Inducible Ablation of HSPA5 Suppresses BCR-ABL1-Driven Leukemia through Massive Accumulation of Reactive Oxygen Species.

Abstract 1976

Poster Board I-999

HSPA5 is a member of the HSP70 family and involved in the folding of proteins including the immunoglobulin m heavy chain in the ER. Of note, the HSPA5 locus at 9q34 is closely located to the ABL1-breakpoint on der9, which arises during the t(9;22)(q34; q11) translocation. Comparing gene expression levels in normal pre-B cells and myeloid progenitors to BCR-ABL1-driven Ph⁺ ALL and CML, we observed that HSPA5 (GRP78; BIP) is strongly upregulated in BCR-ABL1-driven leukemia cells.

For these reasons, we performed a genetic loss-of-function analysis via Cre-mediated deletion of HSPA5 and HSP90B1 to study the function of HSPA5 and its related interaction partner HSP90B1 (GRP94) in BCR-ABL1-driven Ph⁺ ALL and CML. Bone marrow from HSPA5^fl/fl and HSP90B1^fl/fl mice was incubated under B lymphoid (IL7) or myeloid (IL3, IL6, SCF) progenitor cell conditions. After cytokine priming, bone marrow cells were transformed with either p190 or p210 BCR-ABL1 to induce Ph⁺ ALL-like or CML-like leukemia, respectively. Subsequently, HSPA5^fl/fl and HSP90B1^fl/fl Ph⁺ ALL-like and CML-like leukemias were transduced with either Cre-GFP or a GFP empty vector control. Increase or depletion of GFP⁺ cells over time was taken as an indication of the functional relevance of Cre-mediated deletion of HSPA5 and HSP90B1 (confirmed by genomic PCR and Western blot). These experiments showed that Cre-mediated deletion of HSP90B1 had no significant effect on viability and proliferation of leukemia cells and the fraction of Cre-GFP⁺HSP90B1^fl/fl leukemia cells remained stable (deletion confirmed by genomic PCR and Western blot). Interestingly, Cre-GFP⁺HSP90B1^fl/fl leukemia cells showed a strong compensatory upregulation of HSPA5 upon deletion of HSP90B1. To search for subtle effects of HSP90B1-ablation, we studied changes of Imatinib-sensitivity and colony formation. However, deletion of HSP90B1 had no effect on Imatinib-sensitivity and the ability of leukemia cells to form colonies in methylcellulose. Conversely, Cre-mediated deletion of HSPA5 could not be compensated in Ph⁺ ALL-like and CML-like leukemia cells: Two days after Cre-transduction, HSPA5-deleted Ph⁺ ALL cells sharply declined. Cre-mediated deletion of HSPA5 in Ph⁺ ALL cells resulted in G0/G1 cell cycle arrest and cell death in vitro. In HSPA5^fl/fl CML cells, the effect of Cre-mediated HSPA5-ablation was delayed until after day 4. Gene expression analysis showed that HSPA5-deletion resulted in strong upregulation of p21 (CDKN1A) and HSPA5-interacting proteins including ORP150, DNAJC3, HSP90B1 and non-functional HSPA5 transcripts. We next tested whether survival signals from the microenvironment can rescue leukemia cell survival upon Cre-mediated deletion of HSPA5 in vivo. For this experiment, HSPA5^fl/fl Ph⁺ ALL cells were labeled with firefly luciferase and injected into sublethally irradiated NOD/SCID recipients in the presence (Cre-GFP⁺) or absence (GFP⁺) of HSPA5-deletion. In agreement with in vitro observations, HSPA5-deletion significantly extended the latency of leukemia (median survival day 10 vs day 19; n=8 per group; p=0.04). In contrast to in vitro studies, however, all mice injected with HSPA5-deleted Ph⁺ ALL showed progressive leukemia growth and ultimately succumbed to the disease. Genomic PCR confirmed complete HSPA5-deletion in all Cre-GFP-transduced leukemias (n=8).

Since HSPA5-deleted Ph⁺ ALL cells accumulated in the bone marrow, we hypothesize that bone marrow microenvironment factors can rescue survival and proliferation of HSPA5-deficient Ph⁺ ALL cells. While bone marrow stroma-derived SDF1a, the natural ligand of CXCR4 expressed on Ph⁺ ALL cells failed to rescue HSPA5-deficient Ph⁺ ALL cells, we found accumulation of reactive oxygen species (ROS) as the likely cause of cell death following HSPA5-deletion in Ph⁺ ALL: levels of ROS following HSPA5-deletion were as high as ROS levels induced by incubation of non-deleted Ph⁺ ALL cells with 25 mmol/l H₂O₂. Given the low oxygen tension within the bone marrow compared to cell culture conditions, these findings provide a rationale why Ph⁺ ALL cells can survive deletion of HSPA5 within the bone marrow microenvironment (pO₂ 1-7%) but not under in vitro conditions (pO₂ 21%). Since BCR-ABL1 kinase activity results in ROS production, these findings suggest that inhibition of ROS scavengers represents a novel therapeutic approach to BCR-ABL1-driven leukemias.