Azacitidine Down-Regulates Both IL-6 Signalling and NFkB Activity in Human Myeloma Cells.

We have previously demonstrated that Azacitidine (AZA) at clinically achievable concentrations (<10μM) induces apoptosis in a range of genetically heterogeneous human myeloma cell lines (HMCL) and prolongs survival in the murine 5T33 model of systemic multiple myeloma (MM). HMCL U266 demonstrates autocrine IL-6 production, constitutive Stat3 phosphorylation and is resistant to conventional chemotherapy and glucocorticoids. Using U266 we have evaluated further the effect of AZA on human MM cells. AZA treatment (5mM daily) of U266 induced apoptosis at 24, 48 and 72 hours of 28%, 43% and 52%, respectively. Immunoblot analyses showed rapid falls in IL-6R and p-Stat3 (both within 4 hours) prior to evidence of proteolysis (PARP cleavage). ELISA of AZA-treated U266 conditioned media demonstrated a reduction in IL-6 and an immediate and sustained reduction in soluble IL-6R levels, consistent with reduced IL-6 synthesis and IL-6 receptor shedding, respectively. RT-PCR confirmed that the rapid falls in IL-6 and IL6-R were not secondary to reduced transcription with relative increases of 1.7 and 2.4-fold, respectively, by 8 hours post-AZA. Flow cytometry showed an initial rise and subsequent modest fall in both membrane and cytosolic gp130 while RT-PCR and immunoblot analyses showed a rapid fall in both SOCS3 transcriptional activity and protein levels, respectively. Time-course experiments with AZA 5μM and the addition of either physiological (100pg/ml) or supra-physiological (3ng/ml) concentrations of exogenous IL-6 over a 96-hour period demonstrated that 3ng/ml but not 100pg/ml IL-6 was capable of partially restoring Stat3 phosphorylation but that neither concentration of exogenous IL-6 prevented AZA-induced cell death. NFkB activity was evaluated with a luciferase reporter enzyme system following treatment for 24 hours with AZA 1μM, 5μM and 10μM. Marked suppression of NFkB activity was seen at both the 5μM and 10μM dose levels. Flow cytometric evaluation of phospho-IkB and IkB showed a modest decline in the pIkB:IkB ratio over the same period. Finally, combined treatment of U266 with AZA and dexamethasone demonstrated synergistic killing (synergism quotient = 1.2) when AZA was administered prior to dexamethasone but not in the reverse order. We conclude that AZA induces down-regulation of both IL-6 signalling and NFkB activity. Furthermore, our data suggest that the former effect is most likely mediated by rapid AZA-induced inhibition of protein synthesis. The mechanism of NFkB inhibition requires further clarification. AZA represents a potentially novel therapeutic approach to MM.