In This Issue

See article, p. 44.

Chemoprevention trials have shown that androgen deprivation with drugs such as 5α-reductase inhibitors decreases the risk of low-grade prostate cancer. However, the role of antiandrogen therapy in prostate cancer progression is unclear, as the incidence of high-grade prostate cancer was increased in these studies. To further evaluate the effect of androgen deprivation on tumor progression, Jia and colleagues used a preclinical genetically engineered mouse model of high-grade prostatic intraepithelial neoplasia (HG-PIN) precursor lesions that are initiated by loss of PTEN and do not progress to invasive tumors. Androgen deprivation by surgical castration or treatment with the androgen receptor antagonist MDV3100 induced an initial reduction in HG-PIN growth but eventually accelerated the outgrowth of invasive castration-resistant prostate cancer (CRPC) from stable PTEN-deficient PIN lesions in the ventral prostate. An alternative therapeutic strategy, concomitant ablation of the genes encoding the class I phosphoinositide 3-kinase (PI3K) isoforms p110α and p110β, suppressed tumor initiation driven by PTEN loss; pharmacologic PI3K inhibition similarly repressed the growth of HG-PIN and restored normal tissue architecture. In contrast, PTEN-deficient CRPC tumors exhibited enhanced mitogen-activated protein kinase (MAPK) signaling compared with HG-PIN lesions, and combined inhibition of MAPK and PI3K significantly impaired CRPC growth after androgen deprivation. These results suggest that antiandrogens may promote prostate cancer progression but that PI3K inhibition may be an effective targeted therapy for chemoprevention of PTENdeficient prostate cancer if appropriate dosing regimens can be found.

See article, p. 52.

Melanomas are frequently characterized by activating mutations in the BRAF and RAS oncogenes. However, despite the efficacy of the mutant BRAF inhibitor vemurafenib, resistance often occurs, and targeted therapies specific for mutant RAS-driven melanomas have not yet been developed. To address these issues, Held and colleagues performed a high-throughput drug screen of 150 small-molecule inhibitors on a panel of patient-derived melanoma cell lines with mutations in BRAF, RAS, or neither gene. Because most single agents showed variable efficacy across cell lines and did not completely block cell growth, the cytotoxic effect of pairwise drug combinations was evaluated at different concentrations. Clustering analysis identified many agents that interacted synergistically to induce cell death selectively in BRAF-mutant cells, including vemurafenib, AKT inhibitors, and EGF receptor (EGFR)/ERBB inhibitors. Furthermore, dual blockade of EGFR and AKT signaling was sufficient to resensitize vemurafenib-resistant BRAF-mutant cells to this drug and enhance growth inhibition. In contrast, although RAS-mutant cells were more resistant to paired drug treatments than were BRAF-mutant cells, the combination of statins, which inhibit 3-hydroxy-3-methylglutaryl-CoA reductase (HMG-CoA reductase), with pan–cyclin-dependent kinase (CDK) inhibitors increased cytotoxicity and significantly diminished xenograft tumor growth; this effect was mediated in part through loss of NRAS membrane localization and reduced downstream signaling. These results highlight the usefulness of this type of screen to uncover mutation-specific combinatorial therapies and suggest strategies to overcome drug resistance in melanoma.

See article, p. 68.

Inhibition of PARP1 blocks repair of DNA single-strand breaks, which are ultimately converted into double-strand breaks (DSB) during DNA replication. PARP inhibitors induce synthetic lethality in homologous recombination (HR)-defective cells that cannot repair these DSBs and may therefore have promise for cancers with mutations in HR-associated genes such as BRCA1 and BRCA2. Preclinical evaluation of PARP inhibitor response and resistance has identified increased P-glycoprotein drug efflux activity, residual BRCA1/2 activity, and restoration of BRCA1/2 function as potential resistance mechanisms, but other mechanisms may exist. Jaspers and colleagues treated P-glycoprotein–deficient mice bearing mammary tumors caused by a large, irreversible Brca1 deletion with the PARP inhibitor olaparib and observed that stable resistance could still develop in vivo. Interestingly, several olaparib-resistant tumors exhibited reduced 53BP1 protein expression and increased HR foci associated with loss-of-function mutations in the Trp53bp1 gene, consistent with previous work showing that 53BP1 loss partially restores HR in Brca1-deficient cells through inhibition of alternative DNA repair pathways. Short-term use of AZD2461, a PARP inhibitor intended to circumvent P-glycoprotein–mediated resistance, increased overall survival compared with olaparib although several Trp53bp1-mutant refractory tumors developed, but long-term AZD2461 use strongly suppressed the development of refractory tumors. Together, these findings identify 53BP1 loss as a mechanism of PARP inhibitor resistance in vivo and suggest that 53BP1 expression may be a biomarker of drug response while underscoring the potential of preclinical studies to uncover resistance mechanisms and guide the development of next-generation drugs.

See article, p. 82.

