Deep Learning Helps Scientists Keep Track of Cell’s Inner Parts

Deep Learning Helps Scientists Keep Track of Cell’s Inner Parts

Author: Jovana Drinjakovic

Donnelly Centre researchers have developed a deep learning algorithm that can track proteins, to help reveal what makes cells healthy and what goes wrong in disease.

Yeast cells (purple) with DNA-containing nuclei (pink) and a protein (green) that resides in the cell’s waste compartment or vacuole. “We can learn so much by looking at images of cells: how does the protein look under normal conditions and do they look different in cells that carry genetic mutations or when we expose cells to drugs or other chemical reagents? People have tried to manually assess what’s going on with their data but that takes a lot of time,” says Benjamin Grys, a graduate student in molecular genetics and a co-author on the study.

Dubbed DeepLoc, the algorithm can recognize patterns in the cell made by proteins better and much faster than the human eye or previous computer vision-based approaches. In the cover story of the latest issue of Molecular Systems Biology, teams led by Professors Brenda Andrews and Charles Boone of the Donnelly Centre and the Department of Molecular Genetics, also describe DeepLoc’s ability to process images from other labs, illustrating its potential for wider use.

From self-driving cars to computers that can diagnose cancer, artificial intelligence (AI) is shaping the world in ways that are hard to predict, but for cell biologists, the change could not come soon enough. Thanks to new and fully automated microscopes, scientists can collect reams of data faster than they can analyze it.                                 

“Right now, it only takes days to weeks to acquire images of cells and months to years to analyze them. Deep learning will ultimately bring the timescale of this analysis down to the same timescale as the experiments,” says Oren Kraus, a lead co-author on the paper and a graduate student co-supervised by Andrews and Professor Brendan Frey of the Donnelly Centre and the Department of Electrical and Computer Engineering. Andrews, Boone and Frey are also Senior Fellows at the Canadian Institute for Advanced Research.

Similar to other types of AI, in which computers learn to recognize patterns in data, DeepLoc was trained to recognize diverse shapes made by glowing proteins—labeled a fluorescent tag that makes them visible—in cells. But unlike computer vision that requires detailed instructions, DeepLoc learns directly from image pixel data, making it more accurate and faster.

"Deep learning will ultimately bring the timescale of this analysis down to the same timescale as the experiments" - Oren Kraus

Grys and Kraus trained DeepLoc on the teams’ previously published data that shows an area in the cell occupied by more than 4,000 yeast proteins—three quarters of all proteins in yeast. This dataset remains the most complete map showing exact position for a vast majority of proteins in any cell. When it was first released in 2015, the analysis was done with a complex computer vision and machine learning pipeline that took months to complete. DeepLoc crunched the data in a matter of hours.

DeepLoc was able to spot subtle differences between similar images. The initial analysis identified 15 different classes of proteins, each representing distinct neighbourhoods in the cell; DeepLoc identified 22 classes. It was also able to sort cells whose shape changed due to a hormone treatment, a task that the previous pipeline couldn’t complete.

 

The road less traveled - U of T labs tackle neglected parasitic diseases

The road less traveled - U of T labs tackle neglected parasitic diseases

By Jovana Drinjakovic

Researchers partner with pharmaceutical industry to meet a global health challenge

Parasites nearly killed her grandmother, and now Samantha Del Borrello is striking back.

Del Borrello is a graduate student investigating new ways of attacking parasites. She says her “nonna’s” childhood in 1940s rural Italy was plagued by intestinal worms that ravaged her health to the point doctors thought she would die.

“It is crazy to think that I may not be here because of a parasite, and now I am working on preventing the parasites from hurting people. It’s kind of cool,” says Del Borrello, a PhD candidate in the University of Toronto’s Donnelly Centre and the Department of Molecular Genetics.

Internal parasites, which infect the gut, lungs and liver, may not be a major health concern in the developed world, but globally they affect two billion people.

Gut worms infect 880 million children, according to the World Health Organization. Parasitic infections are rampant in the poorest areas, caused by nematode worms like roundworms, whipworms and hookworms. Untreated, these infections typically cause anemia and lethargy, or even death. Children are most vulnerable.

 “Drugs already exist for some parasite infections but resistance is always evolving — we need new ways to attack these complex creatures,” says Andrew Fraser, a professor in the Donnelly Centre, and Del Borrello’s PhD supervisor.

Growing drug resistance comes at a time when the pharmaceutical industry has little incentive to invest in solving health problems that affect poor people who cannot afford treatment.

One way forward is for academic labs to work with pharmaceutical companies to identify promising drugs. Fraser’s work with Janssen, a branch of the pharmaceutical giant Johnson & Johnson, is one example of this kind of collaboration.

Peter Roy, who is also a professor in the Donnelly Centre, says there is also potential for the agriculture industry to play a role in developing new treatments.

“Most of the meat we eat has been treated with anthelmintics, drugs that kill parasitic worms,” says Roy. “If novel anthelmintics are shown to be useful for cows and sheep, then they might become therapies for humans.”

Fraser and Roy, who are also both appointed to the Department of Molecular Genetics, lead research into identifying new anti-parasitic drugs. As their main tool, the researchers are relying on a harmless type of worm called C. elegans, which is also widely used in labs (see inset). Unlike parasites, which live inside a body, lab worms grow on a dish and are easy to work with.

Many parasites make their way to places in the body where there is little oxygen to breathe. In order to survive, they switch to a type of metabolism that’s not fueled by oxygen.

Normally, lab worms need oxygen to live. But Del Borrello and PhD candidate Margot Lautens found a way to trick the lab worm into behaving like a parasite, deep inside the gut. Using drugs, they turned off the worm’s ability to use oxygen, forcing the worm to use parasite-like metabolism. This allowed researchers to study quirky parasite biology in an animal right in front of them.

“The way worms survive in low oxygen is extremely unusual, humans don’t use this process at all. That’s the key. It means that if we can target this unusual metabolic pathway, we should be able to kill the worms without having any impact on the human host,” says Fraser.

