Differentiation and visualization of diverse cellular phenotypic responses in primary high-content screening

High-throughput screening, based on subcellular imaging, has become a powerful tool in lead discovery. Through the generation of high-quality images, not only the specific target signal can be analyzed but also phenotypic changes of the whole cell are recorded. Yet analysis strategies for the exploration of high-content screening results, in a manner that is independent from predefined control phenotypes, are largely missing. The approach presented here is based on a well-established modeling technique, self-organizing maps (SOMs), which uses multiparametric results to group treatments that create similar morphological effects. This report describes a novel visualization of the SOM clustering by using an image of the cells from each node, with the most representative cell highlighted to deploy the phenotype described by each node. The approach has the potential to identify both expected hits and novel cellular phenotypes. Moreover, different chemotypes, which cause the same phenotypic effects, are identified, thus facilitating "scaffold hopping."     

A Network of Genes Antagonistic to the LIN-35 Retinoblastoma Protein of Caenorhabditis elegans

The Caenorhabditis elegans pRb ortholog, LIN-35, functions in a wide range of cellular and developmental processes. This includes a role of LIN-35 in nutrient utilization by the intestine, which it carries out redundantly with SLR-2, a zinc-finger protein. This and other redundant functions of LIN-35 were identified in genetic screens for mutations that display synthetic phenotypes in conjunction with loss of lin-35. To explore the intestinal role of LIN-35, we conducted a genome-wide RNA-interference-feeding screen for suppressors of lin-35; slr-2 early larval arrest. Of the 26 suppressors identified, 17 fall into three functional classes: (1) ribosome biogenesis genes, (2) mitochondrial prohibitins, and (3) chromatin regulators. Further characterization indicates that different categories of suppressors act through distinct molecular mechanisms. We also tested lin-35; slr-2 suppressors, as well as suppressors of the synthetic multivulval phenotype, to determine the spectrum of lin-35-synthetic phenotypes that could be suppressed following inhibition of these genes. We identified 19 genes, most of which are evolutionarily conserved, that can suppress multiple unrelated lin-35-synthetic phenotypes. Our study reveals a network of genes broadly antagonistic to LIN-35 as well as genes specific to the role of LIN-35 in intestinal and vulval development. Suppressors of multiple lin-35 phenotypes may be candidate targets for anticancer therapies. Moreover, screening for suppressors of phenotypically distinct synthetic interactions, which share a common altered gene, may prove to be a novel and effective approach for identifying genes whose activities are most directly relevant to the core functions of the shared gene.     

Targeting mitosis‐regulating genes in cisplatin‐sensitive and ‐resistant melanoma cells: A live‐cell RNAi screen displays differential nucleus‐derived phenotypes

Chemoresistance in malignant melanoma remains an unresolved clinical issue. In the search for novel molecular targets, a live‐cell high‐content RNAi screen based on gene expression data was performed in cisplatin‐sensitive and cisplatin‐resistant MeWo melanoma cells, Mel‐28 cells and a melanocyte cell line. Cells were exposed to 91 siRNAs and distinct nucleus‐derived phenotypes such as cell division, cell death and migration phenotypes were detected by time‐lapse microscopy over 60 h. Using this approach, cisplatin‐sensitive and cisplatin‐resistant melanoma cells were compared by automated image analysis and visual inspection. In cisplatin‐sensitive MeWo melanoma cells, 14 genes were identified that showed distinct phenotype abnormalities after exposure to gene‐specific siRNAs. In cisplatin‐resistant MeWo cells, five genes were detected. Nine genes were detected whose knock‐down led to differential nuclear phenotypes in cisplatin‐sensitive and ‐resistant cells. In Mel‐28 cells, nine genes were identified which induced nuclear phenotypes including all eight genes which were identified in cisplatin‐resistant MeWo cells. An analogous RNAi screen on melanocytes revealed no detectable phenotype abnormalities after RNAi. Pathway analysis showed in cisplatin‐sensitive MeWo cells and Mel‐28 cells an enrichment of at least three genes in major mitotic pathways. We hereby show that siRNA screening may help to identify tumor‐specific genes leading to phenotype abnormalities. These genes may serve as potential therapeutic targets in the treatment of melanoma.     

