Not Just Neighbors

Joshua Tashman couldn’t divorce himself from his passion for engineering as he explored medical schools nearly a decade ago. So, he decided there was only one way to round out his formal education: He’d marry medicine and materials science.

That choice led him to Pittsburgh, where the University of Pittsburgh School of Medicine and Carnegie Mellon University run the Medical Scientist Training Program (MSTP), through which he earned his Ph.D. in bioengineering in 2021 from Carnegie Mellon and his M.D. from Pitt in 2022.

“The real benefit here is that they’re different schools with different expertise,” says Tashman, grateful that he had the chance to learn from and train under brilliant faculty at both universities.

The universities share broad research interests and overlapping expertise in several fields — think neuroscience, robotics and bioengineering — and partner in programs beyond the MSTP.

Carnegie Mellon graduate student Kendra Noneman has a Pitt faculty advisor through the Center for the Neural Basis of Cognition, where researchers at both universities investigate the cognitive and neural mechanisms that give rise to biological intelligence and behavior. Working with J. Patrick Mayo, a PhD assistant professor of ophthalmology, gives her a chance to see clinical research firsthand as she studies where animals are looking based on neurons in the cerebral cortex.

Emily Lopez, another Carnegie Mellon student co-advised by Mayo, was excited that the partnership with Pitt connected her with a broader neuroscience community in Pittsburgh.

“It just increases the amount of and types of research I am exposed to and the people I can meet, which makes for a richer graduate student experience,” says Lopez, whose optogenetics research is part of a collaboration between Mayo and her Carnegie Mellon advisor, Matt Smith, a professor of biomedical engineering and codirector of the Center for the Neural Basis of Cognition. “In all honesty, I tend to lose track of which professors and students are from Pitt and which are from CMU; we are often mixed together at various events and even within labs.”

Together, the two universities are nurturing the next generation of biomedical leaders as their faculty pursue life-changing research.

“I think that over the years, the two universities have collaborated effectively to attract the best students to the city because of their complementary research strengths,” says Saleem Khan, Ph.D. associate dean for graduate studies and academic affairs and professor of microbiology and molecular genetics in Pitt’s School of Medicine. “This is a great thing, because it helps both universities attract and recruit faculty and expands the breadth of their research.”

He says it also gives faculty bargaining power in attracting grant funding. Today, research projects at the School of Medicine involve collaborators from Carnegie Mellon more than any other institution outside of Pitt. In fiscal year 2023, their collaborations included 65 principal investigators and spanned 48 awards with a total worth of over $15.8 million.

Theresa Mayer, vice president for research at Carnegie Mellon, notes that, overall, “the University of Pittsburgh is by far and away CMU’s most frequent and deepest partner in research. Research at CMU is supported from funding awarded in partnership with the University of Pittsburgh more often than through any other source, outside of direct funding awards from major federal agencies like the National Science Foundation and the U.S. Department of Defense.”

And of course, Pitt and Carnegie Mellon share a neighborhood. Many faculty say they can’t immediately name another instance of having one of the country’s best medical schools sitting a few blocks from a top computer science school. In the Boston area, you’ll find a close case in Harvard Medical School and MIT. Yet those schools sit roughly two miles apart.

James Faeder, a Ph.D. Pitt Med associate professor of computational and systems biology and Pitt’s program director for the joint Carnegie Mellon – University of Pittsburgh Ph.D. Program in Computational Biology (CPCB), says there are intangible benefits to Pitt and Carnegie Mellon’s proximity, as well. Faculty from the two universities don’t just work in the same neighborhood, they often live in the same neighborhoods and socialize outside the lab. And on the student side, he points to the pride they take in their joint education:

“Our students view themselves as alumni of both schools.”

The serendipitous geography has made it easy to forge alliances that maximize complementary strengths in medical research. Dozens of Pitt students as well as dozens of Carnegie Mellon students are enrolled in formal programs between the universities.

Examples include the aforementioned MSTP and CPCB programs; the Molecular Biophysics and Structural Biology graduate program, where students can pursue research in disciplines ranging from cellular biophysics to virus structure and nanomachinery; plus the Program in Neural Computation, where computationally minded students can move seamlessly among neuroscience labs at both universities. And there are more informal collaborations than you can shake a pierogi at.

“This is a special part of the Pittsburgh community,” says Douglas Weber, a Ph.D. and the Akhtar and Bhutta Professor in mechanical engineering and neuroscience at Carnegie Mellon. “There are elite institutions that are unable to collaborate between departments. But we’re able to do that here across institutions.”

An emerging generation of physicians and scientists has taken advantage of this city’s extraordinary academic offerings to enrich health care—not only in Pittsburgh, but well beyond. What follows are stories from three of these investigators.

Now in 3D

When Joshua Tashman tells you what he accomplished as part of a research team at Carnegie Mellon, you can understand bioengineering’s appeal to him, especially after he tells you a little about himself.

The self-professed tinkerer used 3D bioprinting to engineer models of the human heart—including one that pumped for weeks.

