Autophagy Inhibition Targets BRAF-Mutant Pediatric CNS Tumors
See article, p. 773.
Genetic or pharmacologic inhibition of autophagy is cytotoxic to BRAFV600E-positive CNS tumor cells.
Chloroquine synergizes with vemurafenib or standard chemotherapy in BRAF-mutant CNS tumor cells.
A child with relapsed BRAF-mutant ganglioglioma responded to chloroquine plus vemurafenib therapy.
Autophagy has been implicated in tumorigenesis given its ability to promote cancer cell survival during stress, and the activating BRAFV600E mutation is known to enhance basal autophagic activity in diverse tumor types. The BRAFV600E mutation has recently been identified in pediatric central nervous system (CNS) tumors, but whether autophagy contributes to BRAFV600E-induced tumorigenesis in these tumor types is not fully understood. Mulcahy Levy and colleagues found that BRAFV600E-positive CNS tumor cell lines exhibited greater starvation-induced autophagy than did BRAF–wild-type CNS tumor cells, raising the possibility that the increased autophagy associated with BRAFV600E might contribute a selective advantage to CNS tumor cells. Indeed, genetic or pharmacologic inhibition of autophagy was cytotoxic to BRAF-mutant CNS tumor cells, but had a minimal effect on survival of BRAF–wild-type cells, indicating that BRAF-mutant CNS tumor cells are dependent on autophagy for survival. Moreover, treatment with the autophagy inhibitor chloroquine improved the effectiveness of both standard chemotherapeutics and the BRAF inhibitor vemurafenib, and showed synergistic activity with vemurafenib in BRAF-mutant CNS tumor cells at clinically achievable doses. Of note, combined chloroquine and vemurafenib treatment overcame vemurafenib resistance in primary BRAFV600E-positive pleomorphic xanthoastrocytoma cells and led to rapid clinical improvement and stabilization of disease in a patient with vemurafenib-refractory BRAFV600E-positive brainstem ganglioglioma that reversed whenever vemurafenib was discontinued. These findings indicating that BRAF-mutant pediatric CNS tumors are autophagy-dependent provide a rationale for combining autophagy inhibitors with BRAF-targeted therapy in patients with relapsed or refractory disease.
Dissemination Is a Barrier to Lung Cancer Metastasis
See article, p. 781.
Kras-mutant, p53-deficient tumor cells are not inherently capable of dissemination in mice.
p53 loss is insufficient to drive dissemination but enables acquisition of required alterations.
Nkx2-1 downregulation and enhanced proliferation in primary tumors precede dissemination.
Dissemination of cancer cells from primary tumors has been shown to occur early during tumor growth in both breast and pancreatic cancer, suggesting that these cancer cells already possess the genetic changes required to initiate the metastatic cascade. However, it is not known whether early premalignant lung lesions are also inherently capable of generating disseminated tumor cells (DTC). To analyze lung cancer cell dissemination, Caswell and colleagues utilized a genetically engineered mouse model of lung cancer driven by Kras mutation in which tumor cells were fluorescently labeled. Intriguingly, few DTCs were detected in Kras-mutant mice with either early-stage hyperplastic lesions or adenocarcinomas. Concomitant loss of p53 was associated with the presence of DTCs in a fraction of mice with late-stage lung tumors, suggesting that p53 inactivation is not sufficient to facilitate dissemination but enables the acquisition of additional genetic changes necessary to trigger dissemination. These DTCs originated from a single primary tumor, indicative of cell-autonomous alterations that confer the ability to disseminate. In addition, DTCs exhibited reduced expression of the prodifferentiation transcription factor Nkx2-1, which was also downregulated in a portion of the parental tumor cells coincident with increased proliferative potential, supporting the notion that loss of Nkx2-1 in a subpopulation of primary tumor cells promotes the generation of DTCs. These results suggest that dissemination is an acquired phenotype in lung adenocarcinoma and a rate-limiting barrier to lung cancer metastasis.
SPSB1-Mediated c-MET Activation Drives Breast Cancer Recurrence
See article, p. 790.
SPSB1 upregulation is necessary and sufficient for breast cancer recurrence in mouse models.
SPSB1 protects breast cancer cells from apoptosis following HER2 inhibition or chemotherapy.
Potentiation of c-MET signaling is required for SPSB1-driven tumor cell survival and recurrence.
