See article by Martins et al., p. 503.

  • Loss of wild-type BRCA1 is often not the first event in BRCA1-mutant breast cancers.

  • Different initiating events are each associated with basal-like and luminal tumors.

  • BRCA1 haploinsufficiency in the mammary epithelium may increase breast cancer risk.

Because germline mutations of the BRCA1 tumor suppressor gene confer a high risk of breast and ovarian cancers, loss of the wild-type BRCA1 allele is thought to be the initiating event in the development of BRCA1-mutant cancers and followed by subsequent genetic events such as PTEN loss or TP53 mutation. Martins and colleagues scored individual cells from the malignant breast tumors of germline BRCA1 mutation carriers for BRCA1 LOH, PTEN loss, or TP53 mutation and determined the order of mutations based on the most frequent single-mutation state among the cells within a tumor. The authors found that BRCA1 LOH was the likely initiating event in only 10 of 55 tumors evaluated; PTEN loss was the most common first event, particularly in basal-like tumors, and TP53 mutation occurred first in the majority of luminal tumors. Furthermore, when BRCA1 LOH appeared to have occurred, a subpopulation of cells within the tumor did not lose the wild-type allele and retained functional BRCA1 protein. Normal breast tissue from patients with germline BRCA1 mutations was characterized by higher cell proliferation, occasional multipolar mitoses, and significantly more cells with extra centrosomes. Together, these findings support a model whereby BRCA1 haploinsufficiency leads to increased cellular proliferation and abnormal mitoses in the mammary epithelium, which facilitates the acquisition of additional mutations and contributes to increased breast cancer risk.

See article by Andraos et al., p. 512.

  • JAK activation loop phosphorylation is increased by ATP-competitive JAK inhibitors.

  • Stabilization of the inactive form of JAK2 inhibits phosphorylation of its activation loop.

  • The inactive JAK2 structure may enable the design of more potent type II inhibitors.

Upon ligand binding to cytokine receptors, Janus kinases (JAK) are activated via auto- or transphosphorylation of activation loop tyrosine residues and subsequently phosphorylate downstream targets such as STAT proteins. Constitutive activation of the JAK–STAT signaling pathway occurs in multiple cancer types and myeloproliferative diseases, and thus has spurred the development of ATP-competitive JAK inhibitors that bind the active form (the type I inhibitor binding mode) and suppress JAK kinase activity. However, Andraos and colleagues demonstrate that, despite inhibition of STAT phosphorylation, several type I JAK inhibitors can lead to an increase of activation loop phosphorylation of multiple JAK family members, thus priming the kinases for rapid reactivation when the inhibitor is no longer present. The authors hypothesized that type II inhibitors that stabilize the inactive kinase conformation would not increase JAK activation loop phosphorylation, but type II JAK inhibitors have not yet been identified. Database mining for compounds that did not potently inhibit active JAK2 in enzymatic assays but suppressed JAK2-dependent cell growth in vitro and possessed structural elements characteristic of type II inhibitors identified the dihydroindole NVP-BBT594. X-ray crystallographic analysis of NVP-BBT594 in complex with JAK2 confirmed that the inhibitor bound the inactive form of the kinase, and NVP-BBT594 not only suppressed STAT phosphorylation but also blocked JAK activation loop phosphorylation in multiple cell lines. Further development of type II JAK inhibitors may therefore lead to improved responses in diseases that are dependent on constitutive JAK activation.

See article by Aarts et al., p. 524.

  • WEE1 inhibitors synergize with gemcitabine to force cells in S-phase to enter mitosis.

  • Premature mitoses are abnormal and result in micronuclei formation and apoptosis.

  • TP53-mutant cells with high cyclin B1 and EZH2 levels are prone to forced mitotic entry.

