Baltimore didn’t have a virus team at his lab, but he did have something else that greatly appealed to Rosenberg. “One of David’s amazing strengths was the freedom that he gave everyone who came to his lab to work on whatever they wanted as long as it fit under a very huge, general umbrella,” she said. His instruction to her upon her arrival was to figure out what she wanted to do. The lack of guidance was daunting yet irresistible to Rosenberg, who was unaccustomed to such openness in an academic lab. And her timing was perfect. A renovation of the MIT Cancer Research Building, supposed to have been completed by Rosenberg’s arrival, was taking longer than expected. Temporarily left without a lab bench, Rosenberg had ample time to search the scientific literature and formulate a plan.
And Rosenberg, despite her lukewarm attitude about reverse transcriptase, had something else for which Baltimore had been searching. She possessed a unique talent in the world of 1970s virology: the ability to grow and manipulate different kinds of cells in culture. In the 1960s and early 1970s, simple RNA viruses were usually studied in cells that came from sources other than the tumors they caused. Examining the ins and outs of how a virus infected a random bunch of cells was definitely interesting. So much was still unknown about viruses, let alone how, exactly, they cause cancer, that any incremental finding mattered. But as long as the cells were not from actual tumors that had grown inside actual animals infected with the virus, the insights remained abstract, an intellectual exercise that, though a valid contribution to science, lacked a gripping connection to cancer in people or animals. Temin’s focus assay, the breakthrough technique from 1958, had made the unseen visible by highlighting abnormal cells against a background of normal cells. But the cells were already cancerous. Rosenberg wanted to study the actual cells from animals infected with a cancer-causing RNA viruses in the process of cancer transformation before, during, and after the time they were infected. She wanted to create a system that would allow her to be the voyeur, watching cells become infected and turn cancerous. “If you really wanted to understand how the virus caused the cancer in the animal, you needed to understand how the virus interacted with that kind of cell,” Rosenberg said. She had to find a way to get cells from a living animal to continue growing outside of their warm-blooded home. And Baltimore knew that if he was to make any progress in manipulating mouse genetics as a way to understand how cancer-causing RNA viruses caused cancer, he had to have a system for studying mouse cells undergoing that transformation outside of the mouse.
Rosenberg needed a virus that would turn cells removed from an animal into cancer cells in tissue culture (that is, in a petri dish, with a neutral medium in which to grow). With all the fur and fat stripped away, Rosenberg might witness the virus committing its crime. The experiment would still be just a model of how tumors developed, but it would be a much more accurate representation. Armed with her green thumb for growing new cell lines, Rosenberg went searching for a good virus candidate. Baltimore, eager to use his discovery of reverse transcriptase to investigate how viruses do their dirty work, encouraged her and joined in the search.
With time on her side, Rosenberg spent hours pouring over the scientific literature. Finally, she stumbled on Abelson’s 1969 paper describing the mouse virus model he’d created. It was exactly what she was looking for: a virus of known origin and a rapidly developing cancer of the blood in mice. She wanted to use the virus to create a system in which to observe cancer transformation. When Rosenberg brought the model to Baltimore’s attention, a look of recognition crept across his face. Why was that name so familiar?
It took them about two minutes to realize why: Baltimore had already caught wind of Abelson’s work because Abelson had joined a lab right upstairs from Baltimore’s when he’d left the NIH. When Rosenberg approached Abelson about using his virus, he was happy to oblige, supplying her with a glass vial of ground-up, filtered tumor extract from a mouse that had been injected with the mysterious virus.
Rosenberg teamed up with a scientist named Chuck Scher, who’d already begun working with the virus. The goal of their work was to figure out which cells the virus infected in mice, find a way to grow those cells in tissue culture, and then watch those cells turn cancerous after exposure to the virus. By doing so, the scientists could, they hoped, chronicle exactly how healthy cells turned malignant.
