NEWS AND VIEWS REVERSE TRANSCRIPTION------transplant, and any future organ trans­ Foamy bubble on plants from primates, should be carefully monitored for foamy transfer as well Robin A. Weiss as other viruses. Parodying Koch's postulates for modern FoR many years we have assumed that all mate that they naturally infect. However, virology, one might propose that if a virus replicate by a uniform major mice transgenic for HFV suffer neurode­ fails as a pathogen, promote it as a vector strategy. Given that, a paper in last week's generation resembling amyotrophic lateral in gene therapy. This is precisely what has Science1 provides pause for thought - it sclerosis6, and virus expression was evident been done for HFY. Schmidt and Reth­ shows that the spumavirus human foamy in the brain of an orang-utan which died wilm 14 deleted the be/2-3 region, while virus (HFV) differs from oncoviruses (such with similar acute neurological symptoms7• Russell and Miller4 substituted marker as mouse leukaemia virus, MLV) and But HFV has not been found in human genes in place of env and bell. With its very (such as HIV) in carrying full­ amyotrophic lateral sclerosis8, and the vari­ broad host range and tropism for cell length, double-stranded DNA genomes in a ous claims that HFV is associated with de types4, HFV may prove to be quite a useful proportion of virus particles. Furthermore, Quervain5 or Graves9 diseases of the thy­ vector. HFV reverse transcriptase and other Pol roid has now been questioned10• Indeed Russell and Miller further noted that proteins are translated from a spliced there is no longer definitive evidence of their foamy virus vectors were able to infect messenger RNA, and lack Gag sequences1•2• human infection, other than in a handful of cells in the stationary phase of the cell cycle, These findings are unexpected and provoca­ and might thus be used to tive; HFV resembles , which deliver genes to non-prolifer­ has a DNA genome, in these respects as ating targets such as neurons. 100nm much as it does retroviruses. Moreover, On the face of it, this finding other reports suggest that HFV is not so appears to run counter to our human after all3, but may serve as a useful recent report15 that HFV repli­ gene vector to non-proliferating cells4 • cation is cell-cycle dependent. Retroviruses have an RNA genome but These apparently contradic­ insert a DNA copy of it into the host chro­ tory results stem largely from mosome following reverse transcription. a different emphasis in inter­ The genomes of the three pretation, however - in both subfamilies ( onco-, lenti- and spuma­ cases, the efficiency of infec­ viruses) each carry at least three genes, tion of stationary cells was gag-pol-env, running 5' to 3', which are approximately 5-10 per cent packaged in virions as an RNA dimer, with that of proliferating cells. Thus cellular transfer RNA acting as a primer HFV infection is less stringent­ for reverse transcription. It was previously ly dependent on mitosis than thought that DNA synthesis is only initiat­ Budding particles of foamy retrovirus, imaged by electron MLV, but is more dependent ed in the cytoplasm of the newly infected microscopy; note the prominent glycoprotein spikes and the than HIV An attractive expla­ cell. Yu et al. 1 show that full-length, DNA preformed spherical cores. As seen through the light micro• nation for HFV infection of proviral genomes, some 12 kilobases long, scope, cell cultures infected with the virus look foamy, non-proliferating cells would are present in 10-15 per cent of the parti­ hence the name - spumaviruses - of the subfamily. be that it occurs through the cles. By contrast, analysis with the poly­ DNA genomes found in 10-15 merase chain reaction indicates that about persons with from contact with per cent of virions (though Yu et al. 1 1 in 100,000 virions of onco- or lentiviruses primates3• HFV was the first 'human' thought that the DNA was not infectious). have associated full-length DNA. Howev­ retrovirus to be described11, yet it is now Perhaps Russell and Miller should check er, the HFV virion DNA does not appear apparent that the HFV genome sequence whether their vectors can infect stationary to be infectious and inhibitors of reverse is indistinguishable from that of one of the cells in the presence of inhibitors of reverse transcriptase indicate that the enzyme is chimpanzee isolates12• Following an inter­ transcriptase. D still required at an early stage of infection. national workshop on foamy viruses in Although foamy viruses have essential­ 1994, at which all laboratories working on Robin A. Weiss is in the Chester Beatty ly the same main gene organization as foamy virus were represented, exchange Laboratories, Institute of Cancer Research, other retroviruses (plus accessory be[ and analysis of samples did not uphold evi­ 237 Fulham Road, London SW3 6JB, UK. genes which lie 3' to env), the finding that dence for human foamy virus in patients pol is translated from its own spliced with autoimmune disease. 1. Yu, S. F., Baldwin, D. N., Gwynn, S. R., Yendapilli, S. & mRNA, independently of gag, is also quite Given that the great apes so frequently Linial, M. L. Science 271, 1579- 1582 (1996). 2. Enssle, J., Jordan, I. , Mauer, B. & Ret hwilm, A. Proc. 7•1 •13, novel. Other retroviruses synthesize Pol harbour foamy viruses it is odd that natn. Acad. Sci. U.S.A. (in the press). products as part of a Gag- Pol fusion pro­ Homo sapiens has escaped endemic infec­ 3. Schweizer, M. et al. AIDS Res. hum. Retrovir. U , tein, translated either by ribosomal -1 tion. The few persons who have become 161-170 (1995). 4. Russell, D. W. & Miller, A. D. J. Viral. 70, 217- 222 (1996). frameshift towards the end of Gag, or by infected with (SFV) 5. Weiss, R. A. Nature 333, 497--498 (1988). use of a suppressor tRNA. Thus Pol is remain well, have a 'full house' of anti­ 6 . Bothe, K. , Aguzzi, A., Lassman, H., Rethwilm, A. & Horak , I. Science 253, 555--557 (1991). incorporated into virions via Gag domains body responses analysed by Western blot, 7. McClure, M. 0. et al. J. Viral. 68, 7124--7130 (1994). in the precursor protein before cleavage and have viral genomes easily detected in 8 . Rosener, M., Kranz, M., Hahn, H., Heeney, J. & Rethwilm, A. J. med. Viral. (in the press). 3• by viral protease. As Yu et al. point out, blood by the polymerase chain reaction 9. Lagaye, S. et al. Proc. natn. Acad. Sci. U.S.A. 89, their results pose the question of how Pol These observations have two implications: 10070- 10074 (1992). is assembled into viral particles. that reports of widespread human infec­ 10. Schweizer, M., Turek, R., Reinhardt , M. & Neumann• Haefelin, D. AIDS Res. hum. Retrovir. 10, 601- 605 When I last wrote about foamy viruses tion are tenuous at most; but that humans (1994). in these columns5 I called them viruses in can nonetheless be infected by SFV, and 11. Achong, B. G., Mansell, P. W. A., Epstein, M. A. & Clifford, P. J. J. gen. Virol. 40, 175- 181 (1971). search of a disease. Although they are the in all probability would become infected 12. Herchenr6der, 0 . et al. Virology 201, 187- 199 (1994). most cytopathic, syncytium-inducing retro­ following xenotransplantation of primate 13. Bieniasz, P. D. et al. Virology 207, 217- 228 (1995). 14. Schmidt , M. & Rethwilm, A. Wology 210, 167- 178 (1995). viruses for many cell types in vitro, disease tissue. Thus the patient with AIDS 15. Bieniasz, P. D. , Weiss, R. A. & McClure, M. 0 . J. Virol. is seldom seen in the many species of pri- who has received a baboon bone-marrow 69, 729 5--7299 (1995). NATURE · VO L 380 · 21 MARCH 1996 201

© 1996 Nature Publishing Group