On entropy and information in gene interaction networks.

MOTIVATION: Modern biological experiments often produce candidate lists of genes presumably related to the studied phenotype. One can ask if the gene list as a whole makes sense in the context of existing knowledge: Are the genes in the list reasonably related to each other or do they look like a random assembly? There are also situations when one wants to know if two or more gene sets are closely related. Gene enrichment tests based on counting the number of genes two sets have in common are adequate if we presume that two genes are related only when they are in fact identical. If by related we mean well connected in the interaction network space, we need a new measure of relatedness for gene sets. RESULTS: We derive entropy, interaction information and mutual information for gene sets on interaction networks, starting from a simple phenomenological model of a living cell. Formally, the model describes a set of interacting linear harmonic oscillators in thermal equilibrium. Because the energy function is a quadratic form of the degrees of freedom, entropy and all other derived information quantities can be calculated exactly. We apply these concepts to estimate the probability that genes from several independent genome-wide association studies are not mutually informative; to estimate the probability that two disjoint canonical metabolic pathways are not mutually informative; and to infer relationships among human diseases based on their gene signatures. We show that the present approach is able to predict observationally validated relationships not detectable by gene enrichment methods. The converse is also true; the two methods are therefore complementary. AVAILABILITY AND IMPLEMENTATION: The functions defined in this paper are available in an R package, gsia, available for download at https://github.com/ucsd-ccbb/gsia.

Identification of transcription factors responsible for dysregulated networks in human osteoarthritis cartilage by global gene expression analysis.

OBJECTIVE: Osteoarthritis (OA) is the most prevalent joint disease. As disease-modifying therapies are not available, novel therapeutic targets need to be discovered and prioritized for their importance in mediating the abnormal phenotype of cells in OA-affected joints. Here, we generated a genome-wide molecular profile of OA to elucidate regulatory mechanisms of OA pathogenesis and to identify possible therapeutic targets using integrative analysis of mRNA-sequencing data obtained from human knee cartilage. DESIGN: RNA-sequencing (RNA-seq) was performed on 18 normal and 20 OA human knee cartilage tissues. RNA-seq datasets were analysed to identify genes, pathways and regulatory networks that were dysregulated in OA. RESULTS: RNA-seq data analysis revealed 1332 differentially expressed (DE) genes between OA and non-OA samples, including known and novel transcription factors (TFs). Pathway analysis identified 15 significantly perturbed pathways in OA with ECM-related, PI3K-Akt, HIF-1, FoxO and circadian rhythm pathways being the most significantly dysregulated. We selected DE TFs that are enriched for regulating DE genes in OA and prioritized these TFs by creating a cartilage-specific interaction subnetwork. This analysis revealed eight TFs, including JUN, Early growth response (EGR)1, JUND, FOSL2, MYC, KLF4, RELA, and FOS that both target large numbers of dysregulated genes in OA and are themselves suppressed in OA. CONCLUSIONS: We identified a novel subnetwork of dysregulated TFs that represent new mediators of abnormal gene expression and promising therapeutic targets in OA.

Dysregulated circadian rhythm pathway in human osteoarthritis: NR1D1 and BMAL1 suppression alters TGF-ß signaling in chondrocytes.

OBJECTIVES: Circadian rhythm (CR) was identified by RNA sequencing as the most dysregulated pathway in human osteoarthritis (OA) in articular cartilage. This study examined circadian rhythmicity in cultured chondrocytes and the role of the CR genes NR1D1 and BMAL1 in regulating chondrocyte functions. METHODS: RNA was extracted from normal and OA-affected human knee cartilage (n = 14 each). Expression levels of NR1D1 and BMAL1 mRNA and protein were assessed by quantitative PCR and immunohistochemistry. Human chondrocytes were synchronized and harvested at regular intervals to examine circadian rhythmicity in RNA and protein expression. Chondrocytes were treated with small interfering RNA (siRNA) for NR1D1 or BMAL1, followed by RNA sequencing and analysis of the effects on the transforming growth factor beta (TGF-ß) pathway. RESULTS: NR1D1 and BMAL1 mRNA and protein levels were significantly reduced in OA compared to normal cartilage. In cultured human chondrocytes, a clear circadian rhythmicity was observed for NR1D1 and BMAL1. Increased BMAL1 expression was observed after knocking down NR1D1, and decreased NR1D1 levels were observed after knocking down BMAL1. Sequencing of RNA from chondrocytes treated with NR1D1 or BMAL1 siRNA identified 330 and 68 significantly different genes, respectively, and this predominantly affected the TGF-ß signaling pathway. CONCLUSIONS: The CR pathway is dysregulated in OA cartilage. Interference with circadian rhythmicity in cultured chondrocytes affects TGF-ß signaling, which is a central pathway in cartilage homeostasis.

