Intragraft Regulatory T Cells in Protocol Biopsies Retain Foxp3 Demethylation and Are Protective Biomarkers for Kidney Graft Outcome

Authors


Oriol Bestard, 35830obm@comb.es

Abstract

Presence of subclinical rejection (SCR) with IF/TA in protocol biopsies of renal allografts has been shown to be an independent predictor factor of graft loss. Also, intragraft Foxp3+ Treg cells in patients with SCR has been suggested to differentiate harmful from potentially protective infiltrates. Nonetheless, whether presence of Foxp3 Treg cells in patients with SCR and IF/TA may potentially protect from a deleterious graft outcome has not yet been evaluated. This is a case-control study in which 37 patients with the diagnosis of SCR and 68 control patients with no cellular infiltrates at 6-month protocol biopsies matched for age and time of transplantation were evaluated. We first confirmed that numbers of intragraft Foxp3-expressing T cells in patients with SCR positively correlates with Foxp3 demethylation at the Treg-specific demethylation region. Patients with SCR without Foxp3+ Treg cells within graft infiltrates showed significantly worse 5-year graft function evolution than patients with SCR and Foxp3+ Treg cells and those without SCR. When presence of SCR and IF/TA were assessed together, presence of Foxp3+ Treg could discriminate a subgroup of patients showing the same graft outcome as patients with a normal biopsy. Thus, presence of Foxp3+ Treg cells in patients with SCR even with IF/TA is associated with a favorable long-term allograft outcome.

Abbreviations: 
Treg

regulatory T cells

SCR

subclinical cellular rejection

IF/TA

interstitial fibrosis and tubular atrophy

MMF

mycophenolate mofetil

TSDR

treg-specific demethylated region

BPAR

biopsy proven acute rejection

PRA

panel reactive alloantibodies

PTC

peritubular capillaries

SCr

serum creatinine

eGFR

estimated glomerular filtrate rate

Introduction

Regulatory T cells (Treg cells) are a well-recognized CD4+ T-cell subset population showing high levels of CD25 and very low levels of the CD127 surface marker and importantly, they display the intranuclear transcription factor Foxp3 which is essential for its development and function, being specific to its cellular lineage (1,2). Foxp3+ Treg cells have the capacity of controlling adaptive effector immune responses and indeed, mutations at the Foxp3 gene induce the development of a severe autoimmune disease (3). In organ transplantation, Foxp3+ Treg cells have been suggested to play a relevant role for the induction and maintenance of peripheral tolerance in many animal models and also in humans (4–7). Interestingly, although it has classically been postulated that alloantigen recognition and alloimmune regulation are mainly carried out in secondary lymphoid organs (8,9) it has also been suggested that this is not always absolutely required (10). In fact, Foxp3+ Treg cells have been found directly within allografts and importantly being crucial for allograft tolerance (11–13). In human renal transplantation increased numbers of intragraft Foxp3+ Treg cells has been suggested to be of importance for predicting a favorable graft outcome in certain clinical settings (14–16). Nonetheless, it has also been shown that more than being a biomarker of tolerance their presence might be merely as a consequence of T-cell activation (17,18). A potential explanation for this discrepancy is the imprecise characterization of such T-cell subset. In fact, although in mice Foxp3 expression is known to be essential for the development and function of Treg cells, in humans Foxp3 expression can also be transiently displayed upon T-cell activation (19–21). Therefore, when Foxp3 is characterized either at the protein or at the transcriptional level far from only assessing ‘true’ Treg cells effector T cells might be identified as well and especially when an ongoing effector alloimmune process such as acute rejection is taking place. Even if functional assays can state whether we are dealing or not with immunosuppressive Treg cells, this approach is time consuming and not always feasible. Noteworthy, a novel approach for a more precise identification of Foxp3 Treg cells has been reported. Indeed, because regulation via differential methylation contributes to the transcriptional control it has recently been demonstrated that demethylation of a conserved region in the first intron of the Foxp3 gene Treg-specific demethylated region (TSDR) constitutes the most reliable criterion for identification of such T-cell subset population (22,23).

Recently, and in agreement with previous reports in animal models of transplantation (7–9) we could show the potential value of the phenotypical characterization of Foxp3+ Treg cells within mononuclear cellular infiltrates in patients with the diagnosis of subclinical acute cellular rejection (SCR; Ref. 14). In fact, presence of Foxp3+ Treg cells among such group of renal transplanted patients was able to differentiate between harmful and potentially protective cellular infiltrates as shown by a better graft function evolution at 2 and 3 years after transplantation regardless of the type of immunosuppression.

