Identification of diverse target RNAs that are functionally regulated by human Pumilio proteins.

Human Pumilio proteins, PUM1 and PUM2, are sequence specific RNA-binding proteins that regulate protein expression. We used RNA-seq, rigorous statistical testing and an experimentally derived fold change cut-off to identify nearly 1000 target RNAs-including mRNAs and non-coding RNAs-that are functionally regulated by PUMs. Bioinformatic analysis defined a PUM Response Element (PRE) that was significantly enriched in transcripts that increased in abundance and matches the PUM RNA-binding consensus. We created a computational model that incorporates PRE position and frequency within an RNA relative to the magnitude of regulation. The model reveals significant correlation of PUM regulation with PREs in 3' untranslated regions (UTRs), coding sequences and non-coding RNAs, but not 5' UTRs. To define direct, high confidence PUM targets, we cross-referenced PUM-regulated RNAs with all PRE-containing RNAs and experimentally defined PUM-bound RNAs. The results define nearly 300 direct targets that include both PUM-repressed and, surprisingly, PUM-activated target RNAs. Annotation enrichment analysis reveal that PUMs regulate genes from multiple signaling pathways and developmental and neurological processes. Moreover, PUM target mRNAs impinge on human disease genes linked to cancer, neurological disorders and cardiovascular disease. These discoveries pave the way for determining how the PUM-dependent regulatory network impacts biological functions and disease states.

Group B streptococcus activates transcriptomic pathways related to premature birth in human extraplacental membranes in vitro.

Streptococcus agalactiae (group B streptococcus [GBS]) infection in pregnant women is the leading cause of infectious neonatal morbidity and mortality in the United States. Although inflammation during infection has been associated with preterm birth, the contribution of GBS to preterm birth is less certain. Moreover, the early mechanisms by which GBS interacts with the gestational tissue to affect adverse pregnancy outcomes are poorly understood. We hypothesized that short-term GBS inoculation activates pathways related to inflammation and premature birth in human extraplacental membranes. We tested this hypothesis using GBS-inoculated human extraplacental membranes in vitro. In agreement with our hypothesis, a microarray-based transcriptomics analysis of gene expression changes in GBS-inoculated membranes revealed that GBS activated pathways related to inflammation and preterm birth with significant gene expression changes occurring as early as 4 h postinoculation. In addition, pathways related to DNA replication and repair were downregulated with GBS treatment. Conclusions based on our transcriptomics data were further supported by responses of prostaglandin E2 (PGE2), and matrix metalloproteinases 1 (MMP1) and 3 (MMP3), all of which are known to be involved in parturition and premature rupture of membranes. These results support our initial hypothesis and provide new information on molecular targets of GBS infection in human extraplacental membranes.

Canalization Leads to Similar Whole Bone Mechanical Function at Maturity in Two Inbred Strains of Mice.

Previously, we showed that cortical mineralization is coordinately adjusted to mechanically offset external bone size differences between A/J (narrow) and C57BL/6J (wide) mouse femora to achieve whole bone strength equivalence at adulthood. The identity of the genes and their interactions that are responsible for establishing this homeostatic state (ie, canalization) remain unknown. We hypothesize that these inbred strains, whose interindividual differences in bone structure and material properties mimic that observed among humans, achieve functional homeostasis by differentially adjusting key molecular pathways regulating external bone size and mineralization throughout growth. The cortices of A/J and C57BL/6J male mouse femora were phenotyped and gene expression levels were assessed across growth (ie, ages 2, 4, 6, 8, 12, 16 weeks). A difference in total cross-sectional area (p < 0.01) and cortical tissue mineral density were apparent between mouse strains by age 2 weeks and maintained at adulthood (p < 0.01). These phenotypic dissimilarities corresponded to gene expression level differences among key regulatory pathways throughout growth. A/J mice had a 1.55- to 7.65-fold greater expression among genes inhibitory to Wnt pathway induction, whereas genes involved in cortical mineralization were largely upregulated 1.50- to 3.77-fold to compensate for their narrow diaphysis. Additionally, both mouse strains showed an upregulation among Wnt pathway antagonists corresponding to the onset of adult ambulation (ie, increased physiological loads). This contrasts with other studies showing an increase in Wnt pathway activation after functionally isolated, experimental in vivo loading regimens. A/J and C57BL/6J long bones provide a model to develop a systems-based approach to identify individual genes and the gene-gene interactions that contribute to trait differences between the strains while being involved in the process by which these traits are coordinately adjusted to establish similar levels of mechanical function, thus providing insight into the process of canalization. © 2017 American Society for Bone and Mineral Research.

