Bone Marrow Stroma Protects BCR-ABL-Negative Myeloproliferative Neoplasms from Ruxolitinib- and Vorinostat-Induced Apoptosis

Background: The classical BCR-ABL-negative myeloproliferative neoplasms (MPN) are characterized by increased proliferation of hematopoietic precursors in the bone marrow resulting in elevated numbers of terminally differentiated cells. Despite the recent advances in the understanding of the biology of MPN, there is still no curative treatment for MPN except for bone marrow transplantation. The discovery of JAK-STAT constitutive activation in the majority of MPN patients led to the development of clinical studies targeting MPN with Ruxolitinib, a JAK1/2-specific inhibitor. However, despite achieving significant reductions in splenomegaly and symptomatic improvement in MPN patients, JAK inhibition failed to induce complete remissions and eradicate the malignant clone. In alternative, histone deacetylase inhibitors have shown success in the treatment of several hematological malignancies but their efficacy in MPN is limited.

Aims: Ruxolitinib and the HDAC inhibitor Vorinostat both fail to eradicate the neoplastic clone in MPN patients. Here, we tested the hypothesis that the bone marrow stroma confers resistance to these drugs by preventing their cytotoxic effects on neoplastic cells.

Material and Methods: MPN patient derived cell lines (SET-2 and HEL) were cultured alone, with HS-5 bone marrow stroma or with HS-5 conditioned medium in the presence of clinically-relevant concentrations of Ruxolitinib and Vorinostat. At different time points, cellular viability was analyzed by staining with Annexin-V/PI and CD45-APC (to distinguish MPN cells from the HS-5 cells). Furthermore, cells were lysed and RNA was extracted for gene expression analysis. Activation of specific signaling pathways was determined by Western-blot using phospho-specific antibodies for effectors of the following signaling pathways: PI3K-Akt-mTOR; MEK-Erk; MAPK-JNK; NF-κB and JAK-STAT.

Results: Treatment of SET-2 and HEL cells with both Vorinostat and Ruxolitinib promoted apoptosis and decreased proliferation. Importantly, apoptosis was significantly abrogated when MPN cells were cultured in the presence of a stromal layer of HS-5 cells or HS-5 conditioned medium. The stroma protective effect was maintained for up to 6 six days and was also dependent on the concentration of the drugs. These effects correlated with altered expression in MPN cells of genes associated with inflammatory processes, apoptosis and proliferation (CDKN1A, IER3, BIRC3, TNFRSF8, TNFRSF9, COX2, IL1B, XIAP), and with the activation of signaling pathways important for cellular homeostasis, such as PI3K-Akt-mTOR, MAPK-JNK, JAK-STAT and NF-κB as shown by increased phosphorylation of Akt (Ser473); GSK3α/β (Ser9/Ser21); S6 (Ser235/236); STAT3/5 (Tyr705/Tyr694) and IKBα (Ser32).

Summary/Conclusions: Overal, we show that bone marrow stroma protects MPN cells from the cytotoxic effects of two clinically effective pharmacological agents of different classes: Vorinostat and Ruxolitinib. This protective effect is likely achieved, through the up-regulation of genes that inhibit apoptosis (BIRC3, TNFRSF8, TNFRSF9) and also through the activation of pro-survival signaling pathways such as PI3K-Akt-mTOR. We did not observe any effects of the bone marrow stroma on proliferation or differentiation, suggesting that the main effect of stromal cells is to prevent apoptosis of the neoplastic cells. Our results identify a possible cell non-autonomous mechanism by which Ruxolitinib fails to eradicate the neoplastic clone in MPN patients. Through the identification of the factors released by the bone marrow stroma, we hope to uncover why the bone marrow protects MPN cells and possibly define novel therapeutic strategies maximizing the use of Ruxolitinib and Vorinostat in MPN.