Coexisting/Coexpressing Genomic Libraries (CoGeL) identify interactions among distantly located genetic loci for developing complex microbial phenotypes

In engineering novel microbial strains for biotechnological applications, beyond a priori identifiable pathways to be engineered, it is becoming increasingly important to develop complex, ill-defined cellular phenotypes. One approach is to screen genomic or metagenomic libraries to identify genes imparting desirable phenotypes, such as tolerance to stressors or novel catabolic programs. Such libraries are limited by their inability to identify interactions among distant genetic loci. To solve this problem, we constructed plasmid- and fosmid-based Escherichia coli Coexisting/Coexpressing Genomic Libraries (CoGeLs). As a proof of principle, four sets of two genes of the l-lysine biosynthesis pathway distantly located on the E. coli chromosome were knocked out. Upon transformation of these auxotrophs with CoGeLs, cells growing without supplementation were found to harbor library inserts containing the knocked-out genes demonstrating the interaction between the two libraries. CoGeLs were also screened to identify genetic loci that work synergistically to create the considerably more complex acid-tolerance phenotype. CoGeL screening identified combination of genes known to enhance acid tolerance (gadBC operon and adiC), but also identified the novel combination of arcZ and recA that greatly enhanced acid tolerance by 9000-fold. arcZ is a small RNA that we show increases pH tolerance alone and together with recA.     

Genomics and genome-wide association studies: an integrative approach to expression QTL mapping

Expression QTL mapping by integrating genome-wide gene expression and genotype data is a promising approach to identifying functional genetic variation, but is hampered by the large number of multiple comparisons inherent in such studies. A novel approach to addressing multiple testing problems in genome-wide family-based association studies is screening candidate markers using heritability or conditional power. We apply these methods in settings in which microarray gene expression data are used as phenotypes, screening for SNPs near the expressed genes. We perform association analyses for phenotypes using a univariate approach. We also perform simulations on trios with large numbers of causal SNPs to determine the optimal number of markers to use in a screen. We demonstrate that our family-based screening approach performs well in the analysis of integrative genomic datasets and that screening using either heritability or conditional power produces similar, though not identical, results.     

Epigenome-wide scans identify differentially methylated regions for age and age-related phenotypes in a healthy ageing population

Age-related changes in DNA methylation have been implicated in cellular senescence and longevity, yet the causes and functional consequences of these variants remain unclear. To elucidate the role of age-related epigenetic changes in healthy ageing and potential longevity, we tested for association between whole-blood DNA methylation patterns in 172 female twins aged 32 to 80 with age and age-related phenotypes. Twin-based DNA methylation levels at 26,690 CpG-sites showed evidence for mean genome-wide heritability of 18%, which was supported by the identification of 1,537 CpG-sites with methylation QTLs in cis at FDR 5%. We performed genome-wide analyses to discover differentially methylated regions (DMRs) for sixteen age-related phenotypes (ap-DMRs) and chronological age (a-DMRs). Epigenome-wide association scans (EWAS) identified age-related phenotype DMRs (ap-DMRs) associated with LDL (STAT5A), lung function (WT1), and maternal longevity (ARL4A, TBX20). In contrast, EWAS for chronological age identified hundreds of predominantly hyper-methylated age DMRs (490 a-DMRs at FDR 5%), of which only one (TBX20) was also associated with an age-related phenotype. Therefore, the majority of age-related changes in DNA methylation are not associated with phenotypic measures of healthy ageing in later life. We replicated a large proportion of a-DMRs in a sample of 44 younger adult MZ twins aged 20 to 61, suggesting that a-DMRs may initiate at an earlier age. We next explored potential genetic and environmental mechanisms underlying a-DMRs and ap-DMRs. Genome-wide overlap across cis-meQTLs, genotype-phenotype associations, and EWAS ap-DMRs identified CpG-sites that had cis-meQTLs with evidence for genotype-phenotype association, where the CpG-site was also an ap-DMR for the same phenotype. Monozygotic twin methylation difference analyses identified one potential environmentally-mediated ap-DMR associated with total cholesterol and LDL (CSMD1). Our results suggest that in a small set of genes DNA methylation may be a candidate mechanism of mediating not only environmental, but also genetic effects on age-related phenotypes.     

