Repurposing ribosome-targeting antibiotics to overcome resistance to venetoclax in acute myeloid leukemia

ABSTRACT

Venetoclax, a selective B-cell lymphoma 2 inhibitor, has shown promise in the treatment of acute myeloid leukemia. However, the development of drug resistance limits its clinical efficacy. In our study, we discovered that ribosome-targeting antibiotics can be repurposed to overcome venetoclax resistance in AML cells through activation of the integrated stress response.

Venetoclax has recently emerged as an exciting therapeutic option for AML patients. It specifically disrupts the interaction between B-cell lymphoma 2 (BCL-2), an antiapoptotic protein, and proapoptotic proteins in the BCL-2 family, such as BCL-2-like protein 11 (BCL2L11, commonly known as BIM) and BCL-2-associated X protein (BAX).¹ Venetoclax treatment of AML cells that are dependent on BCL-2 for survival triggers cell death through the intrinsic apoptotic pathway.² In an early-phase clinical trial, venetoclax monotherapy was found to have limited clinical efficacy.³ However, when used in combination with hypomethylating agents (HMAs), such as azacitidine, or low-dose cytarabine, it achieved significantly higher response rates.^4,5 Nevertheless, the development of drug resistance was still a common problem. The mechanism of resistance has been attributed to increased expression of other antiapoptotic proteins in the BCL-2 family including myeloid cell leukemia 1 (MCL-1) and BCL-2-like protein 1 (BCL2L1, commonly known as BCL-X_L).⁶ Although suppression of these proteins can potentially overcome resistance to venetoclax, the simultaneous inhibition of multiple antiapoptotic proteins may cause severe toxicities to normal tissues, thereby narrowing the therapeutic window.

Given the above issues, we set out to identify additional genes that upon inactivation would overcome resistance to venetoclax in AML. To accomplish this goal, we performed a genome-wide clustered regularly interspaced short palindromic repeats (CRISPR) screen using an AML cell line with acquired resistance to venetoclax.⁷ Through this screen, we identified numerous genes that encode components of the mitochondrial ribosome, suggesting that mitochondrial translation is required to maintain venetoclax resistance. Since antibiotics that target bacterial ribosomes also act on mitochondrial ribosomes,⁸ we reasoned that ribosome-targeting antibiotics should similarly sensitize AML cells to venetoclax. Indeed, multiple ribosome-targeting antibiotics dramatically sensitized resistant AML cell lines and primary cells to venetoclax, with tedizolid and doxycycline showing the greatest effects.

How do ribosome-targeting antibiotics overcome resistance to venetoclax? The simplest explanation would be a reduction in the expression of the antiapoptotic proteins, MCL-1 and BCL-X_L. However, we did not observe changes in their expression or functional activity after treatment with tedizolid. Thus, we reasoned that other mechanisms must be involved. An important clue toward understanding the mechanism surfaced when we performed RNA sequencing analysis on resistant cells treated with venetoclax, tedizolid, or a combination of the two drugs. We found that genes involved in oxidative phosphorylation (OXPHOS) were upregulated in all three conditions. Although this observation was not surprising for cells exposed to tedizolid (presumably as a compensatory response to the suppression of mitochondrial translation), it was unexpected for cells treated with venetoclax. In subsequent experiments, we found that treatment with venetoclax alone suppressed mitochondrial respiration and specifically reduced the stability and activity of complex I (NADH dehydrogenase) in the electron transport chain. These findings uncovered a previously unrecognized role for BCL-2 in maintaining mitochondrial respiration through stabilization of complex I.

How might suppression of mitochondrial respiration by venetoclax explain its synergy with ribosome-targeting antibiotics? One of the well-documented consequences of mitochondrial stress is the activation of the integrated stress response (ISR).⁹ This adaptive response is meant to activate protective pathways that overcome cellular stress. However, if the response becomes excessive or prolonged, it triggers pathways that promote cell death. Given our earlier observations on the impact of each drug on mitochondrial respiration, we hypothesized that perhaps each drug individually may trigger a sublethal response, but when combined, they activate a lethal response. We envisioned it as a synthetic lethal interaction between the two drugs (Figure 1). Indeed, we found that ISR activation was modest after the treatment of either drug alone but was significantly more intense after combination treatment. To prove that this heightened ISR was necessary for cell death, we utilized a tool compound known as integrated stress response inhibitor (ISRIB) which blocks the downstream effects of ISR activation.¹⁰ In line with our hypothesis, the addition of ISRIB completely abrogated the ability of tedizolid to sensitize cells to venetoclax. Moreover, treatment with other tool compounds that directly activate ISR without causing cellular stress was sufficient to overcome resistance to venetoclax, highlighting the central importance of ISR activation in mediating this synergy.

Figure 1.

Potential mechanism for venetoclax and ribosome-targeting antibiotics to synergize in venetoclax-resistant acute myeloid leukemia cells. Venetoclax treatment alone causes mitochondrial stress by inhibiting complex I in the electron transport chain. Treatment with ribosome-targeting antibiotics alone (e.g., tedizolid and doxycycline) also causes mitochondrial stress by blocking mitochondrial ribosomes (mitoribosomes), resulting in inhibition of mitochondrial translation. Each treatment alone results in the inhibition of oxidative phosphorylation (OXPHOS) and subsequent activation of the integrated stress response (ISR) at a level that is not sufficient to trigger cell death. However, the combination of the two drugs causes severe mitochondrial stress and activates a heightened ISR that ultimately triggers cell death. I, II, III, IV, V – complexes in the electron transport chain.

Our next goal was to determine whether this strategy works in vivo. We established cell line-derived (CDX) and patient-derived (PDX) xenograft models of AML and showed that combination treatment with tedizolid and venetoclax had superior anti-leukemic activity compared to single-agent treatments. We further showed that this combination targeted leukemic stem cells through secondary transplantation assays. Venetoclax is currently approved for the treatment of AML in combination with HMAs, such as azacitidine. Thus, we tested whether the addition of tedizolid can further enhance the efficacy of this backbone regimen. Using a PDX model with acquired venetoclax resistance, we found that the triplet combination of venetoclax, azacitidine, and tedizolid was significantly more effective than venetoclax plus azacitidine in reducing leukemic burden in mice with established disease without causing additional toxicities.

In summary, our work demonstrates that ribosome-targeting antibiotics can be repurposed to overcome resistance to venetoclax in AML cells by activating ISR and the triplet combination of venetoclax, azacitidine, and tedizolid has potent anti-leukemic activity in vivo. Our work provides the rational basis for testing this triplet combination in clinical trials.

Funding Statement

This project was supported by the UpCycle Drug Repurposing Grant from the Cancer Research Society, an Operating Grant from the Leukemia and Lymphoma Society of Canada, and a Project Grant from the Canadian Institutes of Health Research (PJT-159521).

Disclosure of potential conflicts of interest

S.M.C has received an honorarium from Celgene and Agios. S.M.C. has received research funding from Agios, Celgene, and AbbVie Pharmaceuticals. D.S. has no potential conflicts of interest to disclose.