Progress in Nucleic Acid Research and Molecular Biology
43:135-204, 1992

Latent Viruses and Mutated
Oncogenes: No Evidence
for Pathogenicity


Department of Molecular and Cell Biology
University of California at Berkeley
Berkeley, California 94720

V. Conclusions

A. Evidence That Latent Viruses and Mutated Cellular Genes Are Pathogenic Is Circumstantial

Compared to classical prototypes, the presumably latent viruses and mutated cellular genes all suffer from activity, infectivity, and specificity gaps: (1) The viruses and genes that are postulated to be pathogenic or oncogenic are each orders of magnitude less active, and their products less abundant, than the transcriptionally active, pathogenic prototypes that have inspired these hypotheses; (2) infections with the viruses do not cause the postulated diseases, and transfections of appropriate cells with mutated cellular genes do not transform normal cells; (3) the hypothetical pathogens are not disease-specific, because (a) the same latent viruses and mutated cellular genes occur in symptomatic and asymptomatic subjects (see Section VI) and because (b) histologically and clinically indistinguishable diseases occur without them.

Clearly, infection without disease involving limited numbers of cells also occurs with classical viral pathogens and is essential for their survival (3, 12). However, it becomes an important deficiency for hypotheses claiming pathogenicity for viruses that are equally inactive and restricted in diseased and healthy carriers.

Therefore, there is only circumstantial or epidemiological evidence for the role of mutated genes in cancer and latent viruses in disease. Indeed, Bishop writes that "... on the basis of circumstantial evidence of considerable variety, damage to diverse proto-onc genes has been implicated in the genesis of human tumors" (7). Varmus pointed out: "Although the dividends of conferring the status of protooncogenes upon these cellular genes have been considerable, it must be acknowledged that the basis for doing so, the genetic definition of v-onc's [retroviral oncogenes], has not been uniformly rigorous" (5). And on the basis of epidemiological correlations, point-mutations in cancer cells are said to be "smoking guns" (272). Yet the same tumors occur without this circumstantial evidence (see Section VI).

Based on just two as yet unconfirmed studies from 1989, Baltimore and Feinberg pointed out that "HIV viremia cannot be said to be 'necessary' for AIDS on the basis of any available data, but the new results are a consistent feature of AIDS" (77). Further, Blattner, Gallo, and Temin (14) argued that "... the strongest evidence that HIV causes AIDS comes from prospective epidemiological studies ..." and Weiss and Jaffe (15) concurred ("the evidence that HIV causes AIDS is epidemiological...."), although Gallo (76) conceded that epidemiology is just "one hell of a good beginning" for proving the virus-AIDS hypothesis. But even this beginning is flawed by the tautological definition of AIDS, which only diagnoses AIDS when antibodies to HIV are found (54) and ignores all AIDS indicator diseases that occur in the absence of HIV, even in AIDS-risk groups (54). For example, half of all American intravenous drug users (55) and 25% of all hemophiliacs (54) are HIV-free, so that their AIDS indicator diseases will not be reportable as AIDS.

The same is true for the epidemiological evidence that HTLV-I causes T-cell leukemia. The leukemia is solely defined by the presence of HTLV-I, although it has been observed in its absence (Section III). In an effort to link human myelopathy with latent HTLV-I, it is proposed that "similarities between HAM (HTLV-I associated myelopathy) and visna [are] the result of still deeper identities" (323). Visna is a neurological disease that occurred in the 1930s in a now extinct strain of sheep. Its cause is believed to have been a latent retrovirus, termed visna virus, that is present in over 50% of healthy sheep in Europe and the U.S. (53). Likewise, there is no controlled study to show that the incidence of cervical cancer is higher in HPV-positive than in matched negative controls (see Section III).

Thus, latent viruses and mutated cellular genes are postulated to be pathogenic only because (i) they structurally resemble active pathogenic viruses or active viral oncogenes, and because (ii) they occur, or are assumed to occur, in the respective diseases more often than in normal tissues (6, 49, 51, 296).

B. Helper Genes and Cofactors to Close the Activity, Infectivity, and Specificity Gaps of Hypothetical Pathogens

Since the latent viruses and mutated cellular genes do not behave like the autonomous pathogens they were originally postulated to be, the original theories have been supplemented by ad hoc hypotheses.

For example, it has been postulated that the long latent periods, ranging up to 55 years, of the hypothetical viral pathogens are necessary for various, unproven cofactors to help the latent viruses to cause disease. Accordingly, the viruses have been termed "slow viruses" or "lentiviruses" (3, 12), although the viruses replicate within a few days and are immunogenic within a few weeks after infection (13, 37, 91, 92). Further, it has been postulated that antiviral antibodies consistently fail to neutralize viruses such as HIV (76, 77) or hepatitis C virus (160), although the respective viruses are almost undetectable for the duration of the diseases they are said to cause. Moreover, helper genes or cofactors are postulated for carcinogenesis by mutated genes that are not transcriptionally activated, including the point-mutated proto-ras genes, the int genes mutated by provirus insertion, or the rearranged proto-myc genes.

In reality, the cofactors are modern euphemisms for new hypotheses. The ad hoc hypotheses all assume second- or third-order mechanisms of pathogenesis relying on unproven cofactors for both latent viruses and mutated cellular genes to cause disease. Moreover, these ad hoc hypotheses all lack appropriate precedents because all available pathogenic viruses and viral oncogenes are helper-independent first-order pathogens. Yet the ad hoc hypotheses are popular because they leave arbitrary but face-saving roles for failing incumbents, which were all originally proposed to cause the diseases by themselves.

In the absence of functional proof, circumstantial and epidemiological evidence is only relevant for causation if the respective viruses and cellular mutations and their hypothetical cofactors are at least disease-specific. This appears not to be the case.