Max Kotlyar

Comparative systems biology of human and mouse as a tool to guide the modeling of human placental pathology

Placental abnormalities are associated with two of the most common and serious complications of human pregnancy, maternal preeclampsia (PE) and fetal intrauterine growth restriction (IUGR), each disorder affecting ∼5% of all pregnancies. An important question for the use of the mouse as a model for studying human disease is the degree of functional conservation of genetic control pathways from human to mouse. The human and mouse placenta show structural similarities, but there have been no systematic attempts to assess their molecular similarities or differences. We collected protein and mRNA expression data through shot‐gun proteomics and microarray expression analysis of the highly vascular exchange region, microdissected from the human and mouse near‐term placenta. Over 7000 ortholog genes were detected with 70% co‐expressed in both species. Close to 90% agreement was found between our human proteomic results and 1649 genes assayed by immunohistochemistry for expression in the human placenta in the Human Protein Atlas. Interestingly, over 80% of genes known to cause placental phenotypes in mouse are co‐expressed in human. Several of these phenotype‐associated proteins form a tight protein–protein interaction network involving 15 known and 34 novel candidate proteins also likely important in placental structure and/or function. The entire data are available as a web‐accessible database to guide the informed development of mouse models to study human disease.

The mammalian-membrane two-hybrid assay (MaMTH) for probing membrane-protein interactions in human cells

Cell signaling, one of key processes in both normal cellular function and disease, is coordinated by numerous interactions between membrane proteins that change in response to stimuli. We present a split ubiquitin–based method for detection of integral membrane protein-protein interactions (PPIs) in human cells, termed mammalian-membrane two-hybrid assay (MaMTH). We show that this technology detects stimulus (hormone or agonist)-dependent and phosphorylation-dependent PPIs. MaMTH can detect changes in PPIs conferred by mutations such as those in oncogenic ErbB receptor variants or by treatment with drugs such as the tyrosine kinase inhibitor erlotinib. Using MaMTH as a screening assay, we identified CRKII as an interactor of oncogenic EGFR(L858R) and showed that CRKII promotes persistent activation of aberrant signaling in non–small cell lung cancer cells. MaMTH is a powerful tool for investigating the dynamic interactomes of human integral membrane proteins.

Fundamentals of protein interaction network mapping

Studying protein interaction networks of all proteins in an organism (“interactomes”) remains one of the major challenges in modern biomedicine. Such information is crucial to understanding cellular pathways and developing effective therapies for the treatment of human diseases. Over the past two decades, diverse biochemical, genetic, and cell biological methods have been developed to map interactomes. In this review, we highlight basic principles of interactome mapping. Specifically, we discuss the strengths and weaknesses of individual assays, how to select a method appropriate for the problem being studied, and provide general guidelines for carrying out the necessary follow‐up analyses. In addition, we discuss computational methods to predict, map, and visualize interactomes, and provide a summary of some of the most important interactome resources. We hope that this review serves as both a useful overview of the field and a guide to help more scientists actively employ these powerful approaches in their research.

Integrated interactions database: tissue-specific view of the human and model organism interactomes

IID (Integrated Interactions Database) is the first database providing tissue-specific protein–protein interactions (PPIs) for model organisms and human. IID covers six species (S. cerevisiae (yeast), C. elegans (worm), D. melonogaster (fly), R. norvegicus (rat), M. musculus (mouse) and H. sapiens (human)) and up to 30 tissues per species. Users query IID by providing a set of proteins or PPIs from any of these organisms, and specifying species and tissues where IID should search for interactions. If query proteins are not from the selected species, IID enables searches across species and tissues automatically by using their orthologs; for example, retrieving interactions in a given tissue, conserved in human and mouse. Interaction data in IID comprises three types of PPI networks: experimentally detected PPIs from major databases, orthologous PPIs and high-confidence computationally predicted PPIs. Interactions are assigned to tissues where their proteins pairs or encoding genes are expressed. IID is a major replacement of the I2D interaction database, with larger PPI networks (a total of 1,566,043 PPIs among 68,831 proteins), tissue annotations for interactions, and new query, analysis and data visualization capabilities. IID is available at http://ophid.utoronto.ca/iid.

In silico prediction of physical protein interactions and characterization of interactome orphans

Protein-protein interactions (PPIs) are useful for understanding signaling cascades, predicting protein function, associating proteins with disease and fathoming drug mechanism of action. Currently, only ∼10% of human PPIs may be known, and about one-third of human proteins have no known interactions. We introduce FpClass, a data mining–based method for proteome-wide PPI prediction. At an estimated false discovery rate of 60%, we predicted 250,498 PPIs among 10,531 human proteins; 10,647 PPIs involved 1,089 proteins without known interactions. We experimentally tested 233 high- and medium-confidence predictions and validated 137 interactions, including seven novel putative interactors of the tumor suppressor p53. Compared to previous PPI prediction methods, FpClass achieved better agreement with experimentally detected PPIs. We provide an online database of annotated PPI predictions (http://ophid.utoronto.ca/fpclass/) and the prediction software (http://www.cs.utoronto.ca/~juris/data/fpclass/).

Network-based characterization of drug-regulated genes, drug targets, and toxicity.

