Protecting plants from the inside out: Some soils show a remarkable ability to suppress disease caused by plant pathogens, an ability that is attributed to plant-associated microbiota. Carrión et al. investigated the role of endophytes, the intimate microbial community found within roots, in fungal disease suppression (see the Perspective by Tringe). The wilt fungus Rhizoctonia solani infects sugar beets, whereupon transcriptional analysis shows that several bacterial endophyte species activate biosynthetic gene clusters to cause disease suppression. These organisms produce antifungal effectors, including enzymes that can digest fungal cell walls, and secondary metabolites, including phenazines, polyketides, and siderophores, which may contribute to the antifungal phenotype. Science , this issue p. 606 ; see also p. 568

Significance: For the 2m DNA to fit into the tiny cell nucleus, it is wrapped around nucleosomes and folded into loops clustering together in domains. Genome function depends on this 3D-organization, especially on-going dynamic processes like transcription. Techniques studying the network of DNA contacts genome-wide have recently revealed this 3D architecture, but the protein factors behind this are not understood. We study two proteins that are known to help form DNA loops: cohesin and CTCC-binding factor (CTCF). Respective depletion and analysis of DNA contacts genome-wide show that CTCF is required to separate neighboring folding domains and keep cohesin in place, whereas cohesin is important for shaping the domains. Consistently, we observe different changes of gene expression.

Current clinical therapy of non-small cell lung cancer depends on histo-pathological classification. This approach poorly predicts clinical outcome for individual patients. Gene expression profiling holds promise to improve clinical stratification, thus paving the way for individualized therapy. A genome-wide gene expression analysis was performed on a cohort of 91 patients. We used 91 tumor- and 65 adjacent normal lung tissue samples. We defined sets of predictor genes (probe sets) with the expression profiles. The power of predictor genes was evaluated using an independent cohort of 96 non-small cell lung cancer- and 6 normal lung samples. We identified a tumor signature of 5 genes that aggregates the 156 tumor and normal samples into the expected groups. We also identified a histology signature of 75 genes, which classifies the samples in the major histological subtypes of non-small cell lung cancer. Correlation analysis identified 17 genes which showed the best association with post-surgery survival time. This signature was used for stratification of all patients in two risk groups. Kaplan-Meier survival curves show that the two groups display a significant difference in post-surgery survival time (p = 5.6E-6). The performance of the signatures was validated using a patient cohort of similar size (Duke University, n = 96). Compared to previously published prognostic signatures for NSCLC, the 17 gene signature performed well on these two cohorts. The gene signatures identified are promising tools for histo-pathological classification of non-small cell lung cancer, and may improve the prediction of clinical outcome.

Epithelial ovarian cancer is a highly heterogeneous disease and remains the most lethal gynaecological malignancy in the Western world. Therapeutic approaches need to account for inter-patient and intra-tumoural heterogeneity and detailed characterization of in vitro models representing the different histological and molecular ovarian cancer subtypes is critical to enable reliable preclinical testing. There are approximately 100 publicly available ovarian cancer cell lines but their cellular and molecular characteristics are largely undescribed. We have characterized 39 ovarian cancer cell lines under uniform conditions for growth characteristics, mRNA/microRNA expression, exon sequencing, drug response for clinically-relevant therapeutics and collated all available information on the original clinical features and site of origin. We tested for statistical associations between the cellular and molecular features of the lines and clinical features. Of the 39 ovarian cancer cell lines, 14 were assigned as high-grade serous, four serous-type, one low-grade serous and 20 non-serous type. Three morphological subtypes: Epithelial (n = 21), Round (n = 7) and Spindle (n = 12) were identified that showed distinct biological and molecular characteristics, including overexpression of cell movement and migration-associated genes in the Spindle subtype. Comparison with the original clinical data showed association of the spindle-like tumours with metastasis, advanced stage, suboptimal debulking and poor prognosis. In addition, the expression profiles of Spindle, Round and Epithelial morphologies clustered with the previously described C1-stromal, C5-mesenchymal and C4 ovarian subtype expression profiles respectively. Comprehensive profiling of 39 ovarian cancer cell lines under controlled, uniform conditions demonstrates clinically relevant cellular and genomic characteristics. This data provides a rational basis for selecting models to develop specific treatment approaches for histological and molecular subtypes of ovarian cancer.