In This Issue

See article, p. 1100.

Cancer-associated fibroblasts (CAF) constitute a major component of the reactive microenvironment and enhance tumor growth and invasion. However, the mechanisms by which cancer cells mediate the reprogramming of quiescent stromal fibroblasts into CAFs are unknown. Mitra and colleagues assessed the contribution of microRNAs (miRNA) to the formation of CAFs from primary human normal fibroblasts derived from the omentum, a principal site of metastasis in serous ovarian carcinoma. Comparison of miRNA expression in normal fibroblasts and fibroblasts cocultured with ovarian cancer cells showed differential expression of a subset of miRNAs. Specifically, upregulation of miR-155 and downregulation of miR-214 and miR-31 were detected in cancer cell–induced and patient-derived CAFs, suggesting that altered regulation of miRNAs promotes CAF formation. In support of this idea, mimicking the cancer cell–driven deregulation of these 3 miRNAs functionally converted normal fibroblasts into tumor-promoting CAFs, whereas reciprocal modulation of these miRNAs reverted CAFs to a normal fibroblast phenotype. In addition, miRNA-driven reprogramming of normal fibroblasts induced gene expression changes similar to those observed in patient-derived CAFs, including elevated secretion of the chemokine (C-C motif) ligand 5 (CCL5), which was directly targeted by miR-214. Moreover, the ability of CAFs to augment tumor growth and invasion was dependent on CCL5-mediated metastasis of tumor cells via binding to its receptor on ovarian cancer cells. These results suggest that targeting miRNAs within the tumor microenvironment may suppress the tumor-promoting functions of CAFs.

See article, p. 1109.

Therapeutic options for liposarcomas are limited, as these tumors are often unresectable or do not respond well to chemotherapeutic agents, emphasizing the need to better predict which patients will benefit from treatment. To investigate whether altered nutrient uptake contributes to treatment outcome, Braas and colleagues performed mass-spectrometry–based metabolomics analysis of 3 patient-derived liposarcoma cell lines. These cells exhibited consumption of cytidine, thymidine, and uridine, suggesting that the nucleoside salvage pathway is active in liposarcomas. Nucleoside uptake was associated with elevated activity of deoxycytidine kinase (dCK), a required enzyme for nucleoside salvage, in cell lines and in approximately 10% of patients with liposarcoma and could be detected in vivo in xenograft tumors via positron emission tomography (PET) imaging using the cytidine-derived tracer 1-(2′-deoxy-2′-[18F]fluoroarabinofuranosyl) cytosine (FAC). Intriguingly, nucleoside salvage activity sensitized liposarcoma cell lines to the cytotoxic effects of gemcitabine, a nucleoside analogue prodrug that is activated by dCK. Furthermore, gemcitabine treatment led to complete regression of liposarcoma xenograft tumors, and this enhanced response to gemcitabine was dependent on dCK expression. These results suggest that detection of this metabolic activity with FAC.PET may identify patients with liposarcoma who will respond to treatment with this chemotherapeutic agent.

See article, p. 1118.

Fibroblast growth factor receptor (FGFR) signaling is frequently hyperactive in human cancers and is an attractive candidate for targeted therapy. Several FGFR inhibitors are currently in clinical development, including NVP-BGJ398, a small-molecule pan-FGFR kinase inhibitor. Because the identification of appropriate patient selection biomarkers is critical to the clinical success of targeted agents, Guagnano and colleagues sought to identify predictors of NVP-BGJ398 sensitivity. To determine which features confer dependence on FGFR signaling and sensitivity to FGFR inhibition, the effect of NVP-BGJ398 on the proliferation of 541 genetically and molecularly annotated cell lines from the Cancer Cell Line Encyclopedia was assayed. A predictive model of NVP-BGJ398 sensitivity identified amplifications, activating mutations, or translocation of FGFR genes or expression of FGF signaling gene signatures as the top predictors of NVP-BGJ398 response. The 32 cell lines sensitive to NVP-BGJ398 were highly enriched for those with a genetic lesion affecting an FGFR compared with insensitive cell lines, and sensitivity to NVP-BGJ398 was confirmed in vivo in xenografts derived from patients with FGFR-amplified tumors. Of note, among the NVP-BGJ398–sensitive cell lines that did not harbor FGFR mutations, several had copy number gains in a gene encoding an FGF ligand. Together, these findings indicate that NVP-BGJ398 may be most effective in a patient population selected for FGF/FGFR genetic alterations and provide an example of how cell line compendia can facilitate the preclinical identification of cellular features associated with drug response.

See article, p. 1134.

PARP inhibition sensitizes cancer cells to chemotherapy and irradiation by suppressing PARP-1-mediated DNA repair. Because prostate cancer is highly chemoresistant and strategies to improve responses to standard treatments like local irradiation and hormone-deprivation therapy are urgently needed, Schiewer and colleagues evaluated the activity of PARP inhibitors on prostate cancer cells. PARP inhibition sensitized both androgen-dependent and castration-resistant prostate cancer (CRPC) cells to ionizing radiation and docetaxel treatment, but single-agent PARP inhibitor treatment also unexpectedly suppressed the proliferation of androgen receptor (AR)-positive cells, suggesting that PARP-1 may affect AR activity in a DNA repair-independent manner. Indeed, PARP-1 bound AR target loci and promoted an open chromatin structure that facilitated AR recruitment and transcriptional activation of AR target genes. Notably, PARP-1 activity was elevated in CRPC and promoted AR binding to target loci in the absence of androgen, indicating that PARP-1 plays a key role in the aberrant ligand-independent AR activity associated with CRPC. PARP inhibition decreased AR target gene expression and blocked the growth and proliferation of primary prostate cancer cells in an ex vivo explant system, and the combination of PARP inhibition and chemical castration significantly slowed xenograft tumor growth. PARP inhibition may therefore be effective as either single-agent therapy or in combination with existing therapies, including androgen deprivation, to treat both early- and late-stage prostate cancers.

See article, p. 1150.

Disseminated breast cancer cells often remain undetected in a dormant state until induced to progress to metastatic disease. Kuznetsov and colleagues investigated the signals that mediate outgrowth of these cells using a human tumor xenograft model in which an instigating luminal breast cancer (LBC) cell line supported tumor formation from otherwise indolent responding human breast cancer cells at a distant site. Mice bearing LBC tumors exhibited an increase in VEGF receptor 2 (VEGFR2)-positive bone marrow cells (BMC), which accumulated in responding tumors and were incorporated into the tumor vasculature. The presence of these VEGFR2-expressing BMCs was required to promote enriched surface expression of CD24 on responding tumor cells, which in turn enhanced the recruitment of circulating platelets to CD24-rich areas within these tumors. Platelets from LBC tumor-bearing mice were sufficient to stimulate angiogenesis in vitro and efficiently absorbed proangiogenic cytokines secreted by LBC tumors in coculture experiments, suggesting that the LBC tumor stimulus and subsequent cytokine release by platelets in the responding tumor are necessary to trigger angiogenesis. Consistent with this idea, inhibition of platelet activity via aspirin treatment significantly impaired responding tumor formation. Importantly, analogous instigating and responding phenotypes were also observed in a subset of primary human luminal breast cancer samples. These findings delineate a systemic cascade of events driven by luminal breast cancers that facilitates indolent tumor progression and that may help to identify patients who will benefit from preventive treatments.

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