Lowered H3K27me3 and DNA hypomethylation define poorly prognostic pediatric posterior fossa ependymomas.

Childhood posterior fossa (PF) ependymomas cause substantial morbidity and mortality. These tumors lack recurrent genetic mutations, but a subset of these ependymomas exhibits CpG island (CpGi) hypermethylation [PF group A (PFA)], implicating epigenetic alterations in their pathogenesis. Further, histological grade does not reliably predict prognosis, highlighting the importance of developing more robust prognostic markers. We discovered global H3K27me3 reduction in a subset of these tumors (PF-ve ependymomas) analogous to H3K27M mutant gliomas. PF-ve tumors exhibited many clinical and biological similarities with PFA ependymomas. Genomic H3K27me3 distribution showed an inverse relationship with CpGi methylation, suggesting that CpGi hypermethylation drives low H3K27me3 in PF-ve ependymomas. Despite CpGi hypermethylation and global H3K27me3 reduction, these tumors showed DNA hypomethylation in the rest of the genome and exhibited increased H3K27me3 genomic enrichment at limited genomic loci similar to H3K27M mutant gliomas. Combined integrative analysis of PF-ve ependymomas with H3K27M gliomas uncovered common epigenetic deregulation of select factors that control radial glial biology, and PF radial glia in early human development exhibited reduced H3K27me3. Finally, H3K27me3 immunostaining served as a biomarker of poor prognosis and delineated radiologically invasive tumors, suggesting that reduced H3K27me3 may be a prognostic indicator in PF ependymomas.

Reversal of the Transcriptome by Prostaglandin E2 during Myofibroblast Dedifferentiation.

Myofibroblasts, the major effector cells in pathologic fibrosis, derive from the differentiation of fibroblasts driven by mediators such as transforming growth factor-ß1 (TGF-ß1) and biomechanical signals. Although the myofibroblast has traditionally been considered a terminally differentiated cell, the lipid mediator prostaglandin E2 (PGE2) has been shown to not only prevent but also reverse myofibroblast differentiation, as characterized by the ability of PGE2 to diminish expression of collagen I and α-smooth muscle actin in established myofibroblasts. Here, we use microarrays to examine the extent of transcriptomic changes that occur during TGF-ß1-induced differentiation and PGE2-induced dedifferentiation of myofibroblasts. Normal primary human adult lung fibroblasts were cultured for 24 hours with or without TGF-ß1 and treated for 48 hours with PGE2. Gene expression levels were assessed from total RNA on the Affymetrix U219 microarray. TGF-ß1 up-regulated 588 genes and down-regulated 689 genes compared with control cells. PGE2 reversed the expression of 363 (62%) of the TGF-ß1-up-regulated genes and 345 (50%) of the TGF-ß1-down-regulated genes. Genes up-regulated by TGF-ß1 and reversed by PGE2 were enriched in annotations for Cell Adhesion, Contractile Fiber, and Actin Binding, whereas genes down-regulated by TGF-ß1 but subsequently reversed by PGE2 were enriched in annotations for Glycoprotein, Polysaccharide Binding, and Regulation of Cell Migration. Surprisingly, the genes whose expression was affected by PGE2 differed between TGF-ß1-induced myofibroblasts and undifferentiated fibroblasts. These data demonstrate the capacity of PGE2 to effect marked global alterations in the transcriptomic program of differentiated myofibroblasts and emphasize the considerable plasticity of these cells.

Longevity Genes Revealed by Integrative Analysis of Isoform-Specific daf-16/FoxO Mutants of Caenorhabditis elegans.