Suppression of REDD1 in osteoarthritis cartilage, a novel mechanism for dysregulated mTOR signaling and defective autophagy.

OBJECTIVE: Aging is a main risk factor for the development of osteoarthritis (OA) and the molecular mechanisms underlying the aging-related changes in articular cartilage include increased mammalian target of rapamycin (mTOR) signaling and defective autophagy. REDD1 is an endogenous inhibitor of mTOR that regulates cellular stress responses. In this study we measured REDD1 expression in normal, aged and OA cartilage and assessed REDD1 function in human and mouse articular chondrocytes. METHODS: REDD1 expression was analyzed in human and mouse articular cartilage by qPCR, western blotting, and immunohistochemistry. For functional studies, REDD1 and TXNIP knockdown or overexpression was performed in chondrocytes in the presence or absence of rapamycin and chloroquine, and mTOR signaling and autophagy were measured by western blotting. REDD1/TXNIP protein interaction was assessed by co-immunoprecipitation experiments. RESULTS: Human and mouse cartilage from normal knee joints expressed high levels of REDD1. REDD1 expression was significantly reduced in aged and OA cartilage. In cultured chondrocytes, REDD1 knockdown increased whereas REDD1 overexpression decreased mTOR signaling. In addition, REDD1 activated autophagy by an mTOR independent mechanism that involved protein/protein interaction with TXNIP. The REDD1/TXNIP complex was required for autophagy activation in chondrocytes. CONCLUSION: The present study shows that REDD1 is highly expressed in normal human articular cartilage and reduced during aging and OA. REDD1 in human chondrocytes negatively regulates mTOR activity and is essential for autophagy activation. Reduced REDD1 expression thus represents a novel mechanism for the increased mTOR activation and defective autophagy observed in OA.

The EndoPredict score provides prognostic information on late distant metastases in ER+/HER2- breast cancer patients.

BACKGROUND: ER+/HER2- breast cancers have a proclivity for late recurrence. A personalised estimate of relapse risk after 5 years of endocrine treatment can improve patient selection for extended hormonal therapy. METHODS: A total of 1702 postmenopausal ER+/HER2- breast cancer patients from two adjuvant phase III trials (ABCSG6, ABCSG8) treated with 5 years of endocrine therapy participated in this study. The multigene test EndoPredict (EP) and the EPclin score (which combines EP with tumour size and nodal status) were predefined in independent training cohorts. All patients were retrospectively assigned to risk categories based on gene expression and on clinical parameters. The primary end point was distant metastasis (DM). Kaplan-Meier method and Cox regression analysis were used in an early (0-5 years) and late time interval (>5 years post diagnosis). RESULTS: EP is a significant, independent, prognostic parameter in the early and late time interval. The expression levels of proliferative and ER signalling genes contribute differentially to the underlying biology of early and late DM. The EPclin stratified 64% of patients at risk after 5 years into a low-risk subgroup with an absolute 1.8% of late DM at 10 years of follow-up. CONCLUSION: The EP test provides additional prognostic information for the identification of early and late DM beyond what can be achieved by combining the commonly used clinical parameters. The EPclin reliably identified a subgroup of patients who have an excellent long-term prognosis after 5 years of endocrine therapy. The side effects of extended therapy should be weighed against this projected outcome.

Chemical induction of silent biosynthetic pathway transcription in Aspergillus niger.

Manipulation of the fungal epigenome is hypothesized to be an effective method for accessing natural products from silent biosynthetic pathways. A library of epigenetic modifiers was tested using the fungus Aspergillus niger to determine the impact of small-molecule inhibitors on reversing the transcriptional suppression of biosynthetic genes involved in polyketide (PKS), non-ribosomal peptide (NRPS), and hybrid PKS-NRPS (HPN) production. Examination of expressed sequence tag libraries from A. niger demonstrated that >70% of its PKS-, NRPS-, and HPN-encoding gene clusters were transcriptionally suppressed under standard laboratory culture conditions. Using a chemical epigenetic methodology, we showed that treatment of A. niger with suberoylanilide hydroxamic acid and 5-azacytidine led to the transcriptional upregulation of many secondary-metabolite-encoding biosynthetic gene clusters. Chemical epigenetic modifiers exhibited positional biases for upregulating chromosomally distal gene clusters. In addition, a phylogenetic-based preference was noted in the upregulation of reducing clade I PKS gene clusters, while reducing clade IV PKS gene clusters were largely unaffected. Manipulating epigenetic features in fungi is a powerful method for accessing the products of silent biosynthetic pathways. Moreover, this approach can be readily incorporated into modern microbial screening operations.