Résumé
Suicide gene therapy clinical approaches using retrovirally transduced donor T-lymphocytes (GMTC) have led to successful alloreactivity modulation following bone marrow transplantation. We have previously shown that our GMTC production method: 12 day ex-vivo culture in the presence of IL-2, entails phenotypical and functional alterations. Such alterations can be limited partly by CD3/CD28 activation before retroviral transduction, plus immunomagnetic selection based on cell surface membrane gene expression, i.e truncated Nerve Growth Factor (ΔNGFR).The present study aims at determining the impact of both initial T-cell activation and transduction/selection on GMTC transcriptional signature, to better understand these phenomena and identify which genes could be targeted by ex-vivo manipulation. We also compare the whole GMTC production process used in our clinical trial [OKT3/IL-2-NeoR (G418 selection)] to another process (CD3/CD28/IL-2-ΔNGFR) that might- based on our previous observations - limit alterations and be used in future clinical trials.Peripheral blood mononuclear cells (PBMC) from 4 healthy donors were activated with either OKT3 or CD3/CD28 and transduced using an experimental retroviral vector carrying 2 selection transgenes, NeoR and ΔNGFR, allowing differential selection. CD4+/CD8+ magnetic cell sorting was applied at the end of the process. The new GeneChip human whole Genome Arrays, which provide comprehensive coverage of 47,000 transcripts (Arrays U133 plus 2.0, Affymetrix) allowed us to study the expression profile of CD8+ GMTC; CD8+ resting cells (RC) or cultured untransduced cells used as controls. Raw data were normalized, filtered and analysed with GeneSpring Software (Silicon Genetics). For each comparison, we chose to explore genes that are present in at least half of all analysed samples (Affymetrix flags), with a 2-fold change expression and p≤0.005.We show that 1/ activation homogenizes the expression profile of GMTC versus RC; 2/ both activation modes affect mainly the same genes (80% similarity) within different biological processes such as cell growth maintenance, cellular communication and signal transduction. Moreover, OKT3 activation upregulates pro-apoptotic genes (such as FADD, BAX) while CD3/CD28 downregulates them. Activation also affects the genes involved in immune response and adhesion e.g. IL-4, 5 & 8, IL-2R, 6R, 7R & 17R, TNF, CXCR3, 4 & 6 or CCR2, 5 & 7; 3/ conversely, transduction/selection affect fewer and different genes (81% dissimilarity between both selection modes). Finally, comparison of the whole processes with RC shows additional effects of different steps than when the steps were take in count separately. Overall, GMTC expression profile is mainly affected by the activation step and less so by the transduction/selection step. Moreover, we identified involved genes with unclassified gene ontology (GO) functions and expressed sequence tag (EST) that warrant further investigations.