Linking gene expression to phenotypes via pathway information

Pathway level subtyping identifies a slow-cycling biological phenotype associated with poor clinical outcomes in colorectal cancer

Molecular stratification using gene-level transcriptional data has identified subtypes with distinctive genotypic and phenotypic traits, as exemplified by the consensus molecular subtypes (CMS) in colorectal cancer (CRC). Here, rather than gene-level data, we make use of gene ontology and biological activation state information for initial molecular class discovery. In doing so, we defined three pathway-derived subtypes (PDS) in CRC: PDS1 tumors, which are canonical/LGR5+ stem-rich, highly proliferative and display good prognosis; PDS2 tumors, which are regenerative/ANXA1+ stem-rich, with elevated stromal and immune tumor microenvironmental lineages; and PDS3 tumors, which represent a previously overlooked slow-cycling subset of tumors within CMS2 with reduced stem populations and increased differentiated lineages, particularly enterocytes and enteroendocrine cells, yet display the worst prognosis in locally advanced disease. These PDS3 phenotypic traits are evident across numerous bulk and single-cell datasets, and demark a series of subtle biological states that are currently under-represented in pre-clinical models and are not identified using existing subtyping classifiers.

Environmental influences on Streptococcus sanguinis membrane vesicle biogenesis

Membrane vesicles are produced by Gram-negative and Gram-positive bacteria. While membrane vesicles are potent elicitors of eukaryotic cells and involved in cell-cell communication, information is scarce about their general biology in the context of community members and the environment. Streptococcus sanguinis, a Gram-positive oral commensal, is prevalent in the oral cavity and well-characterized for its ability to antagonize oral pathobionts. We have found that production and dissemination of membrane vesicles by S. sanguinis is dependent on environmental and community factors. Co-culture with interacting commensal Corynebacterium durum, as well as with the periodontal pathobiont Filifactor alocis had no effect on S. sanguinis vesicle number and size, whereas the periodontal pathobiont Porphyromonas gingivalis abolished S. sanguinis vesicle production. Using both correlation and differential expression analyses to examine the transcriptomic changes underlying vesicle production, we found that differential expression of genes encoding proteins related to the cytoplasmic membrane and peptidoglycan correlate with the abundance of membrane vesicles. Proteomic characterizations of the vesicle cargo identified a variety of proteins, including those predicted to influence host interactions or host immune responses. Cell culture studies of gingival epithelial cells demonstrated that both crude and highly purified membrane vesicles could induce the expression of IL-8, TNF-α, IL-1β, and Gro-α within 6 hours of inoculation at levels comparable to whole cells. Our findings suggest that production of membrane vesicles by S. sanguinis is heavily influenced by community and environmental factors and plays an important role in communication with host cells.

Muscle transcriptome analysis provides new insights into the growth gap between fast- and slow-growing Sinocyclocheilus grahami

Sinocyclocheilus grahami is an economically valuable and famous fish in Yunnan Province, China. However, given its slow growth (40 g/2 years) and large growth differences among individuals, its growth performance needs to be improved for sustainable future use, in which molecular breeding technology can play an important role. In the current study, we conducted muscle transcriptomic analysis to investigate the growth gaps among individuals and the mechanism underlying growth within 14 fast- and 14 slow-growth S. grahami. In total, 1,647 differentially expressed genes (DEGs) were obtained, including 947 up-regulated and 700 down-regulated DEGs in fast-growth group. Most DEGs were significantly enriched in ECM-receptor interaction, starch and sucrose metabolism, glycolysis/gluconeogenesis, pyruvate metabolism, amino acids biosynthesis and metabolism, peroxisome, and PPAR signaling pathway. Some genes related to glycogen degradation, glucose transport, and glycolysis (e.g., adipoq, prkag1, slc2a1, agl, pygm, pgm1, pfkm, gapdh, aldoa, pgk1, pgam2, bpgm, and eno3) were up-regulated, while some genes related to fatty acid degradation and transport (e.g., acox1, acaa1, fabp1b.1, slc27a1, and slc27a2) and amino acid metabolism (e.g., agxt, shmt1, glula, and cth) were down-regulated in the fast-growth group. Weighted gene co-expression network analysis identified col1a1, col1a2, col5a1, col6a2, col10a1, col26a1, bglap, and krt15 as crucial genes for S. grahami growth. Several genes related to bone and muscle growth (e.g., bmp2, bmp3, tgfb1, tgfb2, gdf10, and myog) were also up-regulated in the fast-growth group. These results suggest that fast-growth fish may uptake adequate energy (e.g., glucose, fatty acid, and amino acids) from fodder, with excess energy substances used to synthesize collagen to accelerate bone and muscle growth after normal life activities are maintained. Moreover, energy uptake may be the root cause, while collagen synthesis may be the direct reason for the growth gap between fast- and slow-growth fish. Hence, improving food intake and collagen synthesis may be crucial for accelerating S. grahami growth, and further research is required to fully understand and confirm these associations.