B-cell malignancies commonly harbor MYC translocations that lead to deregulated expression of MYC and its targets, many of which control energy metabolism and cell proliferation. Although direct pharmacologic inhibition of transcription factors such as MYC has proved difficult, targeting downstream pathways may be a feasible alternative. Because MYC overexpression induces upregulation of mTOR complex 1 (mTORC1), a critical mediator of cell growth, Wall and colleagues evaluated the effect of everolimus, a small-molecule mTORC1 inhibitor, on the initiation and maintenance of murine lymphoma driven by the Eμ-Myc transgene that mimics MYC rearrangements found in human disease. Strikingly, everolimus treatment prior to disease onset selectively eliminated premalignant B cells, restored normal B-cell differentiation, and strongly protected against lymphomagenesis, with few Eμ-Myc mice developing disease compared with 100% of placebo-treated mice. Everolimus also induced disease regression and significantly improved survival in mice with established Eμ-Myc lymphomas compared with placebo, although resistant clones eventually arose. Everolimus activity was not associated with increased apoptosis but rather oncogene-induced senescence, implicating mTORC1 in bypass of MYC-induced senescence in B lymphocytes and demonstrating that targeted therapies can exert anticancer activity by activating senescence. Importantly, deletion of p53 in Eμ-Myc mice conferred resistance to everolimus, indicating that this response was dependent on an intact p53 pathway and providing a strong rationale for consideration of MYC rearrangement and p53 status in clinical trials of everolimus in B-cell malignancies.

See article, p. 96.

Defects in EGF receptor (EGFR) internalization and degradation in cancer cells result in sustained signaling that contributes to resistance to EGFR inhibitors, underscoring the need to determine the factors that regulate EGFR trafficking. Sukhanova and colleagues used a network modeling approach to identify the mechanisms by which the cholesterol biosynthesis pathway demethylating enzyme sterol-C4-methyl oxidase-like (SC4MOL), which has been implicated in modulating sensitivity to EGFR antagonists, contributes to drug resistance. Silencing of SC4MOL or its partner NAD(P)-dependent steroid dehydrogenase-like (NSDHL) resulted in accumulation of C4-methylsterol metabolic intermediates and sensitized cancer cells to EGFR inhibitors, indicating a direct role for SC4MOL and its substrates in drug resistance. Network analysis of conserved yeast orthologs predicted a function for SC4MOL in vesicular transport; indeed, SC4MOL or NSDHL inactivation augmented endosomal trafficking and lysosomal degradation of EGFR and its coreceptors, ERBB2 and ERBB3, decreasing EGFR activation and downstream signaling. Similarly, Nsdhl deficiency enhanced platelet-derived growth factor receptor downregulation and impaired EGFR signaling in the skin. This effect was mediated in part by the ADP-ribosylation factor (ARF) 4 and ARF5 endosomal proteins, as suggested by bioinformatics analysis of yeast protein interactions. Importantly, SC4MOL ablation combined with cetuximab prevented xenograft tumor growth, whereas blockade of upstream metabolic enzymes rescued tumor growth and promoted cetuximab resistance in the absence of SC4MOL. These findings suggest that inhibition of these sterol metabolism proteins may improve the efficacy of EGFR inhibitors and limit drug resistance.

See article, p. 112.

Oncogenic activation of RAS signaling via somatic mutation of the KRAS, HRAS, or NRAS genes is one of the most common events in human cancer, but whether the remaining wild-type RAS isoforms support tumorigenesis remains unknown. Young and colleagues observed that cancer cell lines harboring a RAS mutation remained responsive to growth factor signaling, suggesting that RAS-mutant cells are not growth factor independent and that receptor tyrosine kinases promote proliferation by stimulating the remaining 2 wild-type RAS isoforms. Consistent with these possibilities, depletion of the wild-type RAS isoforms blocked EGF-dependent activation of mitogen-activated protein kinase (MAPK) signaling and inhibited cell growth, though to a lesser extent than depletion of the oncogenic isoform. However, silencing of the oncogenic RAS isoform reduced only basal MAPK signaling, indicating that wild-type and oncogenic RAS isoforms have independent, nonoverlapping roles in MAPK regulation. Knockdown of oncogenic RAS also sensitized cells to EGF stimulation, implicating oncogenic RAS in the negative regulation of receptor tyrosine kinase sensitivity, but combining an EGFR inhibitor with oncogenic RAS depletion blocked EGF hypersensitivity and decreased cell proliferation. In addition to providing a rationale for combined inhibition of RAS and EGFR signaling in RAS-mutant tumors based on the activity of wild-type RAS isoforms, these findings provide a potential explanation for why many RAS-mutant tumors are resistant to EGFR inhibitors and for why the efficacy of inhibitors targeting oncogenic RAS activity is limited by feedback activation of EGFR.

Note: In This Issue is written by Cancer Discovery Science Writers. Readers are encouraged to consult the original articles for full details.