Genetics/Genomics | Duke Research Blog

Duke Students Travel to D.C. to Present Findings to FDA

May 6th, 2013

By Nonie Arora

Duke students outside the FDA. Evelyna Kliassov, Ryan Gimple, Jenae Logan, Hiruni Amarasekara, Biqi Zhang, Selina Chen and Akash Shah. Credit: Huntington Willard

Last month, Duke seniors presented findings on noninvasive prenatal testing at the Food and Drug Administration (FDA) in Washington D.C.

The students explained to government officials that noninvasive prenatal testing requires only a blood sample from a pregnant woman. Current tests, such as amniocentesis, involve extracting cells from the placenta or fluid surrounding the fetus.

Instead, with the new technology labs genetically sequence fetal, cell-free DNA from in the mother’s blood to test for certain disorders. The method can detect when a fetus does not have the normal number of chromosomes. Specifically, it can detect abnormalities in chromosomes 13, 18 and 21, which can lead to disorders such as Down’s Syndrome.

The technology can also identify some fetal, sex-linked disorders and certain single-gene mutations. It is reliable after seven weeks of pregnancy, the students reported.

The presentation was a final project of the Genome Sciences & Policy capstone course, which leads to students earning a certificate in the field.

The students said Duke geneticist Hunt Willard and Dr. Robert Cook-Deegan, the professors for the course, chose noninvasive prenatal testing as the capstone topic because it is a new and rapidly growing field.

“Our professors wanted us to have a feel for what it’s like to research technology while it’s happening, while decisions are being made about whether it’s accurate and reliable,” said Biqi Zhang, one of the students in the class.

To investigate the scientific basis for noninvasive prenatal testing, its challenges, the active stakeholders and associated ethical considerations, these students interviewed individuals involved with different aspects of the technology.

“We had to go out and connect with many well-established professionals in related fields. It was exciting to develop skills that you normally don’t inside the classroom,” said Selina Chen, another student in the course.

“We had the opportunity to contact researchers and CEOs of companies to gain a comprehensive understanding of the technology,” Zhang added.

Evelyna Kliassov presenting on cost-effectiveness of noninvasive prenatal testing to the FDA. Credit: Huntington Willard.

The students said that the technology can and will fundamentally alter prenatal medicine. Throughout the semester, they have gained a nuanced understanding of its complexities and the viewpoints of many different stakeholders involved, from technology startup CEOs to primary care physicians.

“It was most exciting being able to go into the real world and see how this technology is being clinically implemented,” said capstone student Ryan Gimple.

“Traveling to the FDA was definitely nerve-wracking, for me at least,” capstone student Hiruni Amarasekara said. “We wanted to present a comprehensive report of the technology so that they could use this information in their decision making process on whether to recommend the test in the future. It was hard to tell what the FDA was thinking as we were presenting our information.”

The FDA has not yet stated a position on use of noninvasive prenatal testing.

Hope for Understanding Ourselves Goes to the Dogs

April 29th, 2013

By Ashley Yeager

Brian Hare and Evan MacLean, co-directors of Duke’s Canine Cognition Center, play with Lilu, a labradoodle. Credit: Ashley Yeager, Duke.

Lilu, a beautiful brown poodle-labradoodle mix, couldn’t sit still. Scents of pizza and peanut butter dog treats and the sights of new people easily distracted her.

The ADD behavior could be one trait that made her fail out of service-dog training.

“Six out of every ten dogs wash out of service training. But it’s hard right now for scientists to understand why,” said Duke evolutionary anthropologist Evan MacLean, co-director of the university’s Canine Cognition Center.

He, along with biological anthropologist Brian Hare and geneticist Misha Angrist spoke about ‘Genes, Brains and Games’ in man’s best friend as part of the Science and Society Journal Club on April 26.

MacLean and Hare explained that dogs have taken on many jobs in human society, acting as everything from pets, to our eyes and ears to being like coal-mine canaries searching for hidden bombs and missing people.

“Dog vocations require different sets of cognitive skills,” MacLean said. He studies military dogs, looking for traits that make them more suited for service tasks than pets like Lilu.

MacLean would ultimately like to identify the genetic components that underlie the characteristics suited for each type of job that a dog might do.

Scientists are interested in correlating dogs’ cognitive traits to their associated genes because the animals are “the most exquisite example of artificial selection,” Angrist said.

In Portuguese water dogs, for example, just six substitutions in individual DNA bases of the dogs explain variations in body size. In humans, nearly every gene could factor into height. It’s the same challenge that makes understanding human cognition and intelligence difficult at the genetic level.

Of course, defining cognition and intelligence at the conceptual level isn’t so clear cut either. “It’s so hard for people, journalists and the general public, to understand multiple intelligences,” Hare said.