Using a different strategy, Roy’s team has already uncovered a treasure trove of potential anti-parasitic compounds. Two years ago, postdoctoral fellow Andrew Burns was part of a team that  uncovered 275 chemical compounds that killed C. elegans. These worm active compounds, dubbed wactives, were then tested on fish and human cells to identify which ones could potentially harm the host.

That team is now teasing apart how wactives work. A new study in PLOS Neglected Tropical Diseases describes how a compound called wact-86 works by blocking an important enzyme in the worm. The next step is to explore whether wactives can clear parasitic infections in larger animals.

Another potential avenue is to work with a pharmaceutical company from the start. To do this, Fraser is working with BIO Ventures for Global Health (BVGH), a Seattle-based non-profit that boosts research in neglected tropical diseases through partnerships between academic labs and the pharmaceutical industry. The organization, among other roles, helps academia and industry share reagents, says Ujwal Sheth, Associate Director at BVGH.

Last month, Fraser signed a deal with Janssen, granting his team rights to the company’s drug collection—a potential chemical gold mine with 80,000 diverse compounds. If they find a medicinally promising compound, Janssen could decide take it on, said Sheth. Or, the BVGH could help connect Fraser with other partners with capacity to develop new medicines, she added.

“The best anthelminthic drug today, ivermectin, was developed in the 1970s as a partnership between an academic lab and a major pharmaceutical company. It’s a great cooperative model to help solve these huge global health problems,” said Fraser.

 

Studies from the Cochrane Lab Outline the characterization of Small Molecule Inhibitors of HIV-1 or Adenovirus Replication that Function Through the Modulation of RNA Processing

Studies from the Cochrane Lab Outline the characterization of Small Molecule Inhibitors of HIV-1 or Adenovirus Replication that Function Through the Modulation of RNA Processing

Dr. Alan Cochrane

Dr. Alan Cochrane

A novel compound that blocks HIV-1 replication inhibits the splicing regulatory function of SRSF10.NAR This study outlines the identification of a novel inhibitor (designated 1C8) of HIV-1 replication that functions through modulation of host splicing factor function, specifically SRSF10. While addition of 1C8 is found to severely impact HIV-1 RNA accumulation, the compound has very limited effects on host RNA processing. The anti-HIV activity of this compound suggests its use as a novel therapeutic to treat this infection.

Identification of Small Molecule Modulators of HIV-1 Tat and Rev Protein Accumulation.  Grosso et al describe the mechanism by which two cardiotonic steroids, digoxin and digitoxin, suppress replication of adenovirus. We demonstrate that these drugs do not affect virus entry or initiation of adenovirus gene expression, but alter the processing of viral RNA to negatively impact the production of other viral proteins and ultimately block virus DNA replication and assembly. The study is of note because it is a repurposing of a drug already approved for use in humans for an infection that currently has few treatment options.

Identification of Small Molecule Modulators of HIV-1 Tat and Rev Protein Accumulation. Retrovirology. Balachandran et al. identify three compounds (designated 791, 833, and 892) that suppress HIV-1 replication through effects on the expression of two essential viral regulatory factors, Tat and Rev. Data presented indicates that the compounds affect expression of these viral factors through changes in protein synthesis/stability. Despite the significant inhibition of HIV-1 replication/gene expression observed in the presence of these compounds, effects on host cell expression are limited. These observations highlight an alternative approach to the control of HIV-1.

Autism Study Reveals 18 New Risk Genes

Autism Study Reveals 18 New Risk Genes

From Dr. Stephen Scherer's lab, the worlds largest whole genome sequencing study in autism reveals 18 new risk genes and describes massive cloud-cased computational resource developed with Google. These findings were published in the March 6 issue of Nature Neuroscience.

Read more

CCM-3 Promotes C. elegans Germline Development by Regulating Vesicle Trafficking Cytokinesis and Polarity

CCM-3 Promotes C. elegans Germline Development by Regulating Vesicle Trafficking Cytokinesis and Polarity

Dr. Brent Derry

Dr. Brent Derry

Cerebral cavernous malformations (CCM) are disorders that cause biological tubes in the brain (i.e., veins and capillaries) to become deformed and leak blood, leading to symptoms that can range from mild headaches to hemorrhagic stroke. This rare disease can occur in people sporadically by unknown mechanisms or by inheritance of mutations in one of three genes (CCM1, CCM2 or CCM3). Patients with mutations in the CCM3 gene have the earliest disease onset (often in childhood) and suffer the greatest lesion burden compared with patients who inherit mutations in CCM1 or CCM2. The mechanism by which CCM3 maintains the integrity of biological tubes is not understood and there are currently no treatments for CCM patients, other than invasive neurosurgery. In this paper, Brent Derry's lab showed that CCM3 functions to maintain the integrity of the C. elegans germline by promoting endocytic recycling of cell surface receptors and membrane to the cytokinetic furrow of dividing cells. The also show that CCM-3, and its associated striatin interacting phosphatase and kinase (STRIPAK) complex, coordinates organization of anillin and non-muscle myosin to generate contractile forces necessary for cytokinesis and assembly of cells into biological tubes. By combining the powerful genetics and cell biology of C. elegans with proteomics methods in collaboration with Molecular Genetics professors Mike Moran and Anne-Claude Gingras they show that association of CCM-3 with its binding partners striatin and the germinal centre kinase GCK-1, dictates its subcellular localization as well as the proper positioning of polarity proteins in dividing embryonic cells and in the developing germline. This work provides new insights into the normal biological functions of CC3/STRIPAK during development that should uncover effective therapeutic targets for treating CCM patients. Towards this goal Derry is also collaborating with Peter Roy (CCBR) to identify small molecules that reverse the germline defects of ccm-3 mutants. This work was supported by grants from the CIHR and a donation from Angioma Alliance Canada.

 New Hope for Parkinson’s as Elusive Proteins Come to Light

New Hope for Parkinson’s as Elusive Proteins Come to Light

Mar 15, 2017

Author:  Jovana Drinjakovic

Study reveals a breadth of new drug targets for neurological conditions and opens the door to a greater understanding of the way in which common medications work.