Tashman, who earned a bachelor’s in mechanical engineering from Cornell University, played an integral role on the Carnegie Mellon team that 3D-bioprinted a functioning tubular heart that is similar to one in embryonic development. They took cardiomyocytes and cardiac fibroblasts—the cells that make up the heart muscle and the ones that produce connective tissue, respectively—that were derived from stem cells and printed them in a tube made from collagen. The tube spontaneously began contracting within a few days of construction and pumped for roughly a month.

Tashman’s most recent work in Pittsburgh was in Pitt Med’s Division of Renal-Electrolyte under the direction of Cary Boyd-Shiwarski (MD ’12, Res ’14, Fel ’16), an assistant professor of medicine. There he investigated how potassium depletion can lead to kidney injury.

Before he headed to Boston over the summer for a Mass General Brigham Combined Residency in pathology, Tashman explained that cell culture generally is two-dimensional, so it has limitations: The lab-grown cells aren’t exposed to the same environment that they are in the body.

“They’re exposed to a mechanical environment and a flow environment that is specific to the geometry and physiology in the human body,” he says of kidney cells. “These cells line a tube in the body with a flow [of water and solutes] passing over them. And these things are very important for the kidney cells to think that they’re in a kidney and that they should behave like kidney cells.”

So, when running experiments on how potassium levels affect the creation and excretion of ammonia, the most common lab model doesn’t reproduce many of the same biomarkers found in the kidney.  “The idea then is, can we 3D print something that allows us to recreate the mechanical and chemical cues?”

Tashman collaborated with Daniel Shiwarski, a Ph.D. and Pitt assistant professor of bioengineering and medicine who also studied at Carnegie Mellon (and is married to Boyd-Shiwarski), to focus on engineering the system as closely as possible so that proximal tubule cells grow on it.

Tashman had teamed up with Daniel Shiwarski before. It was while earning his Ph.D. that he and Shiwarski worked in the group that constructed the heart models — that was in the lab of Adam Feinberg, the Arthur Hamerschlag Career Development Professor in the Departments of Biomedical Engineering and Materials Science and Engineering at Carnegie Mellon. The researchers made computer models of tissue and organs and recreated them with printed biological materials.

Feinberg encouraged the team to spread their newly acquired know-how to biomedical engineers everywhere, Tashman says. “He published open-source papers. So, he would let me make step-by-step guides, and we would give all the designs away so that other people could actually build everything that we designed.

“I’ve seen people on Twitter, for instance, post videos of things that I’ve helped design that they printed out and built themselves. That’s been pretty cool.”

Big Data, Big Opportunities 

Eric Strobl (Ph.D. ’17, M.D. ’19) remembers reading about the human brain in high school and thinking it was the most interesting subject he’d ever encountered. He knew he wanted to study it. But as he got to college, he found himself wondering what he would do with all the knowledge he was accumulating.

Computer science and biomedical informatics gave him an answer. Strobl liked the idea that he could contribute to medical knowledge — and maybe even identify new treatments — by mining big datasets; the approach helps researchers find answers more quickly and economically than clinical trials would. “I was not convinced that the phrase, ‘Correlation does not imply causation,’ was the end of the story,” he says. He wanted to see if stronger conclusions could be drawn from big data, particularly from large sample sizes.

Strobl went on to pursue his M.D. and Ph.D. through the MSTP; he found himself pulled toward an area of study called causal discovery — a way of determining cause-and-effect relationships from big data. Researchers like Strobl’s advisor Shyam Visweswaran, an M.D., Ph.D. Pitt professor of biomedical informatics and a neurologist, develop algorithms to tease out causal connections waiting to be found in medical data like electronic health records — and use those takeaways to come up with new treatments.

As he got deeper into his research, Strobl’s ideas caught the attention of Peter Spirtes, a Carnegie Mellon philosophy professor and one of the originators of causal discovery.

“I was very skeptical of what he was saying, and it made me want to talk to him in more depth,” Spirtes says. “And he convinced me that he did know what he was doing.”

Strobl began collaborating with Spirtes and Kun Zhang, also of Carnegie Mellon’s philosophy department. Many faculty in that department look for ways to apply technical approaches in mathematics, statistics and computer science to big questions about knowledge.

Just as Pitt’s proximity to Carnegie Mellon allowed Strobl to work with some of the foremost experts in the theoretical side of causal discovery, Spirtes and Zhang appreciated the chance to work with those embedded in the world of patients and treatments.

“[Strobl] was really motivated by real problems,” says Zhang. “We’re working on the ideas and the methodologies. So his contribution was essential to our work.”

Strobl’s PhD dissertation would demonstrate new ways of thinking about feedback loops in causal discovery that combined approaches from both biology and computer science. Philosophy and physics treat feedback loops—where a system’s output either spurs on or hinders the system—as if two things caused each other simultaneously, which doesn’t reflect what’s happening in the body.