Unravelling the Impact of Cyclic Mechanical Stretch in Keratoconus-A Transcriptomic Profiling Study

Biomechanical and molecular stresses may contribute to the pathogenesis of keratoconus (KC). We aimed to profile the transcriptomic changes in healthy primary human corneal (HCF) and KC-derived cells (HKC) combined with TGFβ1 treatment and cyclic mechanical stretch (CMS), mimicking the pathophysiological condition in KC. HCFs (n = 4) and HKCs (n = 4) were cultured in flexible-bottom collagen-coated 6-well plates treated with 0, 5, and 10 ng/mL of TGFβ1 with or without 15% CMS (1 cycle/s, 24 h) using a computer-controlled Flexcell FX-6000T Tension system. We used stranded total RNA-Seq to profile expression changes in 48 HCF/HKC samples (100 bp PE, 70-90 million reads per sample), followed by bioinformatics analysis using an established pipeline with Partek Flow software. A multi-factor ANOVA model, including KC, TGFβ1 treatment, and CMS, was used to identify differentially expressed genes (DEGs, |fold change| ≥ 1.5, FDR ≤ 0.1, CPM ≥ 10 in ≥1 sample) in HKCs (n = 24) vs. HCFs (n = 24) and those responsive to TGFβ1 and/or CMS. PANTHER classification system and the DAVID bioinformatics resources were used to identify significantly enriched pathways (FDR ≤ 0.05). Using multi-factorial ANOVA analyses, 479 DEGs were identified in HKCs vs. HCFs including TGFβ1 treatment and CMS as cofactors. Among these DEGs, 199 KC-altered genes were responsive to TGFβ1, thirteen were responsive to CMS, and six were responsive to TGFβ1 and CMS. Pathway analyses using PANTHER and DAVID indicated the enrichment of genes involved in numerous KC-relevant functions, including but not limited to degradation of extracellular matrix, inflammatory response, apoptotic processes, WNT signaling, collagen fibril organization, and cytoskeletal structure organization. TGFβ1-responsive KC DEGs were also enriched in these. CMS-responsive KC-altered genes such as OBSCN, CLU, HDAC5, AK4, ITGA10, and F2RL1 were identified. Some KC-altered genes, such as CLU and F2RL1, were identified to be responsive to both TGFβ1 and CMS. For the first time, our multi-factorial RNA-Seq study has identified many KC-relevant genes and pathways in HKCs with TGFβ1 treatment under CMS, suggesting a potential role of TGFβ1 and biomechanical stretch in KC development.

Overview of gene expression techniques with an emphasis on vitamin D related studies

Each cell controls when and how its genes must be expressed for proper function. Every function in a cell is driven by signaling molecules through various regulatory cascades. Different cells in a multicellular organism may express very different sets of genes, even though they contain the same DNA. The set of genes expressed in a cell determines the set of proteins and functional RNAs it contains, giving it its unique properties. Malfunction in gene expression harms the cell and can lead to the development of various disease conditions. The use of rapid high-throughput gene expression profiling unravels the complexity of human disease at various levels. Peripheral blood mononuclear cells (PBMC) have been used frequently to understand gene expression homeostasis in various disease conditions. However, more studies are required to validate whether PBMC gene expression patterns accurately reflect the expression of other cells or tissues. Vitamin D, which is responsible for a multitude of health consequences, is also an immune modulatory hormone with major biological activities in the innate and adaptive immune systems. Vitamin D exerts its diverse biological effects in target tissues by regulating gene expression and its deficiency, is recognized as a public health problem worldwide. Understanding the genetic factors that affect vitamin D has the potential benefit that it will make it easier to identify individuals who require supplementation. Different technological advances in gene expression can be used to identify and assess the severity of disease and aid in the development of novel therapeutic interventions. This review focuses on different gene expression approaches and various clinical studies of vitamin D to investigate the role of gene expression in identifying the molecular signature of the disease.

