CIFAR launches the Genetic Networks program under the direction of
|Genetics; biochemistry and molecular biology; computational biology and bioinformatics; cell, evolutionary and systems biology; pathology; immunology; biotechnology|
Our genes determine all sorts of things about us, including our predisposition to many diseases. But there are only a small number that we understand well enough to know how a genetic change is propagated through the network of molecular and cellular interactions to ultimately yield a human disease.
If we can get a better understanding of how these interactions work, and of how the effects of genetic perturbation propagate through the system, we will be better able to identify the root causes of many complex genetic diseases like autism, asthma, Alzheimer’s and many cancers.
The Genetic Networks program at CIFAR is charting genetic and molecular interactions to understand how biological systems work and how they fail. The aim is to map complete networks of genetic and molecular interactions, and to use them to decipher the rules of how genes influence one another, what environmental factors alter those interactions, and how the impact of genetic change is propagated through biological systems.
The Genetic Networks Program brings together geneticists with molecular and computational biologists, who work on a wide variety of species, from yeasts, fruit flies, worms and mice to humans. Because evolution has preserved many genes and genetic interactions over millions of years – for instance, humans and yeast share about 40 percent of their genes – research into the simpler organisms can shed light on humans.
This broadly integrated investigation of genetic networks is unique in the world and has helped to recruit eminent researchers and give them the opportunity to connect and collaborate with leading researchers from other countries. Their work is uncovering how networks of molecular interactions mediate the effects of combinations of genetic perturbations, yielding maps from personal genomes to states of health and disease.
The multidisciplinary Genetic Networks program has benefitted greatly from the cutting-edge work of other CIFAR programs. For example, program researchers have used deep learning — a machine learning technique pioneered by CIFAR fellows in the Learning in Machines & Brains program (formerly known as Neural Computation & Adaptive Perception) — to better predict how genetic change affects human disease.
A recent explosion in gene sequencing technologies has identified a massive catalogue of genes. However, we still need to understand how the information that’s encoded in genes, and the vast number of interactions between them, translates into the specific traits and characteristics that are unique to every one of us. By developing methods that predict the direct outcomes, including disease, from an individual’s complex genetic makeup, the program is helping to lay the groundwork that will lead to medicine that is personalized according to each individual’s genome.
Researchers have made progress deciphering the causes of diseases from single genes, like cystic fibrosis and Huntington’s disease. But what’s needed now is a better understanding of how multiple genes mutate, combine, and alter networks of molecular and cellular interactions to cause more complex diseases. Even though there are hundreds of human gene variants, many lead to disease only in specific combinations.
Program members, individually and collaboratively, have made major progress in mapping genetic interactions. Partial network maps now exist for several model organisms, including two different yeast species, and the nematode worm. Work on genetic network mapping in cultured human cells is also underway.
Program members have tested how genetic interactions are conserved between species – a key question for applying knowledge of model systems to more complex organisms, including humans. They found that two distantly related types of yeast separated by a billion years of evolution share three-quarters of their genes and one-third of their genetic interactions. This finding suggests a core genetic network might be common to many even more distant species.
New genetic interaction maps and conservation studies have become starting points for program researchers to address broad evolutionary questions and better understand the genetic basis of many human diseases.
Fellow Chad Myers and Senior Fellow Charles Boone used their analysis of yeast genes to identify two potential targets in humans for a new form of cancer therapy. The team was looking for “synthetic lethal” interactions—two individual mutations, neither of which will kill a cell by itself, but which when present together are lethal. The idea was to find a mutation that causes cancer in humans, and to find one in another gene that is lethal to the cell when combined with the cancer-causing mutation. The identified mutations could allow researchers to design cancer therapies, including drugs, that manipulate expression of the second gene, killing cancer cells but leaving healthy cells unaffected.
Program Director and Senior Fellow Frederick Roth co-led an international research team that carried out the first full-scale mapping of direct physical interactions between human proteins. The reference map of the human ‘interactome’ describes about 14,000 direct interactions between proteins, which is seven times more than any previous map of its kind has uncovered. This map pointed to dozens of new genes that could be involved in cancer.
Senior Fellow and former Program Co-Director Brenda Andrews and Senior Fellow Charles Boone have produced the first global map of protein locations within a eukaryotic cell. This map enhances understanding of protein function – which is still hard to predict just from looking at the DNA sequence – as well as complex protein interactions that underpin disease. After developing a map of protein localization in normal cells, they watched how it changes when cells begin to divide, when they carry a genetic mutation, or when they are under environmental stress.
A team of researchers led by Senior Fellow Eric Shoubridge has discovered a new genetic defect that has been linked to a spectrum of rare and severe neurological disorders. The defect compromises the function of the mitochondrion, which generates energy and ultimately keeps cells alive. Understanding which genes are implicated in these disorders could offer parents options for making reproductive decisions, such as genetically testing embryos or donor cells.
Senior Fellow Brendan Frey has combined the latest in whole genome sequencing with computational techniques to develop the “human splicing code”, an entirely new approach to identifying the genetic determinants of disease. Frey and his colleagues have applied this new approach to identify mutations involved in cancer and neurological disorders.
Their genome-wide analysis has revealed tens of thousands of variants that alter RNA splicing and are embedded in a wide range of known diseases including spinal muscular atrophy, certain cancers, and autism spectrum disorder. The disruption of splicing, a critical step in gene expression, is known to contribute to disease. The study also reveals that the network of genetic interactions in humans covers a much broader area of the genome than some past research has suggested. In 2015, Frey founded a company called Deep Genomics, devoted to commercializing the technology.
Senior Fellow Steve Scherer was one of the first in the world to find that, while autism has genetic roots, people with autism don’t have identical mutations in one or a few genes, as was previously assumed. A collaboration between Scherer and Frey has resulted in a breakthrough in diagnosing autism at a younger age that allows patients to receive therapies earlier. Their study reveals a unifying set of characteristics in the DNA can be woven into a “genetic formula” that helps us calculate which genetic mutations have the highest probability of causing autism, and equally important, which alterations do not have a role. Around 100 genes have been linked to autism so far.
Much of the recent work conducted by program members, particularly that of Scherer and Frey, reflects a new focus on how a better understanding of genetic interactions can lead the way to personalized medicine, in which drugs and therapies can be specifically targeted using the genetic information of the individual.
University of Toronto
University of Washington
University of Toronto
University of Toronto
University of British Columbia
University of Toronto
University of British Columbia
University of Toronto
University of Minnesota
Hospital for Sick Children
CIFAR Senior Fellow Stephen Scherer (The Hospital for Sick Children)
CIFAR Senior Fellow Janet Rossant (The Hospital for Sick Children)
CIFAR Senior Fellows Brendan Frey, Brenda Andrews, Charles Boone and
CIFAR Senior Fellow Donald Moerman (University of British Columbia) leads
Stephen Scherer (Hospital for Sick Children, University of Toronto) in
CIFAR Senior Fellow Timothy Hughes (University of Toronto) and collaborators
CIFAR Senior Fellows Charles Boone and Brenda Andrews’ (both University
CIFAR Senior Fellow Stephen Scherer (The Hospital for Sick Children)
CIFAR Senior Fellow Timothy Hughes (University of Toronto) sequences the
CIFAR Senior Fellow Philip Hieter (University of British Columbia) uses
CIFAR Senior Fellow Don Moerman (University of British Columbia), in
CIFAR Senior Fellow Philip Hieter’s group (University of British Columbia)
CIFAR launches the Genetic Networks program under the direction of Brenda Andrews and Associate Director Frederick P. Roth (both University of Toronto). The program is devoted to discovering how genes interact with one another, which could identify the root causes of many complex genetic diseases and lead to new treatments and preventive measures.
In the first systematic study of its kind, CIFAR Senior Fellows Charles Boone, Brenda Andrews and Timothy Hughes (all University of Toronto) create a synthetic genetic network that examines over 500 interactions between more than half of the genes critical for yeast survival. They identify a previously unknown essential gene, PGA1, which is important for functions of the endoplasmic reticulum, a type of organelle that folds and transports proteins. The study also indicates that even though only approximately 18 per cent of yeast genes are essential for life, most yeast genetic interactions involve at least one essential gene.
CIFAR Senior Fellow Stephen Scherer (The Hospital for Sick Children) and colleagues construct a map of copy number variation — structural variations in the human genome that can cause cells to copy sections of DNA differently, resulting in deletions or duplications. These variations are linked to many diseases, and they are important — and sometimes positive — factors in genetic diversity and evolution. The researchers build the haplotype, map, dubbed HapMap, from the genomes of 270 people with European, African and Asian ancestry. They find 1,447 regions with copy number variations, encompassing hundreds of genes, duplications and locations of the genome that are involved with disease. The analysis shows about two to three per cent of genes vary between populations, and the same gene families are involved. Potential implications are many, but one is the impact that these discoveries will have on the concept of personalized medicine.
CIFAR Senior Fellow Janet Rossant (The Hospital for Sick Children) and collaborators find that signalling by receptors on the cell surface in embryos, known as receptor tyrosine kinases, is an important factor in determining what a cell becomes during development. This paper is highly cited because it implies that stem cells are capable of becoming one of many cell types as they develop — they are pluripotent.
CIFAR Senior Fellows Brendan Frey, Brenda Andrews, Charles Boone and Timothy Hughes (all University of Toronto) propose a new strategy for uncovering the genes and pathways regulated by transcription factors, which are proteins that bind to DNA sequences and regulate various functions such as growth. Even in well-studied organisms such as yeast, scientists don’t know the function or DNA-binding site for about half of the transcription factors that bind to DNA. The researchers use DNA microarray or “gene chip” technology — microscopic DNA spots of sequences — to study transcription factors that cause yeast to grow when they overexpress. They identify for the first time a binding sequence for a transcription factor that helps prevent yeast cells from growing into elongated chains when they are deprived of essential nutrients such as carbon or nitrogen. This abnormality, called pseudohyphal growth, can lead to invasive growth. The finding shows promise for the use of physical interactions and genome sequence to predict genetic interactions.
CIFAR Senior Fellows Brenda Andrews, Charles Boone, Brendan Frey and Timothy Hughes (all University of Toronto) test a collection of yeast mutants for sensitivity to 82 drugs and natural product extracts and compile profiles of chemical-genetic interactions. The cellular response to drugs is quite complex and influenced by our individual genetic make-up, making a greater understanding of interactions and side-effects important for drug discovery and development.
CIFAR Senior Fellow Donald Moerman (University of British Columbia) leads an international effort to develop techniques and tools for systematically mapping mutations in the nematode worm. This simple worm is a nice model for asking if the genetic interactions discovered in budding yeast also take place in a multi-celled organism. Moerman’s group adapts a technique that allows researchers to get a snap shot of the whole genome on a microarray chip. Using this platform, his group is later able to show that several wild-type strains have as much as three per cent of the genes in their genomes deleted compared to the standard model strain that most labs use. Most interesting about the result is that the deleted genes are all related to how an animal senses or deals with its environment, including genes that code for cell sensors or genes involved in innate immunity.
M.R. Jones et al., “Oligonucleotide Array Comparative Genomic Hybridization (oaCGH) based characterization of genetic deficiencies as an aid to gene mapping in Caenorhabditis elegans,” BMC Genomics 8, 402 (Nov. 2007) doi:10.1186/1471-2164-8-402
Stephen Scherer (Hospital for Sick Children, University of Toronto) in CIFAR’s Genetic Networks program makes headlines by collaborating with U.S. biologist Craig Venter to publish a sequenced genome, the first entire DNA makeup of an individual.
A consortium of international scientists including CIFAR Senior Fellow Stephen Scherer (The Hospital for Sick Children) and colleagues use the HapMap to complete the largest genome scan ever attempted in autism research. They find that gene copy number variation can contribute to autism susceptibility. Better understanding of how genes interact to determine the spectrum of autistic disorders in children promises better diagnostics and, ultimately, new therapeutics.
CIFAR Senior Fellow Frederick P. Roth (University of Toronto) and his team evaluate four mathematically distinct definitions of genetic interaction (Product, Additive, Log, and Min). They find the choice among alternative definitions has profound consequences. Although 52 per cent of known synergistic genetic interactions in baking yeast were inferred according to the Min definition, the study finds that both Product and Log definitions are better than Min for identifying functional relationships. The research shows that the Additive and Log definitions, each commonly used in population genetics, lead to differing conclusions related to the selective advantages of sexual reproduction.
