Introduction: Strategies for immunomodulation of xenograft rejection response whilst minimizing long-term immunosuppression need be developed. Polyclonal regulatory T cells (Treg) are suppressive of the xenoimmune responses. However, application of xenoantigen-specific Treg is required to achieve effective suppression and avoid opportunistic infection and malignancy for clinical xenotransplantation. We have previously shown that xenoantigen stimulation enhanced human Treg capacity to suppress the xenogeneic response. However, whether xenoantigen expanded Treg express specific cell surface markers which can be used for separating a particular xenoantigen-specific Treg subset for an effective Treg therapy needs be identified. In this study we investigated whether xenoantigen expanded human Treg express specific cell surface markers and their potential for selection of xenoantigen-specific Treg.
Materials and Methods: Human CD4+CD25+CD127- Treg isolated from healthy donor peripheral blood mononuclear cells (PBMC) were expanded for 3 or 4 cycles with anti-CD3/CD28 beads alone or combined with irradiated porcine PBMC as polyclonally (PlTreg) or xenoantigen stimulated Treg (XnTreg), respectively. Flow cytometry was performed to determine candidate cell surface markers and consequent xenoantigen-specific Treg subset was isolated XnTreg by cell sorting. After sorting, the resulting Treg subset was assessed for their suppressive capacity by mixed lymphocyte reaction (MLR) using irradiated porcine PBMC as xenogeneic stimulating cells, human PBMC as responder cells and autologous XnTreg as suppressing cells.
Results: After 3 and 4 cycles of expansion, although no difference was detected in expression of Treg activation markers CD39, CD45RO and CD15s on both XnTreg and PlTreg, XnTreg exhibited substantially upregulated expression of Treg memory markers HLA-DR and CD27 with a larger proportion of them being HLA-DR+CD27+. The HLA-DR+CD27+ Treg subset separated by cell sorting from XnTreg demonstrated significantly enhanced potency in suppression of proliferating xenoreactive responder cells at ratios of 1:4 through to 1:64, or 1:32 and 1:64 of Treg:responder cells when compared to HLA-DR+CD27+ cell-depleted or unsorted XnTreg, respectively.
Discussion: This study demonstrated that porcine xenoantigen expanded human Treg expressed elevated levels of Treg memory markers HLA-DR and CD27 which could be used for separation of xenoantigen-specific Treg subset. The effect of neutralizing HLA-DR and CD27 on HLA-DR+CD27+ Treg-mediated suppression of the xenogeneic response is being currently under investigation to facilitate their potential as specific markers for selection and enrichment of xenoantigen-specific human Treg. 
Conclusion: Our data suggest that human HLA-DR+CD27+ memory-type Treg are xenoantigen-specific and have potential as an effective immunotherapy in xenotransplantation.