Jamie Snider, Senior Eesearch Associate, and Igor Stagljar,Professor in the Donnelly Centre

Jamie Snider, Senior Eesearch Associate, and Igor Stagljar,Professor in the Donnelly Centre

Ever taken antihistamines? Or heartburn medication? Along with others used for a variety of conditions, from diabetes to high blood pressure to depression, these drugs work by targeting the same class of protein molecules on our cells. They’re the most common type of drug on the market—and in medicine cabinets at home. Yet the available medications are only the tip of the iceberg as a University of Toronto study reveals a large swath of new therapeutic opportunities, including one that could lead to a better treatment for Parkinson’s disease.

Despite representing about a half of prescribed medications worldwide, these compounds target only a sliver of one of the largest—and most elusive—classes of human proteins, called G protein coupled receptors (GPCRs). Tapping into this vast unexplored therapeutic potential has been difficult because available tools weren’t up to the task of surveying the GPCRs on large scale. Enter Professor Igor Stagljar of U of T’s Donnelly Centre.                                                         

“Our cells are made of proteins, which also do most of the work in them. But no protein acts alone and that’s why we have to look at interactions between proteins to understand what’s going on in the cell,” says Stagljar, who is also a professor in the departments of molecular genetics and biochemistry.

Stagljar’s new study, which earned the cover of the March issue of the journal Molecular Systems Biology, is based on a technology called MYTH. Previously developed in the lab, it allows detection of membrane protein interactions as they occur in their natural setting—on the surface of cells. Using MYTH, Stagljar’s team was able to capture almost 1,000 interactions between more than 600 proteins for almost 50 clinically important GPCRs. The largest survey of GPCRs to date, it revealed new associations among proteins involved in neurological disorders, such as motor neuron disease, schizophrenia, and neurodegenerative disorders, as potential targets for new drugs.

One association that stood out involved ADORA2A, a GPCR targeted by Parkinson’s disease drugs. By binding to ADORA2A, these drugs stimulate the release of dopamine, which helps communication between nerve cells to ultimately reduce tremor in patients with Parkinson’s. Stagljar’s team found that ADORA2A associates with another GPCR, called GPR37 or Parkinson’s disease associated receptor, in a way that affects movement in a mouse model of disease. This suggests that a combination of drugs targeting both receptors, may work better in patients.

The study featured on the journal cover The work on Parkinson’s was done in collaboration with Professor Francisco Ciruela’s team at the University of Barcelona in Spain, which will continue to investigate the clinical potential of the enhanced combination therapy involving ADORA2A and GPR37.

“High-throughput studies like ours are going to be major contributors in future drug development. You can look at the cell in the ways we could not do before. We can understand how proteins interconnect better to identify possible reasons why certain drug compounds might be causing side effects and also to predict which targets might potentially be valuable for treating disease,” says Jamie Snider, a senior research associate in the lab and a lead author of the study.

Learn more about Stagljar’s recent work on proteins that play a major role in cancer

To appreciate just how pervasive the 800 or so human GPCRs are, you only need to take a deep breath and look around you. Nestled inside the eye, these proteins detect light and help us see; those in the nose detect scents, while the ones in taste buds let us taste chocolate and other sweet and bitter foods. But these proteins also detect glucose and hormones in the blood, neurotransmitters, or chemicals that help our brain cells communicate, as well as hold cells together ensuring that tissues don’t fall apart. It’s no surprise then, that when GPCRs go awry, this can lead to brain disorders, diabetes, cancer and a host of other diseases.

"Our previous limited knowledge of the GPCRs had already helped us to tremendously improve human health. Think of what we might be able to do if we mapped all these proteins and their interactions" - Professor Igor Stagljar

In the past, scientists would either focus on the GPCR parts that are easily accessible, such as those sticking out on either side of the cell. Or, to study the GPCRs in entirety, they would remove the surrounding membrane, which changes the proteins’ properties. Either way, researchers weren’t getting the full picture of how these proteins work. MYTH and MaMTH, another related technology developed in the lab, have revolutionized the study of membrane proteins, attracting interest from the pharmaceutical industry.

“Our previous limited knowledge of the GPCRs had already helped us to tremendously improve human health. Think of what we might be able to do if we mapped all these proteins and their interactions and then understand the biological importance of those – this would be a huge step forward for biomedicine,” says Stagljar.

U of T Research Unlocks New Data for Cancer Drugs

U of T Research Unlocks New Data for Cancer Drugs

By Jovana Drinjakovic

Dr. Igor Stagljar

Dr. Igor Stagljar

University of Toronto scientists have uncovered more than 300 drug targets in cancer, attracting interest from the pharmaceutical industry looking to develop more precise treatments. Led by Professor Igor Stagljar of U of T’s Donnelly Centre, the study maps interactions between receptor tyrosine kinases (RTKs) and protein tyrosine phosphatases (PTPs) in humans, which can lead to cancer when their functions are disrupted. The highly anticipated study will be featured on the cover of the journal Molecular Cell, available in print on January 19.

Most cancer patients are treated with punishing chemotherapy drugs that have serious side-effects. In the last 15 years, a new generation of "smart" cancer drugs has been developed such as Gleevec, which effectively cures some forms of leukemia. These drugs are designed to target cancer cells with needle-like precision to avoid harming tissue that’s healthy. They do this by blocking proteins called kinases, which include receptor tyrosine kinases (RTKs) that control cell growth. RTKs are often mutated in cancer. However, the existing drugs block only a fraction of RTKs because these proteins have features that have made them notoriously hard to study.

Senior research associate Dr. Zhong Yao was able to carry out the largest study of RTKs to date by mapping their physical interactions with PTPs using methods previously developed in Stagljar’s lab.

“We tested interactions between almost all 58 RTKs and 144 PTPs that exist in human cells. Our map reveals new and surprising ways in which these proteins work together. These insights will help us better understand what goes wrong in cancer in order to develop more effective treatments,” said Stagljar, who is also a professor in U of T’s molecular genetics and biochemistry departments.

Lodged inside the cell’s outer envelope, or membrane, RTKs receive signals from the outside world—a hormone, for example— telling the cell to grow and divide. Normally, their activity is controlled by PTPs, which bind the RTKs and shut them down. This prevents sustained cell division that could lead to cancer.

The RTK’s place in the cell membrane is critical for their function, but it is also what has made them such a tough nut to crack. Traditional methods haven’t been able to capture the often short-lived physical interactions between RTKs and PTPs because the surrounding membrane has to be dissolved, which changes the proteins’ behaviour. Stagljar bridged this gap by developing MYTH and MaMTH, technologies designed precisely for measuring such fleeting interactions between membrane proteins in their natural setting.

The resulting map charts out more than 300 interactions between RTKs and PTPs in human cells, each a potential way to fight cancer. The findings have attracted attention of major pharmaceutical companies, including the pharma giant Genentech, which could lead to future collaborations in drug development.

Stagljar worked with two leading experts in PTP biology: Professor Anne-Claude Gingras of the Lunenfeld-Tanenbaum Research Institute and U of T’s Department of Molecular Genetics, and Professor Benjamin Neel of New York University, who was formerly with U of T and the University Health Network’s Princess Margaret Cancer Center.

“We wanted to show that these two assays we developed in our lab – MYTH and MaMTH – are suitable for studying these two important classes of proteins on such a large scale. The resulting wealth of important data can be used to develop new therapies against various types of cancer,” said Stagljar. “Ultimately, we want to build a map of interactions with all 3,000 or so human membrane proteins, of which at least 500 have direct roles in the onset of many human diseases. This will keep us busy,” he added.

A surprizing finding shines new light on the largest group of human proteins

A surprizing finding shines new light on the largest group of human proteins

By Jovana Drinjakovic

The study could pave the way for more meaningful interpretation of personal genomes.

Toronto scientists have discovered that the largest group of human proteins, which work as genome gatekeepers to control gene activity, are even more diverse in their roles than previously thought. The finding expands our understanding of how proteins “read” the DNA and could lead to a more accurate interpretation of individual genomes.

Donnelly Centre teams, led by Professor Timothy Hughes and University Professor Jack Greenblatt, have shown that proteins called C2H2-zinc fingers (C2H2-ZF) can control gene activity in new and surprizing ways. Reporting in the December issue of Genome Research, the researchers also reveal DNA binding sites for more than a hundred C2H2-ZFs as part of an ongoing effort to decode genome sequences that do not code for genes.

Dr. Tim Hughes & Dr. Jack Greenblatt

Dr. Tim Hughes & Dr. Jack Greenblatt

Despite being the largest group of human proteins—counting 700 members—the C2H2-ZFs are poorly understood partly because their sheer abundance and diversity make them hard to study. Yet knowing how they work is important because they help orchestrate gene activity. Of 20,000 human genes, only a subset is active in the cell at any given time. This subset determines if the cell will, say, build blood, or the brain or go haywire to become cancer.

The C2H2-ZF proteins work by directly binding the DNA to control the genes nearby. Named after their finger-like structures that, aided by zinc ions, clasp the DNA, C2H2-ZFs have previously been thought to act by stifling a wide range of genes. In a previous study that included about 40 C2H2-ZFs, the team showed that each protein recognized a unique DNA snippet as its landing site in the genome, raising the possibility that the rest of the group could be just as diverse.

This was indeed confirmed in the present study in which the teams mapped DNA binding sites, most of which were unique, this time for 131 C2H2-ZF proteins. But they also uncovered a whole new way in which the C2H2-ZF proteins can be regulated to vastly expand their job repertoire in the cell.

In addition to binding the DNA, it turned out that each C2H2-ZF can partner with a motley of other proteins that could potentially tweak its ability to switch genes on and off in a unique way. The finding upended the previous thinking in which C2H2-ZF proteins were seen as limited in their ability to bind other proteins—half of them were thought to interact with a single protein that helps them gag target genes, while the rest lack the usual molecular features that help proteins contact one another.

“Our key finding is that there’s almost as much diversity in the protein-protein interactions as there is in the DNA binding sequences. It tells us that the way the C2H2-ZF proteins work is almost certainly more complicated than we would have expected,” said Hughes, who is also a professor in U of T’s department of molecular genetics and a fellow of the Canadian Institute for Advanced Research (CIFAR).

The kinds of proteins that C2H2-ZFs interact with suggest that their roles go beyond clamping down on genes and may even act to turn genes on or help package DNA inside the cell.

The study also shines light on how the C2H2-ZF evolved to become the largest and most diverse group of proteins we have. The DNA sequences that C2H2-ZF proteins recognize look a lot like they had come from viruses, some of which plagued our mammalian ancestors as long as 100 million years ago. This kind of viral DNA has been called “selfish DNA” because it spreads by inserting itself randomly in a host’s genome.

It is thought that the C2H2-ZF proteins evolved to shut down this selfish DNA, their legion expanding to keep up with new intruders. Once the viral DNA was squashed for good, the C2H2-ZF proteins were able to take on new roles in shutting down mammalian genes. And now, this new data suggest that the C2H2-ZF proteins branched out even more than previously thought to acquire wholly unexpected functions by binding to other proteins.

Knowing how C2H2-ZFs work will give scientists a better handle on predicting which genes they control and how this may relate to disease. So far, mass genome sequencing studies have fallen short from being able to tell one’s risk of common diseases, such as cancer or diabetes, because we still don’t know enough about the meaning of individual differences between genomes.

“Even today, 15 years after the human genome was sequenced, if you give any piece of DNA to a geneticist and ask them what this does, generally they are unable to tell you that. But the more we learn about how human proteins recognize the DNA and what they do, the better our ability will be to interpret genome sequences and say what the significance of the variants is,” said Hughes.

