Abstract 1341

Poster Board I-363

Introduction

Graft versus host disease (GVHD) following haematopoietic stem cell transplantation is often first observed in the skin; a primary target organ of GVHD. GVH related tissue damage in the skin is mainly driven by infiltrating alloreactive cytotoxic effector T cells, facilitated by a cascade of cytokines and chemokines. Our recently published observations showed that addition of regulatory T cells (Treg) suppressed skin GVH tissue damage mediated by alloreactive CD8+ T cells in an in vitro human GVHD skin explant model [1]. The current study investigated the role of Treg in modulating effector T cell infiltration into skin, it's consequence on the severity of skin GVH histopathology and the possible changes of effector T cell production and expression of chemokines and chemokine receptors.

Methods

CD8+ T cells, monocyte derived dendritic cells (mDC) and natural Treg (CD4+CD2highFOXP3+) were generated as previously described [1]. In an in vitro human GVHD skin explant model, CD8+ T cells and ex vivo expanded Treg obtained from buffy coats were used as “donor” cells. mDC and skin biopsies obtained from HLA unmatched unrelated normal volunteers acted as “recipient” tissues. “Donor” CD8+ T cells primed with “recipient” mDC in the presence or absence of Treg were co-cultured with “recipient” skin. The severity of histopathological GVH skin damage was scored as grade 0 to grade IV using a clinically validated scoring system. The number of infiltrating CD8+ T cells in skin was evaluated using immunohistochemistry then correlated to the severity of skin GVH histopathology. The gene expression of selected chemokines and chemokine receptors in alloreactive CD8+ effector T cells was analysed using quantitative RT-PCR. The effector T cell expression of chemokine receptors was assessed using flow cytometry. The secretion of selected chemokines into the culture supernatants was quantified using BD cytometric bead array kit.

Results

The percentage of infiltrating effector T cells in skin was significantly associated with the severity of skin GVH histopathology (2.6±0.8, 12.6±3.1 and 27.2±2.7 for skin sections with GVH histopathology grade I, II and III-IV; p=0.017 and 0.021 respectively). The percentage of skin infiltrating CD8+ T cells was significantly reduced by the presence of Treg (24.8±3.7 vs 11.58±1.8, p=0.011, n=13) which correlated with Treg mediated suppression of skin GVH histopathology (p<0.0001, n=13). The presence of Treg also down-regulated effector T cell expression of chemokine/chemokine receptor genes (CCL3, CCL5, CCR4, CCR5, CXCL10 and CXCL11) that are involved in the recruitment of effector T cells to GVH target tissues. Further analysis indicated a trend toward reduced effector T cell surface expression of CCR4 and CCR5 (31.9±4.3 vs 12.5±2.7; 13.7±1.8 vs 8.8±4.1 respectively, n=3) in the presence of Treg. There was also a reduction in CX3CR1 and cutaneous lymphocyte associated antigen (CLA) (11.1±1.9 vs 2.2±0.6; 34.6%±7.0 vs 11.6±2.2 respectively, n=3). The effector T cell surface expression of CXCR3, CCR2 and CCR10 was very low regardless of the presence or absence of Treg. Release of CCL3, CCL5, CXCL9 and CXCL10 into the culture supernatant was strongly suppressed by the presence of Treg (813±122 vs 77±30; 816±248 vs 405±148; 5517±967 vs 974±540 and 858±209 vs 195±57, p=0.001, 0.065, 0.002 and 0.015 respectively, n=8). However, CXCL11 and CX3CL1 levels were below detectable limits regardless of the presence or absence of Treg.

Conclusions

Ex vivo expanded natural Treg can inhibit CD8+ effector T cell infiltration into skin which correlated with Treg suppression of cytotoxic T cell mediated skin GVH histopathology in an in vitro human GVHD model. This effect may be attributed to a decrease in chemokine and chemokine receptor interactions mediated by Treg. These observations indicate a potential mechanism for Treg mediated GVHD suppression.

[1] X N Wang et al., Transplantation 2009

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution