Think of a cell as a factory. Maybe a factory that makes flat-screen TVs, he suggests. Then along comes a virus, an inert chemical with instructions for completely rewiring the circuitry. 

 

“It changes the job of the people in the factory,” says Dermody. “So now, instead of making flat-screen TVs, the cells make thousands and thousands of iPhones.” 

 

The problem with this shift in production, Dermody says, is that it uses up all of the cells’ basic building blocks, forcing the factories to shut down. One by one, as these factories close, the whole community—or related tissues, like those of the liver or the heart—shuts down. 

 

“How can that be?” Dermody asks. “How does that little inert package recognize a factory and get in there? How does it do the rewiring business? How does the assembly process take place?”

 

These questions drive Dermody’s current research. Trained in virology, Dermody now focuses on teaching and discovery in his roles as chair of pediatrics at Pitt’s School of Medicine and physician in chief and scientific director at UPMC Children’s Hospital of Pittsburgh. 

 

His lab team, notably a couple of doctoral students, has worked to decode weak links in viral replication. The goal? To uncover potential therapeutic targets that would disrupt the replication cycle and inhibit viral infection.

 

Last March in Nature Microbiology, Dermody’s team published findings on a late-stage viral replication process that had not been elucidated previously. They showed that a protein complex in the host cell, called the TRiC chaperonin, guides (or chaperones, if you will) the folding process of the virus’s outer shell, which then creates new viral particles that go on to infect other cells. 

 

In other words, says Dermody, “These are the final steps of the iPhone manufacturing process—the transport of the iPhones to distributors, then the sale of the iPhone, so people have it in their hands.” 

 

This discovery was born out of an “unbiased” genetic screen that was designed by Jonathan Knowlton, with assistance from Paula Zamora—two of Dermody’s students. They used the screen to identify cellular-protein factors that reovirus, a relatively simple virus that is efficient at factory reorganization, requires to replicate. The screening is called unbiased because the researchers have no idea what they will find. 

 

“We just cast a broad net out into the sea of potential host factors,” explains Dermody, “and reeled in what was collected, took a look at the candidates, and ranked them in terms of a priority—which ones were most likely involved in a process of interest and which ones we thought were probably false positives.” 

 

Interestingly, the researchers knew very little about the TRiC chaperonin before viewing the results of this screening. “You could write what we knew about TRiC on your thumbnail,” said Dermody. They called upon Judith Frydman—a biochemist at Stanford who discovered the TRiC chaperonin protein more than 25 years ago—and cell biologist Cristina Risco in Spain to troubleshoot how to find the precise link between TRiC and the viral assembly pathway. “No one had been able to show that before. That was the main contribution of our paper, and why it was published in Nature Microbiology,” says Dermody.

 

Now Dermody and the team are posing three questions they hope will reveal more of a virus’s instruction manual on hostile factory takeovers. One: Do other viruses require TRiC to fold their outer shell? That is, is this process generalizable across all types of manufacturing? Two: Can they uncover the complete assembly pathway used to produce the reovirus particles? And three: As it turns out, TRiC is an essential protein for the host and cannot be a target for antiviral therapy. So, are there other aspects of the assembly line that could be a disruption point for treatment?

“We want to play this out in silicon first,” says Don Burke, an MD and dean of the Graduate School of Public Health who’s also a professor in the School of Medicine. “If we could take the same simulation methods that we’ve developed for contagious epidemics and start using them with the opioid epidemic, we might make some headway.” 

 

That’s the ambitious goal of the Public Health Dynamics Laboratory. It’s gained the endorsement of the Centers for Disease Control and Prevention, which gave the project a two-year $1.5 million grant to teach infectious disease simulators to model and play out solutions to the opioid epidemic in pixels first. 

 

In a lab in the Graduate School of Public Health lives a computer. And in that computer lives a representation of every person, every family, and every community in the United States. It’s accurate to the U.S. Census tract, and knows the gender, race, and socioeconomic status of every simulated person and the relationships between them. The Framework for Reconstructing Epidemiological Dynamics (FRED) was the brainchild of Burke, who is no stranger to tracking an epidemic. An infectious disease physician, Burke cut his epidemiological teeth in the HIV epidemic. With HIV, however, it’s a little easier to simulate how the disease spreads. After all, we know how HIV is transmitted, and we can witness a strain of HIV morph over time, showing us which viruses are related to others in the community and a given strain’s footprints through a population.

 

“Drugs aren’t infectious organisms,” says Burke. “But they do have transmitting properties to them.” 

 

For the last two years, Mark Roberts—an MD, MPP, director of the Public Health Dynamics Laboratory, chair of health policy and management, who’s also a professor of medicine—and his team have been adapting FRED from an infectious disease model to a more general tool for modeling population dynamics. 

 

“It doesn’t necessarily care anymore what specific dynamic it’s modeling,” Roberts says.  

 

So Roberts and his team are trying to pull together available data—that includes figures on opioid prescriptions, opioid overdose deaths, and incidences of injection drug–related diseases and of infants born with neonatal abstinence syndrome.

 

But the rest of the story will be harder to tell. The most common drug implicated in overdose deaths by medical examiners is “unspecified drug”—not very helpful. Law enforcement is not always forthcoming with incarceration data. And efforts to pinpoint the exact mix of street drugs that cause a given overdose are in their infancy. 

 

Yet the researchers already are offering new insight. One finding was featured in Science in September, with Pitt Public Health Dynamics Lab’s Hawre Jalal, an MD/PhD, as the first author. By casting a wide net on drug overdoses generally (including cocaine, methadone, heroin, prescription opioids, and fentanyl), the team learned that deaths from overdoses have been rising exponentially for at least 38 years, with different drugs taking the main stage in a series of subepidemics.

 

Sometimes, building a simulation, Roberts says, “Can direct where we need to do more research.” Plus, he adds, it could also help researchers narrow in on something that’s eluded them for years: What actually causes addiction—and what is just a bystander to the process? 

 

As it is, it will probably take years for the model to accurately simulate the opioid epidemic, and then it will be possible to test the solutions that lawmakers are proposing. It’s daunting, but it’s a project worth doing, says Burke. 

 

“We approach this with humility. This is an imperfect art. Our job is to help policy makers make better decisions, not perfect decisions.”