CIFAR fellows Chad Myers, Olga Troyanskaya (both Princeton University), Timothy Hughes and Frederick P. Roth (both University of Toronto) assemble a collection of data for 21,603 mouse genes and generate predictions about their function. This analysis infers functions for 76 per cent of mouse genes, including 5,000 uncharacterized genes.
CIFAR Senior Fellow Timothy Hughes (University of Toronto) and collaborators use state-of-the-art techniques to catalogue the sequence information of DNA and RNA binding proteins in organisms ranging from yeast to humans. This catalogue forms an essential foundation of our effort to understand transcriptional networks, or the mechanisms that control gene expression, such as interactions between proteins and DNA. The group also invents and applies a new technique that allows us to systematically map the pathways of proteins that control the rate at which DNA is transcribed into RNA in budding yeast. This marks an important step in understanding how all of the parts controlling cells work and how they work together. Genetic Networks fellows argue this is necessary too change medicine, agriculture and biotechnology into disciplines driven by basic knowledge.
G. Badis et al., “Diversity and complexity in DNA recognition by transcription factors,” Science 324 (2009): 1720-1723 doi: 10.1126/science.1162327.
G. Badis et al., “A library of yeast transcription factor motifs reveals a widespread function for Rsc3 in targeting nucleosome exclusion at promoters,” Molecular Cell 32 (Dec. 2008): 878-887 doi: http://dx.doi.org/10.1016/j.molcel.2008.11.020
N. Kaplan et al., “The DNA-encoded nucleosome organization of a eukaryotic genome,” Nature 458 (2009): 362-366 doi:10.1038/nature07667
CIFAR Senior Fellows Charles Boone and Brenda Andrews’ (both University of Toronto) data suggest that approximately 30 per cent of genetic interactions might be the same between distantly related yeast species. CIFAR Senior Fellow Philip Hieter's lab (University of British Columbia) provides direct evidence of this by repeating a set of genetic interactions that disturb the chromosome of yeast in the more complex, multi-cellular roundworm. This study proves that yeast and worms are valuable for studying damaging genetic interactions, such as those that lead to cancer, and for testing new cancer drug targets.
CIFAR Senior Fellow Philip Hieter (University of British Columbia) and his colleagues use data from Genetic Networks fellows to demonstrate that silencing a gene called FEN1 can kill human colorectal cancer cells that have a mutation in a gene that repairs DNA, called FEN1. This result illustrates that yeast is a useful model system for predicting genetic interactions in more complex cells. The research also validates the hypothesis of Nobel Prize winning scientist and former Program Advisor Leland Hartwell, who predicted it was possible to develop drugs that selectively kill cancer cells by targeting the genetic instability that lets them reproduce at an uncontrolled rate — a process called synthetic lethality. Studies also find that mutations in a gene called PARP1 are synthetically lethal with mutations in BRCA1 and BRCA2, two breast cancer genes. Drugs that inhibit PARP1 show sufficient promise to enter Phase 1 clinical trials.
CIFAR Senior Fellow Stephen Scherer (The Hospital for Sick Children) defines genetic factors that account for about 10 per cent of individuals having autism spectrum disorder. Scherer and colleagues show for the first time that several newly discovered autism susceptibility genes link to the same biological pathways and are involved in brain function. Knowing these autism genes are linked, they can begin to develop therapies to target the common pathways involved. This research, which made international news headlines, also impacts other disease studies.
CIFAR Senior Fellow Brendan Frey (University of Toronto) and his research team discover a fundamentally new view of how living cells "read the genome" and use a limited number of genes to generate enormously complex organs such as the brain. Frey describes a discovery of a second level of information hidden in the genome that can account for the exponentially greater complexity required to create a human being. He and his team create a model allowing them to predict how certain genes can be turned off and on in ways that do everything from triggering muscle and brain development to causing neurological disorders. Their “splicing codebook” successfully predicts tens of thousands of deviations in the generation of genetic messages. This major discovery helps explain one of the great mysteries in genetics: how is it possible that human beings have so few genes in comparison to so many simpler organisms?