Taking these preliminary observations into account, and to confirm that Foxp3+ expressing T cells by immunohistochemistry (IMH) among patients with SCR were essentially Treg cells rather than Foxp3+ expressing activated T cells, we first analyzed Foxp3 expression at the TSDR. Subsequently, we were also interested to further investigate whether the initial favorable effect in graft function evolution of Foxp3+ Treg cells infiltrates among patients with SCR would still be maintained in a longer period of follow-up. In addition, we wondered if any different clinical outcome could be detected when comparing these two groups of SCR patients to others with no acute cellular infiltrates at 6-month protocol biopsy. Finally, because the presence of chronic histologic graft lesions (IF/TA) with presence of SCR at 6-month protocol biopsies has been claimed to be an independent risk factor for late graft loss (24,25) we also investigated whether presence of Foxp3 Treg cells within SCR patients could even modulate the clinical outcome when chronic histologic lesions are also present.

Methods

Patients

Since 1988, prospective protocol biopsies are systematically performed in renal transplanted patients from our hospital who gave informed consent. Biopsies were selected from those previously blinded graded accordingly to the Banff’07 criteria (26) in absence of any clinical information. SCR was defined as presence of acute interstitial- and tubular-score ≥1.

Inclusion criteria: patients with a renal transplanted organ, biopsies performed at 6 months after transplantation, serum creatinine < 300 μmol/L, proteinuria < 1 g/24 h, stable renal function (defined as variability of serum creatinine of less than 15% during 2 weeks before and after biopsy) adequate tissue sample (considered as presence of at least 10 glomerular and two arterial sections) and immunosuppression consisting on at least one of the following: cyclosporine/tacrolimus, sirolimus or mycophenolate mofetil.

This is a retrospective, case-control 1:2 study. Nonexhausted paraffin blocks from 6-month protocol biopsies done until 2005 were selected. There were 170 biopsies performed in 170 patients suitable to be included. Among them, 37 cases fit the diagnosis of SCR. From the remaining 133 biopsies, we selected 68 controls with no evidence of acute rejection (acute interstitial and tubular score < 1) matched for age and time of transplantation in a 1:2 relation with the SCR patients. Chronic histologic changes were graded as chronic-interstitial and tubular scores ≥ 2.

Clinical data were obtained from our local transplant database. Graft function was analyzed at the time of biopsy, and every year until 5 years after transplantation by serum creatinine (μmol/L), glomerular filtrate rate (eGFR) estimated by the Cockroft–Gault formula (mL/min) and proteinuria (g/24 h).

Furthermore, eight more randomly selected patients with the diagnosis of biopsy proven acute T-cell rejection (BPAR; all of them showed acute clinical dysfunction, acute interstitial and tubular Banff’07 score ≥ 1 and were all steroid sensible to treatment) were also evaluated for the quantitative DNA methylation analysis.

IMH

IMH assessment of cellular allograft infiltrates of patients with the diagnosis of SCR was performed as previously described (13).

The number of positive FoxP3+ and CD3+ T cells in the cortex, perivascular areas and corticomedullary junction aggregates were first counted by using symmetric square power fields (×400; Leica Geosystems, Barcelona, Spain). Then, the proportion of FoxP3+ Treg among the CD3+ T cells (FoxP3+/CD3+) per field was also scored.

TSDR analysis

For the TSDR analysis we evaluated nine patients with the diagnosis of SCR and presence of Foxp3+ T cells within graft infiltrates using IMH that had enough tissue sample for being further analyzed. We also analyzed eight patients with the diagnosis of BPAR matched for the same Banff’07 acute score to the SCR patients and evaluated for Foxp3 expression using IMH and by the Foxp3 TSDR demethylation assay which was performed as previously reported (Figure 1; Refs. 22,23,27). In brief, for DNA isolation from Formalin-fixed paraffin-embedded (FEPE) tissues up to eight tissue slices were used. Slices were incubated overnight in 50 mmol/L of Tris/HCl (pH 8), 1 mmol/L EDTA, 0.5% Tween 20 containing 2 mg/mL of proteinase K with subsequent incubation at 90°C for 10 min and DNA was purified with Qia-AMP DNA FFPE tissue kit (Qiagen, Valencia, CA, USA) according to the manufacturer's protocol. DNA Preparation from Bisulfite Conversion Genomic DNA was isolated using the DNeasy tissue kit (Qiagen). Bisulfite treatment of genomic DNA was performed according to Wieczorek et al. (27) with minor modifications. Real-time PCR was performed in a final reaction volume of 20 AL using Roche LightCycler 480 Probes Master (Roche Diagnostics, Indianapolis, IN, USA) containing 15 pmol each of methylation or non–methylation-specific forward and reverse primers for TSDR, 5 pmol of hydrolysis probe, 200 ng of lambda DNA (New England Biolabs, Hitchin, Hertfordshire, UK) and 30 ng of bisulfite-treated genomic DNA template or a respective amount of plasmid standard. Each sample was analyzed in triplicate using a LightCycler 480 System (Roche). Cycling conditions consisted of a 95°C preheating step for 10 min and 50 cycles of 95°C for 15 s followed by 1 min at 61°C. Methylation- and demethylation-specific amplification primers for TSDR map to the following chromosomal positions: NCBI36:X:49004163-49004190:1 (forward primers) and NCBI36:X:49004227-49004251:1 (reverse primers). Hydrolysis probes map to chromosomal position NCBI36:X:49004200-49004222:1 (demethylation-specific probe) and NCBI36:X:49004200-49004217:1 (methylation-specific probe). For Plasmid Standardization, PCR products were generated with methyl- and non–methyl-specific primers for FOXP3 TSDR using genomic, bisulfite-treated. DNA fragments were cloned into pCR2.1-TOPO vector, using TOPO TA cloning kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions and verified by sequencing. Plasmids were purified with Qiagen Plasmid Midi Kit, the concentration was determined by Qubit fluorometer (Invitrogen) and diluted to obtain final concentrations of 100, 10, 1 and 0.1 fg representing 20 000, 2000, 200 and 20 plasmid copies as standard for quantitative PCR (qPCR) reactions each for methylated and nonmethylated FOXP3 qPCR assay.