Dormant residual breast cancer cells often persist for long periods following treatment and can give rise to incurable recurrent tumors, emphasizing the need to understand the molecular mechanisms that regulate the outgrowth of these cells. Using genetically engineered mouse models, Feng and colleagues found that SplA/ryanodine receptor domain and SOCS box containing 1 (SPSB1) is upregulated in recurrent mammary tumors and was both necessary and sufficient for tumor recurrence following suppression of the driving HER2/neu oncogene. SPSB1 expression protected both murine and human mammary tumor cells from apoptosis in response to HER2/neu inhibition or treatment with chemotherapeutic agents and was selected for during tumor outgrowth, indicating that SPSB1 confers a growth advantage in residual breast cancer cells. The prosurvival function of SPSB1 was dependent on binding of SPSB1 to c-MET and potentiation of c-MET activity in the absence of HER2/neu expression, as inhibition of c-MET diminished breast cancer cell viability and prevented selection of SPSB1-expressing cells in tumor-bearing mice. Furthermore, elevated SPSB1 expression was associated with basal-like breast cancer and was independently correlated with increased risk of relapse only in patients with increased c-MET expression and activity, suggesting that SPSB1 may contribute to the aggressive phenotype and therapeutic resistance in these tumors via activation of c-MET signaling. These findings define a role for SPSB1-driven c-MET activity in breast cancer recurrence and suggest that targeting SPSB1 may limit tumor relapse.
RINT1 Is a Breast Cancer Predisposition Gene
See article, p. 804.
RAD50-interacting protein 1 (RINT1) mutations were identified in women from multiple-case breast cancer families.
Rare RINT1 variants were enriched in early-onset breast cancer cases compared with unaffected female controls.
Carriers of RINT1 mutations also display a higher incidence of Lynch Syndrome–spectrum cancers.
Women with a family history of breast cancer have a 2- to 3-fold higher risk of developing the disease. Currently, only approximately 50% of familial breast cancers can be attributed to mutations in known cancer susceptibility genes, such as BRCA1 or BRCA2, which suggests that additional genetic mutations may confer hereditary breast cancer risk. Park and colleagues used whole-exome sequencing to screen for previously unrecognized cancer susceptibility genes in a cohort of 89 women with early-onset breast cancer from highly selected families with multiple cases of breast cancer. This approach identified three separate, family-specific mutations in RAD50-interacting protein 1 (RINT1) that were not observed in public databases. In line with these findings, case–control mutation screening showed an enrichment of RINT1 variants that were predicted to be deleterious in women with early-onset breast cancer compared with age-matched controls, and an additional 4 RINT1 mutations were identified in an independent large cohort of multicase breast cancer families. Variants in RINT1 that were likely to be pathogenic included missense mutations, in-frame deletions, and mutations predicted to affect RINT1 splicing. Of note, comparisons of cancer incidence across RINT1 mutation–positive families revealed a significantly higher risk for Lynch Syndrome–spectrum cancers that are associated with DNA mismatch repair defects. These findings implicating RINT1 as a breast cancer susceptibility gene suggest that RINT1 should be added to the list of genes evaluated in genetic testing for hereditary breast cancer.
Transcription States Are Linked to Intrinsic Drug Resistance in Melanoma
See article, p. 816.
Reciprocal MITF and NF-κB activity defines MAPK inhibitor–sensitive and resistant melanomas.
High NF-κB activity suppresses MITF and confers intrinsic resistance to MAPK inhibitors.
Transition to an MITF-low/NF-κB–high state may contribute to acquired MAPK inhibitor resistance.
Inhibitors targeting the MAPK pathway are clinically effective in the majority of patients with BRAF-mutant melanoma; however, a subset of patients fails to respond due to intrinsic resistance mechanisms that remain poorly understood. Konieczkowski and colleagues found that MAPK inhibitor–sensitive and intrinsically resistant melanoma cell lines and tumors were defined by distinct and reciprocal gene expression profiles, suggesting that cell-autonomous differences contribute to RAF and MEK inhibitor drug resistance. Specifically, intrinsically resistant melanomas were characterized by low expression and activity of microphthalmia-associated transcription factor (MITF) and elevated NF-κB pathway signaling (MITF-low/NF-κB–high), whereas MAPK inhibitor–sensitive melanomas were classified as MITF-high/NF-κB–low. Establishment of these two transcriptional states in melanocytes was regulated by the balance between oncogenic MAPK signaling, which activated NF-κB, and sustained MITF expression. Stimulation of NF-κB activity in MAPK inhibitor–sensitive cells induced a transition to the MITF-low/NF-κB–high phenotype by suppressing MITF and was sufficient to confer resistance to MAPK pathway inhibitors, indicative of plasticity between these states. Furthermore, inhibition of BRAF in MITF-high, drug-sensitive cells was associated with a transition to the MITF-low/NF-κB–high state, suggesting that this phenotype correlates with reduced dependence on MAPK signaling and may also contribute to acquired MAPK inhibitor resistance. These results identify transcriptional states that may underlie intrinsic resistance to MAPK inhibition and may help to predict therapeutic responses among patients with BRAF-mutant melanoma.