FoxO transcription factors promote longevity across taxa. How they do so is poorly understood. In the nematode Caenorhabditis elegans, the A- and F-isoforms of the FoxO transcription factor DAF-16 extend life span in the context of reduced DAF-2 insulin-like growth factor receptor (IGFR) signaling. To elucidate the mechanistic basis for DAF-16/FoxO-dependent life span extension, we performed an integrative analysis of isoform-specific daf-16/FoxO mutants. In contrast to previous studies suggesting that DAF-16F plays a more prominent role in life span control than DAF-16A, isoform-specific daf-16/FoxO mutant phenotypes and whole transcriptome profiling revealed a predominant role for DAF-16A over DAF-16F in life span control, stress resistance, and target gene regulation. Integration of these datasets enabled the prioritization of a subset of 92 DAF-16/FoxO target genes for functional interrogation. Among 29 genes tested, two DAF-16A-specific target genes significantly influenced longevity. A loss-of-function mutation in the conserved gene gst-20, which is induced by DAF-16A, reduced life span extension in the context of daf-2/IGFR RNAi without influencing longevity in animals subjected to control RNAi. Therefore, gst-20 promotes DAF-16/FoxO-dependent longevity. Conversely, a loss-of-function mutation in srr-4, a gene encoding a seven-transmembrane-domain receptor family member that is repressed by DAF-16A, extended life span in control animals, indicating that DAF-16/FoxO may extend life span at least in part by reducing srr-4 expression. Our discovery of new longevity genes underscores the efficacy of our integrative strategy while providing a general framework for identifying specific downstream gene regulatory events that contribute substantially to transcription factor functions. As FoxO transcription factors have conserved functions in promoting longevity and may be dysregulated in aging-related diseases, these findings promise to illuminate fundamental principles underlying aging in animals.

Disrupting SUMOylation enhances transcriptional function and ameliorates polyglutamine androgen receptor-mediated disease.

Expansion of the polyglutamine (polyQ) tract within the androgen receptor (AR) causes neuromuscular degeneration in individuals with spinobulbar muscular atrophy (SBMA). PolyQ AR has diminished transcriptional function and exhibits ligand-dependent proteotoxicity, features that have both been implicated in SBMA; however, the extent to which altered AR transcriptional function contributes to pathogenesis remains controversial. Here, we sought to dissociate effects of diminished AR function from polyQ-mediated proteotoxicity by enhancing the transcriptional activity of polyQ AR. To accomplish this, we bypassed the inhibitory effect of AR SUMOylation (where SUMO indicates small ubiquitin-like modifier) by mutating conserved lysines in the polyQ AR that are sites of SUMOylation. We determined that replacement of these residues by arginine enhances polyQ AR activity as a hormone-dependent transcriptional regulator. In a murine model, disruption of polyQ AR SUMOylation rescued exercise endurance and type I muscle fiber atrophy; it also prolonged survival. These changes occurred without overt alterations in polyQ AR expression or aggregation, revealing the favorable trophic support exerted by the ligand-activated receptor. Our findings demonstrate beneficial effects of enhancing the transcriptional function of the ligand-activated polyQ AR and indicate that the SUMOylation pathway may be a potential target for therapeutic intervention in SBMA.

Bisphenol A-associated alterations in genome-wide DNA methylation and gene expression patterns reveal sequence-dependent and non-monotonic effects in human fetal liver.

