Potential of antigen-specific adoptive immunotherapy using peptide/mhc tetramers and microbeads for highyield purification of cytotoxic t lymphocytes from peripheral blood

Adoptive immunotherapy, such as donor lymphocyte infusions (DLI), is a potent approach for the treatment of cancer. Clinically, however, DLI is hampered by toxicity from GVHD. An alternative strategy may be the infusion of antigen-specific donor T lymphocytes, as exemplified by the use of anti-CMV or anti-EB V T cells for complications following allogeneic bone marrow transplantation. Tumor antigen-specific cytotoxic T lymphocyte (CTL) infusions have been more difficult, limited by the lack of widely applicable tumor antigens and the technical demands of isolating and expanding sufficient numbers of antigen-specific T cells. Here, we demonstrate the isolation of high numbers of purified antigen-specific T cells suitable for transfer using recombinant MHC/peptide tetramers and magnetic microbead cell sorting. CTL specific for one of two model antigenic peptides, influenza matrix MI-58 (flu-Mi) or EBV-BMLF1-280, were generated from PBMC of normal donors. CTL were expanded for up to two weeks ex vivo by stimulation with peptide-pulsed PBMC and cytokines. This increased the frequency of flu-M 1 tetramer+ CD3+ CD8+ T cells from <0.01% of PBMC to 6-10%. EBV-BMLF1 tetramer+ CD3+ CD8+ T cells increased from precursor frequencies of 0.02-0.05% of total PBMC to 22-27% (overall mean 814-fold increase). Peptide-specific CTL were sorted using MHC/peptide tetramers conjugated to PE, and anti-PE Miltenyi magnetic MACS microbeads. As determined by flow cytometry, purity ranged from 83-98% tetramer+ cetls, and recovery of antigen-specific CTL ranged between 36-93% (mean 59%). In comparison to different magnetic beads and flow cytometry based cell sorting, this tetramer/ microbead technology was significantly more rapid and efficient. Importantly, magnetic bead sorted CTL were functional as demonstrated by cytotoxicity of peptide-loaded target cells (40-96% lysis at 10:1 E:T ratio) and strong IFNy secretion in response to antigenic stimulation in ELISPOT analysis. Overall, for these two model antigens, up to 25 million antigen-specific CTL could be generated from 50 cc of peripheral blood. Combining non-specific techniques (such as mitogenic or CD3/CD28 stimulation) with this approach may further increase the yield of specific cells. With the ongoing discovery of tumor antigens suitable for clinical therapeutics, the evaluation of antigen-specific lymphocyte infusions, facilitated by tetramer/microbead technology, becomes an important next step in developing anti-tumor immunotherapy.