He explained that at a basic level, cognition is the ability to make inferences, and that when we think of intelligence we think of IQ and standardized tests. These tests, however, measure only one type of intelligence. They don’t measure the ability to empathize, to verbalize a new idea or to put two completely separate ideas together to form a new one, which are other, important facets of intelligence, or really multiple intelligences.

At the Canine Cognition Center, and through the citizen science website Dognition, Hare and MacLean use standardized tests to study the variation in dogs’ intelligence. The tests, unlike the SAT or ACT, “cast a wide net across skills sets dogs could use for different vocations,” Hare said.

Dogs like Lilu, he added, are “really the hope of the world” for understanding cognition.

Not your typical spring break

April 12th, 2013

By Nonie Arora

Students in front of Eisenhower Executive Office Building, Credit: Bob Cook-Deegan

Seventeen Duke students had a taste of science policy over spring break. We traveled to Washington D.C. to meet with influential scientists and policy makers from a variety of different institutions, from the Genetic Alliance to the Office of Science and Technology Policy of the White House.

The trip clarified for many of us what science policy is like in action, and the winding paths that guide people to this career.

The students contributed to a trip blog, on which they discuss experiences such as seeing Bo Obama, the First Dog (!), outside the White House and “sipping the kool-aid” of genome science at the National Human Genome Research Institute.

The trip was sponsored by Focus and the Institute for Genome Sciences & Policy under the direction of professor Bob Cook-Deegan.

Designing Microbial “Factories” Rationally

January 29th, 2013

By Pranali Dalvi

Using microbes to manufacture chemicals is starting to be cheaper and greener than traditional chemistry. And their feedstock is sugar, not oil.

Source: 2010 Agricultural Biotechnology International Conference

On Friday, Dr. Michael Lynch spoke to an engaged audience about how microbes have ushered in a new era in metabolic and genetic engineering. Lynch is the co-founder and CSO of OPX Biotechnologies, a Colorado-based company that makes bio-based chemicals and fuels from microbes. OPXBIO microbes produce fatty acids from hydrogen and carbon dioxide. In turn, the fatty acids are used to make cleaners, detergents, jet fuel, and diesel.

Lynch said it’s easier to understand the genetic circuits and enzymatic pathways of microbes, thanks to much cheaper DNA sequencing. What we still lack though, is an understanding of how to rationally design complex biological systems – likely because we fail to recognize the interplay among an organism’s genotype, phenotype, and environment.

It’s a complex set of factors that go into making phenotypic traits such as color, size, or shape.

“In an industrial setting [phenotypes] are equivalent to metabolism or higher production of the product of interest,” Lynch said. “In a clinical setting, [phenotypes] could be virulence or pathogenesis.”

One approach to understanding how phenotypes are controlled has been through functional genomics.

Let’s say we take a population of wildtype microorganisms and introduce genetic modifications in a controlled way. Next, we selectively screen for the phenotype of interest and compare the sequence of this phenotype to the wildtype to pinpoint the genetic mutations that made the difference.

Comparing phenotypes one at a time is inefficient, though. Lynch wanted to find a way to speed up this process.

“We wanted a process or technology or toolkit that evaluates all of your genes in parallel in a single experiment for the phenotype of interest,” Lynch explained.

Lynch found his inspiration in microbial biofilms, extracellular polysaccharide matrices that grow quickly.

OPXBIO’s Efficiency Directed Genome Engineering (EDGE) technology platform, Source: opxbio.com

Lynch’s studies revealed that microbial cultures grown in enriched media made biofilms, while those in minimal media did not. In a process known as destructional mutagenesis, Lynch and his colleagues then knocked out biofilm-making genes to identify what genes cause the biofilm phenotype in enriched medium but prevent it in minimal medium.

Lynch saw the individual microbial systems as factories that he can genetically modify to produce chemical compounds in biofilms – specifically, 3-hydroxypropionic acid – that can be chemically converted to commercially relevant compounds such as acrylic.

Scientists at OPXBIO have cracked the code for making acrylic from sugar. They give sugar feedstocks to genetically modified bacteria, whose enzymes convert the sugar into acrylic molecules. Acrylic has broad commercial applications in paints, adhesives, diapers, detergents, and even fuel – a $10 billion global market.

What makes humans so unique?

January 18th, 2013

By Pranali Dalvi
Human and chimpanzees are very similar genetically despite the stark differences in their outward appearances. So it must be just a very small portion of human genes that are responsible for everything from our upright posture to our ability to sing. What makes humans so unique?

On Jan. 14, Duke Professor of Biology Greg Wray spoke about his group’s work on the genetic and molecular processes that contribute to our uniquely human physiology and brains as a part of the Computational Biology and Bioinformatics Seminar Series.