PhenCards: a data resource linking human phenotype information to biomedical knowledge

We present PhenCards ( https://phencards.org ), a database and web server intended as a one-stop shop for previously disconnected biomedical knowledge related to human clinical phenotypes. Users can query human phenotype terms or clinical notes. PhenCards obtains relevant disease/phenotype prevalence and co-occurrence, drug, procedural, pathway, literature, grant, and collaborator data. PhenCards recommends the most probable genetic diseases and candidate genes based on phenotype terms from clinical notes. PhenCards facilitates exploration of phenotype, e.g., which drugs cause or are prescribed for patient symptoms, which genes likely cause specific symptoms, and which comorbidities co-occur with phenotypes.

GOcats: A tool for categorizing Gene Ontology into subgraphs of user-defined concepts

Gene Ontology is used extensively in scientific knowledgebases and repositories to organize a wealth of biological information. However, interpreting annotations derived from differential gene lists is often difficult without manually sorting into higher-order categories. To address these issues, we present GOcats, a novel tool that organizes the Gene Ontology (GO) into subgraphs representing user-defined concepts, while ensuring that all appropriate relations are congruent with respect to scoping semantics. We tested GOcats performance using subcellular location categories to mine annotations from GO-utilizing knowledgebases and evaluated their accuracy against immunohistochemistry datasets in the Human Protein Atlas (HPA). In comparison to term categorizations generated from UniProt's controlled vocabulary and from GO slims via OWLTools' Map2Slim, GOcats outperformed these methods in its ability to mimic human-categorized GO term sets. Unlike the other methods, GOcats relies only on an input of basic keywords from the user (e.g. biologist), not a manually compiled or static set of top-level GO terms. Additionally, by identifying and properly defining relations with respect to semantic scope, GOcats can utilize the traditionally problematic relation, has_part, without encountering erroneous term mapping. We applied GOcats in the comparison of HPA-sourced knowledgebase annotations to experimentally-derived annotations provided by HPA directly. During the comparison, GOcats improved correspondence between the annotation sources by adjusting semantic granularity. GOcats enables the creation of custom, GO slim-like filters to map fine-grained gene annotations from gene annotation files to general subcellular compartments without needing to hand-select a set of GO terms for categorization. Moreover, GOcats can customize the level of semantic specificity for annotation categories. Furthermore, GOcats enables a safe and more comprehensive semantic scoping utilization of go-core, allowing for a more complete utilization of information available in GO. Together, these improvements can impact a variety of GO knowledgebase data mining use-cases as well as knowledgebase curation and quality control.

Progress in data interoperability to support computational toxicology and chemical safety evaluation

New approach methodologies (NAMs) in chemical safety evaluation are being explored to address the current public health implications of human environmental exposures to chemicals with limited or no data for assessment. For over a decade since a push toward "Toxicity Testing in the 21st Century," the field has focused on massive data generation efforts to inform computational approaches for preliminary hazard identification, adverse outcome pathways that link molecular initiating events and key events to apical outcomes, and high-throughput approaches to risk-based ratios of bioactivity and exposure to inform relative priority and safety assessment. Projects like the interagency Tox21 program and the US EPA ToxCast program have generated dose-response information on thousands of chemicals, identified and aggregated information from legacy systems, and created tools for access and analysis. The resulting information has been used to develop computational models as viable options for regulatory applications. This progress has introduced challenges in data management that are new, but not unique, to toxicology. Some of the key questions require critical thinking and solutions to promote semantic interoperability, including: (1) identification of bioactivity information from NAMs that might be related to a biological process; (2) identification of legacy hazard information that might be related to a key event or apical outcomes of interest; and, (3) integration of these NAM and traditional data for computational modeling and prediction of complex apical outcomes such as carcinogenesis. This work reviews a number of toxicology-related efforts specifically related to bioactivity and toxicological data interoperability based on the goals established by Findable, Accessible, Interoperable, and Reusable (FAIR) Data Principles. These efforts are essential to enable better integration of NAM and traditional toxicology information to support data-driven toxicology applications.