U of T Researchers Rapidly Identify New Drug Target and Develop Antibodies to Kill Pancreatic Cancer Cells

U of T Researchers Rapidly Identify New Drug Target and Develop Antibodies to Kill Pancreatic Cancer Cells

By Jef Ekins

Researchers at the University of Toronto have developed a process that dramatically cuts the amount of time it takes to create new cancer treatments. Using a breakthrough technology, their study, published this week in Nature Medicine, identified a new potential drug target in a class of pancreatic cancer, and unveiled a new treatment option that exploits genetic faults to destroy cancer cells.

Dr. Sachdev Sidhu

Dr. Sachdev Sidhu

Dr. Jason Moffat

Dr. Jason Moffat

Professors Jason Moffat and Sachdev Sidhu from the Donnelly Centre and the department of molecular genetics, along with Professor Stephane Angers from the Leslie Dan Faculty of Pharmacy, made this discovery using the cutting-edge CRISPR-Cas9 genome editing technology.

Using this revolutionary tool, the teams probed the function of every single gene expressed by pancreatic cancer cells to determine that one of the receptors (Frizzled-5) is essential for the growth of mutant pancreatic cancer cells. Normally, the signaling pathways activated by Frizzled-5 tell cells when to divide, what types of cells to become, and when they should die. When mutated or deregulated, however, they can initiate tumour growth.

"This is the first time that we are able to identify bona fine genetic weaknesses in cancer cells that we can target with drugs, which will not harm healthy tissue," said Moffat.

Having identified the key role that the Frizzled-5 receptor plays in promoting pancreatic cancer growth, the team rapidly developed an antibody drug to inhibit the growth of these cells. The study showed that the antibody proved highly effective in killing the cancer cells in patient-derived samples and shrank tumours in mice without damaging the surrounding healthy cells.

Leveraging the Donnelly Centre’s state-of-the-art platform for custom antibody design, the team was able to create a targeted antibody in months – a fraction of the time it would normally take to develop a safe and effective treatment for a specific cancer.

"Our technology allows us to quickly develop a drug that's tailored to a particular kind of cancer and ready to be tested on people. In this study we show that the approach works for a type of pancreatic cancer, but this is only the tip of the iceberg," said Sidhu.

As part of this study, the team also explored the role of this receptor in colorectal cancer, a form of cancer that shares common features with pancreatic cancer. The results of this study indicate that Frizzled-5 may be a factor across multiple cancer types, broadening the potential use of anti-Frizzled-5 antibodies as a targeted cancer therapy.

“Ultimately, this study revealed genetic vulnerabilities in pancreatic cancer cells that could be exploited through the development of new targeted antibodies to inhibit tumor growth,” said Angers. “By targeting the exact signaling circuit activated in these tumors, these rapidly developed antibodies have shown considerable promise as a cancer treatment. Moreover, the state-of-the-art antibody development platform developed at U of T is a transformational leap forward in our ability to rapidly create exciting new treatments to combat various cancers.”

 

MoGen Alumnus & Postdoc Receives Outstanding Innovation Award

MoGen Alumnus & Postdoc Receives Outstanding Innovation Award

By Jovana Drinjakovic

As a protein engineer, Dr. Wei Zhang gives old molecules new tricks. And now, he’s transformed a single human protein into a virus-crushing arsenal that could lead to long-sought treatments for deadly infections.

This week, Zhang received the Mitacs Award for Outstanding Innovation, for his work on creating molecular antidotes against viruses that cause Middle East Respiratory Syndrome (MERS) and Crimean-Congo Hemorrhagic Fever (Congo Fever). His patent-pending research was crucial to the launch of a new Toronto-based company called Ubiquitech, which will further commercialize his work so it can be used in a clinical setting.

As a postdoctoral research fellow in Professor Jason Moffat’s group in the Donnelly Centre, Zhang already holds a competitive Elevate Fellowship from Mitacs, a nonprofit supporting innovation across the public and private sectors. The fellowship enabled Zhang to start thinking about commercializing his research through a collaboration the with the Centre for Commercialization of Antibodies and Biologics (CCAB), the industry partner for his fellowship, located at the U of T.

“Wei has done a phenomenal job of applying a cutting-edge new technology to the pressing issue of emerging pathogenic viruses. It’s a great example of how an investment by Canada in basic research and talented young scientists can lead to real impacts on human health,” says Dr. Sachdev Sidhu, CEO of CCAB.

After completing a doctorate with a U of T Professor Daniel Durocher, Zhang joined Moffat and Sidhu’s labs to learn the ropes of protein engineering. The two teams had previously developed a powerful technology to quickly create synthetic proteins that could be used as research reagents, or developed further into drugs.

Although Zhang’s work revolves around a single protein, called ubiquitin, its applications are far-reaching. Named after its pervasive presence in every cell on earth, ubiquitin works by attaching to other proteins to help relay signals telling the cell to grow or fight infections, for example. Encoded by genes, proteins do most of the work in the cell. They are built from amino acids, which are stitched together based on the DNA blueprint. Using molecular tricks, scientists can change how a protein behaves by changing its DNA sequence.

And so, through subtle tweaks, Zhang has turned a lone, naturally occurring ubiquitin into a set of tools—synthetic ubiquitin variants (UbV)—which allow him to manipulate the proteins that ubiquitin normally binds to. For example, a ubiquitin variant may spur the activity of the other protein, or it may take it out of action completely. But each variant has a unique target, allowing Zhang to control protein activity with unrivalled precision.

Being able to thwart a protein is particularly useful when dealing with harmful molecules, like those made by bacteria and viruses, for example.

“Ubiquitin-dependent signalling is important in the immune response, and a lot of viruses encode proteins that bind human ubiquitin, which allows them to topple the body’s defense mechanism. One of them is MERS, a respiratory virus similar to SARS that caused a global epidemic in 2002. 14 years later, nothing has come out of clinical trials. While vaccines are being developed for many viruses, there is no treatment in sight for people who are infected,” says Zhang.

MERS emerged in Saudi Arabia in 2012, and it kills almost 40 per cent of those who become infected, making it even more dangerous than SARS. Zhang has engineered a ubiquitin variant which blocks MERS’ ability to evade the immune response. “When we treat the cells infected with MERS with the ubiquitin variant, we can kill the virus in two days,” says Zhang.