Bisphenol A (BPA), a high production volume chemical widely used in consumer products, is an endocrine active compound associated with complex epigenetic responses in animal models and humans. Developmental BPA exposure in mice previously revealed widespread changes in the mouse liver methylome. Here, we undertake the first epigenome-wide analysis of the effect of BPA concentration on human fetal liver DNA methylation. Enzymatic enrichment of genomic DNA for high CG density and methylation followed by next-generation sequencing yielded data for positional methylation across the genome. Comparing three groups of BPA-exposed subjects (n=18; 6 per group), high (35.44-96.76 ng/g), low (3.50 to 5.79 ng/g), and non-detect (<0.83 ng/g), revealed regions of altered methylation. Similar numbers of regions of altered methylations were detected in pairwise comparisons; however, their genomic locations were distinct between the non-detect and low or high BPA groups. In general, BPA levels were positively associated with methylation in CpG islands and negatively associated with methylation in CpG shores, shelves, and repetitive regions. DNA methylation at the SNORD imprinted cluster (15q11q13) illustrated both linear and non-monotonic associations with BPA levels. Integrated methylation and RNA-sequencing gene expression analysis revealed differential regulation of transcription at low BPA levels, as well as expression changes in RNA for ligand-binding proteins as BPA levels increase. BPA levels in human fetal liver tissue are associated with complex linear and non-monotonic as well as sequence-dependent alterations in DNA methylation. Future longitudinal studies are needed to link these changes with altered health risks.

A knockin mouse model of spinocerebellar ataxia type 3 exhibits prominent aggregate pathology and aberrant splicing of the disease gene transcript.

Polyglutamine diseases, including spinocerebellar ataxia type 3 (SCA3), are caused by CAG repeat expansions that encode abnormally long glutamine repeats in the respective disease proteins. While the mechanisms underlying neurodegeneration remain uncertain, evidence supports a proteotoxic role for the mutant protein dictated in part by the specific genetic and protein context. To further define pathogenic mechanisms in SCA3, we generated a mouse model in which a CAG expansion of 82 repeats was inserted into the murine locus by homologous recombination. SCA3 knockin mice exhibit region-specific aggregate pathology marked by intranuclear accumulation of the mutant Atxn3 protein, abundant nuclear inclusions and, in select brain regions, extranuclear aggregates localized to neuritic processes. Knockin mice also display altered splicing of the disease gene, promoting expression of an alternative isoform in which the intron immediately downstream of the CAG repeat is retained. In an independent mouse model expressing the full human ATXN3 disease gene, expression of this alternatively spliced transcript is also enhanced. These results, together with recent findings in other polyglutamine diseases, suggest that CAG repeat expansions can promote aberrant splicing to produce potentially more aggregate-prone isoforms of the disease proteins. This report of a SCA3 knockin mouse expands the repertoire of existing models of SCA3, and underscores the potential contribution of alternative splicing to disease pathogenesis in SCA3 and other polyglutamine disorders.

Lung fibroblasts from patients with idiopathic pulmonary fibrosis exhibit genome-wide differences in DNA methylation compared to fibroblasts from nonfibrotic lung.

Excessive fibroproliferation is a central hallmark of idiopathic pulmonary fibrosis (IPF), a chronic, progressive disorder that results in impaired gas exchange and respiratory failure. Fibroblasts are the key effector cells in IPF, and aberrant expression of multiple genes contributes to their excessive fibroproliferative phenotype. DNA methylation changes are critical to the development of many diseases, but the DNA methylome of IPF fibroblasts has never been characterized. Here, we utilized the HumanMethylation 27 array, which assays the DNA methylation level of 27,568 CpG sites across the genome, to compare the DNA methylation patterns of IPF fibroblasts (n = 6) with those of nonfibrotic patient controls (n = 3) and commercially available normal lung fibroblast cell lines (n = 3). We found that multiple CpG sites across the genome are differentially methylated (as defined by P value less than 0.05 and fold change greater than 2) in IPF fibroblasts compared to fibroblasts from nonfibrotic controls. These methylation differences occurred both in genes recognized to be important in fibroproliferation and extracellular matrix generation, as well as in genes not previously recognized to participate in those processes (including organ morphogenesis and potassium ion channels). We used bisulfite sequencing to independently verify DNA methylation differences in 3 genes (CDKN2B, CARD10, and MGMT); these methylation changes corresponded with differences in gene expression at the mRNA and protein level. These differences in DNA methylation were stable throughout multiple cell passages. DNA methylation differences may thus help to explain a proportion of the differences in gene expression previously observed in studies of IPF fibroblasts. Moreover, significant variability in DNA methylation was observed among individual IPF cell lines, suggesting that differences in DNA methylation may contribute to fibroblast heterogeneity among patients with IPF. These results demonstrate that IPF fibroblasts exhibit global differences in DNA methylation that may contribute to the excessive fibroproliferation associated with this disease.

