Reduced frequency of FOXP3+ CD4+CD25+ regulatory T cells in patients with chronic graft-versus-host disease

Abstract

Chronic graft-versus-host disease (cGVHD) is a major complication of allogeneic hematopoietic stem cell transplantation but the immune mechanisms leading to the diverse clinical manifestations of cGVHD remain unknown. In this study, we examined regulatory T cells (Tregs) in 57 transplant recipients (30 with cGVHD and 27 without active cGVHD) and 26 healthy donors. Phenotypic studies demonstrated decreased frequency of CD4+CD25+ T cells in patients with cGVHD compared with patients without cGVHD (P < .001) and healthy individuals (P < .001). Gene expression of Treg transcription factor FOXP3 was reduced in cGVHD patients compared with patients without cGVHD (P = .009) or healthy donors (P = .01). T-cell receptor excision circle (TREC) assays for the evaluation of thymus activity revealed fewer TRECs in both transplant groups compared with healthy donors (P < .001 and P = .02, respectively) although no difference was observed between patients with or without cGVHD (P = .13). When tested in functional assays, Tregs from both patient cohorts and healthy individuals mediated equivalent levels of suppression. Collectively, these studies indicate that patients with active cGVHD have reduced frequencies of Tregs but the function of these cells remains normal. These findings support the development of new strategies to increase the number of Tregs following allogeneic hematopoietic stem cell transplantation to prevent or correct cGVHD.

Figures

Figure 1.

Decreased frequency of circulating CD4+CD25+ T cells in patients with cGVHD. Percent CD4+CD25+ in PBLs was measured by flow cytometry in patient and donor samples. Values represent total positive cells including CD4+ CD25+dim and CD25+high cells. Box plots define the values for median, range, 25th, and 75th percentiles. One outlier with a value of 37.5% in the group of patients with cGVHD is not shown in the figure. P values were calculated using the Wilcoxon rank-sum test; cGVHD versus no cGVHD, P = .007; cGVHD versus healthy donors, P = .29; no cGVHD versus healthy donors, P < .001.

Figure 2.

Decreased expression levels of FOXP3 in patients with cGVHD. (A) Expression levels of the Treg-specific transcription factor FOXP3 was assessed by quantitative PCR in patient and donor samples, normalized, and reported as a function of total lymphocytes (AU, arbitrary units). Box plots define the values for median, range, 25th, and 75th percentiles. P values were calculated using the Wilcoxon rank-sum test; cGVHD versus no cGVHD, P = .009; cGVHD versus healthy donors, P = .01; no cGVHD versus healthy donors, P = .34. (B) Correlation between CD4+CD25+ phenotypic values and FOXP3 expression levels in patients with cGVHD was calculated using a rank-based Spearman test (P < .001; rs = 0.60). (C) Patients with cGVHD were subdivided according to whether they previously developed aGVHD. Box plots define the values for median, range, 25th, and 75th percentiles. P values were calculated using the Wilcoxon rank-sum test; aGVHD and cGVHD versus no aGVHD and cGVHD, P = .29; aGVHD and cGVHD versus no aGVHD and no cGVHD, P < .001; no aGVHD and cGVHD versus no aGVHD and no cGVHD, P = .005.

Figure 2.

Decreased expression levels of FOXP3 in patients with cGVHD. (A) Expression levels of the Treg-specific transcription factor FOXP3 was assessed by quantitative PCR in patient and donor samples, normalized, and reported as a function of total lymphocytes (AU, arbitrary units). Box plots define the values for median, range, 25th, and 75th percentiles. P values were calculated using the Wilcoxon rank-sum test; cGVHD versus no cGVHD, P = .009; cGVHD versus healthy donors, P = .01; no cGVHD versus healthy donors, P = .34. (B) Correlation between CD4+CD25+ phenotypic values and FOXP3 expression levels in patients with cGVHD was calculated using a rank-based Spearman test (P < .001; rs = 0.60). (C) Patients with cGVHD were subdivided according to whether they previously developed aGVHD. Box plots define the values for median, range, 25th, and 75th percentiles. P values were calculated using the Wilcoxon rank-sum test; aGVHD and cGVHD versus no aGVHD and cGVHD, P = .29; aGVHD and cGVHD versus no aGVHD and no cGVHD, P < .001; no aGVHD and cGVHD versus no aGVHD and no cGVHD, P = .005.

Figure 2.

Decreased expression levels of FOXP3 in patients with cGVHD. (A) Expression levels of the Treg-specific transcription factor FOXP3 was assessed by quantitative PCR in patient and donor samples, normalized, and reported as a function of total lymphocytes (AU, arbitrary units). Box plots define the values for median, range, 25th, and 75th percentiles. P values were calculated using the Wilcoxon rank-sum test; cGVHD versus no cGVHD, P = .009; cGVHD versus healthy donors, P = .01; no cGVHD versus healthy donors, P = .34. (B) Correlation between CD4+CD25+ phenotypic values and FOXP3 expression levels in patients with cGVHD was calculated using a rank-based Spearman test (P < .001; rs = 0.60). (C) Patients with cGVHD were subdivided according to whether they previously developed aGVHD. Box plots define the values for median, range, 25th, and 75th percentiles. P values were calculated using the Wilcoxon rank-sum test; aGVHD and cGVHD versus no aGVHD and cGVHD, P = .29; aGVHD and cGVHD versus no aGVHD and no cGVHD, P < .001; no aGVHD and cGVHD versus no aGVHD and no cGVHD, P = .005.

Figure 3.

Correlation between CD3+CD8+ and CD4+CD25+ T cells in active cGVHD. (A) Percent CD3+CD8+ in PBLs was measured by flow cytometry in patient and donor samples. Lines represent median values. (B) Percent CD4+CD25+ and CD3+CD8+ T cells in total PBLs are shown for patients with cGVHD. Correlation between the 2 variables was calculated using a rank-based Spearman test (rs = -0.81, P < .001).

Figure 4.

Percent CD4+CD25+ in PBLs correlates with Treg suppressive function in cGVHD. (A) CTL reactivity against a pool of class I-restricted common viral peptides was assessed by IFNγ ELISPOT assays before and after CD25+ T-cell depletion using 15 samples collected from patients with cGVHD. Depletion of CD25+ cells was confirmed by flow cytometry for each sample. (B) Percent inhibition was measured for 15 patients with cGVHD and values are plotted together with percent CD4+CD25+. One sample generated a negative value and was interpreted as null. Correlation between the 2 variables was calculated using a rank-based Spearman test (rs = 0.59, P = .02).

Figure 5.

Patient Tregs express normal levels of suppression. CD4+CD25+ Tregs were immunopurified from PBMCs and assessed for their capacity to suppress the proliferation of autologous CD4+CD25- cells at a responder/Treg ratio of 1:1. Percent inhibition of responder cell proliferation due to the presence of Tregs was measured using 17 samples collected from 5 cGVHD patients, 7 patients without cGVHD, and 5 healthy individuals. The solid lines represent the median values.

Figure 6.

Allo-transplant patients have impaired thymus function. Thymic activity was assessed in patient and donor samples using the TREC assay and represented as TREC copy number per 105 CD3+ T cells. Box plots define the values for median, range, 25th, and 75th percentiles. One outlier with a value of 1292 in the group of no-cGVHD patients is not shown in the figure. P values were calculated using the Wilcoxon rank-sum test; cGVHD versus no cGVHD, P = .13; cGVHD versus healthy donors, P < .001; no cGVHD versus healthy donors, P = .02.