Figure 1.

Representative histologic staining of Foxp3+ T cells of a patient undergoing BPAR (A) and a patient with the diagnosis of SCR (B). Respective TSDR and Foxp3+ numbers are also shown in each case. Patient with BPAR (A) showed 11 Foxp3 T-cells/HPF and 2.3% of TSDR, on the other hand, patient with SCR (B) displayed 7.7 Foxp3 T-cells/HPF and 10% of TSDR.

The Treg DNA methylation qPCR assay targets a distinct region in the FOXP3 gene. This region is referred to as TSDR and it is demethylated exclusively in Treg cells but methylated in all other blood cell types or nonhematopoietic cells. Thus, the percental amount of demethylated TSDR template DNA—as quantified via qPCR—corresponds to the percental amount of Treg in a given sample.

Statistics

Results are expressed by mean ± standard deviation. Comparison between groups was performed by means of Chi-square test for categorical data. The one-way analysis of variance or t-test was used for normally distributed data and the nonparametric Kruskall–Wallis or Mann–Whitney U-test for non-normally distributed variables. Renal function expressed either by serum creatinine or eGFR were considered the outcome variable of the study. Graft function was also estimated by the logistic–regression curve analysis considering event when eGFR ≤ 40 mL/min and significance was analyzed by the log-rank test. To decide which was the most sensible significant cut off GFR value to be used as event in the logistic–regression analyses, we made a sensibility/specificity ROC curve test using different values of GFR (30, 40 and 50 mL/min). The most sensible value was a GFR of 40 mL/min (which had a 74% of sensibility, IC 95%= 0.52–0.93 and p = 0.04). Univariate and multivariate Cox model was used to evaluate risk factors for eGFR ≤ 40 mL/min. All p values were two-tailed and the statistical significance level was defined as p < 0.05. Rate of loss of graft function (eGFR) was done considering it as a quantitative variable for each patient from the 6-month protocol biopsy until 5-year of follow-up.

Results

Patient baseline characteristics

Clinical data  Main clinical and demographic characteristics at the time of the protocol biopsy are depicted in Table 1.

Table 1.  Baseline demographics data and clinical characteristics at the time of protocol biopsy
 SCR with Foxp3+ Treg cells N = 25SCR without Foxp3+ Treg cells N = 12NoSCR N = 68p-Value
  1. 1p = 0.041; 2p < 0.001; 3p < 0.001.

Donor age (years, mean ± SD)39 ± 1243 ± 1545 ± 18NS
Donor gender (M/F)17/88/446/23NS
Recipient age (years, mean ± SD)45 ± 1242 ± 1751 ± 15NS
Recipient gender (M/F)18/78/439/29NS
Cause ESRD   NS
 Glomerulonephritis8.004.0020.00 
 Diabetes1.001.00 
 Unknown10.002.0026.00 
 APKD1.002.006.00 
 Interstitial3.001.0011.00 
 Nephrosclerosis2.003.004.00 
No. trasplant (1/2)23/210/260/8NS
HLA mismatches A, B, DR (mean ± SD)3.2 ± 0.82.9 ± 0.83.34 ± 1.09NS
Cold ischemia time (h, mean ± SD)19.9 ± 4.523.6 ± 7.019.8 ± 6.6NS
DGF (no/yes)17/87/548/20NS
BPAR (no/yes)16/910/258/10NS
Immunossuppresion    
 Induction therapy (no/yes)6/1918/430/380.04
 mTor-inhibitor based (no/yes)11/14212/055/13<0.001
 Non–mTor-inhibitor based (no/yes)16/93/914/543<0.001
Serum Creatinine (μmol/L; mean ± SD)140 ± 74138 ± 41127 ± 40NS
eGFR (mL/min; mean ± SD)55 ± 1550 ± 1254 ± 16NS
Proteinuria (g/24 h; mean ± SD)0.55 ± 1.00.58 ± 1.10.39 ± 0.92NS
Acute Banff’05 score (BLc/IA/IB)15/9/19/2/10NS
Chronic Banff’05 score (No, Yes)14/115/746/22NS
CD3+ T cells/high-power field (mean ± SD)12.9 .± 8.114.6 ± 10.00NS

Among the 37 patients with the diagnosis of SCR, none had presence of C4d in the peritubular capillaries (PTC) whereas only one showed positivity for C4d in the PTC within the 68 patients without SCR. This patient did not have detectable circulating alloantibodies at that moment assessed by flow PRA.

