BORTEZOMIB LOWERS THE STAT3 PATHOGENETIC ROLE IN T-LGLL NEUTROPENIA DEVELOPMENT BY RESTORING MIR146B EXPRESSION

Background: Symptomatic or severe neutropenia is the most common indication for treatment in T-Large Granular Lymphocytes Leukemia (T-LGLL) patients. To date, the therapeutic strategy mostly relies on immunosuppressive agents, however the ORR is quite low (<60%). A targeted therapy for neutropenic patients has not been developed yet, due to the lack of knowledge about the pathogenesis of this cytopenia. One of the most supported hypotheses relies on Fas Ligand (FasL) involvement, since high levels of this protein, accounting for regulation of neutrophils lifespan, were detected in sera of neutropenic patients. Of note, we recently found that the increased FasL transcription was induced by the high STAT3 activation that specifically characterizes T-LGLs of neutropenic patients. We also showed that a microRNA(miR)146b down-regulation, that represents a signature of these patients, led to the translation of Human antigen R (HuR), an essential FasL mRNA stabilizer. Several drugs are under pre-clinical evaluation to improve the therapeutic possibilities for T-LGLL. Among these, the proteasome inhibitor Bortezomib (Bz) has been demonstrated to affect in-vitro leukemic LGL viability by increasing the levels of the pro-apoptotic protein BID and by inhibiting the constitutive activation of NF-kB pathway. Despite its promising effects, up to date Bz has never been tested in T-LGLL patients and its effects on STAT3-axis have never been investigated. Aims: Since in different neoplastic conditions miR146b expression is epigenetically silenced and in malignant T-lymphocytes the DNA Methyl-Transferase 1 (DNMT1) is reported to be STAT3-induced, our work aims to investigate the potential linkage between STAT3 and miR146b and to evaluate whether Bz might prevent STAT3 pathogenetic role in neutropenia development. Methods: T-LGL were immunomagnetically purified from peripheral blood of untreated T-LGLL patients. The methylation status of miR146b promoter was assessed by Methylated DNA Immuno-Precipitation (MeDIP). T-LGL were treated for 72h-2.5 µM 5-aza-2’-deoxycytidine (DAC), an inhibitor of methyl-transferase activity, or 24h-15 µM STATTIC, a STAT3 inhibitor, or 48h-5 µM Bz. Transcriptional and protein expression were evaluated by RTq-PCR and western blot assays, respectively. Results: MeDIP analysis highlighted a significantly higher level of 5-methyl-cytosine in −687/−496 bp (+141%) and −149/+98 bp (+58%) regions upstream the miR-146b transcription start site in T-LGL of neutropenic as compared to non-neutropenic patients. DAC treatment restored miR146b primary transcript expression in T-LGL of neutropenic patients, suggesting that it was epigenetically silenced. By specifically inhibiting STAT3 activity, we showed a 58% DNMT1 transcriptional decrease in the STATTIC treated T-LGL, as compared to controls, providing proof of a molecular linkage, likely mediated by DNMT1, between STAT3 activation and miR146b promoter methylation. Regarding Bz evaluation, a reduction of STAT3 activation (70%) and of DNMT1 transcription (67%) was observed in treated vs untreated conditions. Among miR146b downstream targets, a decrease of HuR protein (44%) and, thus, of FasL mRNA (43%) was observed in Bz-treated T-LGL, consistent with the restoration of miR-146b expression. Summary/Conclusion: Our data demonstrate the presence of a STAT3- miR46b-FasL axis in neutropenic T-LGLL patients. We also show for the first time that Bz down-regulates STAT3 activation in in-vitro treated leukemic T-LGL, offering the rationale for testing Bz in clinical trials to manage this high-risk subset of patients