Spiegelman's synthetic RNA experiments led him, ultimately, back to the study of cancer, the topic that had long hovered in the background of his research work. He was impressed by the considerable fund of information accumulated concerning the so-called B-type and C-type tumor viruses, which had been implicated in a variety of cancers in a wide spectrum of animals. Working with his synthetic mutant RNA, he had wondered if perhaps such a system could generate RNA molecules that would compete with the virus RNA for replicating enzyme, interrupting its synthesis and arresting the infection. Testing this idea with RNA tumor viruses might also lead to better understanding of human cancer, and he wanted to make a more direct contribution to solving the mysteries of that disease. In 1969, he left the University of Illinois to direct the Institute of Cancer Research at Columbia University's College of Physicians and Surgeons in New York.
Spiegelman thought it improbable that the etiologies of human cancers and animal cancers would be radically different. Nevertheless, evidence for viral agents in human tumors was quite fragmentary at that time. He turned to molecular hybridization and other techniques of molecular biology to find them, and began looking at animal tumor viruses that might serve as probes to examine human tumors.
In May of 1970, just as Spiegelman had got his new lab organized, Howard Temin and David Baltimore announced that, working independently, they had found (within Rous sarcoma virus and mouse leukemia virus, respectively) an enzyme capable of making DNA from a RNA template--a reverse transcriptase. This meant that RNA tumor viruses infecting a cell contained an enzyme that could take over the cell's genetic machinery and produce a new, tumor-virus-directed DNA; when the cell reproduced, its offspring would be tumor cells instead of normal ones. It was just the sort of enzyme that Spiegelman had tried--and failed--to find in RNA phages many years before. And its discovery overturned the "central dogma" of molecular biology, which stated that genetic transcription went in one direction only--from DNA to RNA.
Spiegelman immediately switched most of his research team to studying reverse transcriptase. The group soon demonstrated that eight other RNA tumor viruses carried the reverse transcriptase. They then proved that the DNA synthesized by the enzyme was in fact a complementary copy of the viral RNA in each case. Just as important, they found that the reverse transcriptase they had isolated and purified from avian myeloblastosis virus was non-specific. That is, if added to RNA from a virus unrelated to its source virus, it would generate DNA complementary to that RNA. Spiegelman subsequently showed that he could synthesize a DNA copy of virtually any RNA with reverse transcriptase. This in turn enabled him to develop molecular probes to use in the search for tumor viruses, using the DNA or fragments of it to detect RNA with hybridization. Finally, he developed the "simultaneous detection test," a sensitive and dependable method for detecting virus particles in tumor tissue. This technique identified, in one step, the presence of the virus RNA, the reverse transcriptase, and the DNA synthesized from the RNA. It added a promising new dimension to human tumor virology, allowing researchers to examine human cancers for possible viral agents even in cases for which no animal model existed--which was most cases.
With these tools in hand, Spiegelman planned to use reverse transcriptase to make DNA copies of a tumor virus's RNA. These copies would act like "magnets" when exposed to any human cancer cells that did contain viral RNA. The DNA that had been produced initially from the viral RNA would readily hybridize with any viral RNA present in the cell, thus "pulling out" from the cell the evidence that it harbored a virus. For his first "magnet" he looked to a recent study of breast cancer incidence among Parsi women in India. The researchers had found that the women also had in their breast milk a high incidence of virus particles similar to the virus that causes mammary cancer in mice. Spiegelman reasoned that if human breast cancer involved a virus, and if it was at all similar to the mouse tumor virus, the mouse virus DNA might be attracted to the virus RNA molecules found in the Indian women. And indeed, when tested, DNA copies of mouse mammary tumor virus hybridized neatly with the tumor RNA molecules within a large proportion of breast cancer tissue. Spiegelman also showed that it did not hybridize with the RNA of any other human tumor tested. In further studies, he achieved the same success in hybridizing with human leukemias, and even with solid tumors such as sarcoma (tumors of the connective tissue) and even with Hodgkin's lymphoma. His team later worked with carcinomas of the lung, colon, and ovary, using the same method. This research held out the promise that techniques to cure cancer in laboratory animals would have some direct relevance to the treatment of cancer in humans.
In these same studies, Spiegelman and his group tested and threw serious doubt on the "oncogene" hypothesis formulated several years earlier by Robert Huebner and George Todaro. This hypothesis proposed that all cells contained oncogenes, regions of DNA that were normally repressed. When these genes were activated by a virus, a chemical carcinogen, or radiation, they expressed themselves by coding for a "transforming protein." When this occurred, a cell could become a tumor cell, even if no viruses could be recovered from it. Using another new research technique called "recycling," Spiegelman's team found that the DNA of leukemias and lymphomas contained certain sequences not found in the DNA of normal tissues. They confirmed this discovery in an investigation of identical twins, which demonstrated that the DNA in white blood cells of a twin with leukemia contained these abnormal sequences, but the DNA of a twin without leukemia did not. If the malignant information were "built-in" it should appear in both twins.
From 1975 to 1983, Spiegelman and his team continued to explore these ideas, including their applications to enhanced chemotherapy, and to develop a blood test for breast cancer that detected the presence of viral-related proteins. The test was granted a patent shortly after Speigelman's death in January 1983.
Subsequent research did not fulfill the hopes of Spiegelman and others that viruses would prove to be a common cause of cancer. They are the culprits in only a few human cancers, such as cervical carcinoma. And, while the first oncogene hypothesis was discredited by the early 1980s, later investigators, notably Harold Varmus and Michael Bishop, found that the normal human genome does indeed contain a number of genes with the ability to undergo mutation and direct malignant cell growth. Nevertheless, the innovative techniques Spiegelman developed and the intricate investigations he carried out were invaluable to the progress of molecular biology during his lifetime. And although he didn't achieve the goal of a cancer cure, the knowledge that he and others gained in their cancer virus research during the 1970s proved crucial to the rapid identification and characterization of the AIDS virus when it emerged in the early 1980s, and to the development of the drugs which control it.