Search

Heterogeneity of the tumor microenvironment (TME) is one of the major causes of treatment resistance in breast cancer. Among TME components, nervous system role in clinical outcome has been underestimated. Identifying neuronal signatures associated with treatment response will help to characterize neuronal influence on tumor progression and identify new treatment targets. The search for hormonotherapy-predictive biomarkers was implemented by supervised machine learning (ML) analysis on merged transcriptomics datasets from public databases. ML-derived genes were investigated by pathway enrichment analysis, and potential gene signatures were curated by removing the variables that were not strictly nervous system specific. The predictive and prognostic abilities of the generated signatures were examined by Cox models, in the initial cohort and seven external cohorts. Generated signature performances were compared with 14 other published signatures, in both the initial and external cohorts. Underlying biological mechanisms were explored using deconvolution tools (CIBERSORTx and xCell). Our pipeline generated two nervous system-related signatures of 24 genes and 97 genes (NervSign24 and NervSign97). These signatures were prognostic and hormonotherapy-predictive, but not chemotherapy-predictive. When comparing their predictive performance with 14 published risk signatures in six hormonotherapy-treated cohorts, NervSign97 and NervSign24 were the two best performers. Pathway enrichment score and deconvolution analysis identified brain neural progenitor presence and perineural invasion as nervous system-related mechanisms positively associated with NervSign97 and poor clinical prognosis in hormonotherapy-treated patients. Transcriptomic profiling has identified two nervous system–related signatures that were validated in clinical samples as hormonotherapy-predictive signatures, meriting further exploration of neuronal component involvement in tumor progression. Significance: The development of personalized and precision medicine is the future of cancer therapy. With only two gene expression signatures approved by FDA for breast cancer, we are in need of new ones that can reliably stratify patients for optimal treatment. This study provides two hormonotherapy-predictive and prognostic signatures that are related to nervous system in TME. It highlights tumor neuronal components as potential new targets for breast cancer therapy.

Heterogeneity of the tumor microenvironment (TME) is one of the major causes of treatment resistance in breast cancer. Among TME components, nervous system role in clinical outcome has been underestimated. Identifying neuronal signatures associated with treatment response will help to characterize neuronal influence on tumor progression and identify new treatment targets. The search for hormonotherapy-predictive biomarkers was implemented by supervised machine learning (ML) analysis on merged transcriptomics datasets from public databases. ML-derived genes were investigated by pathway enrichment analysis, and potential gene signatures were curated by removing the variables that were not strictly nervous system specific. The predictive and prognostic abilities of the generated signatures were examined by Cox models, in the initial cohort and seven external cohorts. Generated signature performances were compared with 14 other published signatures, in both the initial and external cohorts. Underlying biological mechanisms were explored using deconvolution tools (CIBERSORTx and xCell). Our pipeline generated two nervous system-related signatures of 24 genes and 97 genes (NervSign24 and NervSign97). These signatures were prognostic and hormonotherapy-predictive, but not chemotherapy-predictive. When comparing their predictive performance with 14 published risk signatures in six hormonotherapy-treated cohorts, NervSign97 and NervSign24 were the two best performers. Pathway enrichment score and deconvolution analysis identified brain neural progenitor presence and perineural invasion as nervous system-related mechanisms positively associated with NervSign97 and poor clinical prognosis in hormonotherapy-treated patients. Transcriptomic profiling has identified two nervous system–related signatures that were validated in clinical samples as hormonotherapy-predictive signatures, meriting further exploration of neuronal component involvement in tumor progression.Significance:The development of personalized and precision medicine is the future of cancer therapy. With only two gene expression signatures approved by FDA for breast cancer, we are in need of new ones that can reliably stratify patients for optimal treatment. This study provides two hormonotherapy-predictive and prognostic signatures that are related to nervous system in TME. It highlights tumor neuronal components as potential new targets for breast cancer therapy.

