Background:

Venetoclax (VEN), a specific BCL2 inhibitor, exerts anti-leukemic effects in various high-risk (HR) B-ALL subtypes, such as ALL with mixed lineage leukemia (MLL) gene rearrangements (MLLr ALL) (PMID: 26711339), Philadelphia chromosome-positive (Ph+) (PMID 30546081) or hypodiploid B-ALL (PMID 30862722). Nevertheless, despite high rationale for targeting BCL2 in these subtypes of B-ALL, VEN monotherapy is not effective enough to completely eliminate leukemic cells. For this reason identification of other drugs that could sensitize leukemic cells to VEN may become beneficial treatment strategy in HR ALL. Previously, we showed that the enzymes of the thioredoxin system are upregulated in primary B-ALL cells and that auranofin (AUR), a thioredoxin reductase inhibitor, effectively kills leukemic cells in vitro and in vivo. Importantly, elements of the thioredoxin system are not only balancing redox homeostasis within the cells, but may also interact with other pathways, including anti-apoptotic signaling. Considering above, we hypothesized that AUR may potentiate VEN efficacy in HR B-ALL.

Methods:

To evaluate cytostatic/cytotoxic effects of VEN+AUR combination by MTT assay and propidium iodide (PI)-staining we used HR B-ALL cell lines, including SEM (MLLr ALL), BV-173 (Ph+ ALL) and NALM-16 (hypodiploid ALL). Patient derived xenograft cells (PDX) were generated through long-term propagation of primary B-ALL samples in immune-deficient NSG mice. Ex vivo drug testing in co-culture system was performed using primary bone marrow-derived mesenchymal stem cells (BM-MSC) and murine stromal OP9 cell line. NOXA genomic knockout (KO) in SEM cells was established by CRISPR/Cas9 system. Chromatin accessibility within PMAIP1 gene (encodes for NOXA) was detected using ATAC-seq.

Results:

We observed that AUR sensitizes HR B-ALL cell lines to VEN, as determined by MTT and PI-staining. Further, we mimicked the bone marrow support of stromal cells towards B-ALL and evaluated its impact on the response to VEN+AUR. For this reason we employed an ex vivo co-culture system of B-ALL PDX cells with primary BM-MSC or an OP9 cell line. In all tested PDX samples representing diagnostic/relapsed MLLr ALL (n=8), Ph+ ALL (n=2) and Ph-like ALL (n=2) we observed synergistic effect of this combination (Fig. 1A). Next, we determined the efficacy of VEN+AUR combination in vivo using a PDX model of MLLr B-ALL. We observed that administration of VEN+AUR diminished the progression of leukemia during a 3 week-long treatment more effectively than any single drug alone, which reflected in longer survival of NSG mice (Fig. 1B). Subsequently, we aimed to uncover the mechanism responsible for the synergistic action of VEN+AUR. In cells treated with both drugs we observed enhanced caspase activation and changes in the levels of BCL2 family proteins involved in apoptotic signaling. In particular, we found that AUR strongly upregulates a pro-apoptotic NOXA protein, both in HR B-ALL cell lines and in MLLr ALL PDX samples (Fig. 1C). To evaluate whether NOXA induction is functionally relevant for the cell death mediated by VEN+AUR, we generated SEM cells with a NOXA genomic KO. Lack of NOXA significantly abolished VEN-single agent as well as VEN+AUR combination cytotoxicity, demonstrating its dependence on NOXA expression (Fig. 1D). We then showed that NOXA is regulated at the transcriptional level, as co-treatment with AUR and the transcription inhibitor, actinomycin D, abolished AUR-mediated NOXA induction at mRNA and protein levels in SEM cells. Additionally, to test whether AUR-treatment itself provokes changes in chromatin accessibility within the NOXA encoding gene (PMAIP1) we performed ATAC-seq. We observed a clear increase in accessibility at PMAIP1 in response to AUR, which correlated with transcriptional induction of NOXA. Moreover, ChIP-qPCR revealed that increased ATAC peaks within PMAIP1 were associated with an increase in H3 lysine 27 acetylation (H3K27ac) - an epigenetic mark associated with open chromatin conformation.

Conclusions:

Our results demonstrate that FDA-approved drug, AUR, is a promising candidate to be used in combination with VEN for the therapy of HR B-ALL subtypes. Importantly, NOXA induction by AUR plays a central role in the VEN+AUR synergistic cytotoxicity. More studies elucidating the mechanism of NOXA upregulation by AUR are underway.

Disclosures

Milne:OxStem Oncology (OSO), a subsidiary company of OxStem Ltd.: Other: Founding shareholder .

Author notes

*

Asterisk with author names denotes non-ASH members.