The list of anti-viral ubiquitins could become long. Zhang has already created a variant that’s effective against Congo Fever virus, which causes internal bleeding and kills almost half of those infected. But what’s most exciting about Zhang’s approach is that it can be applied to any viral protein that binds ubiquitin. “We are able to quickly—in less than one month—generate inhibitors for any ubiquitin-binding proteins in the virus,” says Zhang. This means that Zhang’s work could also be applied to viruses such as SARS, Zika, and Ebola, as well as to preventing viral damage to food crops and animals.

“Wei has fully captured the potential of protein engineering technology with his research. He is poised to turn basic science into applications that could help people,” says Moffat.

Why bad genes aren’t always bad news

Why bad genes aren’t always bad news

By Jovana Drinjakovic

Arrays of mutant yeast strains in a Petri dish. The faster the cells grow, the bigger the size of colonies (dots). 

Arrays of mutant yeast strains in a Petri dish. The faster the cells grow, the bigger the size of colonies (dots). 

The study paves the way for understanding how some people stay healthy despite having disease-causing mutations.

We usually think of mutations as errors in our genes that will make us sick. But not all errors are bad, and some can even cancel out the fallout of those mutations known to cause disease. While little has been known about this process — called genetic suppression — that will soon change as University of Toronto researchers uncover the general rules behind it.

Teams led by Professors Brenda Andrews, Charles Boone and Frederick Roth of the Donnelly Centre and the Department of Molecular Genetics, in collaboration with Professor Chad Myers, of the University of Minnesota-Twin Cities, have compiled the first comprehensive set of suppressive mutations in a cell, as reported in Science today. The four researchers are members of the Genetic Networks program of the Canadian Institute for Advanced Research. Their findings could help explain how suppressive mutations combine with disease-causing mutations to soften the blow or even prevent disease.

This curious bit of biology has only come to light as more healthy people have had their genomes sequenced. Among them are a few extremely lucky folks who remain healthy despite carrying catastrophic mutations that cause debilitating disorders, such as Cystic Fibrosis or Fanconi anemia.

How could this be?

“We don’t really understand why some people with damaging mutations get the disease and some don’t. Some of this could be due to environment, but a lot of could be due to the presence of other mutations that are suppressing the effects of the first mutation,” said Roth, who is also a Senior Scientist at Sinai Health System’s Lunenfeld-Tanenbaum Research Institute.

Imagine being stuck in a room with a broken thermostat and it’s getting too hot. To cool down, you could fix the thermostat—or you could just break a window. Genetic suppression essentially “breaks the window” to keep cells healthy despite damaging mutations. And it opens a new way of understanding, and maybe even treating, genetic disorders.

“If we know the genes in which these suppressive mutations occur, then we can understand how they relate to the disease-causing genes and that may guide future drug development,” said Dr. Jolanda van Leeuwen, a postdoctoral fellow in the Boone lab and one of the scientists who spearheaded the work.

But finding these mutations is not easy; it’s a proverbial needle in the haystack. A suppressive mutation could, in theory, be any one of the hundreds of thousands of misspellings in the DNA, scattered across the 20,000 human genes, which make every genome unique. To test them all would be impractical.

“A study like this has never been done on a global scale. And since it is not possible to do these experiments in humans, we used yeast as a model organism, in which we can know exactly how mutations affect the cell’s health,” said Van Leeuwen. With only 6,000 genes, yeast cells are a simpler version of our own, yet the same basic rules of genetics apply to both.

The teams took a two-pronged approach. On the one hand, they analyzed all published data on known suppressive relationships between yeast genes. While informative, these results were inevitably skewed towards the most popular genes — the ones that scientists have already studied in detail. Which is why Van Leeuwen and colleagues also carried out an unbiased analysis by measuring how well the cells grew when they carried a damaging mutation on its own, or in combination with another mutation. Because harmful mutations slow down cell growth, any improvement in growth rate was thanks to the suppressive mutation in a second gene. These experiments revealed hundreds of suppressor mutations for the known damaging mutations.

Importantly, regardless of the approach, the data point to the same conclusion. To find suppressor genes, we often don’t need to look far from the genes with damaging mutations. These genes tend to have similar roles in the cell — either because their protein products are physically located in the same place, or because they work in the same molecular pathway.

“We’ve uncovered fundamental principles of genetic suppression and show that damaging mutations and their suppressors are generally found in genes that are functionally related. Instead of looking for a needle in the haystack, we can now narrow down our focus when searching for suppressors of genetic disorders in humans. We’ve gone from a search area spanning 20,000 genes to hundreds, or even dozens. That’s a big step forward,” said Boone.

Press contact: Jovana Drinjakovic, +1 416 543 7820, Jovana.drinjakovic@gmail.com

Landmark Map Reveals the Genetic Wiring of Cellular Life

Landmark Map Reveals the Genetic Wiring of Cellular Life

By Jovana Drinjakovic

The new map breaks away from the old way of studying genes one at a time and shows how genes interact in groups to shed light on the genetic roots of diseases.

Donnelly Centre researchers have created the first map that shows the global genetic interaction network of a cell. It begins to explain how thousands of genes coordinate with one another to orchestrate cellular life.

Dr. Brenda Andrews

Dr. Brenda Andrews

The study was led by U of T Professors Brenda Andrews and Charles Boone, and Professor Chad Myers of the University of Minnesota-Twin Cities. It opens the door to a new way of exploring how genes contribute to disease, with a potential for developing finely-tuned therapies. The findings are published in the journal Science on September 23.

“We’ve created a reference guide for how to chart genetic interactions in a cell. We can now tell what kind of properties to look for in searching for highly connected genes in human genetic networks with the potential to impact genetic diseases,” said Dr. Michael Costanzo, a Research Associate in the Boone lab and one of the researchers who spearheaded the study.

The study took 15 years to complete and adds to Andrews’ rich scientific legacy for which she was awarded a Companion of the Order of Canada – the highest civilian honour in the country.