ENU mutagenesis in mice identifies candidate genes for hypogonadism

Genome-wide mutagenesis was performed in mice to identify candidate genes for male infertility, for which the predominant causes remain idiopathic. Mice were mutagenized using N-ethyl-N-nitrosourea (ENU), bred, and screened for phenotypes associated with the male urogenital system. Fifteen heritable lines were isolated and chromosomal loci were assigned using low-density genome-wide SNP arrays. Ten of the 15 lines were pursued further using higher-resolution SNP analysis to narrow the candidate gene regions. Exon sequencing of candidate genes identified mutations in mice with cystic kidneys (Bicc1), cryptorchidism (Rxfp2), restricted germ cell deficiency (Plk4), and severe germ cell deficiency (Prdm9). In two other lines with severe hypogonadism, candidate sequencing failed to identify mutations, suggesting defects in genes with previously undocumented roles in gonadal function. These genomic intervals were sequenced in their entirety and a candidate mutation was identified in SnrpE in one of the two lines. The line harboring the SnrpE variant retains substantial spermatogenesis despite small testis size, an unusual phenotype. In addition to the reproductive defects, heritable phenotypes were observed in mice with ataxia (Myo5a), tremors (Pmp22), growth retardation (unknown gene), and hydrocephalus (unknown gene). These results demonstrate that the ENU screen is an effective tool for identifying potential causes of male infertility.     

Genome-Wide Pathway Association Studies of Multiple Correlated Quantitative Phenotypes Using Principle Component Analyses

Genome-wide pathway association studies provide novel insight into the biological mechanism underlying complex diseases. Current pathway association studies primarily focus on single important disease phenotype, which is sometimes insufficient to characterize the clinical manifestations of complex diseases. We present a multi-phenotypes pathway association study(MPPAS) approach using principle component analysis(PCA). In our approach, PCA is first applied to multiple correlated quantitative phenotypes for extracting a set of orthogonal phenotypic components. The extracted phenotypic components are then used for pathway association analysis instead of original quantitative phenotypes. Four statistics were proposed for PCA-based MPPAS in this study. Simulations using the real data from the HapMap project were conducted to evaluate the power and type I error rates of PCA-based MPPAS under various scenarios considering sample sizes, additive and interactive genetic effects. A real genome-wide association study data set of bone mineral density (BMD) at hip and spine were also analyzed by PCA-based MPPAS. Simulation studies illustrated the performance of PCA-based MPPAS for identifying the causal pathways underlying complex diseases. Genome-wide MPPAS of BMD detected associations between BMD and KENNY_CTNNB1_TARGETS_UP as well as LONGEVITYPATHWAY pathways in this study. We aim to provide a applicable MPPAS approach, which may help to gain deep understanding the potential biological mechanism of association results for complex diseases.     

Cellular phenotype recognition for high-content RNA interference genome-wide screening

Genome-wide, cell-based screens using high-content screening (HCS) techniques and automated fluorescence microscopy generate thousands of high-content images that contain an enormous wealth of cell biological information. Such screens are key to the analysis of basic cell biological principles, such as control of cell cycle and cell morphology. However, these screens will ultimately only shed light on human disease mechanisms and potential cures if the analysis can keep up with the generation of data. A fundamental step toward automated analysis of high-content screening is to construct a robust platform for automatic cellular phenotype identification. The authors present a framework, consisting of microscopic image segmentation and analysis components, for automatic recognition of cellular phenotypes in the context of the Rho family of small GTPases. To implicate genes involved in Rac signaling, RNA interference (RNAi) was used to perturb gene functions, and the corresponding cellular phenotypes were analyzed for changes. The data used in the experiments are high-content, 3-channel, fluorescence microscopy images of Drosophila Kc167 cultured cells stained with markers that allow visualization of DNA, polymerized actin filaments, and the constitutively activated Rho protein Rac(V12). The performance of this approach was tested using a cellular database that contained more than 1000 samples of 3 predefined cellular phenotypes, and the generalization error was estimated using a cross-validation technique. Moreover, the authors applied this approach to analyze the whole high-content fluorescence images of Drosophila cells for further HCS-based gene function analysis.     

