Walking the Tightrope: Balancing Delicate Inflammation Response to Eradicate Acute Myeloid Leukemia

Inflammation contributes to multiple cancer hallmark capabilities through both cell-extrinsic and cell-intrinsic cues. Inflammation stimulates the production of growth and/or proangiogenic factors that sustain proliferative signaling, restrain cell death mechanisms, and/or silence anticancer immune responses (1, 2). Whether tumor cell–intrinsic inflam­matory factors might constitute tissue-specific cancer vulnerabilities per se remains largely unknown.

The hematopoietic system, owing to its singular role in regulating immune cell production and tissue repair, is particularly poised to respond to inflammatory signals caused by environmental insults such as infection or injury. All blood lineages are involved in the initiation, progression, and resolution of inflammatory responses, which often result in an increased production of myeloid cells and platelets (3). Upon the restoration of normal tissue homeostasis, inflammation needs to be quickly suppressed; failure to do so can trigger chronic inflammation that contributes to the acquisition of premature aging phenotypes and myeloid neoplasms. For instance, soluble factors such as TNF, IFNs, and IL6 can support aging-related hematologic neoplasms such as myeloproliferative neoplasms (MPN), myelodysplastic syndromes, and acute myeloid leukemia (AML). Likewise, defects in TET2, one of the most commonly mutated genes in these myeloid diseases, reportedly enhance the production of inflammatory cytokines (e.g., IL6) in both mice (4) and humans, where it is described to play prominent roles in atherosclerosis (5). Motivated by these intriguing observations, the seminal work performed by Ellegast and colleagues (6) in this issue of Cancer Discovery establishes a thorough framework for the identification of inflammation-related and AML-specific cell-intrinsic vulnerabilities, followed by the in-depth characterization of the mechanisms underpinning their function as AML dependencies.

Using gene expression data sets from three large AML cohorts queried by single-sample gene set enrichment analysis, Ellegast and colleagues established that nearly half of all patient samples were enriched for immune- and inflammatory-related gene sets. Strikingly, these primary AML samples were also enriched for monocytic lineage signatures and were accordingly overrepresented among the M4/M5 French–American–British classification (FAB) subtypes, suggesting a critical functional interrelation between the molecular machineries involved in inflammation and monocytic differentiation. In light of these findings, they elegantly mined three orthogonal high-throughput pooled CRISPR–Cas9- and short hairpin RNA (shRNA)–based screening studies in order to nominate a core set of five gene dependencies that satisfied the essential prerequisite of being (i) functionally enriched within the monocytic lineage, thereby making them ideal AML-specific vulnerabilities, and (ii) potentially involved in the regulation of cell-intrinsic inflammatory processes, thereby making them selective liabilities of leukemic cells over normal hematopoietic counterparts. Among these candidate genes, they pinpointed two mediators of monocytic differentiation, MYB and SPI1/PU.1, previously identified as regulators of cellular responses to inflammatory cues. Specifically, IL1 has been shown to trigger SPI1/PU.1-dependent cell-cycle restriction in a subset of hematopoietic stem cells (7), whereas MYB was described to negatively affect IL1α expression, thereby restricting NF-κB activity and the inflammatory circuit (8). In addition, this core set of AML-selective vulnerabilities includes a still unexplored gene, IRF2BP2, whose role as an AML dependency was subsequently evaluated using IRF2BP2-directed single-guide RNAs (sgRNA) and shRNAs in a large collection of AML cell lines and patient-derived xenografts exhibiting monocytic features. In these models, IRF2BP2 suppression led to a profound reduction in AML cell growth and leukemia burden in vivo through induction of apoptosis. Although pyroptosis and necroptosis are typically considered the primary forms of cell death caused by inflammation, they were not observed in the context of IRF2BP2 suppression (9). Interesting follow-up studies could be designed to uncover the potential existence of an AML-restricted inflammatory circuit linking the specific markers of monocytic differentiation—namely, SPI1/PU.1 and MYB—to IRF2BP2.

A remarkable feature of this study stems from the orthogonal deployment of various innovative functional technologies combined to comprehensively interrogate IRF2BP2 function in AML. AML cells with a complete loss of endogenous IRF2BP2 (generated by virtue of CRISPR-directed editing) were modified to express an exogenously engineered IRF2BP2 variant, FKBP12^F36V-tagged IRF2BP2, that can be rapidly degraded by the use of a small dTAG-13 molecule (dTAG-13 binds to the FKBP12^F36V tag to recruit the proteasome machinery and mediate its selective degradation). This cutting-edge system allowed for acute and brief IRF2BP2 depletion, which facilitated deep epigenetic- and transcriptomic-based mechanistic studies within time frames (as short as 6 hours after dTAG-13 supplementation) that would be incompatible with other approaches. Indeed, ablation using genetic tools often requires longer kinetics and is thus more likely to trigger confounding transcriptional programs that may mask subtle epigenetic and transcriptional variations, such as those involving inflammatory responses. The dTAG13 system enabled the authors to establish that IRF2BP2 is a repressor of NF-kB–mediated TNFα signaling. IRF2BP2 suppression leads to the upregulation of IL1β, which in turn promotes NF-κB nuclear translocation and activation, prompting AML cell death through activation of the caspase-8 and caspase-3 apoptotic cascade (Fig. 1).

Collectively, these results suggest that AML progresses on a delicate knife-edge. The intrinsic inflammatory response machinery inherent to the monocytic lineage of origin of AML blasts provides them with an exquisite capability to survive and proliferate in their milieu despite inflammatory signals that might otherwise threaten their expansion. As a direct consequence, this creates an innate dependency of AML blasts on a finely tuned inflammatory response machinery that needs to be tightly controlled to avoid AML collapse. A heightened inflammatory response by suppression of IRF2BP2 tips the delicate balance of AML blasts, perhaps just enough to topple them from their tightrope into cell death. This observation is reminiscent of the role of PARP1, which participates in the control of DNA repair processes and cancer cell survival upon moderate DNA damage. Although PARP1 inhibition exhibits mild anticancer activities in MPN, its suppression in the context of overwhelming DNA damage triggered by ruxolitinib-induced deficiencies in double-strand break repair pathways leads to massive MPN cell death (10).

This study was primarily designed to identify and validate IRF2BP2 as a cell-intrinsic inflammatory-related AML liability using experimental conditions, such as ex vivo culture or immunodeficient mouse models, which are destitute of immune system components. Nonetheless, the inflammatory circuit is also a fundamental mechanism reported to be a mediator of immune response through a process involving many molecular and cellular components consisting of lipid inflammatory mediators and cytokines including the NF-κB modulators, IFNγ, TNFα, IL1, as well as chemokines, which initiate and sustain a constant dialogue between bone marrow cellular components, leukemic cells, and immune cells. The work by Ellegast and colleagues thereby opens stimulating new avenues for the discovery of additional, and not yet defined, immune cell–related factors that may also contribute to mediate the antileukemia activity elicited by IRF2BP2 suppression.

The critical function played by IRF2BP2 in the control of AML growth described in this study illustrates the need for more research exploring how the rewiring of the inflammatory circuitry affects AML progression. Another path forward might be the development of new agents selectively targeting this transcriptional regulator, which could preferentially eliminate AML blasts while sparing their normal bone marrow counterparts whose inflammatory machinery might not be as much overawed.

H. Medyouf reports grants from the Hessen State Ministry for Higher Education, Research, and the Arts during the conduct of the study, as well as personal fees from Affimed outside the submitted work. A. Puissant reports financial support from grants from the ERC Starting Program (758848), INCA PLBIO (PLBIO20-246), and Amgen Innovations.