Day after day, Rosenberg and Scher would test out different cells taken from uninfected mice. Would red blood cells infected with the virus continue reproducing in the semisolid medium of the petri dish? What about cells from their lymph nodes? The search involved more than simply sucking out cells with a syringe and squirting them onto a bit of inert agar. Rosenberg had to find not only the right cells to use from the mouse, but also the right process for moving them from the mouse into the culture, and for infecting them in vitro—that is, outside of the living organism—without killing them.
After several grueling months of this crash course in mouse research—Rosenberg had scarcely touched a lab animal before—and no cells cooperating with their wishes, Rosenberg and Scher tried a microscopic extraction of liver cells from mouse fetuses. After surgically removing the fetus from its mother, they dissected the embryo and transported a sample of cells from its liver onto the culture dish.
plate 4
It worked. The cells became infected and continued to divide, and the virus continued to replicate. Like an impressionable child, these embryonic cells proved susceptible to the virus outside of the protective shield of the womb. And as they began to replicate, they turned cancerous. Bone marrow cells from mature mice, they later found, did likewise. Rosenberg had created a transformation system, as it was called, for the Abelson virus.
Now, Rosenberg and Baltimore had a way to observe normal mouse cells turning malignant. She would soon be joined by other postdocs in the Baltimore lab eager to parse the mechanism behind cancer transformation. What protein was responsible for catapulting cancer into action? What was the interplay between the viral genome and the mouse genome? These were all valid questions, and the Abelson virus had provided the perfect model for answering them, a replica of what went on inside cells. Rosenberg and her colleagues at the Baltimore lab weren’t driven to understand the Abelson virus per se; after all, that virus caused an anomalous cancer in mice and wasn’t relevant to human cancer—or so they thought. Simply, the virus gave them a way to watch the process. They thought that whatever process they observed in the cancer transformation system would apply to other RNA viruses that caused cancer, maybe all of them.
THAT WAS IN 1975, two years after Janet Rowley had shaken the foundations of genetics with her discovery of chromosomes swapping pieces. In the meantime, Erikson was making his way toward identifying the protein coded for by src, the key gene behind the Rous sarcoma virus. Bishop and Varmus were about to inform the world that the gene on that virus, known to cause cancer in chickens, came from the chicken cell itself. The inner workings of the Philadelphia chromosome, the search for the protein product encoded by src, and the search for origin of that src had now been joined by yet another unrelated pursuit. Or so it seemed at the time.
10
_______
A FUNNY NEW PROTEIN
Rosenberg’s transformation system was considered a major breakthrough at the Baltimore lab. Everyone there wondered what it might reveal about how cancer occurred. The possibility of uncovering the cellular mechanisms tapped by the virus to set a malignancy in motion was incredibly exciting. With Rosenberg having succeeded in creating a way to study the Abelson virus outside of a mouse, Owen Witte, another Baltimore lab postdoc, now began searching for the protein responsible for the transformation from healthy cell to cancer cell. The question guiding his search was the same one that Ray Erikson and Tony Hunter had asked: What protein was inducing cancer in the mice? But rather than starting with oncogenes contained in viruses, Witte had turned his attention to antibodies, his area of expertise.
Antibodies are proteins made by the immun
e system that can recognize foreign invaders and trigger an attack against them. The creation of experimental antibodies for lab research had been ongoing for some time. Antibodies automatically gravitate toward their designated antigens. So if a scientist wants to know whether a particular substance is present in a cell, an antibody can be used to track it down, like a metal detector scanning for gold lost underneath the sand.
The trick was making the antibody. Scientists knew the body naturally produced antibodies to foreign invaders. If an unrecognized protein is injected into an animal, the animal will likely produce an antibody to that protein. If the animal makes the antibody—it doesn’t always happen—then the antibody can be extracted, manufactured on a commercial scale, and used to search for that same protein elsewhere.
