Follow Arthur Horwich into his first-floor laboratory at Yale’s Boyer Center for Molecular Medicine, and you will quickly see where status ranks among his priorities. Dressed for comfort in khaki corduroys and a worn North Face fleece, the 61-year-old medical geneticist drops his bag on a cluttered corner desk and reaches for a tray of DNA samples left on the lab bench in an overnight experiment. He grins from behind round wire glasses and an Einsteinian moustache. “I’m a little stunted,” he apologizes as he completes a task normally reserved for students and technicians. “I still function as a postdoc.”
Loathe to distance himself from data, Horwich keeps four microscopes (three dissecting, one confocal) in his small office down the hall. Chin-high piles of manuscripts are stacked against the far wall; he is on the editorial boards of three scientific journals. Photos of his wife, three grown children, and lab members, along with a former patient’s yellowed thank-you letter, decorate his bulletin board. What’s missing is any evidence of the many honors he’s received—including the prestigious Albert Lasker Basic Medical Research Award, which he shared in 2011 with Franz-Ulrich Hartl of the Max Planck Institute of Biochemistry—for his game-changing contributions to our understanding of protein folding.
“In science, it’s not what did you do for me 20 years ago, it’s what have you done for me today,” says a cheerful Horwich, who has been an HHMI investigator since 1990. Today his full attention is on a mystery of biology gone wrong that has eluded scientists for decades and a disease that takes thousands of human lives every year. His goal, pursued with relentless open-mindedness, is nothing short of a cure.
The View from the Bench
On a chilly Sunday afternoon in April 1939, New York Yankees’ first baseman Lou Gehrig went to bat in the Bronx for the last time. Having played a record-setting 2,130 consecutive games, Gehrig’s power and once uncanny aim were fading visibly. He struck out. Inside his body, clumps of misfolded proteins wrecked the nerve cells that for 35 years had faithfully sent messages from his brain to the muscles operating his arms, legs, and lungs. Just a couple of months after that game, Gehrig was diagnosed with the neurodegenerative disease that would thereafter be linked with his name; he lived just over two more years.
These days, the prognosis for a patient diagnosed with amyotrophic lateral sclerosis (ALS) remains grim. For Horwich, a sad reminder of that fact came in 2003 when his children’s beloved tennis coach began limping on the court and struggled to communicate. Like Gehrig at the time of his diagnosis, this man was in his 30s and otherwise healthy; he died a year after his ALS diagnosis.
“It really affected me,” says Horwich, whose career has straddled clinical medicine and basic science. Having spent decades studying cellular machines called chaperones that help proteins fold properly into their useful forms, Horwich turned his attention to a scientific problem with a human face: “Why, in ALS and other neurodegenerative diseases, are chaperones failing to do their jobs?” he asked.
And so five years ago, Horwich transformed a laboratory dedicated to the biology and kinetics of protein folding—a field in which he is widely recognized as a pioneer—to a multidisciplinary combat zone against a brutal human disease. “I felt a deep obligation to go in this direction,” he says.
Revisiting a Classic
Although that decision required Horwich to become expert in disciplines in which he had little experience—mouse genetics, stem cell biology, neuroscience—it was also a natural extension of his past work.
Proteins are made from chains of amino acids that fold in intricate, specific ways into three-dimensional structures. The physical shape of a protein, once folded, governs its behavior. The process sometimes goes awry, however, and misfolded proteins are associated with a number of neurodegenerative and other diseases. In the 1950s, American biochemist Christian Anfinsen unfolded a protein—a common mammalian enzyme—in a test tube and found that it spontaneously refolded into its useful conformation. His conclusion, and that of most scientists who conducted protein research in the following three decades, was that proteins don’t need help from the cell to get into shape.
Page 2 of 4
Horwich was just a boy, growing up in a western suburb of Chicago, when Anfinsen conducted his famous experiment. But he remembers the day he learned about it, the day Anfinsen received the Nobel Prize in 1972. Horwich was an undergraduate at Brown University, having joined the first class of students in a program that combined an undergraduate degree with a medical degree. That evening, he says, “we went to the lab and looked for every protein we could find to duplicate Anfinsen’s experiment, it was so astonishing.” Taken as he was by Anfinsen’s discovery, Horwich could not have predicted that decades later his own work would forever amend it.
Horwich completed medical school, graduating first in his class at Brown, followed by a residency in pediatrics at Yale School of Medicine. He loved the human contact that came with clinical medicine but knew the life of a physician wouldn’t satisfy his curiosity. Lured by the California climate and a chance to study a particular virus, polyoma, with Walter Eckhart and Tony Hunter, he took a postdoctoral research position at the Salk Institute before returning to Yale in 1981. There, in the lab of Leon Rosenberg, Horwich began his work on a human enzyme—ornithine transcarbamylase (OTC)—that would prove key to overturning the dogma, grounded in Anfinsen’s work, that proteins fold on their own.