CIFAR Senior Fellows Chad Myers (University of Minnesota), Brenda Andrews and Charles Boone (both University of Toronto) are part of a team that examines the reactions of 5.4 million gene pairs in baker’s yeast. They identify the characteristics and behaviours of genes, or phenotypes, that occur when an essential cellular function collapses. They assemble the information into a network of the cell’s genetic landscape. It is the first genome-scale map of genetic interactions for a eukaryotic cell. A major implication of this work is explaining the genetic component of human diseases.
The research of Senior Fellow Stephen Scherer (The Hospital for Sick Children) and his team use the latest technologies in genome analysis, including whole genome sequencing, to find autism risk genes. They also analyze the products of these genes — proteins — to see how their function in the body might be modulated to benefit the patient. The latter ‘functional’ approach increasingly relies on understanding the genetic networks involved in brain function. To learn more about this research, please see this Spectrum story.
CIFAR Senior Fellow Timothy Hughes (University of Toronto) sequences the genome of Cannabis Sativa. The presence of cannabinoids, a group of more than 100 natural products that accumulate mostly in female flowers, gives cannabis its unique pharmacological properties. However, scientists don’t completely understand how cannabinoid biosynthesis works. The Hughes lab resequences two hemp strains and examines gene expression in flowers of both marijuana and hemp. The researchers identify numerous genes and transcripts involved in cannabinoid biosynthesis. They also note that one key enzyme, THCA synthase, expresses only in marijuana and another, cannabidiolic acid synthase, expresses only in hemp. These genome and transcriptome sequences are useful for understanding and controlling the molecular differences between marijuana and hemp, for identifying cannabinoid biosynthetic enzymes, and for mapping the origins and genealogy of modern hemp and marijuana varieties.
CIFAR Senior Fellows Charles Boone and Brenda Andrews (University of Toronto) analyze a network of genetic interactions involving dosage suppression, the term for when one gene overexpresses and rescues another, mutated gene. They find that dosage suppression is a unique genetic interaction that sheds light on the relationships certain genes have with others. Combinations of genes with particular variations are major contributors to human genetic disorders, and genetic interactions based on increased gene dosage are highly relevant to our understanding of diseases such as cancer.
CIFAR Senior Fellow Frederick Roth (University of Toronto) and collaborators conduct a systematic study of proteins in DNA tumour viruses to identify cancer genes. They use two technologies to discover interactions between the host cell and the viral proteins that take control of them, coaxing them to replicate at the uncontrolled rates seen in cancer. They also study how a single viral gene transfected into human cells affects gene expression. They find that systematic analysis is a useful method to identify cancer genes, equal to approaches such as sequencing the genomes of tumours.
CIFAR Senior Fellow Philip Hieter (University of British Columbia) uses a cross-species genetic approach, based on the pioneering work of Charles Boone and Brenda Andrews, to identify potential cancer drug targets by finding negative genetic interactions within mutants of the protein complex cohesin. The Hieter group uses baker’s yeast and C. elegans as model organisms and finds that that cohesin mutants require the function of proteins that regulate the replication fork, a structure that develops as DNA replicates. In human cells, Hieter shows Poly-ADP Ribose Polymerases (PARPs) repair stalls replication forks and are the target of a class of anti-tumour chemotherapeutics called PARP inhibitors, which are under clinical trials for treatment of breast cancer carrying mutations in BRCA1. PARP inhibitors may be effective for the treatment of tumours containing cohesin mutations.
CIFAR Senior Fellow Frederick Roth (University of Toronto) describes a new procedure he calls green monster technology, for constructing yeast strains carrying dozens of chosen mutations. Rather than using traditional nutritional or drug resistance markers, the Roth lab disrupts genes of interest using a green fluorescent protein (GFP) marker. GFP encodes a fluorescent protein and, unlike traditional markers that can be used only once per strain, GFP can be used repeatedly to disrupt multiple genes in the same strain background. As a pilot project, Roth’s group disrupts all 16 ATP-binding cassette transporters in a single yeast strain to generate a broadly drug-sensitive strain. The green monster technology is potentially applicable to assembling other engineered genetic alterations in different species and thus provides an additional strategy for examining more complex interactions that involve more than one gene.