Just as societies in the world are organized from countries down to local communities, the genes in cells operate in hierarchical networks to organize cellular life. If we are to understand what 20,000 human genes do, we must first find out how they are connected to each other.

Studies in yeast cells first showed the need to look farther than a gene’s individual effect to understand its role. With 6,000 genes, many of which are also found in humans, yeast cells are relatively simple but powerful stand-ins for human cells. Over a decade ago, an international consortium of scientists first deleted every yeast gene, one by one. They were surprised to find that only one in five were essential for survival. It wasn’t until the last year that advances in gene editing technology allowed scientists to tackle the equivalent question in human cells. It revealed the same answer: a mere fraction of genes are essential in human cells too.

These findings suggested most genes are “buffered” to protect the cells from mutations and environmental stresses. To understand how this buffering works, scientists had to ask if cells can survive upon losing more than one gene at a time, and they had to test millions of gene pairs. Andrews, Boone and Myers led the pioneering work in yeast cells by deleting two genes at a time, in all possible pairwise combinations, to find gene pairs that are essential for survival. This called for custom-built robots and a state-of-the-art automated pipeline to analyze almost all of the mind-blowing 18 million different combinations.

The yeast map identified genes that work together in a cell. It shows how, if a gene function is lost, there’s another gene in the genome to fill its role. Consider a bicycle analogy: a wheel is akin to an essential gene – without it, you couldn’t ride the bike. Front brakes? Well, as long as the back brakes are working, you might be able to get by. But if you were to lose both sets of brakes, you are heading for trouble. Geneticists would say that front and back brakes are “synthetic lethal,” meaning that losing both, but neither by themselves, spells doom. Synthetic lethal gene pairs are relatively rare, but because they tend to control the same process in the cell, they reveal important information about genes we don’t know much about – for example, scientists can predict what an unexplored gene does in the cell, simply based on its genetic interaction patterns.

It’s becoming increasingly clear that human genes also have one or more functional backups. So instead of searching for single genes underlying diseases, we could instead be looking for gene pairs. That is a huge challenge because it means examining about 200 million possible gene pairs in the human genome for association with a disease.

Fortunately, with the know-how from the yeast map, researchers can now begin to map genetic interactions in human cells, and even expand it to different cell types. Together with whole-genome sequences and health parameters measured by new personal devices, it should finally become possible to find combinations of genes that underlie human physiology and disease.

Dr. Charlie Boone

Dr. Charlie Boone

“Without our many years of genetic network analysis with yeast, you wouldn’t have known the extent to which genetic interactions drive cellular life or how to begin mapping a global genetic network in human cells. We have tested the method to completion in a model system to provide the proof of principle for how to approach this problem in human cells. There’s no doubt it will work and generate a wealth of new information,” said Boone, who is also a professor in U of T’s molecular genetics department, a Senior Fellow and a co-director of the Genetic Networks program at the Canadian Institute for Advanced Research (CIFAR) and holds Canada Research Chair in Proteomics, Bioinformatics and Functional Genomics.

The concept of synthetic lethality is already changing cancer treatment because of its potential to identify drug targets that exist only in tumour cells. Cancer cells differ from normal cells in that they have scrambled genomes, littered with mutations. They’re like a bicycle without a set of brakes. If scientists could find the highly vulnerable back-up genes in cancer, they could target specific drugs at them to destroy only the cells that are sick, leaving the healthy ones untouched.

Seeing is believing - New molecular structure sheds light on DNA repair

Seeing is believing - New molecular structure sheds light on DNA repair

August 09, 2016
By Jovana Drinjakovic

If you want to know what the Moon really looks like, you should go visit. Yet most of us will only ever know it as a shiny circle in the sky. It’s like that for biologists. When they look at molecules, they see blobs because most microscopes aren’t powerful enough to discern any structural detail.

But now, thanks to a collaboration between Mount Sinai Hospital’s Lunenfeld-Tanenbaum Research Institute (LTRI), part of Sinai Health System, and The Hospital for Sick Children (SickKids), Toronto scientists have caught a glimpse of the molecules involved in DNA repair — a process that counteracts DNA damage caused by, for example, radiation or chemicals. Published in the latest issue of Nature, the finding deepens our understanding of how cells see and respond to DNA damage and opens new avenues of research into the process that guards us from mutations that could lead to cancer and other diseases.

Dr. Daniel Durocher

Dr. Daniel Durocher

Teams led by Drs. Daniel Durocher and Frank Sicheri, of the LTRI, and John Rubinstein, of SickKids, used Electron Cryomicroscopy (Cryo-EM) to see how a protein called 53BP1, a key component of DNA repair machinery, recognizes histones, little balls of proteins that organize our DNA. Cryo-EM is gaining popularity as a high-resolution imaging method of choice, allowing scientists to see atoms that build our body’s molecules.

“This work helps us answer how cells recognize and respond to the alarm caused by damage to our DNA. We were able to see, with this powerful microscope, that 53BP1 decodes the chemical signal that is triggered by DNA damage,” says Durocher, who is also a Professor in the University of Toronto’s Department of Molecular Genetics and Canada Research Chair in Molecular Genetics of the DNA Damage Response.

DNA damage is inevitable – whether it’s harmful products of our own metabolism or environmental factors, all these can cause unwanted genetic changes. Luckily, the cells have evolved efficient repair mechanisms. In one of them, 53BP1 quickly binds to the site of damage and recruits other components of the repair machinery.

Inside the cells, the DNA is tightly wound up around the histones. Should a piece of DNA get broken by, say radiation, the nearby histones become swiftly adorned with chemical tags, which signal to 53BP1 that help is needed. But how 53BP1, or other cellular machineries, read this “histone code” has been a long-standing question in biology.

“We knew that certain modifications happen near the damaged DNA and we knew that 53BP1 was recruited to the sites of damage, but we really had no idea how 53BP1 recognized those modifications,” says Durocher.