As shown in Table 1, there was no statistical difference between the three evaluated groups regarding relevant clinical, demographic and histologic characteristics such as acute rejection, graft function (both serum creatinine or eGFR and proteinuria. Acute Banff’07 histologic lesions were not different between the SCR patients. Likewise, no differences were found regarding acute histologic scores with the global T cell (CD3+) infiltration between FoxP3+ and FoxP3– patients. Moreover, no differences were observed regarding the chronic graft damage depicted by the presence or absence of interstitial fibrosis and tubular atrophy (IF/TA) following the Banff’07 classification between the three group of patients.

Immunosuppression

Induction therapy was used in 62 of 105 (59%). Thymoglobulin (rATG; Genzyme, Madrid, Spain) was given in 33 patients and Basiliximab (Simulect, Novartis, Basel, Switzerland) in 29 patients. Maintenance immunosuppression was based either on calcineurin inhibitors in 72 patients or on Sirolimus (SRL; Rapamune, Wyeth, Madrid, Spain) in 27 patients. Thymoglobulin was given as induction therapy in 23 patients who were on SRL and to 10 patients receiving a CNI-based regimen. Basiliximab was given to 5 patients receiving SRL and to 24 patients on CNI. There were no patients receiving the combination of a CNI and SRL.

Receiving SRL was significantly associated to the advent of Foxp3+ Treg cells among cellular infiltrates (p < 0.001 between SCR with Foxp3+ and SCR without Foxp3+ Treg cells and p < 0.001 between SCR with Foxp3+ and no SCR). Also, receiving induction therapy was significantly associated to presence of Foxp3+ Treg cells among cellular infiltrates (p < 0.041 between SCR with Foxp3+ and SCR without Foxp3+ Treg cells and p = NS between SCR with Foxp3+ and no SCR).

Although the combination of rATG and SRL was significantly associated to SCR with Foxp3+ Treg cells within graft infiltrates (p = 0.006 between SCR with Foxp3+ and SCR without Foxp3+ Treg cells and p < 0.001 between SCR with Foxp3+ and no SCR), the combination of induction therapy with anti-CD25 monoclonal antibodies and a CNI drug was significantly associated with no mononuclear cell infiltrates at 6-month protocol biopsy (p = 0.029; Table 2).

Table 2.  Immunosuppressive regimens and presence of FoxP3+ Treg
Immunosuppressive regimensSCR with Foxp3+ Treg cells N = 25SCR without Foxp3+ Treg cells N = 12NoSCR N = 68p Value
  1. 1p = 0.006; 2p < 0.029.

 rATG + SRL (yes/no)13/1210/1210/58<0.001
 rATG + CNI (yes/no)2/232/106/62NS
 AntiCD25 + SRL (yes/no)1/240/124/64NS
 Anti CD25 + CNI (yes/no)3/221/1120/4820.029
 No induction + CNI (yes/no)6/198/429/39NS

Comparison of Foxp3+ Treg assessment between Foxp3 TSDR demethylation assay and IMH in renal transplant patients with SCR or BPAR

Among patients with SCR an average of 3.13 Foxp3+ cells/HPF (range 0.1–7.67 Foxp3+ cells/HPF) was observed by IMH (Figure 2A) whereas an average of demethylation of 2.1% (ranging from 0.6% to 10%) was observed using the Foxp3 TSDR demethylation assay (Figure 2B). A positive correlation was observed between the assessment of Foxp3+ expressing cells by IMH and Foxp3 TSDR demethylation analysis (r2= 0.03, p < 0.005; Figure 2C). Regarding patients with BPAR, significantly higher number of Foxp3+ cells/HPF were counted (average 5.6 Foxp3+ cells/HPF; range 1.8–11.3 Foxp3+ cells/HPF) as compared to patients with SCR (p = 0.046; Figure 2A). Conversely, a lower percentage of Foxp3 TSDR demethylation was observed among BPAR (an average of 1.1% ranging from 0.2% to 3%) as compared to SCR patients (p = 0.37; Figure 2B). No correlation between Foxp3+ cells by IMH and Foxp3 TSDR demethylation could be detected (Figure 2D). Similar numbers of CD3+ T lymphocytes were counted in the SCR and BPAR groups. When the proportion of Foxp3+ expressing cells among the global CD3+ T-cell infiltrate was analyzed, a significantly higher Foxp3+ Treg proportion was observed in patients with SCR as compared to patients with BPAR when assessed by TSDR (Figure 2E). Conversely, no statistical significant differences were seen regarding the ratio of Foxp3+ Treg cells among the global T-cell infiltrate between groups when assessed by IMH (data not shown).