CIFAR Senior Fellow Don Moerman (University of British Columbia), in collaboration with Robert Waterston, publishes the Million Mutation Project. The resource is a sequence analysis of 2,000 mutated genetic strains of the roundworm C. elegans, the culmination of a large scale project undertaken over four years. The project had its inception at a CIFAR meeting, and it is expected to have a dramatic impact on the worm community and beyond as it provides about 1.4 million sequenced mutations covering the 20,000 genes of this organism. It will provide researchers with strains that are mutated in essentially any gene of interest, opening the door for researchers to more readily perform sophisticated and large-scale genetic interaction studies in a multicellular animal.
CIFAR Senior Fellow Philip Hieter’s group (University of British Columbia) studies genetic networks involving the versions of human cancer genes with unstable genetic material carried by yeast, focusing on essential genes. This study identifies candidate genes and processes that, if disrupted, could potentially kill human cells that carry cancer-causing mutations, via a “synthetic lethal” genetic interaction. The discovery of genetic interactions in model systems can lead to important new insights into human diseases such as cancer.
Van Pel et al., “Saccharomyces cerevisiae genetics predicts candidate therapeutic genetic interactions at the mammalian replication fork,” G3: Genes Genomes Genetics, 3:273-82 (2013) doi: 10.1534/g3.112.004754
Van Pel et al., “An Evolutionarily Conserved Synthetic Lethal Interaction Network Identifies FEN1 as a Broad-spectrum Target for Anticancer Therapeutic Development,” PLoS Genetics 9(1):e1003254 (2013) doi: 10.1371/journal.pgen.1003254
CIFAR Senior Fellow Timothy R. Hughes and his team, including CIFAR Fellow Andrew Fraser and Senior Fellow Brendan Frey (all University of Toronto) use a powerful technique they invented to explore the patterns of behaviour in RNA that encourage regulatory proteins to bind, after which they influence gene expression. In a landmark study published in Nature, Hughes’ team reveals new rules about how RNA binding proteins influence gene regulation, both in normal cells and in human disease. The discovery involves the identification of a new role for well-studied protein called RBFOX1 in regulating RNAs involved in controlling the function of brain cells or neurons. Hughes’ work reveals that levels of this protein are reduced in the brains of autism spectrum disorder patients, suggesting that it may play a role in gene regulation for those with autism. This example illustrates how the compendium of RNA binding protein motifs can affect human disease.
CIFAR Senior Fellow Timothy Hughes (University of Toronto) and collaborators work towards the assembly and analysis of a large compendium of DNA "motifs" — patterns in DNA that repeat and are thought to accomplish a biological function. They choose the motifs that are bound by transcription factors. They find that many transcription factors are similar across vast evolutionary distances, while others are unique to individual species. They broadly sample parts of protein sequences with motifs that have an affinity to DNA, called DNA-binding domains (DBDs), from multiple groups of organisms with cells that have nuclei (eukaryotes). Hughes and colleagues determine DNA sequence preferences for more than 1,000 transcription factors from 131 diverse eukaryotes. These data are a powerful resource for mapping transcriptional networks across eukaryotes.
CIFAR Senior Fellow Philip Hieter (University of British Columbia) and his team introduce cancer-associated mutations into yeast and nematode worms to understand how to manipulate genetic networks to improve anti-cancer therapies. Many tumours can be targeted with DNA-damaging therapies that are specifically toxic to cancer cells. However, some tumours are not sensitive to safe doses of DNA-damaging chemotherapeutics. Hieter proposes that scientists could weaken cancer cells by knocking down a second gene that would sensitize the cancer cell to the toxic effect of a DNA-damaging drug. Working with yeast allows him to discover, from thousands of potential genetic interactions, 14 genes that when knocked out sensitized cells with a cancer-associated mutation to a DNA-damaging chemotherapeutic. At least one of the combinations also works in the nematode worm, suggesting that the approach may work in more complex organisms. The team names the process synthetic cytotoxicity.
Researchers have exposed the weak points of several types of cancer using a gene-editing technique that can turn genes on...
by Kate Lunau, Maclean’s Cancer begins when the genes in a cell are mutated, causing cells to multiply like a...
by Matt McFarland, Washington Post Twelve years ago was a low point for Brendan Frey. His wife at the time...
The start-up company Deep Genomics is on a mission to predict the consequences of genomic changes by developing new deep learning technologies.