Supplementary Table S1: R package referencesSupplementary Table S2: Merged Affymetrix cohort patient characteristicsSupplementary Table S3: Validation cohort patient characteristicsSupplementary Table S4: Algorithm parameter optimizationSupplementary Figure S1: Patient selection and algorithm pipeline configurationSupplementary Figure S2: Generation of NervSign24Supplementary Figure S3: ML-generated gene lists for the first signature and second signature generationsSupplementary Figure S4: Supplementary information related to Figure 2Supplementary Figure S5: Supplementary information related to Figure 3Supplementary Figure S6: Univariate Cox analyses in cohorts of patients with early or late recurrenceSupplementary Figure S7: Supplementary information related to Figure 4Supplementary Figure S8: Multivariable and univariate Cox analyses in merged Affymetrix, METABRIC and Saal cohortsSupplementary Figure S9: Correlation between nervous system mechanisms and NervSign97Supplementary Figure S10: Kaplan-Meier analysis of previously published biomarkers

Supplementary Figures 1-7 and Supplementary Table 1

PDF file - 43K, HDIs do not induce endogenous ABCG2 expression in Flip-In-293 cells

PDF file - 52K, Ubiquitination state of ABCG2

PDF file - 65K

PDF file - 47K, Effect of inhibition of autophagy on ABCG2 expression

PDF file - 26K, Chaperone expression after HDI treatment

PDF file - 82K

PDF file - 29K, Impact of CHX and actinomycin D on HDI-induced ABCG2 surface overexpression

PDF file - 26K, CHX does not affect histone and �-tubulin acetylation induced by HDIs

We are taking advantage of the launch of the latest version (v4.6) of our web-based data mining tool "breast cancer gene-expression miner" (bc-GenExMiner) to take stock of its position within the oncology research landscape and to present an activity report ten years after its establishment (http://bcgenex.ico.unicancer.fr). bc-GenExMiner is an open-access, user-friendly tool for statistical mining on breast tumor transcriptomes, annotated with more than 20 clinicopathologic and molecular characteristics. The database comprises more than 16,000 patients from 64 cohorts - including TCGA, METABRIC and SCAN-B - for whom several thousands of genes have been quantified by microarrays or RNA-seq. Correlation, expression and prognostic analyses are available for targeted, exhaustive or customized explorations of queried genes. bc-GenExMiner facilitates the validation, investigation, and prioritization of discoveries and hypotheses on genes of interest. It allows users to analyse large databases, create data visualizations, and obtain robust statistical analysis, thereby accelerating biomarker discovery. Ten years after its launch, judging by the number of visits, analyses, and scientific citations of bc-GenExMiner, we conclude that this web resource serves its purpose in the international scientific community working in breast cancer research, with a never-ending rise in its use., Nous profitons de l’occasion de la mise en ligne de la dernière version (v.4.6) de l’outil web breast cancer gene-expression miner (bc-GenExMiner) pour rappeler sa place dans le paysage de la recherche en oncologie et réaliser un bilan d’activité, dix ans après sa création (http://bcgenex.ico.unicancer.fr). bc-GenExMiner est un outil web de fouille statistique de données d’expressions géniques issues du criblage du transcriptome de tumeurs du sein annotées par une vingtaine de critères clinicopathologiques et moléculaires. Sa base de données inclut plus de 16 000 patientes issues de 64 cohortes, dont les cohortes METABRIC, SCAN-B et TCGA, pour lesquelles l’expression de plusieurs milliers de gènes a été mesurée par des puces à ADN ou par RNAseq. Il est composé de trois modules : « Corrélation », « Expression » et « Pronostic », qui permettent différentes sortes d’analyses. bc-GenExMiner offre aux chercheurs des possibilités de validation, d’exploration, de hiérarchisation d’hypothèses et de découverte concernant leurs gènes d’intérêt. Il permet de s’autonomiser par rapport à l’analyse de grandes bases de données, la production de figures, d’obtenir des résultats robustes et ainsi de gagner du temps pour la découverte de biomarqueurs. Dix ans après sa mise en ligne, à en juger par le nombre de visites, d’analyses et de citations de bc-GenExMiner dans des articles scientifiques, nous pouvons conclure que cet outil web sert la communauté scientifique internationale engagée dans la recherche sur le cancer du sein, et qu’il est de plus en plus utilisé.

