A Molecular Scalpel

Summer 2018
MOTOR is a tumor registry and tissue bank at Pitt designed to spur discovery in pediatric osteosarcoma research. ABOVE: A cell population grows from a registry-derived primary tumor sample.
When orthopaedic oncologist Kurt Weiss tells a family that their child has osteosarcoma (OS), he gives them these words to live by: “treatable and curable.” As evidence, he shares his own story. A defensive end on his high school football team, Weiss was diagnosed with OS at the age of 15 and treated at UPMC. In addition to 25 operations, including amputation of his right leg, Weiss  had metastases to his lungs, which pushes a child’s five-year survival prospects from 75 percent to less than 30 percent. “I say, If we find disease in the kid’s lungs, is it time to freak out?” says Weiss. “The answer is no. I pull down the top of my shirt and show them the scar on my chest. It’s obviously not good, but it’s not insurmountable. There’s nothing that will change our goal of curing a kid.”
Osteosarcoma survival rates haven’t budged since 1990, when Weiss participated in the clinical trial he credits with his own remission. And so in addition to his work as a clinician treating children and adults with both primary and metastatic bone cancer, the associate professor of orthopaedic surgery founded Pitt’s Musculoskeletal Oncology Laboratory to pursue basic research into the pathogenesis of osteosarcoma, the third-most common cancer diagnosis among people younger than 20. In 2011, soon after he joined the Pitt faculty, Weiss (Res ’08) and colleagues established MOTOR, a tumor registry and tissue bank designed to spur discovery. 
“Lots of people have patient databases,” says Weiss. “But it’s being able to capture patient samples that makes us unique. And it makes us strong, because you can take those biological specimens, and [those patients’ outcomes], and you can use that combination to ask intelligent questions.”
Weiss has focused his own inquiry on the role of aldehyde dehydrogenase (ALDH), a stem cell factor that, combined with a cell signaling pathway that most multicellular organisms have, called Notch1, facilitates metastatic invasion of the lungs by osteosarcoma. After he’d mapped out the basic mechanisms using cells cultured from patient samples, he used a mouse model to test his theories. “But none of my patients are mice,” says Weiss, so he went back to MOTOR to analyze the correlation of ALDH in tissue samples with each patient’s disease trajectory. “The level of ALDH expression in cultured cells perfectly correlated with metastasis,” he says. “It seems like we might be on the right track.” 
Weiss at the Grand Canyon in 1998.
More recently, he’s begun investigating whether known ALDH inhibitors can slow metastasis. In 2015, Sarcoma published his finding that retinol (vitamin A) altered the expression of metastasis-related genes and decreased the proliferation, invasion capacity, and resistance to oxidative stress of metastatic cells. 
In a study now under way, Weiss is testing the effect of disulfiram, a compound approved by the FDA as a treatment for alcohol addiction (marketed under the brand name Antabuse). If the compound works, it would mean a level of precision targeting of cancer cells that’s a far cry from the current standard treatment, broad-spectrum chemotherapy that ravages every fast-dividing cell in a kid’s body. 
“It’s napalm,” says Weiss. “I want a molecular scalpel.”
Steffi Oesterreich is professor of pharmacology and chemical biology at Pitt and a member of UPMC Hillman Cancer Center. She teamed up with Weiss in 2013 to add samples from people with breast cancer—which can metastasize to the bones—to MOTOR. In 2016, Clinical Cancer Research published an analysis showing that mutations in the estrogen receptor gene ESR1, which is implicated in resistance to endocrine therapy, were over-represented in metastatic tissue. 
“Kurt’s studies are very comprehensive—from the nitty-gritty mechanistic details to cells, to mice, to treating patients,” says Oesterreich. “He’s obviously super busy clinically, but he’s absolutely excited about research.”

Split Revision

A new understanding of how the brain controls motor function


It turns out that the central sulcus does not neatly divide brain function, as the textbooks say. (Illustration by Michelle Leveille)