Proteins do not exert their effects in isolation of one another, but interact together in complex networks. In recent years, sophisticated methods have been developed to leverage protein-protein interaction (PPI) network structure to improve several stages of the drug discovery process. Network-based methods have been applied to predict drug targets, drug side effects, and new therapeutic indications. In this paper we have two aims. First, we review the past contributions of network approaches and methods to drug discovery, and discuss their limitations and possible future directions. Second, we show how past work can be generalized to gain a more complete understanding of how drugs perturb networks. Previous network-based characterizations of drug effects focused on the small number of known drug targets, i.e., direct binding partners of drugs. However, drugs affect many more genes than their targets - they can profoundly affect the cells transcriptome. For the first time, we use networks to characterize genes that are differentially regulated by drugs. We found that drug-regulated genes differed from drug targets in terms of functional annotations, cellular localizations, and topological properties. Drug targets mainly included receptors on the plasma membrane, down-regulated genes were largely in the nucleus and were enriched for DNA binding, and genes lacking drug relationships were enriched in the extracellular region. Network topology analysis indicated several significant graph properties, including high degree and betweenness for the drug targets and drug-regulated genes, though possibly due to network biases. Topological analysis also showed that proteins of down-regulated genes appear to be frequently involved in complexes. Analyzing network distances between regulated genes, we found that genes regulated by structurally similar drugs were significantly closer than genes regulated by dissimilar drugs. Finally, network centrality of a drugs differentially regulated genes correlated significantly with drug toxicity.

Inferring the functions of longevity genes with modular subnetwork biomarkers of Caenorhabditis elegans aging

A central goal of biogerontology is to identify robust gene-expression biomarkers of aging. Here we develop a method where the biomarkers are networks of genes selected based on age-dependent activity and a graph-theoretic property called modularity. Tested on Caenorhabditis elegans, our algorithm yields better biomarkers than previous methods - they are more conserved across studies and better predictors of age. We apply these modular biomarkers to assign novel aging-related functions to poorly characterized longevity genes.

Predicting Protein-Protein Interactions by Association Mining

Identifying protein-protein interactions is a key problem in molecular biology. Currently, interactions cannot be reliably predicted on a proteome-wide scale but direct and indirect evidence for interactions is increasingly available from high-throughput interaction detection methods, gene expression microarrays, and protein annotation projects. In this paper we propose an association mining approach to integrating these diverse types of evidence. We apply this approach to a number of datasets consisting of interacting and non-interacting protein pairs annotated with different types of evidence. We identify patterns that distinguish interacting and non-interacting protein pairs, and use these patterns to assign a confidence level to proposed interactions.

A Global Analysis of the Receptor Tyrosine Kinase-Protein Phosphatase Interactome

Receptor tyrosine kinases (RTKs) and protein phosphatases comprise protein families that play crucial roles in cell signaling. We used two protein-protein interaction (PPI) approaches, the membrane yeast two-hybrid (MYTH) and the mammalian membrane two-hybrid (MaMTH), to map the PPIs between human RTKs and phosphatases. The resulting RTK-phosphatase interactome reveals a considerable number of previously unidentified interactions and suggests specific roles for different phosphatase families. Additionally, the differential PPIs of some protein tyrosine phosphatases (PTPs) and their mutants suggest diverse mechanisms of these PTPs in the regulation of RTK signaling. We further found that PTPRH and PTPRB directly dephosphorylate EGFR and repress its downstream signaling. By contrast, PTPRA plays a dual role in EGFR signaling: besides facilitating EGFR dephosphorylation, it enhances downstream ERK signaling by activating SRC. This comprehensive RTK-phosphatase interactome study provides a broad and deep view of RTK signaling.

Prioritizing Therapeutics for Lung Cancer: An Integrative Meta-analysis of Cancer Gene Signatures and Chemogenomic Data

Repurposing FDA-approved drugs with the aid of gene signatures of disease can accelerate the development of new therapeutics. A major challenge to developing reliable drug predictions is heterogeneity. Different gene signatures of the same disease or drug treatment often show poor overlap across studies, as a consequence of both biological and technical variability, and this can affect the quality and reproducibility of computational drug predictions. Existing algorithms for signature-based drug repurposing use only individual signatures as input. But for many diseases, there are dozens of signatures in the public domain. Methods that exploit all available transcriptional knowledge on a disease should produce improved drug predictions. Here, we adapt an established meta-analysis framework to address the problem of drug repurposing using an ensemble of disease signatures. Our computational pipeline takes as input a collection of disease signatures, and outputs a list of drugs predicted to consistently reverse pathological gene changes. We apply our method to conduct the largest and most systematic repurposing study on lung cancer transcriptomes, using 21 signatures. We show that scaling up transcriptional knowledge significantly increases the reproducibility of top drug hits, from 44% to 78%. We extensively characterize drug hits in silico, demonstrating that they slow growth significantly in nine lung cancer cell lines from the NCI-60 collection, and identify CALM1 and PLA2G4A as promising drug targets for lung cancer. Our meta-analysis pipeline is general, and applicable to any disease context; it can be applied to improve the results of signature-based drug repurposing by leveraging the large number of disease signatures in the public domain.