Figure 2.

Foxp3 analysis using IMH and quantitative DNA methylation (TSDR). (A) Foxp3 expression in patients with SCR and BPAR. When using IMH, among patients with SCR an average of 3.13 ± 1.6 Foxp3+ cells/HPF (range 0.1–7.67 Foxp3+ cells/HPF) was observed. Among patients with BPAR, 5.6 ± 3 Foxp3+ cells/HPF (range 1.8–11.3 Foxp3+ cells/HPF) was detected (p = 0.046). (B) Percentage of Foxp3 demethylation in patients with SCR and BPAR. When the concentration of Treg cells using the Foxp3 TSDR demethylation assay was used, an average of demethylation of 2.1 ± 3.4% (ranging from 0.6% to 10%) among patients with SCR was observed, whereas an average of 1.1 ± 0.6% (ranging from 0.2% to 3%) in patients with BPAR was seen (p = 0.37). (C) Foxp3 correlation between IMH and TSDR analysis in SCR patients. A positive correlation was observed between the assessment of Foxp3+ expressing cells by IMH and Foxp3 TSDR demethylation analysis (r2= 0.03, p < 0.005) in patients with SCR. (D) Foxp3 correlation between IMH and TSDR analysis in BPAR patients. No correlation between Foxp3+ cells by IMH and Foxp3 TSDR demethylation could be detected among patients with BPAR. (E) Percentage of Foxp3 demethylated/CD3 T cells between patients with SCR and BPAR. The proportion of Foxp3+ expressing cells among the global CD3+ T-cell infiltrate was significantly higher among patients with SCR as compared to patients with BPAR when assessed by TSDR (7.6 ± 5.2 vs. 3.1 ± 0.8, p < 0.005).

Five-year patient evolution according to the study groups

At the end of the follow-up (5 years), eight patients (7.6%) had passed away (three in the group of SCR and presence of Foxp3+ Treg cells, one in the group of SCR without Foxp3+ Treg cells and four more patients in the group without SCR, p = NS). Causes of death were mainly because of cardiovascular disease (50%), malignancies (12.5%) and infections (37.5%). Seven patients (6.6%) had lost their graft at the end of the follow-up (three in the group with SCR and presence of Foxp3+ Treg cells, one in the group with SCR without Foxp3+ Treg cells and three patients in the group without SCR, p = NS). Causes of graft loss were IF/TA in 54%, acute rejection in 16%, transplant glomerulopathy in 28% and 14% because of relapse of primary renal disease).

Five-year renal allograft outcome between SCR patients with presence or absence of Foxp3+ Treg cells and patients without SCR

As shown in Figure 2, patients with SCR without evidence of Foxp3+ Treg cells had significantly worse graft function evolution both serum creatinine (Figure 3A) and eGFR (Figure 3B) than the other two groups at 3, 4 and 5 years of follow-up. No statistical significant differences were observed in graft function between patients without SCR and patients with SCR with Foxp3+ Treg cells during the 5-year follow-up. Moreover, the percentage of Foxp3+ Treg among the global CD3+ T-cell infiltrate (CD3+) was positively correlated with a better 5-year graft function (r =–0.526, p = 0.03 for serum creatinine and r = 0.538, p = 0.03 for eGFR; Figure 4).

Figure 3.

Five-year graft function evolution among patients with and without SCR. (A) Patients without SCR and those with SCR and presence of Foxp3+ Treg cells had significantly better serum creatinine than patients with SCR without evidence of Foxp3+ Treg cells at 3, 4 and 5 years of follow-up [131.78 ± 56.06 and 120.27 ± 34.74 vs. 168.97 ± 49.72 (p = 0.021), 136.17 ± 67.66 and 123.71 ± 51.26 vs. 183.3 ± 75.7 (p = 0.022),148.55 ± 104.90 and 113.21 ± 31.48 vs. 199.8 ± 86.02 (p = 0.002), respectively]. (B) Patients without SCR and those with SCR and presence of Foxp3+ Treg cells had significantly better eGFR than patients with SCR without evidence of Foxp3+ Treg cells at 3, 4 and 5 years of follow-up [58.1 ± 18 and 53.9 ± 18.4 vs. 41.9 ± 13.7 (p < 0.05), 58.3 ± 20.5 and 53 ± 18.36 vs. 38.4 ± 17.2 (p < 0.05), 63.15 ± 19.6 and 52.9 ± 20.7 vs. 34.8 ± 16.7 (p < 0.05), respectively].

Figure 4.

Percentage of Foxp3+ Treg cells/CD3+ T cells and graft function at 5 years. The percentage of Foxp3+ Treg among the total CD3+ cells (Foxp3+/CD3+ T cells) in the 25 patients with presence of Foxp3+ Treg infiltrating the graft was also positively correlated to the eGFR (mL/min) at 5 years after transplantation (r = 0.538, p = 0.03).