Four decades ago, a neuroscientist at Johns Hopkins University put forth a revolutionary new idea—and took a lot of heat for it. Some said he lacked evidence.  “When he proposed his command hypothesis, Vernon Mountcastle got a lot of grief,” says the University of Pittsburgh’s Peter Strick, a PhD, Thomas Detre Professor of Neuroscience, Distinguished Professor and chair of neurobiology, and scientific director of the Brain Institute. 
It might have taken 40 years, but Strick and colleagues finally vindicated Mountcastle and validated his work. The original hypothesis? That the central sulcus—a prominent fold in the middle of the brain (right about where a tiara might sit, between the anterior and the posterior portions)—does not, in fact, neatly divide the brain between motor control in the front and somatosensory in the back. (The somatosensory encompasses our senses of touch, position and movement, and immediate physical surroundings.) That clean division of labor was in all good anatomy and physiology textbooks at the time.
Instead, Mountcastle proposed that a region called the posterior parietal cortex, an area behind the central sulcus, had its own direct link to the spinal cord, giving it a major role in motor control and hand movement. 
This was a striking new theory, Strick explains. Mountcastle said that the posterior parietal cortex wasn’t just involved in perception—it also had a command role. “This was just a dramatic change in people’s notion of the function of the cortical area,” says Strick.
He met Mountcastle as a young postdoctoral fellow, and years later, Strick independently conceived of a way to define the cortical areas that connect directly to the spinal cord to command motor function. 
Since then, Strick has found eight different pathways that control motor neurons—from the premotor and motor cortex in front of the central sulcus to some areas behind the central sulcus, including the posterior parietal cortex, as Mountcastle postulated years ago. 
“This is an important aspect of motor control—that there isn’t just one route to Rome,” says Strick. These different pathways may be important for some people with brain or spinal cord injuries.
For example, Strick explains, “If you have a stroke, and it’s limited to one of these cortical areas, there are other cortical areas that have access to the spinal cord. This can play a role in compensating for the stroke and promoting recovery of motor function.” 
In addition, implanted electrodes in the posterior parietal cortex of the brain have shown promise in sending signals to prosthetic devices, giving people who are paralyzed the ability to move again. 
While advances in the clinic continue, Strick is actively researching these pathways, digging deeper to understand: Why do we have so many pathways to control motor neurons? What does each of them do? And how do they do it differently? Why are multiple systems dealing with the task of motor control?
Mountcastle lived to be 96. Fortunately, he was able to learn about Strick’s experimental techniques—they kept in contact throughout the years. 
However, Strick regrets that Mountcastle was not able to see his hypothesis borne out in the pages of Proceedings of the National Academy of Sciences, in a paper penned by Strick last year.
“I wanted to send him the paper before I submitted it,” he says. “It could have been titled ‘Mountcastle Was Right!’ He would have enjoyed the results.”


The Tortoise and the Hare

How cancer wins the race in MCV infection


The Chang-Moore team was the first to observe Merkel cell virus replicating in a lab. Here, their mutant virus (left, in red and yellow) infects a new crop of cells (blue, on the right).

Yuan Chang and Patrick Moore are virus-hunting superstars. Ten years ago, they discovered—not their first, but their second—cancer-causing virus. (Only seven are known.) The virus is called Merkel cell polyomavirus (MCV), and it is responsible for 80 percent of Merkel cell carcinoma cases. Although this cancer is rare—only about 1,500 cases per year at the time of the virus’s discovery in 2008—Merkel cell carcinoma is on the rise, and it’s one of the deadliest forms of skin cancer. 
To learn how this virus works, researchers in the Chang-Moore lab tried to infect cells in a petri dish with the virus’s DNA, thinking they could observe MCV’s life cycle. But it didn’t work. Despite the crew’s best efforts, “no one [had] ever seen MCV replicate” by infecting cells with the viral genome, says Moore. “It’s just always been dead, and we didn’t know why.” 
Then, a researcher in the Chang-Moore lab named Hyun Jin Kwun, who now has her own virus lab at Penn State, noticed that the protein responsible for MCV replication—called LTag (large T antigen)—has several tags that signal for its own breakdown. She mutated the protein to remove those tags and tried to infect cells with the virus again. And then, it worked. “If we make these mutations,” says Moore, “we see—boom!—[the virus] is actually replicating, and now it can be transmitted.” By eliminating these degradation signals on LTag, he says, “I can do a better job of making a replicating virus than Father Darwin. Why?” 
It turns out that no one in the lab had ever caught the virus replicating because, without LTag, the virus lies dormant, or latent, in cells. Unlike, say, a highly contagious virus like influenza, which replicates and moves on to the next cell at a rapid pace, MCV “stays there like a little rock,” says Moore. It only replicates when the host cell copies its genome—which takes time. Kwun’s mutation of the virus gave LTag the freedom to do its thing, which is prying open DNA and allowing it to be copied. That’s when the slow-mo switched to fast-forward.
Understanding how latency hides MCV in cells is key, says Moore, because “most people are infected with this virus, and it’s harbored in their skin. Once you’re infected it’s a lifelong infection.” But most of us don’t have Merkel cell carcinoma, he assures, “because it’s a silent infection.”  
The team found that cancer arises through a series of unfortunate events: Stress in the host signals can stop the LTag breakdown signals (this is like what Kwun did in the mutants), and the virus begins replicating fast. As the virus copies itself, it can make mistakes. “You get incomplete replication that occurs, … genetic fragments that occur,” says Moore. The host human cells “stick those fragments into chromosomes to try and deal” with the extra DNA. If the right fragment is knit into the human genome, the cell becomes cancerous. These findings were reported in PNAS in May 2017. 
For most of us, this series of unfortunate events doesn’t happen. The cancerous DNA fragment isn’t knit into the genome, or a savvy immune system recognizes funny business in a cell. But the bodies of immunosuppressed patients, like those getting an organ transplant or people with AIDS, aren’t surveilling for cells that behave oddly. So if a cancerous integration of the viral DNA occurs, their defenses are down. That’s how many Merkel cell carcinoma cases start.
Merkel cell virus images reprinted with permission. “Protein-mediated viral latency is a novel mechanism for Merkel cell polyomavirus persistence,” Hyun Jin Kwun, Yuan Chang, Patrick S. Moore. Proceedings of the National Academy of Sciences May 2017, 114 (20) E4040-E4047; DOI: 10.1073/pnas.1703879114