Functional genomics complements quantitative genetics in identifying disease-gene associations

An ultimate goal of genetic research is to understand the connection between genotype and phenotype in order to improve the diagnosis and treatment of diseases. The quantitative genetics field has developed a suite of statistical methods to associate genetic loci with diseases and phenotypes, including quantitative trait loci (QTL) linkage mapping and genome-wide association studies (GWAS). However, each of these approaches have technical and biological shortcomings. For example, the amount of heritable variation explained by GWAS is often surprisingly small and the resolution of many QTL linkage mapping studies is poor. The predictive power and interpretation of QTL and GWAS results are consequently limited. In this study, we propose a complementary approach to quantitative genetics by interrogating the vast amount of high-throughput genomic data in model organisms to functionally associate genes with phenotypes and diseases. Our algorithm combines the genome-wide functional relationship network for the laboratory mouse and a state-of-the-art machine learning method. We demonstrate the superior accuracy of this algorithm through predicting genes associated with each of 1157 diverse phenotype ontology terms. Comparison between our prediction results and a meta-analysis of quantitative genetic studies reveals both overlapping candidates and distinct, accurate predictions uniquely identified by our approach. Focusing on bone mineral density (BMD), a phenotype related to osteoporotic fracture, we experimentally validated two of our novel predictions (not observed in any previous GWAS/QTL studies) and found significant bone density defects for both Timp2 and Abcg8 deficient mice. Our results suggest that the integration of functional genomics data into networks, which itself is informative of protein function and interactions, can successfully be utilized as a complementary approach to quantitative genetics to predict disease risks. All supplementary material is available at http://cbfg.jax.org/phenotype.     

B-cell non-Hodgkin lymphoma: Selective vulnerability to PIKFYVE inhibition

We identified the PIKFYVE inhibitor apilimod as a potent and selective cytotoxic agent against B-cell non-Hodgkin lymphoma (B-NHL). Our data robustly establish PIKFYVE as the target through which apilimod kills B-NHL cells and show that apilimod-induced death in B-NHL is mediated by broad disruption of lysosome homeostasis characterized by lysosomal swelling, TFEB nuclear translocation, impaired maturation of lysosomal enzymes and incomplete autophagosome clearance. Furthermore, through genome-wide CRISPR knockout screening, we identified specific lysosomal genes (TFEB, CLCN7, OSTM1 and SNX10) as critical determinants of apilimod-induced cytotoxicity. Together these data highlight disruption of lysosome homeostasis through PIKFYVE inhibition as a novel anticancer mechanism in B-NHL and potentially other cancers.     

Phenome-Wide Association Study (PheWAS) for Detection of Pleiotropy within the Population Architecture using Genomics and Epidemiology (PAGE) Network

Using a phenome-wide association study (PheWAS) approach, we comprehensively tested genetic variants for association with phenotypes available for 70,061 study participants in the Population Architecture using Genomics and Epidemiology (PAGE) network. Our aim was to better characterize the genetic architecture of complex traits and identify novel pleiotropic relationships. This PheWAS drew on five population-based studies representing four major racial/ethnic groups (European Americans (EA), African Americans (AA), Hispanics/Mexican-Americans, and Asian/Pacific Islanders) in PAGE, each site with measurements for multiple traits, associated laboratory measures, and intermediate biomarkers. A total of 83 single nucleotide polymorphisms (SNPs) identified by genome-wide association studies (GWAS) were genotyped across two or more PAGE study sites. Comprehensive tests of association, stratified by race/ethnicity, were performed, encompassing 4,706 phenotypes mapped to 105 phenotype-classes, and association results were compared across study sites. A total of 111 PheWAS results had significant associations for two or more PAGE study sites with consistent direction of effect with a significance threshold of p<0.01 for the same racial/ethnic group, SNP, and phenotype-class. Among results identified for SNPs previously associated with phenotypes such as lipid traits, type 2 diabetes, and body mass index, 52 replicated previously published genotype–phenotype associations, 26 represented phenotypes closely related to previously known genotype–phenotype associations, and 33 represented potentially novel genotype–phenotype associations with pleiotropic effects. The majority of the potentially novel results were for single PheWAS phenotype-classes, for example, for CDKN2A/B rs1333049 (previously associated with type 2 diabetes in EA) a PheWAS association was identified for hemoglobin levels in AA. Of note, however, GALNT2 rs2144300 (previously associated with high-density lipoprotein cholesterol levels in EA) had multiple potentially novel PheWAS associations, with hypertension related phenotypes in AA and with serum calcium levels and coronary artery disease phenotypes in EA. PheWAS identifies associations for hypothesis generation and exploration of the genetic architecture of complex traits.     

