In Vitro System for Generation of Leukemia-Specific Cytotoxic T Cells for Adoptive Immunotherapy.

Background: Adoptive T cell-based immunotherapy holds promise as an additive treatment option for leukemia, but its development has been impeded by the lack of reproducible methods for generating therapeutic numbers of antigen-specific CD8+ cytotoxic T lymphocytes (CTLs). We hypothesized that tumor associated antigens from whole leukemia cell lysates can efficiently pre-expand and activate an antigen-specific T cell precursor population that can be further expanded to obtain large amounts of tumor-specifc CTLs for therapeutic use.

Methods:A specific and physiologic T cell activation step using tumor lysate-pulsed bone marrow derived dendritic cells (DC) was combined with a nonspecific large scale expansion system with anti-CD3/anti-CD28-coated microspheres. In order to validate the system we transduced the murine leukemia cell line C1498 with a defined protein (OVA) serving as a surrogate tumor antigen for analyzing anti-tumor CTL responses in vitro and in vivo.

Results: Using T cell receptor (TCR)-transgenic OT-I mice, with a TCR specific for H-2K^b and OVA peptide, we could demonstrate, that bone marrow derived DC pulsed with cell lysates from C1498-OVA cells can process and present leukemia cell derived antigens effectively to T cells in vitro in a MHC-I-restricted fashion. Naive T cells within OT-I derived splenocytes could be activated and expanded, generating specific killing against C1498-OVA but not C1498 leukemia cells. Upon adoptive transfer these CTLs mediated potent anti-leukemic effects in tumor bearing animals. Over 80% of mice that received a lethal challenge with C1498-OVA cells by tail vein injection could be rescued after receiving 40 x 10⁶ OT-I CTLs on the following day. In contrast, mice receiving identical treatment after receiving a lethal dose of mock-transfected C1498 cells all died within 47 days as did C1498-OVA challenged mice after PBS treatment only (p < 0.005). In order to simulate the more realistic setting of a naïve T cell population containing a small percentage of antigen-specific T cell precursors only, small amounts of OT-I derived splenocytes were mixed with wild type splenocytes at a ratio of 1:10. Whereas the percentage of C1498-OVA-specific CTLs increased to 5% after priming with C1498-OVA cell lysate-pulsed DC, the frequency dropped to 1.5% after priming with C1498 lysate-pulsed DCs. We then asked the question whether higher yields of leukemia specific CTLs could be generated in our in vitro system by immunizing potential T cell donors with tumor antigen-presenting DC. Syngeneic wild type mice were immunized with C1498-OVA lysate-pulsed DC. Controls received C1498 lysate-pulsed DC. Using the immunized mice as splenocyte donors and the Vα2β5 TCR as a surrogate marker to identify OVA-reactive T cells from wild type mice, the yield of Vα2β5 positive CTLs was in average 3.5 fold higher after only 8 days of repriming and further expansion with anti-CD3/anti-CD28-coated microspheres in vitro.

Conclusion: High numbers of CTLs specific against leukemia-associated antigens can be generated and expanded within a relatively short time span of 9–13 days in vitro and mediate robust antigen-specific anti-leukemic effects upon adoptive transfer into leukemia bearing animals. This can be achieved by combining a priming step with tumor cell lysate pulsed DC followed by unspecific large scale expansion using anti-CD3/anti-CD28-coated microspheres.