The Development of Retrovirology as Intellectual History

Publication Details

Tools of Analysis

Thomas Kuhn, in his classical 1962 essay on the structure of scientific revolutions, laid out the criteria and the tools for a historical analysis of scientific developments (Kuhn 1962). The fundamental idea behind Kuhn's method is that scientific development is not gradual and accumulative but that periods of steady advance referred to as normal science are punctuated by discontinuities, scientific revolutions, and changes of the ruling paradigms. These are not mere sudden quantitative advances in knowledge, they are replacements of the elementary framework, the paradigms, under which a scientific field operates. Scientific revolutions can be major, affecting our view of the world, but they can also be less encompassing, and some may interest only the scientists working in a defined field. Whatever the extent of a paradigm change, what all scientists have in common is that they shape the character of a field and define the problems that should be worked on. Successive paradigms are largely incompatible. Much of the old is, however, accepted into the new era, although often reinterpreted to comply with the new paradigm. Paradigm changes do not happen spontaneously but are caused by crisis. A crisis may be acute, in which the old paradigms fail to lead to solutions of significant and pressing current problems, or it may be chronic, smoldering, manifested mainly in a lack of substantive progress.

Scientific revolutions are rarely accepted without opposition. The new paradigm is bound to be questioned and attacked. In times of change, such criticism provides a valuable service to the growth of science, but once a new paradigm is generally adopted, scientists form a community bound by that shared foundation. A persistent opponent remains outside that community for purely intellectual reasons. The opposing arguments, coming from different premises, are not understood or not taken seriously, or both. A lifelong opponent risks oblivion.

Scientific revolutions are accompanied by yet another phenomenon. They are often preceded by partial anticipatory discoveries. These are either too incomplete and unconvincing or out of phase with the preoccupations of the field and thus fail to have the impact required to change the paradigm.

Paradigm Changes in Retrovirology

During the history of retrovirology, paradigm changes can be linked to the following discoveries: (1) the isolation of the first retroviruses, (2) the development of the focus assay for RSV, (3) the discovery of reverse transcription, (4) the derivation of retroviral oncogenes from the cellular genome, and (5) the isolation of human retroviruses. Incompatibility between new and old paradigms in retrovirology is sometimes but not always evident, but there is no question that after each paradigm change, the new paradigm determines the character of the field and the problems that retrovirologists find attractive to work on. Most paradigm changes in retrovirology were preceded by anticipatory discoveries, and all were accompanied and followed by resistance of varying intensities.

The isolation of the first oncogenic retroviruses by Ellermann and Bang (1908) and by Rous (1911) created the founding paradigm of the field. These discoveries were of absolute novelty, without anticipatory predecessors, and in many respects, they were far ahead of medical science of their times. Yet they did not have the initial impact that hindsight accords them; they were met by skepticism and a kind of resistance that was more insidious and persistent than ardent and open. The principal reasons at the time were that the status of leukemia as a cancer of the blood-forming tissues was still in doubt; the chicken was considered too unrelated to man to provide models for human disease; and because cancer in humans is not contagious, its transmissibility in birds was not believed to be relevant in any way to the human situation. The process of overcoming opposition to the paradigm of viral oncogenesis was an unusually lengthy one. Leukemia was subsequently recognized as a special form of cancer, and lack of contagiousness and absence of infectious virus from a tumor were found not to rule out viral etiology. Virally induced tumors are not a peculiar disease of birds but occur also in mammals as Bittner (1936) was able to show. But more than four decades passed from Ellermann and Bang and Rous to the isolation of murine leukemia virus by Ludwik Gross (1951, 1957). The pioneering work of Gross catalyzed the field, winning young converts to the expanding search for tumor viruses. The fertile and exciting period of biological exploration that followed uncovered most of the oncogenic viruses to which retrovirologists have devoted their efforts and which continue to pose challenging problems today.

The search for new viruses was embodied in the founding paradigm of retrovirology. The virus hunters in the second half of the century were distant disciples of Rous and of Ellermann and Bang. But viral isolation, transmission, and description of disease could not sustain the momentum of the field. An opening to a different level was needed. This was not an acute crisis, but a chronic one; nevertheless, it demanded fundamental change. The opening came with a new paradigm that could be called the “one cell, one virus paradigm.” It was derived from the focus assay for RSV developed by Temin and Rubin (1958). The intellectual environment from which it emerged was that of the Biology Division of the California Institute of Technology in the 1950s, the cradle of modern virology. Max Delbrück, a founder of molecular biology, was there as was Renato Dulbecco who, together with Marguerite Vogt, had devised the plaque assay for cytocidal animal viruses (Dulbecco and Vogt 1953). It would be a mistake to look at the focus assay as just a method to determine viral concentrations. The focus assay deals with infection at the single cell–single virus level. This emphasis on the elementary participants of infection reflects the same reductionist discipline that was the hallmark of the Delbrück and Dulbecco schools of virology. In the focus assay, this principle was extended from infection to oncogenic transformation. It made transformation a quantifiable event of a genetic nature.

