Gamma Delta T Cells (GDTCs) to Eliminate Minimal Residual Disease (MRD) in Chronic Myeloid Leukemia (CML) : A Pre-Clinical Model

Abstract

GDTCs constitute a small subset of T-lymphocytes (<10%) involved in tumor immune surveillance and are an attractive option for adoptive immunotherapy. We are especially interested in their potential role in eliminating MRD in CML. As a first step, we determined whether or not GDTCs were part of the CML clone. 17–50 metaphases of GDTCs from each of 3 patients with documented Ph+ BCR-ABL+ CML were karyotyped and found to be Ph(−). Results were confirmed by fluorescence in situ hybridization analysis and quantitative RT-PCR for BCR-ABL. Using Miltenyi Biotec positive selection (MACS), we obtained high yield and purity of GDTCs isolated from healthy donor-derived peripheral blood mononuclear cells (PBMCs). 2.4 × 106 (± 8.6×105) GDTCs (n=13) were isolated and expanded over 1000-fold in vitro using optimized cell culture conditions. Commercial sera were screened; the chosen serum supported up to 5 times greater expansion than autologous human serum, as had been used for CML-patient-derived GDTCs. 1mg/mL concanavalin A proved the optimal mitogen, with a 2524-fold GDTC expansion compared to PHA (1764-fold) and the anti-CD3 antibody, OKT3 (267-fold). In all cases, IL-2 and IL-4 (both 10 ng/mL) supplemented culture media; neither additional IL-7 (10 ng/mL) nor IL-15 (10 ng/mL), alone or combined, improved expansion of GDTCs. The initial post-isolation purity of 97.9% (± 0.9%) as assessed by flow cytometric (FACS) analysis was maintained for culture duration. We observed that expanded GDTCs obtained from both healthy donors and Ph(+) CML patients exhibited specific cytotoxic activity against Ph+ lines, K562 and EM-2. Expanded GDTCs from Ph(+) CML patients were co-incubated with K-562 at effector:target (E:T) ratios of 1:1 and 10:1 in triplicate for 3 and 24 hours. Percent cytotoxicity, defined as % reduction in colony-forming cells (CFC) after co-incubation of GDTCs compared to CFC generated by targets alone, was comparable at both ratios: ~50% at 3 hours, decreasing to 39% at 24 hours. GDTCs isolated from 9 healthy donors were expanded in vitro and co-incubated with K562 or EM-2 at E:T of 20:1. Apoptotic cells were identified by a FACS cytotoxicity assay based on Annexin V staining. After 4 hours co-culture, K562 were 27% ± 2% and EM-2 45% ± 3% apoptotic. K562 cultured alone did not show increased Annexin V staining. Comparable results were obtained by standard chromium release assay. Simultaneous incubation of autologous PBMC with GDTCs did not yield significant PBMC Annexin V staining, confirming selective cytotoxicity of GDTCs against leukemic cells. Results were verified by incubating GDTCs from 3 healthy donors with autologous PBMCs mixed with 1.0%, 0.5%, or 0.1% K562eGFP+ cells at 20:1 E:T for 0 or 4 hours. Co-incubation with GDTCs resulted in significant and similar reduction in K562eGFP+ colonies at all ratios (96.5%, 96% and 97%, respectively). GFP-negative colony numbers were maintained after 4 hours co-incubation with GDTCs indicating specificity against K562eGFP. Similarly, GDTC co-incubation with autologous PBMCs generated 105% BFU-E and 117% CFU-GM colonies, comparing CFC generated from t=4 to t=0 hours co-culture, indicating lack of cytotoxicity against autologous committed hematopoietic progenitors. We confirmed cytotoxic efficacy of GDTCs in vitro by RT-PCR. 16 hours co-incubation of GDTCs with EM-2 resulted in a 2 log reduction in BCR-ABL transcripts. We have established a xenogeneic mouse model of CML with fluorescent and bioluminescent leukemic cells (K562eGFPluc) that we visualize via in vivo imaging system (IVIS®) technology. Both K562eGFPluc and CFSE-labelled GDTCs engraft in NODscidIL2Rg−/− mice. We are investigating cytotoxicity of GDTCs against CML MRD in this pre-clinical animal model.