Effect of trichostatin a and 5'-azacytidine on transgene reactivation in U937 transduced cells

In mammals, methylation of DNA within regulatory sites and histone deacetylase recruitment in transcriptional repressing domains are involved in the loss of the expression of retroviral DNA or repeat arrays transferred in cells for therapeutic purposes. Various investigation results suggest that methylation/deacetylation events are modulated by extracellular and cytoplasmic signal transduction pathways closely involved in regulating cell differentiation. To analyse gene silencing mechanisms and assess if potential pharmacological treatment affects gene silencing kinetics we transduced U937 myelomonocytic cells with a bicistronic retroviral construct carrying the herpes simplex virus thymidine kinase (HSV-TK) and beta-galactosidase (Lac-Z) genes. This vector can be employed in vivo and in vitro to render transduced cell populations susceptible to ganciclovir (GCV). We verified the effect of the histone deacetylase inhibitor Trichostatin A (TSA) alone or combined with 5'-azacytidine (5'aza-C) on transcription downmodulation. Our results indicate that in our in vitro model TSA is able to reactivate transgene expression, more efficiently and with quicker kinetics (12-24h) than 5'aza-C (36-48 h). The effect is dose dependent (between 1 and 50 nM), with no relevant toxicity. Treatment with both drugs is synergistic in gene reactivation in terms of extension and persistence, with low toxicity and no relevant differentiating effects. The cells in which transgene expression has been reactivated undergo progressive silencing, but once weekly drug treatment can maintain high transgene expression levels for more than 90 days with no evidence of selection. The results obtained by treating U937 transduced clones with TSA and/or 5'aza-C together with IL-3, G-CSF or GM-CSF cytokines suggest that transduced U937 differentiation levels do not affect basal expression, but render these cells more responsive to reactivation by TSA or TSA plus 5'aza-C, but not to 5'aza-C alone. In conclusion, the results suggest that in vitro inhibition of histone deacetylase by TSA can interfere with gene silencing mechanisms affecting 5' Moloney murine leukaemia virus long terminal repeat (MoMuLV-LTR) driven transgene expression thus providing the rationale for TSA and/or 5'aza-C administration in animal models for the translation on gene therapy applications.