RAS-mutant cancers are among the most refractory to treatment. Apart from new G12C genotype targeted therapies, strategies to kill RAS-mutant cells by directly targeting RAS or its downstream effectors have been mostly unsuccessful, mainly due to pathway redundancy and heterogeneities in RAS-induced phenotypes. Here we identified a RAS-phenotype that can be targeted by the histone deacetylase inhibitor (HDACi) romidepsin. We showed that the hyperacetylation induced by romidepsin depleted acetyl-CoA, the cell donor substrate for acetylation, and led to metabolic stress and death in KRAS-activated cells. Elastic net analysis on transcriptomics from a 608-cell panel confirmed that HDACi sensitivity was linked to a difference in profiles in two pathways involved in acetyl-CoA metabolism. The analysis of a clinical dataset confirmed that perturbation of the two acetyl-CoA pathways were correlated with HDACi sensitivity in patients treated with belinostat. Our analysis suggests the potential utility of a RAS-associated acetyl-CoA phenotype to sharpen treatment choices for RAS-activated tumors.

Histone deacetylase inhibitors (HDACi) induce hyperacetylation of histones by blocking HDAC catalytic sites. Despite regulatory approvals in hematological malignancies, limited solid tumor clinical activity has constrained their potential, arguing for better understanding of mechanisms of action (MOA). Multiple activities of HDACis have been demonstrated, dependent on cell context, beyond the canonical induction of gene expression. Here, using a clinically relevant exposure duration, we established DNA damage as the dominant signature using the NCI-60 cell line database and then focused on the mechanism by which hyperacetylation induces DNA damage. We identified accumulation of DNA–RNA hybrids (R-loops) following romidepsin-induced histone hyperacetylation, with single-stranded DNA (ssDNA) breaks detected by single-cell electrophoresis. Our data suggest that transcription-coupled base excision repair (BER) is involved in resolving ssDNA breaks that, when overwhelmed, evolve to lethal dsDNA breaks. We show that inhibition of BER proteins such as PARP will increase dsDNA breaks in this context. These studies establish accumulation of R-loops as a consequence of romidepsin-mediated histone hyperacetylation. We believe that the insights provided will inform design of more effective combination therapy with HDACis for treatment of solid tumors. Implications: Key HDAC inhibitor mechanisms of action remain unknown; we identify accumulation of DNA–RNA hybrids (R-loops) due to chromatin hyperacetylation that provokes single-stranded DNA damage as a first step toward cell death., Part of this work was supported by the Center for Cancer Research, Intramural Research Program of the National Cancer Institute; (1ZIABC010621–12; to S.E. Bates) and (BC-006161 and BC-006150; to W.C. Reinhold and Y. Pommier), and the European Research Council (ERC2014 AdG669898 TARLOOP; to A. Aguilera).

Background: The mechanism of action of romidepsin and other histone deacetylase inhibitors is still not fully explained. Our goal was to gain a mechanistic understanding of the RAS-linked phenotype associated with romidepsin sensitivity. Methods: The NCI60 dataset was screened for molecular clues to romidepsin sensitivity. Histone acetylation, DNA damage, ROS production, metabolic state (real-time measurement and metabolomics), and gene expression alterations (transcriptomics) were determined in KRAS-WT versus KRAS-mutant cell groups. The search for biomarkers in response to HDACi was implemented by supervised machine learning analysis on a 608-cell transcriptomic dataset and validated in a clinical dataset. Results: Romidepsin treatment induced depletion in acetyl-CoA in all tested cell lines, which led to oxidative stress, metabolic stress, and increased death—particularly in KRAS-mutant cell lines. Romidepsin-induced stresses and death were rescued by acetyl-CoA replenishment. Two acetyl-CoA gene expression signatures associated with HDACi sensitivity were derived from machine learning analysis in the CCLE (Cancer Cell Line Encyclopedia) cell panel. Signatures were then validated in the training cohort for seven HDACi, and in an independent 13-patient cohort treated with belinostat. Conclusions: Our study reveals the importance of acetyl-CoA metabolism in HDAC sensitivity, and it highlights acetyl-CoA generation pathways as potential targets to combine with HDACi.