OTC facilitates the conversion of ammonia to urea in human cells, neutralizing waste. In newborns with a rare, X-linked genetic mutation, OTC deficiency can cause ammonia to accumulate. Seemingly normal at birth, the infants can fall into a coma within days. Rosenberg, Horwich, and colleagues cloned and sequenced the gene responsible for OTC, ultimately developing a genetic test that allowed patients with a family history of the disease to determine whether a fetus carries the often lethal mutation.
Capturing a Complex Picture
Horwich established his own lab at Yale and was soon joined by Krystyna Furtak, his technical “right hand” to this day, and graduate student Ming Cheng. They had learned that to do its job, OTC must first be delivered to the mitochondria, the oval organelles that supply energy to cells. Curious about how this happens, the group inserted the human OTC gene into yeast. They then created mutant forms of the yeast, looking for failures in the OTC pathway that might help them decipher the steps involved.
By then it was known that proteins generally can’t enter mitochondria in their bulky three-dimensional forms. To pass through tiny entryways in the mitochondrial membranes, they must first unfold. Once inside, before they can do their jobs, they must fold up again. One evening in 1987, after a day of examining mutant yeast for variations, Horwich and Cheng, looked across the lab bench at each other and asked a question no one else had: might the OTC protein need help folding after it has entered the mitochondrial chamber?
Anfinsen’s experiment, after all, had been conducted in the protected isolation of a test tube; the environment inside a cell is a cacophony of enzymes, chemicals, and tiny protein machines. Some of these machines were known to help refold proteins under stressful conditions that disrupt their shape, such as heat. What if such machines were also required for a protein to fold under normal conditions? Horwich and Cheng decided to look for a mutant yeast strain in which OTC made it into the mitochondria but didn’t fold properly once there. That, they guessed, would signal the absence of any such protein-folding machine if it existed.
Within days, Cheng had identified such a mutant, which seemed to confirm their hypothesis. Hartl, an expert on how proteins are imported into the mitochondria, then at the University of Munich, heard about these experiments and called Horwich to see if he needed assistance on the biochemistry side of the problem. “And of course we did,” laughs Horwich. “We were three people in a new lab who had little or no experience with yeast biochemistry and were stumbling our way around."
Page 3 of 4
Horwich shared the mutant strain, called mif4, and heard back from an excited Hartl two weeks later. “He said, ‘you’re absolutely right,’” remembers Horwich. “His work suggested that whatever was going wrong with this mutant, it had something to do with polypeptide chain folding inside the mitochondrial matrix.”
The mutation carried by this strain, it turned out, was in a gene that encodes a mitochondrial protein whose production is amped up when cells overheat—one of the class of protein machines thought to help refold proteins under stressful conditions. For that reason it had been dubbed heat shock protein 60, or Hsp60. Versions of this protein are found in nearly every cell of every organism in the world. The surprising thing was that Hsp60, in Horwich’s mutant yeast, appeared necessary for protein folding under normal conditions.
The notion that Hsp60 could be necessary for protein folding provided “enormous focus” for the group, according to Wayne Fenton, a senior member of Horwich’s lab. The researchers probed a related protein found in bacteria, called GroEL, from every angle possible to understand how it worked. “If we didn’t have the tools or the skills to answer the question we needed to answer, then we figured someone else did and we found a collaborator,” says Fenton. “Usually they would teach us a thing or two, and we’d say, ‘hey, this isn’t so hard,’ and then start doing it ourselves. Art and I share that point of view.”
A biochemist by training, Fenton was a research scientist in the Rosenberg lab at Yale when Horwich joined as a postdoc. He ran his own lab for a time but lacked patience for the administrative duties that came along with being a principal investigator. When Horwich established his own lab, sights set on OTC, Fenton joined up and has been working at Horwich’s side ever since. Both avid sailors, the two men are friends as well as colleagues.
In whatever direction the project needs him to go to become expert, he does that. It’s one of the hallmarks of a great scientist, following the trail wherever it leads.
In the 1980s and early ’90s, Horwich, Fenton, Cheng, and a growing group of collaborators focused their attention on GroEL. In a painstaking three-year effort with Yale crystallographer Paul Sigler, who died in 2000, the group elucidated GroEL’s molecular structure by x-ray crystallography. Working with Helen Saibil at Birkbeck College in London, the team captured further details of the machine in various states via electron microscopy.