“Humans are not the best model organisms since there is a limit to what you can do genetically and mentally. You can’t really make a human knockout (but sometimes, nature makes it for you),” Wray said.

Still, humans are immensely important to study for practical reasons. We have uniquely human courses of disease in part due to our physiological, cognitive, and mechanical properties. Also, we’re just intrinsically curious about our own bodies.

According to Wray, the answer to human uniqueness is our regulome, the genes, mRNAs, proteins, and metabolites that regulate which genes are turned on when.

This graph shows the two major shifts in diet (meat-rich diet and grain-based diet) that likely contributed to our divergence from chimpanzees and thus differential gene expression. Source: Greg Wray

One prevailing hypothesis is that human forerunners likely began diverging from chimps about 2 million years ago when we took on a meat-rich diet in the savannah. The ancestors of chimpanzees retreated to the rainforest to eat a diet consisting mostly of fruits. Our meat-rich diet seems to coincide with an increase in brain size. And today we metabolize fats much differently than chimpanzees.

Wray’s lab studied the effects of dietary changes on five tissue samples – the cerebral cortex and cerebellum of the brain, liver, fat, and skeletal muscle. What seems to have changed in chimps versus humans are genes related to neural functioning, development, and metabolism. For instance, 31 of 61 genes involved in insulin signaling are operated differently in chimps and humans. These differences in gene expression may also explain why humans are uniquely susceptible to diet-related illnesses like type II diabetes.

On the other hand, genes involved in the transcription, translation and replication of DNA, RNA processing and protein localization haven’t changed in chimps versus humans.

Fat cells also behave differently in humans versus chimps. Wray’s lab took adult stem cells from adipose tissue in both chimps and humans and challenged them with either more oleic acid (the main fatty acid in a meat-heavy diet) or more linoleic acid (the dominant fatty acid in a grain-based diet). The enzymes involved in fatty acid synthesis were more common in human adipose tissues. Wray believes that the increased fatty acid synthesis is probably responsible for building and fueling a larger human brain.

Another major shift in diet occurred during the agricultural revolution, which introduced omega-6 fatty acids into our diet along with pro-inflammatory compounds. Wray explains that the increase in grains from the shift in diet likely contributes to chronic pro-inflammatory diseases in humans, such as atherosclerosis.

“Understanding our metabolic history from an evolutionary context can potentially give us insight into some pretty prominent health concerns,” says Wray.

A Call For Action: Genetic Testing Before Prescriptions

January 12th, 2013

By Prachiti Dalvi

Structure of Codeine

Codeine is an opioid pain medication; but if you are a poor metabolizer of a particular enzyme (CYP2D6), you will experience no pain relief from this drug. However, if your doctor could administer something called pharmacogenetic testing, she would know to simply give you morphine (an active metabolite of codeine) instead. For now, this kind of testing isn’t available.

Mary Relling, PharmD

Mary V. Relling, PharmD, the Chair of Pharmaceutical Sciences at St. Jude’s Children Hospital spoke about the need to implement pharmacogenetic testing on Thursday, January 10. A number of tests have recently emerged that are ready for prime time. When we know that some drugs may have adverse effects for people with particular genetic phenotypes, it is unethical to prescribe these drugs without knowing the patient’s genetic status.

However, Relling said there are a number of barriers to integrating pharmacogenetic tests into clinical care: fragmentation of our healthcare system, a focus on sick-care rather than disease prevention, a lack of evidence for clinical utility or cost-effectiveness, complex underlying lab results, and a lack of a centralized system for recording patient information.

The best way to break through these barriers is to conduct testing preemptively, Relling said. We can simply take drop of blood when the baby is born and run genetic tests. “Genetic tests are lifetime results. It makes sense to have it in the background, just as we know a patient’s age, weight, sex, etc.,” Relling said. The barriers discussed above can be avoided to a certain extent at St. Jude’s because they have adopted a team approach to patient care and a 100% electronic system for recording patient records.

The growing affordability of genotyping makes using preemptive pharmacogenetic testing more feasible, she said. The cost of sequencing one or two genes in the past will now produce results for 225 genes. Two years ago, the Clinical Pharmacogenetics Implementation Consortium (CPIC) studied how to migrate pharmacogenetic testing from the laboratory into routine patient care. They looked for gene-drug pairs associated with potential risks of life-threatening toxicity, serious adverse effects, or lack of effectiveness. Eleven of the genes CPIC determined met the threshold for high-risk were found to have profound effects on 33 drugs.

Relling said approximately 48% of patients receiving drugs at St. Jude’s received orders for at least one of those pharmacogenetically high-risk medications.

She said the question now is how to use genetic test results rather than whether a genetic test should be ordered. In the coming years, we will have to address how to maintain the fine balance of providing the clinician with enough information to treat the patient and overwhelming the patient with genetic testing results that are difficult to interpret.