// Julian C. Bahr 1, * , Robert W. Robey 1, 2, * , Victoria Luchenko 1 , Agnes Basseville 1 , Arup R. Chakraborty 1 , Hanna Kozlowski 2 , Gary T. Pauly 3 , Paresma Patel 3 , Joel P. Schneider 3 , Michael M. Gottesman 2 , Susan E. Bates 1, 4 1 Developmental Therapeutics Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA 2 Laboratory of Cell Biology, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA 3 Chemical Biology Laboratory, National Cancer Institute, Frederick Cancer Research Center, Frederick, MD 21702, USA 4 Columbia University Medical Center, Division of Hematology/Oncology, New York, NY 10032, USA * These authors contributed equally to this work Correspondence to: Susan E. Bates, email: seb2227@cumc.columbia.edu Keywords: romidepsin, Ras mutation, apoptosis, MEK inhibitor, AKT inhibitor Received: May 27, 2016 Accepted: September 01, 2016 Published: September 13, 2016 ABSTRACT We previously demonstrated activation of the mitogen-activated protein kinase (MAPK) pathway in a series of romidepsin-selected T-cell lymphoma cell lines as a mechanism of resistance to the histone deacetylase inhibitor (HDI), romidepsin. As Ras mutation leads to activation of both the MAPK and the phosphoinositide 3-kinase (PI3K) pathway, we examined whether combining romidepsin with small molecule pathway inhibitors would lead to increased apoptosis in cancers harboring Ras mutations. We treated 18 Ras mutant or wild-type cell lines with romidepsin in the presence of a MEK inhibitor (PD-0325901) and/or an AKT inhibitor (MK-2206) and examined apoptosis by flow cytometry. A short-term treatment schedule of romidepsin (25 ng/ml for 6 h) was used to more closely model clinical administration. Romidepsin in combination with a MEK and an AKT inhibitor induced apoptosis preferentially in cells harboring mutant versus wild-type Ras (69.1% vs. 21.1%, p < 0.0001). Similar results were found in a subset of cell lines when belinostat was combined with the MEK and AKT inhibitors and when romidepsin was combined with the dual extracellular signaling-related kinase (ERK)/PI3K inhibitor, D-87503, which inhibited both the MAPK and PI3K pathways at 5–10 μM. The observed apoptosis was caspase-dependent and required Bak and Bax expression. Cells with wild-type or mutant Ras treated with romidepsin alone or in combination with the MEK inhibitor displayed increased expression of proapoptotic Bim. We thus conclude that cancers bearing Ras mutations, such as pancreatic cancer, can be targeted by the combination of an HDI and a dual inhibitor of the MAPK and PI3K pathways.

Because 5-year survival rates for patients with metastatic melanoma remain below 25%, there is continued need for new therapeutic approaches. For some tumors, pharmacologic ascorbate treatment may have a beneficial antitumor effect and may work synergistically with standard chemotherapeutics. To investigate this possibility in melanoma, we examined the effect of pharmacologic ascorbate on B16-F10 cells. Murine models were employed to compare tumor size following treatment with ascorbate, and the chemotherapeutic agents dacarbazine or valproic acid, alone or in combination with ascorbate. Results indicated that nearly all melanoma cell lines were susceptible to ascorbate-mediated cytotoxicity. Compared to saline controls, pharmacologic ascorbate decreased tumor size in both C57BL/6 (P < 0.0001) and NOD-scid tumor bearing mice (P < 0.0001). Pharmacologic ascorbate was superior or equivalent to dacarbazine as an antitumor agent. Synergy was not apparent when ascorbate was combined with either dacarbazine or valproic acid; the latter combination may have additional toxicities. Pharmacologic ascorbate induced DNA damage in melanoma cells, as evidenced by increased phosphorylation of the histone variant, H2A.X. Differences were not evident in tumor samples from C57BL/6 mice treated with pharmacologic ascorbate compared to tumors from saline-treated controls. Together, these results suggest that pharmacologic ascorbate has a cytotoxic effect against melanoma that is largely independent of lymphocytic immune functions and that continued investigation of pharmacologic ascorbate in cancer treatment is warranted.

