The Emperor of All Maladies: A Biography of Cancer

The biochemist Arthur Kornberg joked that biology was like the man who lost his keys at home but looked for them under the lamp in the street because that’s where it was brightest. In the world of cancer, Rous’s study of sarcoma virus was such a bright spot. He felt firmly that viruses were the only cause of cancer. Therefore, cancer researchers split into three opposing camps: virologists, led by Rous; epidemiologists such as Doll and Hill; and the successors of Theodor Boveri, who thought that genes might cause cancer but lacked the data to prove it.

In 1956, a young virologist by the name of Howard Temin came to Cal Tech to study genetics and moved into the lab of Renato Dulbecco to study Rous sarcoma virus. He succeeded in growing cancer in a petri dish, pushing the cells to grow uncontrollably. He found that Rous sarcoma virus affected the cell’s DNA and changed its genetic composition. He observed that a virus’s genes could attach itself to the genes in a cell, but he could not explain how virus genes that started as RNA could then, in its copy, go back to DNA. Temin worked in Madison along with a postdoc student named Satoshi Mizutani, who found that the Rous virus was able to convert RNA to DNA. As Temin presented his findings at a cancer conference in Houston in 1970, the importance of his work dawned on the audience. Another virologist at MIT, David Baltimore, also found that RNA could become DNA. Both Temin and Baltimore published their results, which postulated that RNA turned into DNA in viruses and then made RNA copies in an endlessly repeating cycle.

Some cancer researchers, such as Sol Spiegelman at Columbia, were convinced that this explained the mechanistic process of cancer growth. Temin’s work suggested that an RNA virus could come into a cell, make a DNA copy of its genes, and then become part of the cell’s genome. This process unleashed “pathological mitosis” (355), or cancer. Spiegelman rushed to find retroviruses in cancer, and he found them in almost every human cancer he looked at. However, later experiments found that saw viruses that did not exist. Only one cancer, a rare form of leukemia, was the result of a retrovirus. Instead, scientists later found retroviruses to cause HIV.

Temin then proposed that the virus was merely the messenger that entered a cell and that researchers needed to find the origin of the message itself. In the 1970s in California, virologists Steve Martin, Peter Vogt, and Peter Duesberg made mutants of the Rous virus and found the single gene that caused cancer, which they called “src,” or “sarc.” They believed that RSV had acquired this abnormal gene, or oncogene (a gene that could cause cancer), during its evolution.

Ray Erikson at the University of Colorado found that src encoded a protein that functioned to change other proteins by attaching a phosphate group to these proteins. Scientists called these proteins “kinases,” and they worked as “master switches within a cell” (358), capable of turning on the protein’s function. One kinase could turn on another, amplifying the signal at each step and producing a signal to the cell to divide. Rous sarcoma virus resulted in cancer by bringing a gene called src into the cell that encoded for a “hyperactive overexuberant kinase” (359).

At the University of California in San Francisco (UCSF), virologist J. Michael Bishop, along with Harold Varmus, tried to find out the origins of the src gene—a quest that other scientists called “the hunting of the sarc” (360). Working with Deborah Spector and Dominique Stehelin, they found homologues of src across the bird kingdom and in cows, sheep, and humans. Hideasburo Hanafusa, a Japanese virologist at Rockefeller University, found that the viral form of src carried mutations that changed its function. It was an overactive kinase that promoted cell division, while the cellular src, though it had the same kinase activity, was regulated so it did not get out of control. A radical theory emerged: src was endogenous to the cell and had come out of regular src. The virus was a messenger for a cancer cell that had brought the cancer to the genome. Viruses cause cancer, but they originate in the cell itself, not from outside. Cancer cells come from the genome, not from without. As Varmus, who, along with Bishop, won the Nobel Prize in 1989, said, “In our adventures, we have only seen our monster more clearly and described his scales and fangs in new ways—ways that reveal a cancer cell to be, like Grendel, a distorted version of our normal selves” (363). 

Varmus and Bishop’s theory finally provided a comprehensive theory of what causes cancer, explaining how soot, radiation, and cigarette smoke all mutate and activate oncogenes. Scientists then pieced together what human cancer genes look like. In 1973, Janet Rowley, a hematologist in Chicago, traced a mutated cell in CML, or chronic myelogenous leukemia (CML) that pathologists Peter Nowell and David Hungerford discovered in Philadelphia. She found a particular translocation, or misplacement, of chromosomes in every case of CML. Cancer was a form of organized chaos in chromosomes, as it followed a pattern.

Geneticist Alfred Knudson found that retinoblastoma, the eye disease de Gouvêa found in Brazil, had an inherited and a sporadic form. He found that children with inherited retinoblastoma developed the disease more quickly, as it only required one genetic change, while those with the sporadic form developed it more slowly, as it required two genetic changes. In the sporadic form, both copies of the gene in the cell had mutate, and it develops later because two different mutations must occur in this case. In the inherited form of retinoblastoma, children already have one defective copy of the gene, but they only need one mutation, so they develop cancer more quickly. With two mutations, src activates cell division, while the normal gene suppresses cell function. It must inactivate before cell division can occur.

Researchers realized cancer genes came in two types. The first, “positive” genes such as src, are mutated forms of normal genes, and they are like a “jammed accelerator” (369), unleashing cell division. The other type is a negative gene such as Rb, which normally suppresses growth but has “missing breaks” (381). 