Preclinical studies have suggested that inhibitors of WEE1, a cell cycle regulatory kinase that prevents mitotic entry, sensitize cancer cells to chemotherapy or radiation by allowing cells with damaged DNA to bypass the G2–M checkpoint and enter mitosis. Aarts and colleagues present evidence that WEE1 inhibition can also abrogate the intra–S-phase checkpoint and force cancer cells into mitosis before the completion of DNA replication. MK-1775, a small-molecule WEE1 inhibitor, synergized with gemcitabine, a chemotherapeutic agent that inhibits DNA synthesis and leads to S-phase arrest, to specifically inhibit the growth of a subset of TP53-mutant breast cancer cell lines. The addition of MK-1775 to S-phase–arrested, TP53-mutant cells forced a significant portion of the cells to enter mitosis without completing DNA synthesis, which led to abnormal mitoses characterized by premature chromosome condensation and disorganized mitotic spindles. Forced mitotic entry of gemcitabine-treated cells due to WEE1 inhibition ultimately resulted in a significant increase in micronuclei formation and induction of apoptosis in vitro, and combined MK-1775 and gemcitabine treatment more effectively inhibited tumor xenograft growth than either treatment alone. As the authors observed that TP53-mutant breast cancer cells expressing high levels of mitotic cyclins and the Polycomb protein EZH2 were particularly dependent on WEE1 to prevent mitotic entry in gemcitabine-arrested cells, combining WEE1 inhibitors with inhibitors of DNA synthesis may be particularly effective in the triple-negative and basal-like subtypes that most frequently display these molecular characteristics.

See article by Jahid et al., p. 540.

  • Overexpression of miR-23a is specifically observed in early-stage colorectal cancer.

  • Increased miR-23a expression correlates with downregulation of its predicted target genes.

  • miR-23a knockdown inhibits cell motility and invasion and suppresses metastasis.

Adenomas (polyps) and preinvasive adenocarcinomas (carcinomas in situ) of the colon are slow-growing neoplasms that take several years to develop into colorectal cancer. However, once formed, colorectal cancer spreads through the colon wall and metastasizes in a relatively short time. Using high-resolution tiling arrays to compare murine adenomas and invasive colorectal cancers, Jahid and colleagues identified a recurrent amplicon in invasive tumors that led to overexpression of 2 microRNAs (miRNA): miR-23a and miR-27a. Analysis of human preinvasive tumors, locally invasive colorectal cancers (stage I/II), and metastatic colorectal cancers (stage III/IV) revealed that miR-23a was specifically upregulated in early-stage colorectal cancers compared with preinvasive or metastatic tumors, whereas miR-27a was upregulated in both early- and late-stage colorectal cancers. miR-27a was required for colorectal cancer cell proliferation and tumor formation in vivo, but its sustained upregulation suggested a more general role in tumor progression. Consistent with its expression pattern indicating a role in the acquisition of an invasive phenotype, miR-23a knockdown suppressed cell motility and invasion and reduced the metastasis of tail vein–injected tumor cells. Mechanistically, miR-23a was required for filopodia formation and SRC activation, potentially through downregulation of its target metastasis suppressor 1 (MTSS1). Downregulation of other predicted miR-23a target genes was also specifically observed in genomewide expression profiles of early-stage colorectal cancer samples, further implicating miR-23a–dependent gene regulation in the transition from indolent to invasive colorectal cancer.

See article by Corno et al., p. 554.

  • NSCs derived from the mouse postnatal hindbrain are similar to medulloblastoma CSCs.

  • Ptch/p53-mutant CSC gene signatures reflect different human medulloblastoma subtypes.

  • Reactivation of the developmental regulator Ebf3 promotes medulloblastomagenesis.

Medulloblastoma is believed to arise from neural stem or progenitor cells, but it remains unclear which precursor populations are tumor-initiating cells. Additionally, the features that distinguish normal neural stem cells (NSC) from medulloblastoma cancer stem cells (CSC) are poorly understood. Corno and colleagues examined whether murine postnatal hindbrain cell populations represent NSCs that, upon oncogenic transformation, might give rise to medulloblastomas occurring in late adolescence or early adulthood. NSCs could be derived from the fourth ventricle (IVv) postnatally, and the gene expression profile of IVv NSCs was generally more similar to that of CSCs isolated from Ptch/p53-mutant medulloblastomas than cells derived from the subventricular zone, a known source of stem and progenitor cells in the adult forebrain. The genes shared by different CSC populations were specifically enriched in the desmoplastic and SHH human medulloblastoma subgroups, whereas the genes exclusive to Ptch+/−p53−/− CSCs were enriched in the WNT subgroup. This analysis also identified upregulation of early B-cell factor 3 (Ebf3), a regulator of proliferation and differentiation in the developing cerebellum, as a distinguishing feature of medulloblastoma CSCs. CSCs in which Ebf family members were knocked down gave rise to significantly smaller tumors with a less malignant phenotype, and enforced expression of Ebf3 in CSCs promoted tumorigenesis. Together, these findings establish the IVv as a potential source of postnatal NSCs and suggest that the molecular analysis of mouse NSC and CSC populations can provide insight into human medulloblastomagenesis.

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