Zaprinast Blocks 2-Hydroxyglutarate Production by Inhibiting Glutaminase
See article, p. 828.
A fluorimetric assay was used in a screen for drugs that reduce cellular levels of 2-hydroxyglutarate.
Zaprinast blocks 2-hydroxyglutarate production through noncompetitive inhibition of glutaminase.
IDH1-mutant and glutamine-addicted cancer cells are both sensitive to Zaprinast.
Several cancers such as gliomas and acute myeloid leukemias harbor recurrent neomorphic mutations in isocitrate dehydrogenase 1 or 2 (IDH1/2) that cause the enzyme to convert α-ketoglutarate (αKG) to 2-hydroxyglutarate (2HG), an oncometabolite that modulates αKG-dependent DNA and histone demethylases and promotes cellular transformation. To identify drugs that can reduce 2HG production, Elhammali and colleagues screened a library of compounds using a high-throughput fluorimetric assay that measured 2HG levels in IDH1-mutant cells. The most potent compound identified by the screen was Zaprinast, a known inhibitor of phosphodiesterase type 5 (PDE5). However, Zaprinast did not reduce 2HG levels through inhibition of PDE5 but instead acted through an off-target effect on glutaminase, an enzyme that operates upstream of mutant IDH. Zaprinast noncompetitively inhibited glutaminase, preventing the enzyme from metabolizing glutamine into glutamate, a precursor of the mutant IDH substrate αKG. Accordingly, Zaprinast reduced DNA and histone methylation and prevented soft-agar colony formation in IDH1-mutant cells, suggesting that this compound could reverse the mutant IDH phenotype. Additionally, Zaprinast increased levels of reactive oxygen species, enhanced susceptibility to oxidative damage, and reduced growth in pancreatic ductal adenocarcinoma cells that are dependent upon glutamine metabolism. Although clinically useful dosages of Zaprinast may not be achievable given its higher potency against PDE5 than glutaminase, these findings raise the possibility that Zaprinast or more glutaminase-selective Zaprinast derivatives may have activity in IDH-mutant or glutamine-addicted cancers.
pRB Regulates Genome Stability in a Dosage-Sensitive Manner
See article, p. 840.
A pRB–E2F1–condensin II complex regulates DNA replication at pericentromeric chromatin.
Loss of a single RB1 allele induces replication stress and defects in chromosome segregation.
RB1 haploinsufficiency results in aneuploidy that may contribute to tumor formation.
Disruption of the tumor suppressor gene RB1 conforms to the classical two-hit model, in which mutation followed by LOH promotes tumor formation. In addition to its role in cell-cycle progression, the RB protein (pRB) has also been suggested to participate in DNA replication, DNA repair, and chromosome condensation, but the mechanism by which pRB regulates genome stability remains unclear. Coschi and colleagues found that loss of pRB or expression of a mutant pRB enhanced γH2AX foci formation and triggered aberrant DNA replication, particularly at major satellite repeats within pericentromeric chromatin, indicative of replication stress. Replication of pericentromeres was regulated by formation of a complex between pRB, E2F1, and condensin II at major satellite repeats. Intriguingly, loss of a single Rb1 allele reduced condensin II recruitment to pericentromeres, induced γH2AX deposition to a level similar to that of Rb1-deficient cells, and resulted in mitotic errors and chromosome structure defects, suggesting that Rb1 is haploinsufficient for maintenance of genome stability. Consistent with this idea, γH2AX foci and mitotic defects were also enhanced in normal RB1+/− fibroblasts from patients with hereditary retinoblastoma, and RB1+/− cancer cell lines of mesenchymal origin harbored increased chromosomal abnormalities, similar to RB1−/− cells. Furthermore, tumors isolated from mice heterozygous for mutant Rb1 exhibited increased chromosomal gains and losses. These results identify a gene dosage–dependent function of pRB in suppressing genome instability and suggest that disruption of this function contributes to aneuploidy in cancer.
Note: In This Issue is written by Cancer Discovery Science Writers. Readers are encouraged to consult the original articles for full details.