So Strobl developed a new algorithm that outperformed a commonly used one in its predictions. He also took into account the different stages of a disease that may be reflected in data from a sample. The dissertation won the Drs. S. Sutton Hamilton MSTP Scholar Award, recognizing Strobl’s contribution to scientific literature.  

Strobl also published papers that mathematically defined and identified root causes of disease from data with the goal of applying the methods to psychosis. Psychosis is complex—which lends itself well to Strobl’s approaches. They’re helpful in sorting through the manifold factors involved in a disease.

Strobl is now a psychiatry resident at Vanderbilt University Medical Center, but he won’t be away from Pitt for long. In July 2024, he’ll join Pitt’s faculty as an assistant professor of biomedical informatics with a secondary appointment in psychiatry. Visweswaran is excited that his onetime advisee will be the Department of Biomedical Informatics’ first faculty member with a clinical appointment, helping to apply their work to pressing, real-world problems. He also hopes Strobl’s arrival will mark the beginning of a renewed interest in causal inference at Pitt, which they expect to expand on with future hires.

Strobl’s research will use genomic data for causal inference to address autism and neurodevelopmental disorders. He’ll continue to work with Carnegie Mellon’s Spirtes and Zhang to fine-tune his algorithms. Pitt, he says, offers a collaborative community that brings together his different interests.

“It’s very rare to have a lot of expertise in both medicine and causal inference in the same place,” he says. “It’s also rare, I think, to have a combination of people who are interested in biology and clinical medicine. Usually, the communities are very separate.”

Coinfection Collaboration

Janie French (Ph.D. ’23) first became interested in microbiology after a harrowing spring break with her rowing team at the University of Wisconsin, Madison. On a training trip in Oak Ridge, Tennessee, the whole team came down with norovirus. Having no idea at first what was causing the collective gastrointestinal distress was the scariest part. “Knowing what we had was the first step to feeling better about it,” French says. (Thankfully, they made it back home and healthy after a few days of misery and a canceled race.)

The experience, along with an undergraduate class in virology, set her on the path toward a PhD in microbiology and immunology. French is now a medical science liaison for HeathTrackRx, a company that performs diagnostics for infectious diseases. As a graduate student, she worked in the lab of Seema Lakdawala, then a Ph.D. assistant professor at Pitt who studies influenza A, which has been the culprit behind some of history’s worst pandemics and regular seasonal epidemics.

“When I was rotating [labs], I really liked the translational aspect of Seema’s work,” French says. Studying ways to mitigate transmission, “you didn’t have to search very hard to connect our work to improved health outcomes.” (Lakdawala has since moved to Emory University.)

French found herself interested in studying how the body battles two infections at once, with a project focused on coinfection of influenza virus with the bacteria Streptococcus pneumoniae. The latter, a major cause of bacterial pneumonia and other infections, has been known to worsen death rates during flu outbreaks, including the 2009 H1N1 pandemic. French wanted to understand more about how and why that happens.

“As a flu lab studying transmission, we wanted to know whether the pathogens truly were synergistic. And if they were, what exactly was going on there?” French says.

To learn whether and how these infections were collaborating to wreak havoc, French had to do some collaborating of her own. She turned to an expert on S. pneumoniae: N. Luisa Hiller, the Eberly Family Career Development Associate Professor of Biological Sciences at Carnegie Mellon. 

The team also included Pitt’s Valerie Le Sage, a Ph.D. research assistant professor, and Lakdawala, as well as Karina Mueller Brown, a Ph.D. who was a Carnegie Mellon graduate student at the time. Together, they looked at coinfection in ferrets, which are naturally susceptible to flu, have a respiratory tract more similar to humans than mice do, and cough and sneeze like humans.

As they reported in a paper published in 2022, they found in ferret models that getting the bacterial infection after the virus led to more severe symptoms. But, surprisingly, it didn’t increase transmission of the virus or the viral load — they actually found less virus in nasal washes of coinfected animals.

Why? One clue might be in mucus crusts the researchers noticed on the coinfected ferrets’ noses, not seen during flu or pneumonia infection alone. More study is needed, French says, but “it’s possible the virus is getting stuck in mucus deeper within the nasal passages.” It’s likely that coinfection prompts a heightened immune response.

In a paper published in 2023, they looked at aerosols expelled by coinfected ferrets through their coughs and sneezes. Surprisingly, says Hiller, it appears that when both pathogens are in the same released droplets, they can influence how long each persists outside the body.

The joint expertise was critical to the work — and Carnegie Mellon’s proximity to Pitt made it easy to exchange samples, get feedback and share equipment.

“We couldn’t have done the project without them,” says French.

“And from the starting experiments, it blossomed into additional questions we were interested in asking and future directions we could take.”

Says Hiller, “The right collaborations drive scientific discoveries.” 

Photography by Aimee Obidzinski/University of Pittsburgh and Erin O. Smith/Vanderbilt University Medical Center | Silhouettes by Frank Harris
This story appeared in the Winter '23-'24 issue of Pitt Med Magazine.