Gene ORGANizer: linking genes to the organs they affect

One of the biggest challenges in studying how genes work is understanding their effect on the physiology and anatomy of the body. Existing tools try to address this using indirect features, such as expression levels and biochemical pathways. Here, we present Gene ORGANizer (geneorganizer.huji.ac.il), a phenotype-based tool that directly links human genes to the body parts they affect. It is built upon an exhaustive curated database that links >7000 genes to ∼150 anatomical parts using >150 000 gene-organ associations. The tool offers user-friendly platforms to analyze the anatomical effects of individual genes, and identify trends within groups of genes. We demonstrate how Gene ORGANizer can be used to make new discoveries, showing that chromosome X is enriched with genes affecting facial features, that positive selection targets genes with more constrained phenotypic effects, and more. We expect Gene ORGANizer to be useful in a variety of evolutionary, medical and molecular studies aimed at understanding the phenotypic effects of genes.

Muscle transcriptome resource for growth, lipid metabolism and immune system in Hilsa shad, Tenualosa ilisha

The information on the genes involved in muscle growth, lipid metabolism and immune systems would help to understand the mechanisms during the spawning migration in Hilsa shad, which in turn would be useful in its future domestication process. The primary objective of this study was to generate the transcriptome profile of its muscle through RNA seq. The total RNA was isolated and library was prepared from muscle tissue of Tenualosa ilisha, which was collected from Padma River at Farakka, India. The prepared library was then sequenced by Illumina HiSeq platform, HiSeq 2000, using paired-end strategy. A total of 8.68 GB of pair-end reads of muscle transcriptome was generated, and 43,384,267 pair-end reads were assembled into 3,04,233 contigs, of which 23.99% of assembled contigs has length ≥ 150 bp. The total GO terms were categorised into cellular component, molecular function and biological process through PANTHER database. Fifty-three genes related to muscle growth were identified and genes in different pathways were: 75 in PI3/AKT, 46 in mTOR, 76 in MAPK signalling, 24 in Janus kinase-signal transducer and activator of transcription, 45 in AMPK and 27 in cGMP pathways. This study also mined the genes involved in lipid metabolism, in which glycerophospholipid metabolism contained highest number of genes (32) and four were found to be involved in fatty acid biosynthesis. There were 58 immune related genes found, in which 31 were under innate and 27 under adaptive immunity. The present study included a large genomic resource of T. ilisha muscle generated through RNAseq, which revealed the essential dataset for our understanding of regulatory processes, specifically during the seasonal spawning migration. As Hilsa is a slow growing fish, the genes identified for muscle growth provided the basic information to study myogenesis. In addition, genes identified for lipid metabolism and immune system would provide resources for lipid synthesis and understanding of Hilsa defense mechanisms, respectively.

Novel application of normalized pointwise mutual information (NPMI) to mine biomedical literature for gene sets associated with disease: use case in breast carcinogenesis

Advances in technology within biomedical sciences have led to an inundation of data across many fields, raising new challenges in how best to integrate and analyze these resources. For example, rapid chemical screening programs like the US Environmental Protection Agency's ToxCast and the collaborative effort, Tox21, have produced massive amounts of information on putative chemical mechanisms where assay targets are identified as genes; however, systematically linking these hypothesized mechanisms with in vivo toxicity endpoints like disease outcomes remains problematic. Herein we present a novel use of normalized pointwise mutual information (NPMI) to mine biomedical literature for gene associations with biological concepts as represented by Medical Subject Headings (MeSH terms) in PubMed. Resources that tag genes to articles were integrated, then cross-species orthologs were identified using UniRef50 clusters. MeSH term frequency was normalized to reflect the MeSH tree structure, and then the resulting GeneID-MeSH associations were ranked using NPMI. The resulting network, called Entity MeSH Co-occurrence Network (EMCON), is a scalable resource for the identification and ranking of genes for a given topic of interest. The utility of EMCON was evaluated with the use case of breast carcinogenesis. Topics relevant to breast carcinogenesis were used to query EMCON and retrieve genes important to each topic. A breast cancer gene set was compiled through expert literature review (ELR) to assess performance of the search results. We found that the results from EMCON ranked the breast cancer genes from ELR higher than randomly selected genes with a recall of 0.98. Precision of the top five genes for selected topics was calculated as 0.87. This work demonstrates that EMCON can be used to link in vitro results to possible biological outcomes, thus aiding in generation of testable hypotheses for furthering understanding of biological function and the contribution of chemical exposures to disease.

