Phenotype, Function and Expansion of Regulatory T Cells in the Cynomolgus Macaque (Macaca fascicularis)

Regulatory T cell (Treg) therapy provides an alternative approach to chronic immunosuppression and has shown promising results in murine models. Translating this approach into the clinic depends on developing efficient methods with which to expand human or non-human primate Tregs in vitro.

At the Columbia Center for Translational Immunology (CCTI) we have started a pre-clinical bone marrow transplant program using the Cynomolgus macaque (CM) monkey model involving Tregs to promote tolerance to tissues and organs. We first characterized CM Tregs, which have similar phenotype when compared to human Tregs. 3.67% of CD4 T cells in the CM co-express high levels of CD25 and FoxP3 (n=15) and 3.53% express CD25 and are negative for CD127, a commonly used Treg marker. Similar to human Tregs, CM Tregs include naïve or resting Tregs (1.63%) and activated effector Tregs (2.4%) based on CD45RA expression (n=7). In vitroCM CD25hiCD127- (FoxP3+) Tregs were found to be similarly suppressive as human Tregs, supporting the clinical relevance of the CM model.

We subsequently aimed to expand polyclonal regulatory T cells (pTregs) in vitrofor use in our transplant studies. We developed four protocols for the expansion of pTregs over 21-56 days. 50-100 thousand CD4+CD8-CD25hi or CD4+CD8-CD25hiCD127- Tregs were isolated from 20-30mL of blood and cultured with IL-2, αCD3, rapamycin and either donor PBMCs, artificial APCs (L-cells, murine fibroblast expressing human CD80, CD58 and CD32) or both. 1000-10,000 fold expansion was reliably achieved when artificial APCs were used. In contrast, when donor PBMCs (without artificial APCs) were used as the source of APCs, 10-100 fold fewer Tregs were obtained. Regardless of protocol, pTregs successfully prevented the proliferation of bead (CD2CD3CD28) stimulated self PBMCs with up to 50% suppression at a 1:32 ratio of Tregs:PBMCs.  

Contamination of CD8+ T cells was observed despite our efforts to gate out CD8 cells during sorting. Re-sorting of the Treg cultures early (day 14-21 post isolation) successfully eliminated the CD8s. Re-sorting inherently caused a significant loss of Tregs, leading to a longer culture period. Because Treg burnout was of concern, we assessed the function and phenotype of cells maintained in cultures >50 days and observed that Tregs with early high FoxP3 expression maintained function and phenotype.

Finally, demonstration of fitness of cryopreserved pTregs is necessary for clinical application. Our preliminary studies show that Tregs can be re-cultured and re-expanded without loss of suppression or phenotype. In summary, we have successfully established several protocols for Treg expansion in the Cynomolgus macaque that can be used in pre-clinical studies of transplant tolerance induction.