A bioinformatics approach reveals novel interactions of the OVOL transcription factors in the regulation of epithelial - mesenchymal cell reprogramming and cancer progression.

BACKGROUND: Mesenchymal to Epithelial Transition (MET) plasticity is critical to cancer progression, and we recently showed that the OVOL transcription factors (TFs) are critical regulators of MET. Results of that work also posed the hypothesis that the OVOLs impact MET in a range of cancers. We now test this hypothesis by developing a model, OVOL Induced MET (OI-MET), and sub-model (OI-MET-TF), to characterize differential gene expression in MET common to prostate cancer (PC) and breast cancer (BC). RESULTS: In the OI-MET model, we identified 739 genes differentially expressed in both the PC and BC models. For this gene set, we found significant enrichment of annotation for BC, PC, cancer, and MET, as well as regulation of gene expression by AP1, STAT1, STAT3, and NFKB1. Focusing on the target genes for these four TFs plus the OVOLs, we produced the OI-MET-TF sub-model, which shows even greater enrichment for these annotations, plus significant evidence of cooperation among these five TFs. Based on known gene/drug interactions, we prioritized targets in the OI-MET-TF network for follow-on analysis, emphasizing the clinical relevance of this work. Reflecting these results back to the OI-MET model, we found that binding motifs for the TF pair AP1/MYC are more frequent than expected and that the AP1/MYC pair is significantly enriched in binding in cancer models, relative to non-cancer models, in these promoters. This effect is seen in both MET models (solid tumors) and in non-MET models (leukemia). These results are consistent with our hypothesis that the OVOLs impact cancer susceptibility by regulating MET, and extend the hypothesis to include mechanisms not specific to MET. CONCLUSIONS: We find significant evidence of the OVOL, AP1, STAT1, STAT3, and NFKB1 TFs having important roles in MET, and more broadly in cancer. We prioritize known gene/drug targets for follow-up in the clinic, and we show that the AP1/MYC TF pair is a strong candidate for intervention.

H2A.B facilitates transcription elongation at methylated CpG loci.

H2A.B is a unique histone H2A variant that only exists in mammals. Here we found that H2A.B is ubiquitously expressed in major organs. Genome-wide analysis of H2A.B in mouse ES cells shows that H2A.B is associated with methylated DNA in gene body regions. Moreover, H2A.B-enriched gene loci are actively transcribed. One typical example is that H2A.B is enriched in a set of differentially methylated regions at imprinted loci and facilitates transcription elongation. These results suggest that H2A.B positively regulates transcription elongation by overcoming DNA methylation in the transcribed region. It provides a novel mechanism by which transcription is regulated at DNA hypermethylated regions.

Transcription factors OVOL1 and OVOL2 induce the mesenchymal to epithelial transition in human cancer.

Cell plasticity regulated by the balance between the mesenchymal to epithelial transition (MET) and the opposite program, EMT, is critical in the metastatic cascade. Several transcription factors (TFs) are known to regulate EMT, though the mechanisms of MET remain unclear. We demonstrate a novel function of two TFs, OVOL1 and OVOL2, as critical inducers of MET in human cancers. Our findings indicate that the OVOL-TFs control MET through a regulatory feedback loop with EMT-inducing TF ZEB1, and the regulation of mRNA splicing by inducing Epithelial Splicing Regulatory Protein 1 (ESRP1). Using mouse prostate tumor models we show that expression of OVOL-TFs in mesenchymal prostate cancer cells attenuates their metastatic potential. The role of OVOL-TFs as inducers of MET is further supported by expression analyses in 917 cancer cell lines, suggesting their role as crucial regulators of epithelial-mesenchymal cell plasticity in cancer.

Prostaglandin E2 increases fibroblast gene-specific and global DNA methylation via increased DNA methyltransferase expression.