Witte had become an expert at preparing antibodies while he was earning his PhD at Stanford University, studying other viruses that cause leukemia. After that, he returned to his New England roots as a medical intern at one of Harvard University’s hospitals. His mother was dying of cancer, however, and Witte found he couldn’t stomach such heavy exposure to illness. “It was not an easy time to be taking care of other sick people,” he recalled. As it turned out, Witte preferred the bench to the bedside. “If you’re taking care of sick people dying of horrible diseases, you don’t really look forward to the next day as much as you do in the lab, where, frankly, every day is a potential for discovery.” He left his clinical training and went to David Baltimore’s lab in 1976, working across the bench from Steve Goff, who had also joined the lab as a postdoc. Considering his deep knowledge of protein chemistry, Witte’s arrival could not have been timed any better. “It was a good time to do something I liked rather than something I was supposed to do,” Witte said.
Witte began working with the Abelson virus after Rosenberg succeeded in creating the transformation system, which exposed the virus and the mouse cells. Witte thought he could use this system and his antibody techniques to find the culprit protein encoded by the oncogene in the virus.
His first step was to make an antibody. This antibody would help him identify proteins from the cells one by one so that he could take a closer look. The obvious problem here was that Witte didn’t know what protein he was looking for—after all, that was the question driving his research—so how could he make an antibody against it?
He decided to use an antibody targeted against a protein called Gag, the product of the gag gene. Witte knew that gag was present in the Moloney virus, the one that Herb Abelson had initially injected into the chemically amputated mice. Since the Abelson virus was derived from the Moloney virus, the Abelson virus would most likely have the gag gene too. Witte also knew that viruses, including Abelson, have very few genes, just four or five, and therefore very few proteins. So all he had to do was work through them one at a time to find the protein responsible for triggering cancer. He had a one-in-four or -five chance of choosing the key protein when he began with Gag. He could have started with another protein. The choice of Gag was a matter of convenience; it was the only one for which he had an antibody—as good a place to start as any.
Witte took an Abelson-infected cell—rendered that way inside Rosenberg’s transformation system—and extracted all of its proteins. Then he added a tiny bit of Gag antibody. Witte knew that the antibody would immediately bind to its target. If the Gag protein was present in those cells, the Gag antibody would latch onto it. If Witte’s antibody found its target protein, then he knew the approach was working, and he could begin homing in on which protein was behind the transformation of cells from normal to cancerous.
The experiment worked. The Gag antibody bound to a protein inside the mouse cells that had been infected with the cancer-inducing Abelson virus. Witte then coaxed the antibody and its target protein away from the other proteins in the mix by spinning the entire extract through a centrifuge. With the mixture split up into its various components, like shaken salad dressing settling into stripes of oil, vinegar, and spices, Witte could rinse away everything but the Gag antibody-and-protein pairing. He got rid of the antibody using detergent, leaving just the antigen, the Gag protein.
At the end of this process, Witte knew he’d obtained a purified Gag protein. Now he wanted to look more deeply inside the protein, to investigate all its molecules for any sign of a cancer-inducing tendency. Witte put the protein into a gel through which an electric current was run. The technique, called gel electrophoresis, caused the protein to separate into its component molecules. The result was surprising; the protein that Witte had extracted was not what he’d expected to find. It was Gag, but something else was going on. “It was a funny new protein,” said Goff, who by then had joined the Baltimore lab as a postdoc. Funny because it was really big. The weight of molecules is typically represented in daltons, with the average protein weighing around 55 kilodaltons. This one weighed about 120 kilodaltons.
Witte and Rosenberg worked as a team to investigate the protein further. They had already made mutant strains of the Abelson virus that replicated inside cells but did not render those cells cancerous. When they injected the Gag antibody into those cells, they could see that the protein it attacked was smaller than the protein Witte had found in cells infected with the original strain of Abelson. The smaller size was much more in keeping with how they expected a protein to look.
