Most of us have heard of Simian virus 40 (SV40), a non-enveloped DNA virus that is best known for its capacity to transform tissue-culture cells. SV40 enters mammalian cells by an unusual route: it binds to MHC class I molecules at the cell surface and is subsequently taken up by non-clathrin-coated vesicles by a mechanism that is dependent on caveolin and cholesterol. Early electron-microscopy studies showed the endocytosed virus accumulating in the endoplasmic reticulum (ER). At the time, this observation was confusing for two reasons: ultimately the virus needs to enter the nucleus, where uncoating and replication occur, and it was unclear how it would manage to reach this destination from the ER lumen; this raised significant doubt as to whether the viral population in the ER was in any way connected to productive infection of the cell. In addition, there was no precedent for a connection between the ER and the endocytic pathway.We now know that a few plant and bacterial toxins such as cholera toxin enter cells by endocytosis and travel retrogradely through the secretory pathway all the way to the ER; here, the toxins camouflage themselves as misfolded proteins and exit through the protein-conducting channel in the ER membrane into the cytosol, where they inactivate their target molecules and thus kill the cell. Can a virus do the same thing?A paper by Parton and colleagues suggest that this might be the case [1xInhibitors of COP-mediated transport and cholera toxin action inhibit Simian Virus 40 infection. Richards, A.A. et al. Mol. Biol. Cell. 2002; 13: 1750–1764Crossref | PubMed | Scopus (76)See all References][1]. The authors investigated the effects of incubation at 20°C, and agents that interfere with trafficking mediated by coatomer COPI-and COPII-coated vesicles, on SV40 infection and cholera toxin toxicity and find that, in all cases, both processes are equally affected. This suggests that cholera toxin and SV40 indeed use the same route to the ER and that this route includes early endosomes, the trans-Golgi network and the Golgi complex itself. Intriguingly, the authors found that a modified dipeptide that blocks cholera oxin poisoning at a post-Golgi step also prevents SV40 infection. Incubation with the peptide depletes intracellular calcium stores, and it has been suggested that it directly affects the channels that mediate cholera toxin transport from the ER to the cytosol.So, how does a virus resemble a toxin or a misfolded protein? And if it really uses this disguise to exit from the ER, how does SV40 with a diameter of 45 nm make it through the protein translocation channel, which, even by the most generous estimate, has a maximal opening of 8 nm? There is some evidence that, at the plasma membrane, SV40 can ‘recruit’ caveolin to surround the viral particles before entry. Perhaps it can do the same with protein-translocation channel subunits and thus generate a giant pore in the ER membrane. The subdomains of the ER in which the virus accumulates are continuous with the rough ER but not studded with ribosomes, which primarily bind to channels engaged in protein translocation. So either this region of the ER is devoid of protein-translocation channels or the channels in this region have all become SV40 exits. Time and more experimentation will tell.