When analyzing only patients treated with a CNI-based regimen, those with presence of Foxp3+ Treg cells also showed better graft function evolution than those without at 4 and 5 years after transplantation (51 ± 14 mL/min vs. 38.5 ± 16.7 mL/min, p = 0.09 and 53.1 ± 9.3 mL/min vs. 34.8 ± 16.7 mL/min, p = 0.03 at 4 and 5 years, respectively).

Subsequently, to investigate the impact of such different histologic settings on graft outcome we first performed a death-censored logistic–regression analysis among the three groups of patients considering event graft loss or achieving a significant decrease of renal function depicted as an eGFR < 40 mL/min, similarly as it had been previously evaluated (14). Furthermore, we analyzed the eGFR between 6 months and 5 years after transplantation among the three groups.

As shown in Figure 5A, patients with SCR without evidence of Foxp3+ Treg cells had significantly higher risk of event as compared to the other two groups. The univariate and multivariate Cox regression analysis to assess clinical, histologic and analytical data associated to the achievement of eGFR < 40 mL/min was also performed. Although in the univariate and multivariate Cox regression analysis neither immunosuppression nor acute rejection were associated with graft function, only 6-month serum creatinine, donor age and presence of Foxp3+ Treg cells were independent predictors of achieving better eGFR than 40 mL/min (data not shown).

Figure 5.

Loss of graft function among patients with or without SCR and Foxp3+ Treg cells from 6 months until 5 years after transplantation. (A) Logistic–regression curve estimates achieving eGFR < 40 mL/min per 1.73 m2 in patients with SCR and presence or absence of Foxp3+ Treg cells and patients without SCR (log-rank, p = 0.03). (B) Rate of loss of graft function (eGFR) between 6 months and 5 years after transplantation. Patients with SCR without evidence of Foxp3+ Treg cells experienced a significant loss of graft function during the first 5 years after transplantation. Conversely, patients with noSCR and patients with SCR and presence of Foxp3+ Treg cells maintained a similar 6-month graft function (n = 10, SCR FOXP3––12.3 ± 4.1; n = 19 SCR FOXP3++5.8 ± 2.9; n = 61 noSCR +0.17 ± 0.19 mL/min, p = 0.026).

Next, the eGFR between 6 months and 5 years after transplantation was analyzed between groups. As shown in Figure 5B, patients with SCR without evidence of Foxp3+ Treg cells experienced a significant eGFR during the first 5 years after transplantation. Conversely, patients with noSCR and those with SCR and presence of Foxp3+ Treg cells maintained the same 6-month graft function (n = 10, SCRFOXP3––12.3 ± 4.1; n = 19 SCRFOXP3++5.8 ± 2.9; n = 61 noSCR + 0.17 ± 0.19 mL/min, p = 0.026).

Five-year graft function evolution and acute and chronic histologic lesions

We next analyzed if presence of acute and/or chronic histologic lesions at 6-month protocol biopsies would lead to different graft outcome.

SCR and IF/TA

We first analyzed graft outcome considering all patients with SCR in a same group, regardless they displayed Foxp3+ Treg cells or not within graft infiltrates. Therefore, there were patients with SCR with or without IF/TA (SCR+ and IF/TA+, n = 17; SCR and IF/TA–, n = 20) and patients without SCR that showed or not IF/TA (no SCR and IF/TA+, n = 46; no SCR and IF/TA–, n = 22). The incidence of IF/TA among patients without SCR was 32% whereas 49% of patients with SCR did not show chronic lesions (p = NS).

Main baseline characteristics among these four groups of patients at the time of the protocol biopsy were not statistically different between groups. When we performed a Kaplan–Meier logistic–regression analysis considering event achieving a poorer graft function less than 40 mL/min at 5 years after transplantation we observed that patients with SCR and IF/TA+ displayed significantly higher risk of achieving event than the others (log-rank, p < 0.0001). Also, univariate and multivariate Cox regression analysis taking into account relevant clinical data showed that only 6-month graft function (both serum creatinine and eGFR) and SCR were independent predictors factors of worse graft outcome (data not shown).

SCR with or without Foxp3+ Treg cells and IF/TA

We next evaluated the 5-year clinical outcome considering presence or absence of IF/TA and presence or absence of SCR but taking also into account whether they showed or not Foxp3+ Treg cells among graft infiltrates (noSCR IF/TA+, n = 46; noSCR IF/TA–, n = 22; Foxp3+ SCR IF/TA+, n = 10; Foxp3+ SCR IF/TA–, n = 15; Foxp3– SCR IFTA+, n = 7; Foxp3– SCR IF/TA–, n = 5).

