hum Immunol
2000; 61: 863.
30. Hubacek JA, Lacha J. Alternative method for diagnosis of two polymorphisms in the human transforming growth factor-beta1 by PCR-mediated double site-directed mutagenesis. Clin Chim Acta
2000; 295: 187.
31. The International HapMap Project. Nature
2003; 426: 789.
32. Morton NE. Linkage disequilibrium maps and association mapping. J Clin Invest
2005; 115: 1425.
33. Lee YH, Harley JB, Nath SK. Meta-analysis of TNF-alpha promoter −308 A/G polymorphism and SLE susceptibility. Eur J Hum Genet
2006; 14: 364.
34. Balding J, Kane D, Livingstone W, et al. Cytokine gene polymorphisms: association with psoriatic arthritis susceptibility and severity. Arthritis Rheum
2003; 48: 1408.
35. Modi WS, Goedert JJ, Strathdee S, et al. MCP-1-MCP-3-Eotaxin gene cluster influences HIV-1 transmission. AIDS
2003; 17: 2357.
36. Liu H, Chao D, Nakayama EE, et al. Polymorphism in RANTES chemokine promoter affects HIV-1 disease progression. Proc Natl Acad Sci U S A
1999; 96: 4581.
37. Khalil MS, El Nahas AM, Blakemore AI. Transforming growth factor-beta1 SNPs: genetic and phenotypic correlations in progressive kidney insufficiency. Nephron Exp Nephrol
2005; 101: e31.
38. Tinckam K, Rush D, Hutchinson I, et al. The relative importance of cytokine gene polymorphisms in the development of early and late acute rejection and six-month renal allograft pathology. Transplantation
2005; 79: 836.
39. Melk A, Henne T, Kollmar T, et al. Cytokine single nucleotide polymorphisms and intrarenal gene expression in chronic allograft nephropathy in children. Kidney Int
2003; 64: 314.
40. Alakulppi NS, Kyllonen LE, Jantti VT, et al. Cytokine gene polymorphisms and risks of acute rejection and delayed graft function
after kidney transplantation. Transplantation
2004; 78: 1422.
41. Dmitrienko S, Hoar DI, Balshaw R, Keown PA. Immune response gene polymorphisms in renal transplant recipients. Transplantation
2005; 80: 1773.
42. Cooper DN, Nussbaum RL, Krawczak M. Proposed guidelines for papers describing DNA polymorphism-disease associations. Hum Genet
2002; 110: 207.
43. Hattersley AT, McCarthy MI. What makes a good genetic association study? Lancet
2005; 366: 1315.
44. Lee CG, Homer RJ, Cohn L, et al. Transgenic overexpression of interleukin (IL)-10 in the lung causes mucus metaplasia, tissue inflammation, and airway remodeling via IL-13-dependent and -independent pathways. J Biol Chem
2002; 277: 35466.
45. Altman DG MD, Bryant TN, Gardner MJ, eds. Statistics with Confidence BJM Books. BJM Books. Bristol, UK: Blackwell Publishing Ltd, 2000: 163.
46. Miettinen OS, Cook EF. Confounding: essence and detection. Am J Epidemiol
1981; 114: 593.
47. Kirkwood BSJ. Essential medical statistics. 2nd ed. Boston: Blackwell Science; 2003: 426.
48. Marshall SE, McLaren AJ, Haldar NA, et al. The impact of recipient cytokine genotype on acute rejection after renal transplantation. Transplantation
2000; 70: 1485.
49. Wacholder S, Chanock S, Garcia-Closas M, et al. Assessing the probability that a positive report is false: An approach for molecular epidemiology studies. J Natl Cancer Inst
2004; 96: 434.