Transgenic Expression of Inducible Caspase9 Suicide Gene for In Vivo Elimination of Antigen Specific Cytotoxic T Cells (CTLs) Engineered To Produce Cytokines.

Transgenic expression of cytokines such as IL2 and IL15 by tumor specific cytotoxic T cells (CTLs) can improve the survival and expansion of these cells within the tumor microenvironment. Using our model of tumor antigen-Epstein-Barr virus-specific CTLs (EBV-CTLs), we found that transgenic production of either cytokine could sustain CTL expansion ex vivo, and improve their anti-tumor effect in vivo in a SCID mouse model. Although the proliferation of cytokine gene transduced CTLs remained antigen dependent, clinical application of this approach likely requires the inclusion of a suicide gene to deal with the potential development of T-cell mutants with autonomous, antigen-independent, growth. We constructed a novel suicide gene based on the inducible caspase9 gene [Straathof et al Blood 2005:105;4247] and showed that EBV-CTLs expressing this suicide gene can be eliminated after exposure to the small molecule chemical inducer of dimerization (CID) AP20187. However, it has been suggested that the anti-apoptotic effects of transgenic cytokine expression would counteract the pro-apoptotic action of icaspase9. To discover the general applicability of our inducible suicide gene approach, we generated tricistronic retroviral vectors encoding full length hIL2 or hIL15, a truncated human CD34 molecule (DCD34) as a marker, and the inducible Caspase9 suicide gene. The three genes were linked using 2A peptide cleavable sequences, which allows equal expression of the 3 transgenes. Established EBV-CTLs were transduced either with IL2.DCD34.iCasp9 or IL15.DCD34.iCasp9 or DCD34 empty vector. After stimulation with the antigen (LCLs) at a E:T ratio of 1:1, we observed significant expansion of CTLs/IL2.DCD34.iCasp9 (221 fold expansion, range 84–452) and IL15.DCD34.iCasp9 (182, range 48–355) after 28 days of culture, while CTLs/DCD34 maintained without cytokines had <2 fold expansion. Cytokine release after antigen stimulation was 6.4±3.4 pg/mL and 157±99pg/mL for IL15 and IL2, respectively. Transgenic CTLs maintained the same phenotype as control T cells (>90% CD3+/CD8+) and remained polyclonal as assessed by staining for the TCRVb repertoire. To evaluate the cytotoxic capacity of the suicide gene, control and transgenic CTLs were incubated with CID AP20187 at 20nM. After 24 hours the survival of transgenic CTLs was estimated from residual CD34+ expression. For both CTLs/IL2.DCD34.iCasp9 and CTLs/IL15.DCD34.iCasp9 the percentage of CD34+ cells was reduced by >1.5 logs. Moreover, the few residual cells were CD34^dim, and produced no measurable cytokine release as measured by ELISA. To assess activity in vivo, we used an LCL-engrafted SCID mouse model, injected with CTLs transgenic for the firefly-luciferase gene. These cells were tracked in vivo using the Xenogen-IVIS bioluminescence system. We found that CTLs transgenic for either IL2 or IL15 both migrated to the tumor site, and had increased expansion compared to DCD34 control CTLs. Mice were then treated with 2 −3 doses of CID AP20187 (50 mg/mice i.p). The bioluminescence signal diminished to background levels by 48hr, suggesting efficient elimination of transgenic CTLs. In conclusion, these data indicate that induction of a transgenic caspase9 suicide gene can effectively destroy adoptively transferred T cells, even when these lymphocytes are expressing anti-apoptotic cytokines such as IL2 and IL15. This approach may increase the safety margin of gene modified CTL therapies.