It is only a slight exaggeration to say that the focus assay was as much a philosophical statement as it was an innovative technique. The underlying method had actually been described earlier in broad outline but had failed to move the field (Manaker and Groupé 1956). The focus assay marked a true change of paradigm because it provided a new framework and set the criteria for the selection of problems that avant garde virologists worked on; it must be considered as the primary activator for the exponential growth of retrovirology that began in the 1960s. Not only did it start the cell biology of retroviruses, but, in doing so, it also defined the major molecular themes of the field. Opposition to the new paradigm was passive, consisting of a persistence in the older approaches; occasionally, these still generated significant contributions, as, for example, the extensive investigations of the leukemia viruses of chickens and mice for which focal assays either were not available or required more fastidious hematopoietic cell cultures (Beard 1963a,b; Rossi and Friend 1967; Tkaczevski et al. 1968).

The study of infection at the cellular level rapidly generated a solid body of knowledge on viral replication, virus-induced transformation, and genetics that serves as the biological foundation of retrovirology. But questions and problems were soon discovered that could not be answered or solved within the outlines of the existing paradigms of virology and of biology in general. These problems included the absence of replicative RNA forms from cells infected with an RNA virus, the sensitivity of an RNA viral infection to inhibitors of DNA synthesis and of DNA transcription, and the genetic stability of the transformed state, all inexplicable with existing paradigms.

Retrovirology in the late 1960s was a field in crisis. Temin's provirus hypothesis of 1964 addressed the enigma, but reaction to this proposal extended from mild disbelief to ardent opposition (Temin 1964). The revolutionary explanation of the riddle of retroviral replication offered by this hypothesis required nothing less than the repeal of the Central Dogma of molecular biology, the hierarchical transfer of genetic information from DNA to RNA to protein. It is only fair to say that the evidence for the provirus hypothesis prior to the discovery of reverse transcriptase was fragmentary and not convincing. Most workers continued to search for evidence of more conventional genetic strategies, citing possible nonspecific effects of the inhibitors that had suggested involvement of DNA in retroviral infection and pointing out the weakness of the data on a DNA provirus. All the more admirable is Temin's vision, courage, and sheer tenacity through the seven years it took to come up with the final proof for RNA-dependent DNA synthesis, the key element of the provirus hypothesis.

When this proof was finally presented, the effect was remarkable. Seldom has there been such an instantaneous and complete conversion of a field as that triggered in retrovirology by the discovery of reverse transcriptase (Baltimore 1970; Temin and Mizutani 1970). This was a paradigm change that most decidedly deserves to be called a scientific revolution with implications that go far beyond the narrow confines of the field. The transfer of genetic information from RNA to DNA has since been recognized as a general principle of evolution operative throughout biology. For Retrovirology, the provirus hypothesis became the guiding beacon, and it remains the paradigm from which the molecular work in the field is derived.

The change of paradigm following the discovery of reverse transcriptase was preceded by anticipatory data and proposals that may have prepared the ground for but did not trigger the revolution. The idea of a provirus had been discussed in the literature, and the discovery of X-ray-induced murine leukemia that was transmissible by cell-free extracts prompted comparison to the radiation-induced induction of lysogenic phage (Lieberman and Kaplan 1959; Lwoff 1960; Latarjet and Duplan 1962). Closer to the critical juncture, in 1967, a virus-coded RNA polymerase was discovered in vaccinia virus particles, demonstrating that viruses can carry essential nucleic-acid-synthesizing enzymes required for the initiation of viral growth (Kates and McAuslan 1967). In 1969, the early block in Retroviral replication caused by inhibitors of DNA synthesis was found to be specific for the infecting virus (Duesberg and Vogt 1969). In cells preinfected with one retrovirus, a DNA inhibitor could still prevent infection by another retrovirus, suggesting that the DNA synthesized by the first virus did not abolish the requirement for DNA synthesis by the second virus. All of these observations can now be convincingly interpreted on the basis of the provirus hypothesis. At the time they first appeared, their impact was limited, failing to resolve the crisis in the field.