This lecture was a part of the Genomics and Personalized Medicine Forum sponsored by the Duke Institute for Genome Sciences and Policy (IGSP).

A Passion for Research

November 19th, 2012

By Prachiti Dalvi

Akash Shah, Trinity ’13

“Research enables me to think about a question that excites me and helps patients,” says Trinity senior Akash Shah. A biology major, philosophy minor, and a candidate for the Genome Sciences and Policy Certificate, Akash became interested in genomics as a freshman in the Genome Focus. Originally from Fullerton, CA, Akash was drawn to Duke because of its its immense biomedical research enterprise. He also loved the fact that at Duke, the medical school, law school, and business school were on the same campus as the undergraduate campus.

Intrigued by the research his professor Dr. Hunt Willard was conducting, he asked to get involved. His work in Dr. Willard’s lab dealt with artificial human chromosomes. More specifically, he was working with others in the lab to identify which regions of the chromosome would be deleted when transformed into human cells.

Now, Akash works in the Nevins Lab, where he looks at candidate genes in the epidermal growth factor receptor (EGFR) pathway: an important pathway in many cancers. When growth factors bind to the external portion of the receptor, the receptor becomes activated. Side effects of receptor activation include tumor growth and metastasis. When scientists target genes associated with this pathway, they can increase tumor cells’ sensitivity to pathway inhibitors and better prevent tumor cell reproduction.

The advent of computational genomics has allowed for major advances in the field. Fifteen to twenty years ago, cloning genes was considered a PhD project, and now, it is something an undergraduate can do.

Akash’s favorite aspect of research at Duke is its collaborative nature. Faculty members work with another and across departments. His research is not limited to labs at Duke. In fact, he as also worked with professors at UCLA and Harvard. The culture of research varies from one university to the next; thus, Shah encourages undergraduates to do research at different institutions. “It gives you a chance to succeed in different cultures.”

When he is not in the lab, Akash enjoys playing cricket and exploring local restaurants with friends. During his time at Duke, he has been involved with numerous organizations, and has become an integral part of the Genome Research and Education Society (GRES). During his sophomore year, he founded a program in which undergraduates shadowed other undergraduates doing genomics research. In order to make research more accessible to undergraduates, Akash has helped organize career talks, including MD/PhD information sessions. After graduating from Duke in the spring, Akash hopes to begin medical school, and eventually pursue a career in academic medicine so he can continue conducting research. He has worked extensively in cancer genomics research and hopes to explore cancer stem cells in the future.

When the Genome Gets Personal

November 15th, 2012

By Nonie Arora

It has been almost ten years since the first draft sequence of the human genome was completed in 2003, and some patients are starting to see benefits in clinic.

Dean Nancy Andrews

Dr. Nancy Andrews, Dean of the School of Medicine, recently spoke to undergraduate students about “When the Genome Gets Personal” over a hearty dinner of chicken stuffed with goat cheese, rice pilaf, and caramelized brussel sprouts. She was the latest guest in the 2012 Chautauqua West Lecture Series.

Andrews explained how DNA sequencing analysis can lead to a new diagnosis for patients. Even if the disease is not treatable, a diagnosis can mean a lot to patients and families, said Andrews.

“We are pushing boundaries between taking care of patients and doing research. The lines are blurry,” she said. Researchers want to sequence patient DNA to find causes for genetic diseases and, at times, to help individual patients who don’t have a diagnosis, according to Andrews.

She said it is easy to find variations in DNA sequence, but much, much harder to know how to interpret the changes. One of the tricky situations researchers face is telling parents or patients what they have found when they are not certain of the finding’s significance. She chairs a committee at Duke that is working on standards to help guide researchers to know what to report and how to design informed consent forms.

Andrews said DNA sequencing is already being used in clinical care: about $5 billion a year is spent on clinical sequencing. However, this sequencing is highly focused on genes relevant to the clinical situation; insurance companies will not yet support exploratory whole-genome sequencing. Andrews pointed out that there is a potential for exploitation by private for-profit companies with DNA sequencing capability, which may overstate their claims or capabilities.

Complicated scenarios can arise when sequencing is done in families. Among other issues, “There is a very real possibility of learning dad is not the biological father,” Andrews said.

Example of a pedigree generated from discussion of family history with patients, modified from Wikimedia Commons

Andrews said that clinical geneticists are going to need algorithms for interpreting sequence data and standard principles for revealing information for patients. These are under development at Duke and across the country.

Ultimately, Andrews thinks that personalized medicine “shouldn’t just be about genetics and genomics but [it should] also incorporate many other types of clinical data, including imaging studies and patient preferences, as well as a deep understanding of environmental factors.”