Although alterations in DNA methylation patterns have been associated with specific diseases and environmental exposures, the mediators and signaling pathways that direct these changes remain understudied. The bioactive lipid mediator prostaglandin E(2) (PGE(2)) has been shown to exert a myriad of effects on cell survival, proliferation, and differentiation. Here, we report that PGE(2) also signals to increase global DNA methylation and DNA methylation machinery in fibroblasts. HumanMethylation27 BeadChip array analysis of primary fetal (IMR-90) and adult lung fibroblasts identified multiple genes that were hypermethylated in response to PGE(2). PGE(2), compared with nontreated controls, increased expression and activity (EC(50)∼10(7) M) of one specific isoform of DNA methyltransferase, DNMT3a. Silencing of DNMT3a negated the ability of PGE(2) to increase DNMT activity. The increase in DNMT3a expression was mediated by PGE(2) signaling via its E prostanoid 2 receptor and the second messenger cAMP. PGE(2), compared with the untreated control, increased the expression and activity of Sp1 and Sp3 (EC(50)∼3×10(7) M), transcription factors known to increase DNMT3a expression, and inhibition of these transcription factors abrogated the PGE(2) increase of DNMT3a expression. These findings were specific to fibroblasts, as PGE(2) decreased DNMT1 and DNMT3a expression in RAW macrophages. Taken together, these findings establish that DNA methylation is regulated by a ubiquitous bioactive endogenous mediator. Given that PGE(2) biosynthesis is modulated by environmental toxins, various disease states, and commonly used pharmacological agents, these findings uncover a novel mechanism by which alterations in DNA methylation patterns may arise in association with disease and certain environmental exposures.

Convergence of genetic influences in comorbidity.

BACKGROUND: Predisposition to complex diseases is explained in part by genetic variation, and complex diseases are frequently comorbid, consistent with pleiotropic genetic variation influencing comorbidity. Genome Wide Association (GWA) studies typically assess association between SNPs and a single-disease phenotype. Fisher meta-analysis combines evidence of association from single-disease GWA studies, assuming that each study is an independent test of the same hypothesis. The Rank Product (RP) method overcomes limitations posed by Fisher assumptions, though RP was not designed for GWA data. METHODS: We modified RP to accommodate GWA data, and we call it modRP. Using p-values output from GWA studies, we aggregate evidence for association between SNPs and related phenotypes. To assess significance, RP randomly samples the observed ranks to develop the null distribution of the RP statistic, and then places the observed RPs into the null distribution. ModRP eliminates the effect of linkage disequilibrium and controls for differences in power at tested SNPs, to meet RP assumptions in application to GWA data. RESULTS: After validating modRP based on both positive and negative control studies, we searched for pleiotropic influences on comorbid substance use disorders in a novel study, and found two SNPs to be significantly associated with comorbid cocaine, opium, and nicotine dependence. Placing these SNPs into biological context, we developed a protein network modeling the interaction of cocaine, nicotine, and opium with these variants. CONCLUSIONS: ModRP is a novel approach to identifying pleiotropic genetic influences on comorbid complex diseases. It can be used to assess association for related phenotypes where raw data is unavailable or inappropriate for analysis using other approaches. The method is conceptually simple and produces statistically significant, biologically relevant results.

A genetic network model of cellular responses to lithium treatment and cocaine abuse in bipolar disorder.

BACKGROUND: Lithium is an effective treatment for Bipolar Disorder (BD) and significantly reduces suicide risk, though the molecular basis of lithium's effectiveness is not well understood. We seek to improve our understanding of this effectiveness by posing hypotheses based on new experimental data as well as published data, testing these hypotheses in silico, and posing new hypotheses for validation in future studies. We initially hypothesized a gene-by-environment interaction where lithium, acting as an environmental influence, impacts signal transduction pathways leading to differential expression of genes important in the etiology of BD mania. RESULTS: Using microarray and rt-QPCR assays, we identified candidate genes that are differentially expressed with lithium treatment. We used a systems biology approach to identify interactions among these candidate genes and develop a network of genes that interact with the differentially expressed candidates. Notably, we also identified cocaine as having a potential influence on the network, consistent with the observed high rate of comorbidity for BD and cocaine abuse. The resulting network represents a novel hypothesis on how multiple genetic influences on bipolar disorder are impacted by both lithium treatment and cocaine use. Testing this network for association with BD and related phenotypes, we find that it is significantly over-represented for genes that participate in signal transduction, consistent with our hypothesized-gene-by environment interaction. In addition, it models related pharmacogenomic, psychiatric, and chemical dependence phenotypes. CONCLUSIONS: We offer a network model of gene-by-environment interaction associated with lithium's effectiveness in treating BD mania, as well as the observed high rate of comorbidity of BD and cocaine abuse. We identified drug targets within this network that represent immediate candidates for therapeutic drug testing. Posing novel hypotheses for validation in future work, we prioritized SNPs near genes in the network based on functional annotation. We also developed a "concept signature" for the genes in the network and identified additional candidate genes that may influence the system because they are significantly associated with the signature.