To answer this question, Dr. Marcus Wilson, a postdoctoral researcher in Durocher’s team, teamed up with Dr. Samir Benlekbir from Rubinstein’s group that is among the world-leading in Cryo-EM. Thanks to recent technological improvements, Cryo-EM now stands almost neck-to-neck with the gold standard, but technically challenging, method of X-ray crystallography.

Unlike X crystallography, Cryo-EM requires a small amount of material that is relatively easy to prepare. It works by snap-freezing a small volume of liquid containing the sample so that molecular complexes, including the whole lot — the DNA, wound up around the chemically-tagged histones, with 53BP1 bound to them — are randomly arranged in a thin layer of ice. Then, the complexes are irradiated with an electron beam to capture their images from all possible angles. Finally, software sifts through tens of thousands of snapshots to compute a high-resolution 3D model of the entire molecular structure. It worked so well that it caught the researchers by surprise when they realized they were able to see parts of molecules as small as half a nanometer (that’s half a billionth of a metre!) in size.

“Many in the field will be surprised by how high a resolution we got with our particular hardware - near the X crystallography resolution – and at a sufficient resolution to answer many biological questions about how 53BP1 recognizes the damaged site on the DNA,” says Rubinstein, who is also a U of T Professor at the Departments of Biochemistry and Medical Biophysics and Canada Research Chair in Electron Cryomicroscopy.

The long-awaited 3D model not only lets scientists see exactly how 53BP1 is nestled between the DNA and the histones, but it also underscores the importance of the histone code. If even a single tag is missing, or is in a wrong place, 53BP1 will not bind it.

“This is the first time anyone’s seen how a component of the DNA repair machinery reads and interprets the histone code. It’s a really exciting time for EM and it is allowing us to see how the machinery of the cell works with unprecedented clarity,” says Wilson.

This research was primarily supported by operating grants from the Canadian Institutes of Health Research and the Krembil Foundation.

Ref.: http://www.nature.com/nature/journal/vaop/ncurrent/full/nature18951.html

Kim Lab - A More Powerful Way to Develop Therapeutics?

Kim Lab - A More Powerful Way to Develop Therapeutics?

Author: Heidi Singer

Dr. Philip Kim

Dr. Philip Kim

A University of Toronto scientist has developed a new method for identifying the raw ingredients necessary to build ‘biologics’, a powerful class of medications that has revolutionized treatment of diseases like rheumatoid arthritis and some cancers.

Biologics are a type of drug that results from the high-tech manipulation of our own proteins, as opposed to more traditional drugs built from synthetic chemicals. Because of their success so far, scientists are racing to create new biologics – and now, a University of Toronto researcher has developed a way to make that process more powerful.

Philip M. Kim, an associate professor in U of T’s Donnelly Centre for Cellular and Biomolecular Research, combined high-tech computer simulation and high-throughput laboratory experiments to create what he hopes will be the most effective way to discover the proteins that are key to new biologics. His research was published online in the journal Science Advances on July 20, 2016.

“A large fraction of new therapeutics these days involve engineered proteins that latch onto a drug target, for instance on a cancer cell,” says Kim, also of the departments of Molecular Genetics and Computer Science. “Finding a protein that effectively binds to a target can feel like looking for a needle in a haystack. Our method should open up new opportunities to find those key proteins – and make a major impact on the development of new biologics.”

Under the traditional approach to developing a biologic, researchers identify a protein of interest and then test billions of variants, either randomly generated or from a natural source, hoping to find an effective binder. But these methods allow very little control over where and how the protein performs this crucial function on its target – a major factor in its effectiveness.

Kim and his team took a different approach. They used a computer to simulate the binding process, and then designed proteins that would work on the target. This type of theoretical approach has been in development for several decades, but is still not effective enough. So Kim combined the best of both methods. Instead of randomly creating massive libraries of variants, as with the traditional approach, he used computer modelling to generate a smaller, but intelligently designed repertoire of variants. Designing each variant allows for the tight control of all its properties, in contrast to conventional approaches.

“We showed that this method gives you binders that are somewhat stronger than what you get with the conventional approach,” says Kim. “The much smaller library also solves many technical problems, and we can screen for new, previously unscreenable, targets. It’s a very exciting time for cancer research, and for biologics.”

For Kim, the next step is to produce proteins that are important to certain types of cancer, but have not been screened before due to the difficulty producing them.


Via Faculty of Medicine News, University of Toronto

Miller Lab - How the fingertip is teaching scientists about tissue repair

Miller Lab - How the fingertip is teaching scientists about tissue repair

Dr. Freda Miller

Dr. Freda Miller

When a newt loses a limb due to injury, it simply grows back. Mammals are not as fortunate as evolution has left us without this useful regenerative capacity. One exception however, is the fingertip which regenerates from the distal tip (farthest end of finger) to the nailbed in both mice and humans. How this occurs has been largely unknown.

A new SickKids-led study suggests that nerve-derived cells may play a key role in regulating tissue repair and regeneration in the distal tip and throughout the body.

The research team found that after digit tip injury in adult mice, cells that initially create nerves in the digit tip return to their original regenerative state and secrete growth factors that promote regeneration of the injured fingertip.  When this response was prevented, nail and bone regeneration was impaired.

“We have been interested for a long time in how stem cells contribute to the ability of adult tissues to repair themselves, and digit tip regeneration is the ultimate example of repair. In this case, we found that it's not just a single tissue, like skin, repairing itself, but many tissues being repaired in a highly organized fashion,” says Dr. Freda Miller, principal investigator of the study and Senior Scientist in Neurosciences & Mental Health at SickKids.

Miller and her team were surprised to find that nerves contain a population of cells that after injury migrate into the injured tissues and promote tissue repair. She adds that these results demonstrate that this is true of digit tip regeneration, of skin repair, and that it may be important for tissue repair throughout the body too.

This research was supported by the Canadian Institutes of Health Research (CIHR), an Ontario Stem Cell Initiative fellowship, an Ontario Institute of Regenerative Medicine fellowship and SickKids Foundation.

For more information please see the full paper published in the June 30 online edition of Cell Stem Cell.


Published June 30, 2016 by the Hospital for Sick Children