Soon the techniques and tools derived from reverse transcription and from viral genetics were applied to a molecular study of viral oncogenesis. The turning point in this area was the production of the src probe and the ensuing demonstration that src is a cellular gene (Stéhelin et al. 1976b). This discovery was generalized to include all retroviral oncogenes, so that “cellular origin” became a defining criterion for such genes. With these developments, oncogenes ceased to be a domain of virology and gained far broader, general significance. The change in paradigm did not herald a new phase of retrovirology but created a new branch of science altogether, the genetics of cellular growth control. Nor was this change prompted by an acute crisis in the field of retrovirology; oncogenes had already been shown to be irrelevant for viral growth and survival. The need to clarify the nature and actions of oncogenes was not a virological problem but one of cancer research. In the process of defining oncogenes and tracing their provenance, the meaning of the term “oncogene” underwent a remarkable change. It was originally coined in 1960 to designate hypothetical oncogenic information of endogenous retroviral genomes (Huebner and Todaro 1969). The expression was then applied to the oncogenes that genetics and nucleic acid analysis had delineated in exogenous retroviruses and that had also gone by the name onc genes. With the lineage of these genes traced back to the cell genome, the meaning of “oncogene” expanded to include cellular genes; in the literature, the prefix “c-” is used for the cellular version of the gene and “v-” is used for the viral version of the gene (Coffin et al. 1981). The startling genetic findings on the nature of retroviral oncogenes were followed by the identification of the Src protein (Brugge and Erickson 1977). The regulatory potential of this protein became evident when its then novel enzymatic function—tyrosine-specific protein kinase—was recognized (Hunter and Sefton 1980). This discovery proved to be seminal for the field of cellular signaling and for integrating src and other kinase oncogenes into growth regulatory pathways.

Today, the study of oncogenes constitutes a large segment of cancer research. Cancer research has been changed into a genetic science, and the origins of this metamorphosis can be found in virology, going back to the introduction of the focus assay. A broad consensus among scientists regards cancer as resulting from an accumulation of somatic and, more rarely, germ-line mutations in growth regulatory genes. Oncogenes are the positive regulators of cell growth and division; tumor suppressors are negative regulators. The idea of somatic mutation acting as a cause of cancer goes back to the developmental biologist and geneticist Theodore Boveri. In his astonishingly prescient publication on the origin of tumors, Boveri argued from his observation of frequent chromosomal aberrations in tumor cells that somatic mutation must be the root cause of cancer (Boveri 1914). For more than eight decades, this proposal has been a subject of discussion and controversy, with wide swings in the extent of its acceptance among cancer researchers until its recent prevalence. It is interesting that Peyton Rous, whose seminal discovery coincided with Boveri's work, fervently disputed a universal role of mutation in oncogenesis because he failed to recognize the mutagenic potential of viruses (Rous 1959). Despite the widespread acceptance of the paradigm, the genetic origin of cancer still has its opponents. It has been pointed out that some aspects of cancer biology are difficult to explain by discrete genetic changes in the tumor cell (Rubin 1985, 1994; Farber and Rubin 1991). More strident opposition to the paradigm of the cellular oncogene has invoked the argument that the cellular and viral versions of oncogenes are fundamentally different, divergent in both structure and function (Duesberg 1987; Duesberg and Schwartz 1992). At present, these contrary views appear to be more provocative than persuasive. Future research will decide whether some of them have any lasting merit.

The most recent paradigm change in retrovirology followed the discovery of human retroviruses, most notably HIV. It was preceded by a crisis not of retrovirology but of public health, caused by the advancing AIDS epidemic. Human retroviruses did not change the basic intellectual framework of the field, although the long history of negative and artifactual observations created resistance to the idea of retroviral diseases in humans. Human retroviruses did profoundly alter the emphasis of the discipline by determining which virus retrovirologists study: Contemporary retrovirology is largely devoted to HIV. This dominance of HIV derives from its etiological role in AIDS, a role that has been established through seroepidemiological and viral isolation data, through prospective studies on infected individuals, and indirectly through the ability of molecular clones of closely related animal lentiviruses to cause an identical disease in their hosts (Weiss 1993). The HIV-AIDS connection is still being attacked in much publicized exposés, but these arguments are unconvincing and have little effect on research in the field (Duesberg 1991). HIV poses intellectually novel and unique problems to research on all levels of analysis: transmission and pathogenesis in the patient, cell biology and genetics of infection, and molecular mechanisms of replication, of viral gene expression, and of virion assembly and release. Research in all of these areas contributes to the attainment of the most urgent goals: a protective vaccine and an effective therapy for HIV infection.