The incidence of IF/TA in patients with SCR and presence of Foxp3+ Treg cells was 40% whereas in patients with SCR without Foxp3+ Treg cells was 59% (p = NS). When all these groups were analyzed at the time of the protocol biopsy for different clinical and analytical data such as donor and recipient age and gender, cause of ESRD, number of transplants, HLA missmatches, incidence of delayed graft function, cold ischemia time, BPAR, graft function or type of immunosuppression at the time of the protocol biopsy and no differences could be observed among all groups (data not shown).

Subsequently, we investigated the eGFR between 6 months and 5 years after transplantation in each of these groups (Figure 6). As shown, patients with SCR and absence of Foxp3+ Treg cells and especially those with IF/TA+ showed a significant loss in graft function. Conversely, those patients with SCR and presence of Foxp3+ Treg cells and even with IF/TA significantly preserved their graft function similarly to patients with a normal biopsy (n = 11, FOXP3+ IFTA++4.66 ± 2.9; n = 8, FOXP3+ IFTA–+1.15 ± 4.09; n = 4, FOXP3-IFTA––4.1 ± 2.11; n = 6, FOXP3-IFTA+–18.85 ± 7.38; n = 18, CONTROL IFTA +–1.78 ± 3.2; n = 43, CONTROL IFTA––0.8 ± 2.1 mL/min; p = 0.04).

Figure 6.

Loss of graft function between patients with or without SCR, IF/TA and Foxp3+ Treg cells from 6 months until 5 years after transplantation. The rate of loss of graft function (eGFR) between 12 months and 5 years after transplantation was significantly higher among patients with SCR and absence of Foxp3+ Treg cells and especially in those with IF/TA, as compared to patients with SCR and presence of Foxp3+ Treg cells, even with IF/TA and to patients with a normal biopsy (FOXP3+IFTA– vs. FOXP3–IFTA+, p = 0.017 and CONTROL IFTA– vs. FOXP3–IFTA+, p = 0.052).

To overcome the limitations of defining the extent of inflammation and presence of Treg cells as categorical variables we further analyzed them as continuous variables in a multivariate model. As shown in Table 3, only the 6 month serum creatinine, the amount of Treg infiltration and the degree of IF/TA remained as independent predictor factors of graft outcome.

Table 3.  Variables associated with eGFR < 40 mL/min by multivariate Cox regression analysis adjusting for timing of the protocol biopsy among the three evaluated groups
 Multivariate analysis
RRCI 95%p-Value
IF/TA grade I0.1340.039–0.4600.0014
IF/TA grade II0.1660.050–0.5570.0036
6-month SCr1.0041.002–1.007<0.0001
Number Foxp3+/HPF0.6690.462–0.9690.0333
Number CD3+ T cells/HPF1.0370.995–1.0790.0820

Discussion

Differentiation between different cellular graft infiltrates in clinically stable renal allografts would be of significant interest to predict graft outcome. Here, we first show that Foxp3 expression within mononuclear cellular graft infiltrates in patients with stable kidney graft function can be appropriately considered Treg cells as showed by the positive correlation with the demethylation degree at TSDR. Furthermore, we confirm preliminary data suggesting the deleterious effect on graft outcome of absence of Treg cells among patients with the diagnosis of SCR in a longer term follow-up and more strikingly the presence of Foxp3+ Treg cells within graft infiltrates seem to even modulate the negative impact of chronic histologic damage on graft outcome as shown by the preservation of graft function over time.

In the last years, different approaches have been focused in the evaluation and characterization of Foxp3+ Treg cells within renal allograft infiltrates to better understand the functional effect of lymphocyte infiltrates within renal allografts and even potentially be used as a biomarker of graft outcome. Nonetheless, and differently to what has been suggested in animal models discrepant interpretations of such issue has been pointed out within the transplant community when dealing with humans. In fact, a major concern here is whether the assessment of Treg cells through the expression of Foxp3 is accurately enough to identify Treg cells. This issue is of relevance, as Foxp3 expression does not always seem to be an exact approach for Treg quantification in humans due to the absence of cell type specificity in some settings (19–21). However, because demethylation of the Foxp3 TSDR has been shown to be exclusive to Treg cells and methylated in all other blood cell types or nonhematopoietic cells (22,23) the percental amount of demethylated TSDR template DNA as quantified via qPCR corresponds to the percental amount of Treg thus, being a reliable tool to precisely identify Treg cells (27).