Modeling complex genetic and environmental influences on comorbid bipolar disorder with tobacco use disorder.

BACKGROUND: Comorbidity of psychiatric and substance use disorders represents a significant complication in the clinical course of both disorders. Bipolar Disorder (BD) is a psychiatric disorder characterized by severe mood swings, ranging from mania to depression, and up to a 70% rate of comorbid Tobacco Use Disorder (TUD). We found epidemiological evidence consistent with a common underlying etiology for BD and TUD, as well as evidence of both genetic and environmental influences on BD and TUD. Therefore, we hypothesized a common underlying genetic etiology, interacting with nicotine exposure, influencing susceptibility to both BD and TUD. METHODS: Using meta-analysis, we compared TUD rates for BD patients and the general population. We identified candidate genes showing statistically significant, replicated, evidence of association with both BD and TUD. We assessed commonality among these candidate genes and hypothesized broader, multi-gene network influences on the comorbidity. Using Fisher Exact tests we tested our hypothesized genetic networks for association with the comorbidity, then compared the inferences drawn with those derived from the commonality assessment. Finally, we prioritized candidate SNPs for validation. RESULTS: We estimate risk for TUD among BD patients at 2.4 times that of the general population. We found three candidate genes associated with both BD and TUD (COMT, SLC6A3, and SLC6A4) and commonality analysis suggests that these genes interact in predisposing psychiatric and substance use disorders. We identified a 69 gene network that influences neurotransmitter signaling and shows significant over-representation of genes associated with BD and TUD, as well as genes differentially expressed with exposure to tobacco smoke. Twenty four of these genes are known drug targets. CONCLUSIONS: This work highlights novel bioinformatics resources and demonstrates the effectiveness of using an integrated bioinformatics approach to improve our understanding of complex disease etiology. We illustrate the development and testing of hypotheses for a comorbidity predisposed by both genetic and environmental influences. Consistent with our hypothesis, the selected network models multiple interacting genetic influences on comorbid BD with TUD, as well as the environmental influence of nicotine. This network nominates candidate genes for validation and drug testing, and we offer a panel of SNPs prioritized for follow-up.

Identifying hypothetical genetic influences on complex disease phenotypes.

BACKGROUND: Statistical interactions between disease-associated loci of complex genetic diseases suggest that genes from these regions are involved in a common mechanism impacting, or impacted by, the disease. The computational problem we address is to discover relationships among genes from these interacting regions that may explain the observed statistical interaction and the role of these genes in the disease phenotype. RESULTS: We describe a heuristic algorithm for generating hypothetical gene relationships from loci associated with a complex disease phenotype. This approach, called Prioritizing Disease Genes by Analysis of Common Elements (PDG-ACE), mines biomedical keywords from text descriptions of genes and uses them to relate genes close to disease-associated loci. A keyword common to, and significantly over-represented in, a pair of gene descriptions may represent a preliminary hypothesis about the biological relationship between the genes, and suggest the role the genes play in the disease phenotype. CONCLUSION: Our experimentation shows that the approach finds previously published relationships, while failing to find relationships that don't exist. The results also indicate that the approach is robust to differences in keyword vocabulary. We outline a brief case study in which results from a recently published Type 2 Diabetes association study are used to identify potential hypotheses.