Karl Popper, a philosopher of science, said that theories generate risky predictions that scientists can then test. By the late 1970s, Varmus and Bishop’s theory generated the risky prediction that precursors to oncogenes exist in all normal cells, therefore they could detect these proto-oncogenes in cancer cells.

The biologist Robert Weinberg at MIT had an epiphany when walking in Boston in a snowstorm in 1978. If activated oncogenes existed in cells, transferring the cancer genes to normal cells would cause the normal cells to divide. Chiaho Shih, a graduate student in Weinberg’s lab, found cells from a patient at Dana-Farber Cancer Institute—a man named Earl Jensen, a smoker who died of bladder cancer—and transferred his cells into normal human cells. Foci, indicating cancer, formed. In 1982, Weinberg, Mariano Barbacid in Spain, and Michael Wigler at Cold Spring Harbor Lab in New York all published their results, showing that they had isolated a fragment of DNA that included a gene called ras from cancer cells. Ras was, like src, in all cells, but in cancer cells, the mutated form of ras caused uncontrolled cell division.

Now, scientists had to isolate the suppressor gene that Knudson predicted and show that both copies of it are in an inactivated state in retinoblastoma. The problem was, as Weinberg wrote, “How can one capture genes that behave like ghosts […]?” (376). Although scientists knew that the Rb gene was located on chromosome 13, isolating one gene from thousands on the chromosome seemed impossible.

Thad Dryja, an ophthalmologist at Massachusetts Eye and Ear Infirmary, believed that retinoblastoma was caused when the Rb gene was inactivated through mutation, and that mutation was probably a deletion of the gene. He finally found a tumor that had a blank space in both chromosomes. He found the piece of DNA missing in tumor cells and now had to find it present in normal cells. To do so, he worked alongside a molecular geneticist named Steve Friend in Weinberg’s lab to find the Rb gene in cells. Later, scientists found mutated Rb in lung, bone, breast, and other adult cancers. The retinoblastoma gene encodes a protein, also called Rb, that binds to other proteins and holds them in its molecular “pocket” (380). When the cell divides, a phosphate group tags Rb and forces open the molecular floodgates.

Between 1983 and 1993, scientists identified other oncogenes and anti-oncogenes (or suppressor genes) in human cancers. Varmus and Bishop’s theory that oncogenes came from normal genes proved true in different forms of cancer, and it was also widely found to be true that tumor suppressors needed to inactivate in both chromosomes.

In the late 1980s, geneticists also found genes that predisposed people to different spectrums of cancers such as BRCA-1, a gene that predisposes people to breast and ovarian cancer. Purists still complained that oncogenes did not meet the final criteria for being a disease—re-creating the disease in a host. In the mid-1980s, Philip Leder and his team at Harvard introduced the c-myc gene into mice embryos, and then created the “OncoMouse” (383). The team made sure the genes only affected breast tissue. The mice developed small cancers, usually late in life and generally only after pregnancy, suggesting that hormones played a role in transforming breast tissue. The genes themselves did not lead to the development of cancer; instead, there had to be another event. After these experiments, the scientists quelled the debate over the causation of cancer, as they had successfully transferred cancer to another host. 

Bert Vogelstein at Johns Hopkins Hospital spent nearly two decades trying to figure out the number of genetic changes required to start the process of cancer. He collected cells from patients who were in various stages of colon cancer. Across patients, he found that the stages of cancer were marked by the same transitions in genetic changes. As Mukherjee writes, Vogelstein had proved “that such a discrete genetic march existed” (386).

In the 1980s, scientists discovered many proto-oncogenes and tumor suppressor genes. The question was why these genes didn’t explode into cancer all the time. First, mutations had to activate the oncogenes, which didn’t occur often. Second, tumor suppressor genes need to be inactivated, but there are two copies of each, making this event even rarer. Vogelstein also figured out that activating or inactivating a gene is just the first step. Instead, there had to be many mutations carried out many times.

Researchers discovered that proto-oncogenes and tumor suppressor genes sit at the hubs of activated signaling pathways in the cell. The way in which the pathways intersects with other pathways also determines cancer activity. Judah Folkman at Children’s Hospital in Boston found that certain signaling pathways, ras included, could also result in making neighboring blood vessels grow. In that way, tumors could have their own blood supply, a process Folkman called tumor angiogenesis. Folkman’s colleague at Harvard, Stan Korsmeyer, found that other pathways prevented cell death. Therefore, cancer was not just genetic in its origins but in all its behavior, as pathways gone haywire determined its behavior and sustained it.

By the early 1990s, cancer biologists began to model how cancer develops from changes in genes. The author discusses how a particle of asbestos infects a man’s normal lung cell, creates a mass of cells, and undergoes a mutation in the ras gene. As the man smokes, carcinogens collide with the ras-mutated cells. Then, when an X-ray hits another cell in the mass of created cells, a second mutation inactivates another tumor suppressor gene. Now, the mutated cells row and activate pathways. Mutant cells create other cells, some of which develop motility and the ability to live in the bone. The man receives a diagnosis of metastatic cancer, and the cells mutate to develop resistance to his chemotherapeutic drugs. This man, who died at age 76, was the first to die in Mukherjee’s care during his fellowship at Massachusetts General Hospital.

In 2000, cancer biologists Robert Weinberg and Douglas Hanahan published an article called “The Hallmark of Cancer.” The article explained six rules for the behavior of most human cancers, including “self-sufficiency in growth signals” (391) and “insensitivity to growth-inhibitory signals” (391). The task was now to build on this understanding to create cures.