Large Scale Identification of Genes Involved in Cell Surface Biosynthesis and Architecture in Saccharomyces Cerevisiae

The sequenced yeast genome offers a unique resource for the analysis of eukaryotic cell function and enables genome-wide screens for genes involved in cellular processes. We have identified genes involved in cell surface assembly by screening transposon-mutagenized cells for altered sensitivity to calcofluor white, followed by supplementary screens to further characterize mutant phenotypes. The mutated genes were directly retrieved from genomic DNA and then matched uniquely to a gene in the yeast genome database. Eighty-two genes with apparent perturbation of the cell surface were identified, with mutations in 65 of them displaying at least one further cell surface phenotype in addition to their modified sensitivity to calcofluor. Fifty of these genes were previously known, 17 encoded proteins whose function could be anticipated through sequence homology or previously recognized phenotypes and 15 genes had no previously known phenotype.     

High-throughput screening of cellular features using high-resolution light-microscopy; application for profiling drug effects on cell adhesion

High-resolution light-microscopy and high-throughput screening are two essential methodologies for characterizing cellular phenotypes. Optimally combining these methodologies in cell-based screening to test detailed molecular and cellular responses to multiple perturbations constitutes a major challenge. Here we describe the development and application of a screening microscope platform that automatically acquires and interprets sub-micron resolution images at fast rates. The analysis pipeline is based on the quantification of multiple subcellular features and statistical comparisons of their distributions in treated vs. control cells. Using this platform, we screened 2200 natural extracts for their effects on the fine structure and organization of focal adhesions. This screen identified 15 effective extracts whose fractionation and characterization were further analyzed using the same microscope system. The significance of combining resolution, throughput and multi-parametric analyses for biomedical research and drug discovery is discussed.     

Towards a mutant map of the mouse – new models of neurological, behavioural, deafness, bone, renal and blood disorders

With the completion of the first draft of the human genome sequence, the next major challenge is assigning function to genes. One approach is genome-wide random chemical mutagenesis, followed by screening for mutant phenotypes of interest and subsequent mapping and identification of the mutated genes in question. We (a consortium made up of GlaxoSmithKline, the MRC Mammalian Genetics Unit and Mouse Genome Centre, Harwell, Imperial College, London, and the Royal London Hospital) have used ENU mutagenesis in the mouse for the rapid generation of novel mutant phenotypes for use as animal models of human disease and for gene function assignment (Nolan et al., 2000). As of 2003, 35,000 mice have been produced to date in a genome-wide screen for dominant mutations and screened using a variety of screening protocols. Nearly 200 mutants have been confirmed as heritable and added to the mouse mutant catalogue and, overall, we can extrapolate that we have recovered over 700 mutants from the screening programme. For further information on the project and details of the data, see http://www.mgu.har.mrc.ac.uk/mutabase.     

Integrative analysis of the mitochondrial proteome in yeast

In this study yeast mitochondria were used as a model system to apply, evaluate, and integrate different genomic approaches to define the proteins of an organelle. Liquid chromatography mass spectrometry applied to purified mitochondria identified 546 proteins. By expression analysis and comparison to other proteome studies, we demonstrate that the proteomic approach identifies primarily highly abundant proteins. By expanding our evaluation to other types of genomic approaches, including systematic deletion phenotype screening, expression profiling, subcellular localization studies, protein interaction analyses, and computational predictions, we show that an integration of approaches moves beyond the limitations of any single approach. We report the success of each approach by benchmarking it against a reference set of known mitochondrial proteins, and predict approximately 700 proteins associated with the mitochondrial organelle from the integration of 22 datasets. We show that a combination of complementary approaches like deletion phenotype screening and mass spectrometry can identify over 75% of the known mitochondrial proteome. These findings have implications for choosing optimal genome-wide approaches for the study of other cellular systems, including organelles and pathways in various species. Furthermore, our systematic identification of genes involved in mitochondrial function and biogenesis in yeast expands the candidate genes available for mapping Mendelian and complex mitochondrial disorders in humans.