The Achilles Heel of Malignant Rhabdoid Tumors

In a previous issue of Cancer Research, Howard and colleagues identified MDM2 and MDM4 as drug targets in malignant rhabdoid tumors (MRT; ref. 1). This finding lays a foundation for designing effective targeted therapies for this aggressive tumor. Malignant rhabdoid tumors (MRT) are rare pediatric tumors that mainly arise in the kidney, brain, or soft tissues (2). Current treatment options for MRTs are surgery, chemotherapy, and radiation, none of which are particularly effective. Targeted therapies are therefore much needed to improve the survival and quality of life of patients with MRT. Finding actionable mutations in MRTs is challenging as recent genome-wide sequencing studies revealed few mutations in these tumors except for loss-of-function mutations in the SMARCB1/SNF5/INI1/BAF47 gene (2). SMARCB1 encodes a core subunit of the mammalian SWI/SNF chromatin remodeling complex; subunits of this complex are mutated in approximately 20% of all human cancers (3). Although SMARCB1 is a bona fide tumor suppressor in MRT, the mechanism by which SMARCB1 deficiency drives MRT is not well understood, however, gene expression changes related to cell-cycle dysregulation and stem cell differentiation have been implicated. To date, an important question in the field is how the mutation of SMARCB1, a hallmark of MRT, can be exploited therapeutically.

To identify targetable vulnerabilities in SMARCB1-mutant MRTs, Howard and colleagues turned to unbiased RNAi and CRISPR library screening. Loss-of-function screens of this type are powerful tools for discovering genotype-specific dependencies in cancer cells. These approaches can identify synthetic lethal interactions, which may serve as drug targets, particularly in cancer cells that are primarily driven by tumor suppressor loss. Taking advantage of Project Achilles, a large-scale effort that has analyzed hundreds of cancer cell lines from different tissue types with genome-wide RNAi and CRISPR screens to identify their functional vulnerabilities (4), the authors discovered that MRT cells are particularly sensitive to genetic and pharmacologic suppression of MDM2 and MDM4 compared with other cancer cell lines.

Unlike other solid tumors, the majority of MRTs are wild-type (WT) for the TP53 tumor suppressor gene and retain a functional p53 pathway (5, 6). MDM2 and MDM4 are both negative regulators of p53; MDM2 binds to and degrades p53 while MDM4 sequesters p53 (7). MDM2 and MDM4 amplifications are commonly seen in human tumors including pediatric tumors such as osteosarcoma, however, their genetic alternations are rare in MRTs (5, 6). MDM2 and MDM4 also possess p53-independent activities (7). In their screen, Howard and colleagues also identified two other negative regulators of p53, PPM1D/WIP1 and USP7/TEF1/HAUSP. This suggested that dysregulation of p53 activity was the underlying mechanism of MDM2 and MDM4 dependency in MRT cells. Interestingly, they found that MRT cells are more sensitive to the genetic ablation of these negative regulators of p53 than other TP53 WT cancer cell lines, suggesting MRTs might be highly sensitive to p53-targeted therapies. Indeed, the authors showed that two structurally unrelated compounds, the small-molecule MDM2 inhibitor idasanutlin and the dual MDM2- and MDM4-targeted stapled peptide ATSP-7041, had a greater effect on the viability of MRT cells than other TP53 WT cancer cell lines. In MRT cells, both inhibitors induced the expression of p53 target genes and deletion of TP53 resulted in drug resistance. Thus, the cytotoxicity of these inhibitors is due to the activation of p53 in MRT cells. MDM2 and MDM4 inhibition led to both apoptosis and cell-cycle arrest in MRT cells. Interestingly, the cytostatic response in some MRT cells was associated with an irreversible, senescent-like phenotype. Thus, activation of p53 can serve as an effective means to eliminate MRT cells.

Mechanistically, Howard and colleagues demonstrated that SMARCB1 loss has an upstream role in the dysregulation of the proapoptotic activity of p53 and this effect appears reversible. The authors restored the expression of SMARCB1 under an inducible promoter in SMARCB1-deficient MRT cells and observed decreased sensitivity to idasanutlin and ATSP-7041. The restoration of SMARCB1 appears to preferentially diminish the effect of these drugs on the induction of proapoptotic p53 target genes including BBC3/PUMA and TP53I3/PIG3. It is known that p53 can regulate cell cycle and proapoptotic genes through distinct mechanisms. Given that SMARCB1 is a core subunit of the SWI/SNF complex, it is possible that this complex, through its chromatin remodeling activity, contributes to the differential regulation of p53 target genes. It remains unclear why SMARCB1 emerges as a dominant tumor suppressor driver in lieu of p53 in MRT, given that loss of SMARCB1 could sensitize MRT cells toward p53-induced apoptosis. In other types of cancer, mutations in SWI/SNF subunits are also frequently found to be mutually exclusive with TP53 mutations (8). These observations suggest that SWI/SNF and p53 might be regulating the same pathway during tumorigenesis. One speculative idea would be that loss of SWI/SNF activity leads to global chromatin changes that strike a good balance between promoting cell proliferation and shaping cell fate specification (3). This would spare the need for p53 mutation. However, the fitness cost of such mutational choice would be an increased chromatin accessibility of p53-regulated proapoptotic genes and, consequently, heightened sensitivity to p53-induced cell death.

To explore the translational potential of their findings, Howard and colleagues treated an MRT xenograft model using idasanutlin and ATSP-7041. Both compounds significantly attenuated tumor growth and extended survival. Remarkably, idasanutlin caused complete and durable response in half of the mice at a dose that is well tolerated. In tumors treated with either drug, both decreased proliferation and increased apoptosis was observed, although it was unclear whether senescence also contributed to the durable response. Using a 13-gene signature that is associated with MDM2 sensitivity, the authors showed that human primary MRT samples scored higher for this signature than both normal tissue and other pediatric tumors. Together, these results suggest that MDM2 and MDM4 could be an Achilles heel for MRTs. A previous study by The Pediatric Preclinical Testing Program also found that MDM2 inhibitors inhibited the proliferation of MRT cells (9). The study by Howard and colleagues has provided both new mechanistic insights into the synthetic lethality between SMARCB1 and the MDM2–p53 axis and a strong scientific rationale for why MRT might be particularly sensitive to MDM2 and MDM4 inhibitors. Idasanutlin is currently being tested in clinical trials for several cancer types but not specifically for MRT. The findings by Howard and colleagues warrant further clinical investigation of idasanutlin and other MDM2 and MDM4 inhibitors in MRT to evaluate their efficacy in this rare but lethal form of childhood cancer.

The content is solely the responsibility of the authors and does not necessarily represent the official views or policies of the National Institutes of Health and the Department of Health and Human Services. The mention of trade names, commercial products or organizations does not imply endorsement from the US Government.

No potential conflicts of interest were disclosed.