Taking advantage of a previous study in which we analyzed the impact of Foxp3+ expressing T cells infiltrating renal allografts in patients with the diagnosis of SCR (14), we show that these Foxp3+ expressing T cells assessed by IMH are appropriately characterized Treg cells as there is a positive correlation between numbers of Foxp3+ expressing T cells and Treg frequencies measured by Foxp3 TSDR qPCR. Interestingly, in patients with acute renal dysfunction due to BPAR this correlation could not be observed suggesting that in this clinical setting a significant fraction of Foxp3+ expressing T-cells assessed by IMH rather than being Treg cells, these cells might be transiently activated nonregulatory effector T cells. These observations could explain to some extend the opposite results shown by other groups in patients undergoing BPAR. In this setting, the presence of intragraft Foxp3+ expressing T cells has been highly correlated to histologic acute Banff score lesions and even to pathogenesis-based transcript sets of T-cell activation (17,18). These significant differences are in line with relevant evidences in animal models showing that under severe inflammatory conditions, Treg cells may not be able to develop their suppressive activity and more importantly the achievement of allograft tolerance or rejection depends upon the functional balance of graft-protecting Treg cells to alloreactive graft T effector cells (13,28). Moreover, it has also been shown that proinflammatory cytokines such as IL-6 inhibits Foxp3 transcription by inducing methylation of an upstream Foxp3 CpG island enhancer (29). Accordingly, in our two groups of evaluated patients by the quantitative DNA methylation analyses the percentage of Treg cells to the global CD3+ T cells infiltrating the graft was significantly higher among patients with SCR than in those with BPAR. Although unfortunately, we could not study the predominant cytokine profile to qualitatively explore the inflammatory degree in both clinical situations our findings suggest a different biological effect of Treg cells. On the one hand, during an active alloimmune effector response such as BPAR leading to acute graft dysfunction Treg cells might act as an intrinsic part of the whole alloimmune activation process per se but on the other, under lower inflammatory conditions such in some patients with the diagnosis of SCR Treg cells could illustrate a feature of allorecognition or tolerance of the transplant organ maintaining a local immunologic homeostasis. Nonetheless, this hypothesis deserves further mechanistic investigation.

Here, we further confirm that the initial positive effect of presence of Foxp3+ Treg cells with SCR on graft outcome is maintained over time as compared to those with SCR and absence of Foxp3+ Treg cells and showing the same graft function evolution than patients without any cellular infiltrate at 6-month protocol biopsies. Interestingly, the combination of rATG and SRL was significantly associated with presence of Foxp3+Treg cells in patients with SCR whereas anti-CD25 monoclonal antibodies with a CNI-based regimen was associated to not having any acute cellular infiltrate. These results are in consonance with previous reports showing the higher incidence of acute cellular graft infiltrates in SRL-based treated patients (24), although a more accurate identification of such cellular infiltrates shows that SRL-associated infiltrates are predominantly Foxp3+ Treg. However, among our CNI-treated patients the same positive effect on graft function evolution is observed suggesting that CNI may not completely hamper Foxp3+ Treg function (30,31).

Moreover, not only the presence of Treg cells, but also its proportion among the global T-cell infiltrate seems to be of relevance to favor allograft acceptance as shown by the positive correlation with the 5-year graft function. Similarly, Taflin et al. (32) interestingly showed that percentages of Foxp3+ Treg cells within the global CD4+ T-cell infiltrate in patients with SCR were significantly higher than in patients with AR and it negatively correlated to the intensity of interstitial fibrosis and graft function at the time of biopsy. Likewise, in our study when chronic histologic lesions were evaluated, those patients with no chronic damage had a significantly higher percentage of Foxp3+/CD3+ T cells than those with any chronic histologic damage.

Previous reports have shown that presence of SCR and IF/TA in protocol biopsies has a deleterious impact for graft outcome (24,25). Therefore, we wonder if discrimination of SCR patients depending on whether they displayed or not Foxp3+ Treg cells could show different graft outcome. Although a rather low number of patients were obtained, we observed that presence of Foxp3+ Treg cells in patients with SCR, even with IF/TA is not associated with a significant loss of renal function within the first 5 years after transplantation as compared to those with absence of Foxp3+ Treg cells or no acute cellular infiltrate regardless of their chronic allograft damage. This data suggests that the detrimental effect of chronic histologic lesions on outcome might be modulated by the presence of Foxp3+ Treg cells within cellular graft infiltrates. Similar evidence has been pointed out among renal transplant patients with inflamed fibrosis, in which low percentages of Foxp3+/CD3+ T cells was shown to be an independent indicator of bad outcome depicted as either graft loss or persistent deterioration of graft function (33). Nonetheless, because both IF/TA lesions and Foxp3+ T-cell infiltration independently predict outcome in the multivariate model probably due to the small sample size of our study it remains difficult to state whether Foxp3+ cells infiltrating the graft have a causative role or reflect an epiphenomenon of inflammation.

In conclusion, we confirm presence of bona fide Treg cells that maintain Foxp3 demethylation at the TSDR level within clinically stable acute cellular infiltrates in some renal transplant recipients and this feature seems to be associated to a favorable graft outcome evolution. These preliminary data should be duplicated in larger number of patients and different centers to potentially design a prospective, randomized study for immunosuppression minimization.

Acknowledgments

This study was supported by a grant from the Ministerio de Sanidad y Consumo Instituto Carlos III (PI07/0688, FIS 08/2008) y Fundación Mutua Madrileña (2008–2011).

L. Cuñetti and R. Valdez received a grant from the ISN-SALMASI family fellowship program.

Disclosure

This paper was neither prepared nor funded by any commercial organization.

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

Ancillary