Snapped together in the cell with a protein cap called GroES, the GroEL complex is shaped like a hollow bullet whose tip pops on and off to chauffeur in unfolded proteins and close them off from the cellular environment. Safely inside, the proteins fold on their own, as Anfinsen predicted. But without GroEL, the proteins clump together before they have a chance to fold, and the cell dies. The scientists named the complex a chaperonin and placed it in a class of proteins—called chaperones—believed to assist other proteins in their tasks.
“From the time of Anfinsen we thought that proteins fold and that’s all there is to it,” says Richard Lifton, an HHMI investigator and Horwich’s department head at Yale. “Art’s work completely overturned that paradigm because actually, no, if you didn’t have these chaperonin machines, the proteins would come crashing down.”
Anfinsen’s principles, Horwich hastens to point out, were alive and well throughout these experiments. Horwich’s belief has always been that chaperonins facilitate a process that happens quite naturally by creating a sheltered “changing room” for proteins much like Anfinsen’s test tube.
An Obligation to Evolve
Horwich and his collaborators still study GroEL—most recently using electron microscopy with Saibil and collaborators in California to capture images of proteins moving through the machine—but with the larger mystery of its structure and function solved, the ALS work now occupies most of his time.
Page 4 of 4
“While we were busy trying to understand the mechanism of chaperones,” Horwich says, “a whole industrious group of people was learning that many common neurodegenerative diseases are associated with protein misfolding and aggregation. It was shocking to see that a number of these aggregates are occurring in the cytosol—the cell’s watery interior—where there should be a good supply of chaperones.”
In ALS, the aggregates form almost exclusively in the spindly motor neurons that extend down the body’s limbs and operate the muscles. Despite years of study, no one knows why this happens or whether the aggregates are a cause or a symptom of the disease. Tackling those questions requires the kind of experimental openness Horwich is known for. “We just have to look at it at all levels,” he says. “Electron microscopy, light microscopy, whole animal, embryonic stem cells if they’ll work, genomics.”
On a given day, you might find Fenton bent over a microscope, his long white beard dangling as he examines fledgling motor neurons cultured from stem cells. In the next room, postdoctoral researcher Urmi Bandyopadhyay might be examining a spinal cord cross-section from a mouse in the last stages of the murine equivalent of ALS. She guides a fine laser to cut out and capture, for further study, the protein clumps associated with the disease.
Making the Rounds
If it’s early in the morning, Horwich can be found in the basement of the building next door, making his daily rounds.
Most ALS cases in people are sporadic, meaning that the family of the affected person has no history of the disease. But roughly 10 percent of cases are familial, inherited from one generation to the next. Scientists have identified a number of genes associated with familial forms of ALS, and Horwich has homed in on one of them, called SOD1, to study a mouse model for the disease. He begins every day by checking in on his colony, many members of which have been bred to express mutations in SOD1.
The SOD1 gene codes for an abundant protein—it accounts for roughly 1 percent of the proteins in a cell’s cytosol—whose precise role in ALS is unknown. Horwich bred mice with a mutant form of the SOD1 protein, called G85R, that cannot fold properly and causes features of disease like those associated with ALS in humans, including partial paralysis and clumped proteins inside motor neurons. The mutation appears to cause a gain of function, not the loss of one: delete the gene entirely, and the animals survive.
How an animal could survive without a protein normally so abundant is one of the many questions in the overall ALS puzzle that Horwich’s lab is pursuing. Pinpointing the function gained as a result of the mutation associated with the disease is another. With Lifton, Horwich is also sequencing the protein-coding portions of genomes of ALS patients. In doing so, the scientists hope to explain the role of genetics in sporadic forms of the disease.
“There hasn’t been a lot of ambiguity in Art’s lab as to what the mission has been,” says Lifton. “In whatever direction the project needs him to go to become expert, he does that. It’s one of the hallmarks of a great scientist, following the trail wherever it leads.”
But beyond elucidating the basic biology of ALS, Horwich is not shy about his ultimate goal. “If, when my children’s tennis coach first started limping we had known what to do to arrest the process associated with aggregation of these proteins, he would still be alive,” he says. “That would be the dream.”
To that end, Horwich has treated some of his mutant mice with experimental therapies; in others, he is investigating how variations in the SOD1 gene affect the progression of disease. He is also exploring whether it’s possible to get cells to ramp up their production of chaperone proteins, amplifying their ability to capture free-floating unfolded proteins that might otherwise clump together.
It’s too soon to say whether any of these approaches will yield a viable treatment. For Horwich, who in the GroEL days grew accustomed to experiments that could be conceived and completed in a single day, the work is maddeningly slow. Still, his characteristic optimism is on full display.
“I keep going in the mouse room and hoping that we will be able to see an animal whose disease improves because of something we can track,” he says. But “whether we’re successful at trying this huge thing that we’re trying to tackle or not, I’m still going to come to work. I’m still going to tinker side by side with my group.”