Pancreatic cancer remains a difficult-to-treat malignancy, yet nab-paclitaxel plus gemcitabine prolongs survival. Closer examination of the mechanism of action of nab-paclitaxel hints at a role for targeting KRAS. We discuss how nab-paclitaxel may be active in pancreatic cancer and how this informs the way forward to better treat patients with pancreatic cancer.

ABCG2 is an ATP-binding cassette (ABC) half-transporter localized to the cell membrane and important in normal physiology, in normal tissue protection including in the maternal-fetal barrier and the blood brain barrier, and potentially in multidrug resistance. Dimerization is required for function. Polymorphic variants have been described that impair trafficking to the cell surface, and one, Q141K, has been associated with a higher incidence of gout in individuals who are carriers. Working in concert with ABCB1 (P-glycoprotein), ABCG2 is expressed at high levels in the vascular endothelium of the brain and pumps xenobiotics and chemotherapeutics back into the bloodstream. ABCG2 expression is also found at high levels in some cancer types including pancreatic and liver cancers. Whether its expression in these tumor types contributes to drug resistance via drug efflux and reduction in drug accumulation is a question that remains unanswered. Studies of drug accumulation, such as through imaging solid tumors in patients, are needed to answer this question.

Ras pathway mutations are very common in cancer, with the highest prevalence for RAS itself, found in about 30% of tumors. Despite multiple efforts, targeting the RAS pathway has not yet been amenable to therapeutic intervention. Taking a different approach, our laboratory has focused on combination treatment with romidepsin, an HDAC inhibitor approved by the FDA to treat T-cell lymphoma. Our group indeed demonstrated perturbation of RAS downstream pathways after romidepsin treatment, by analyzing gene expression in samples obtained from patients treated with romidepsin for T-cell lymphoma (Chakraborty et al, Blood 2013, 121: 4115). The NCI-60 panel was screened with romidepsin plus or minus RAS pathway downstream inhibitors (MEK and/or Akt inhibitors). We found that mutations in RTK/RAS/BRAF correlated with romidepsin sensitivity (R = 0.472), with cell death ranging from 40% to 95% after 48h (assayed by annexin V labeling). Three KRAS mutant and three KRAS WT cell lines were then chosen to investigate the mechanism of this selectivity. Romidepsin induced histone acetylation and DNA damage from 30min treatment, and an increase in reactive oxygen species (ROS) after 10h exposure, independent of KRAS status. Only KRAS mutant cells showed mitochondrial membrane depolarization, cytochrome c release, and apoptosis, all starting about 24h after the beginning of treatment. Interestingly, cytoplasmic - but not mitochondrial - ROS scavengers rescued KRAS mutant cells from death, suggesting that ROS is required but not sufficient for romidepsin-induced apoptosis. Since Son et al (Nature 2013, 496: 101) reported that glutamine metabolism was reprogrammed by oncogenic KRAS to maintain redox balance, we explored this pathway using the Seahorse XF96 analyzer. We observed that, while all cell lines utilized glutamine for oxidative phosphorylation (OXPHOS) before drug exposure, romidepsin impaired the ability of only KRAS mutant cells to use glutamine for OXPHOS. Metabolomic analysis confirmed reduced glutamine use in KRAS mutant cells after drug treatment, as well as an inability to replenish their reduced glutathione stock. Moreover, this analysis also highlighted a romidepsin-induced drop in nucleotide synthesis in KRAS mutant cells, potentially leading to impaired DNA repair. Accordingly, adding back nucleosides to media blunted the effect of romidepsin on cell growth in KRAS mutant, but not in KRAS WT cells. From these results, we conclude that romidepsin causes cell damage through DNA damage and subsequent ROS release. And while KRAS WT cells are able to scavenge ROS and repair DNA breaks, KRAS mutant cells cannot. We hypothesize that the reprogramming of glutamine metabolism, which supports redox maintenance and may support nucleoside synthesis, makes KRAS mutant cells particularly sensitive to romidepsin treatment. Citation Format: Agnes Basseville, Carole Sourbier, Robert W. Robey, Dan L. Sackett, W. Marston Linehan, Susan E. Bates. Metabolic reprogramming in KRAS mutant cancer cells may cause sensitivity to the histone deacetylase (HDAC) inhibitor romidepsin. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1771. doi:10.1158/1538-7445.AM2015-1771