Witte knew that the difference in size was significant. The size of a protein is determined by the number of amino acid sequences within. The mouse cells infected with the cancerous strain of Abelson had larger Gag proteins than those infected with the noncancerous strain. The amino acid sequences that accounted for the difference had to be responsible for triggering cancer transformation. What’s more, the number of amino acids required to make up that difference in size meant that there had to be another protein attached to Gag. The “funny new protein” was really two proteins in one. Witte and Rosenberg dubbed the new protein—the extra bit attached to Gag—Abl, pronounced like “able,” short for the virus to which it was connected. The combination protein that Witte had first extracted from the Abelson-infected mice became known as Gag/Abl.
It was an incredible coincidence that the Abl protein, the one responsible for triggering cancer in the Abelson-infected mice, happened to be stuck to Gag. It was only because he’d used the Gag antibody to fish out the Gag protein that Witte came across Abl. If it had not been fused to Gag, he would have missed it. Now, just like Erikson and Hunter when they found the proteins they’d sought, Witte wanted to identify the type of protein he’d found.
By then, Ray Erikson had discovered that the protein encoded by the src gene, the one that the Rous sarcoma virus used to cause cancer, encoded a kinase. Well aware of the pronouncement, Witte and everyone else in the Baltimore lab wondered if the excavated protein would prove to be the same kind of enzyme. With the reports from Bishop and Varmus about the cellular origin of oncogenes, and then from Erikson, the kinase was starting to seem like an almost obvious mechanism for inducing cancer in a host.
Over the years, scientists had found a way to use ATP, the basic unit of fuel inside cells, to determine the function of a protein. Witte knew how kinases worked: They took a phosphate from ATP and bound it to a protein. The fleeting presence of a phosphate triggered the protein into action, launching a cascade of signals within the cell that resulted in all manner of events: metabolism, division, blood cell production, and on and on.
Witte incubated the Gag/Abl protein with ATP in a test tube. If it was a kinase, then it would start grabbing phosphates from ATP and sticking them onto other substances. Sure enough, this was exactly what happened. Gag/Abl was a kinase.
Baltimore, still Witte’s supervisor, knew the next logical step was to check whether the Abl kinase attached phosphates to serine or threonine, the two most common amino acid binding sites. Identifying this landing platform was part of the routine for kinase research. But when Witte ran the standard experiment to confirm which amino acid was in play, he q
uickly realized it was neither.
At almost this same moment, Tony Hunter made his discovery about the tyrosine kinase. The kinase involved in cancer induced by the polyomavirus he’d been studying phosphorylated not serine or threonine, but tyrosine, a little-known amino acid present in far smaller amounts than the other two. The splash made by Hunter’s finding should have turned Baltimore and Witte’s attention to that amino acid right away. Ray Erikson was quick to pick up Hunter’s lead. He checked Src to see if it might phosphorylate tyrosine. His report that Src was a tyrosine kinase was published in 1980.
But the clue eluded Baltimore for the strangest of reasons. Other concurrent work in his lab had uncovered a link between poliovirus and tyrosine. Witte’s kinase work came so close on the heels of that link that Baltimore assumed tyrosine could not be important in both research arenas. What were the odds that two ongoing efforts in his lab could both involve this obscure amino acid? “I put the possibility that this was tyrosine at a very low level, just because of the adage that lightning doesn’t strike twice,” said Baltimore. Finally, when tests for every other possible amino acid proved negative, they turned to tyrosine. The result was immediately clear: Abl—the protein encoded by the oncogene present in the Abelson virus—was a tyrosine kinase. With the discoveries that Src and now Abl were tyrosine kinases, the world of cancer research began to light upon the notion that tyrosine kinases—that is, kinases that bind phosphates to tyrosine as a way to power up a protein—might be fundamental to malignancies.
Still, although there were whiffs of some greater significance in the air, the research was not yet connected to cancer in people. “Is there any human disease like this?” Goff asked himself. “No idea. Not a clue.” Witte was similarly clueless. “Whether it would have anything to do with human cancer, whether it would be a drug target, that was beyond the state of our knowledge at the time,” he said.
The Philadelphia Chromosome Page 8