Molecular studies of phenotype variation in canine RPGR-XLPRA1

PURPOSE

Canine X-linked progressive retinal atrophy 1 (XLPRA1) caused by a mutation in retinitis pigmentosa (RP) GTPase regulator (RPGR) exon ORF15 showed significant variability in disease onset in a colony of dogs that all inherited the same mutant X chromosome. Defective protein trafficking has been detected in XLPRA1 before any discernible degeneration of the photoreceptors. We hypothesized that the severity of the photoreceptor degeneration in affected dogs may be associated with defects in genes involved in ciliary trafficking. To this end, we examined six genes as potential disease modifiers. We also examined the expression levels of 24 genes involved in ciliary trafficking (seven), visual pathway (five), neuronal maintenance genes (six), and cellular stress response (six) to evaluate their possible involvement in early stages of the disease.

METHODS

Samples from a pedigree derived from a single XLPRA1-affected male dog outcrossed to unrelated healthy mix-bred or purebred females were used for immunohistochemistry (IHC), western blot, mutational and haplotype analysis, and gene expression (GE). Cell-specific markers were used to examine retinal remodeling in the disease. Single nucleotide polymorphisms (SNPs) spanning the entire RPGR interacting and protein trafficking genes (RAB8A, RPGRIP1L, CEP290, CC2D2A, DFNB31, and RAB11B) were genotyped in the pedigree. Quantitative real-time PCR (qRT-PCR) was used to examine the expression of a total of 24 genes, including the six genes listed.

RESULTS

Examination of cryosections from XLPRA1-affected animals of similar age (3-4 years) with different disease severity phenotype revealed mislocalization of opsins and upregulation of the Müller cell gliosis marker GFAP. Four to ten haplotypes per gene were identified in RAB8A, RPGRIP1L, CEP290, CC2D2A, DFNB31, and RAB11B for further assessment as potential genetic modifiers of XLPRA1. No correlation was found between the haplotypes and disease severity. During mutational analysis, several new variants, including a single intronic mutation in RAB8A and three mutations in exon 3 of DFNB31 were described (c.970G>A (V324I), c.978T>C (G326=), and c.985G>A (A329T)). Expression analysis of stress response genes in 16-week-old predisease XLPRA1 retinas revealed upregulation of GFAP but not HSPA5, DDIT3, HSPA4, HSP90B1, or HIF1A. Western blot analysis confirmed GFAP upregulation. In the same predisease group, no significant differences were found in the expression of 18 selected genes (RHO, OPN1LW, OPN1MW, RLBP1, RPGRORF15, RAB8A, RPGRIP1L, CEP290, CC2D2A, DFNB31, RAB11B, CRX, RCVRN, PVALB, CALB1, FGFR1, NTRK2, and NTRK3) involved in neuronal function.

CONCLUSIONS

Lack of association between haplotypes of RAB8A, RPGRIP1L, CEP290, CC2D2A, DFNB31, and RAB11B and the disease phenotype suggests that these genes are not genetic modifiers of XLPRA1. Upregulation of GFAP, an established indicator of the Müller cell gliosis, manifests as an important early feature of the disease.

Special issue on bio-ontologies and phenotypes

The bio-ontologies and phenotypes special issue includes eight papers selected from the 11 papers presented at the Bio-Ontologies SIG (Special Interest Group) and the Phenotype Day at ISMB (Intelligent Systems for Molecular Biology) conference in Boston in 2014. The selected papers span a wide range of topics including the automated re-use and update of ontologies, quality assessment of ontological resources, and the systematic description of phenotype variation, driven by manual, semi- and fully automatic means.

The digital revolution in phenotyping

Phenotypes have gained increased notoriety in the clinical and biological domain owing to their application in numerous areas such as the discovery of disease genes and drug targets, phylogenetics and pharmacogenomics. Phenotypes, defined as observable characteristics of organisms, can be seen as one of the bridges that lead to a translation of experimental findings into clinical applications and thereby support 'bench to bedside' efforts. However, to build this translational bridge, a common and universal understanding of phenotypes is required that goes beyond domain-specific definitions. To achieve this ambitious goal, a digital revolution is ongoing that enables the encoding of data in computer-readable formats and the data storage in specialized repositories, ready for integration, enabling translational research. While phenome research is an ongoing endeavor, the true potential hidden in the currently available data still needs to be unlocked, offering exciting opportunities for the forthcoming years. Here, we provide insights into the state-of-the-art in digital phenotyping, by means of representing, acquiring and analyzing phenotype data. In addition, we provide visions of this field for future research work that could enable better applications of phenotype data.