When Rockefeller University's Tom Tuschl published a report in Science last year revealing that the Epstein-Barr virus expresses several microRNAs, he and his colleagues became the first to show that viruses can encode for these small RNAs. The exact roles of those microRNAs, however, were unknown.
Building on Tuschl's findings, a group from the University of California, San Francisco, the University of Pennsylvania, and the University of Pittsburgh have identified microRNAs encoded by the simian virus 40 and shown how the virus uses them to improve its chances of successful infection. The findings appeared in last week's issue of Nature.
"Tuschl certainly found the first viral microRNA, but it wasn't clear what the EBV microRNAs did," Chris Sullivan, the Nature paper's first author and a postdoc in the lab of UCSF researcher Don Ganem, told RNAi News this week.
The team's research represents "the first good example of a viral microRNA's function" and is the "best characterized example of a viral microRNA having a function," he said. "It's made late in infection to optimize viral replication efficiency, and at least in vitro we know it's able to regulate cytotoxic T-cell-mediated lysis — almost certainly that what happens in vivo, too."
That these auto-regulatory miRNAs exist "indicates that viruses can use the host RNAi machinery, which is often speculated to have evolved as an antiviral mechanism, to generate small RNAs that serve their own purposes," the researchers noted in the Nature paper.
SV40, Sullivan said, was a perfect virus to research because it's "a smaller virus, so … there wouldn't be as many nucleotides to scan and it would be much easier to screen all our good candidates by Northern blot. In fact, we only had two [miRNAs], so that was a lot easier then working in larger viruses."
Another reason SV40 was used was because of earlier reports indicating that there "were small RNAs being made … in different locales of the virus," he said. "They weren't consistent with microRNAs, but they were consistent with pre-microRNAs.
"Also, SV40 is a really cool virus in itself," Sullivan added. "Before the era of molecular biology, this was an abundant source of DNA, and so it was widely studied. You could also take cells, and you could program them to make new transcripts, and you could study splicing or protein processing. So to find a microRNA in SV40 would, in itself, I think, be very interesting," he said.
To start their search for SV40 miRNAs, Sullivan and his colleagues developed computer algorithms, which have since been updated and enhanced, designed to predict likely miRNA precursors, and then performed a screen in silico for these RNAs.
"We dumped the 5,200 nucleotides into this program that basically predicts hairpin structures," he explained, adding that "it [also] has other criteria that are more common to pre-microRNAs than other hairpins."
Only two good candidate microRNAs were found, one of which, if it were a microRNA, "was below the level of detection on our Northern blot," Sullivan said. The other, however, was highly abundant.
"Once I realized I had the microRNA, then it was nuts-and-bolts molecular biology to figure out when it was made, what its target was, and what its biological function was," he said.
According to the Nature paper, the SV40 miRNAs "accumulate at late times in the infection, are perfectly complementary to early viral mRNAs, and target those mRNAs for cleavage. This reduces the expression of viral T antigens but does not reduce the yield of infectious virus relative to that generated by a mutant lacking" SV40 miRNAs.
The paper notes that "wild-type SV40-infected cells are less sensitive than the mutant to lysis by cytotoxic T-cells, and trigger less cytokine production by such cells. Thus, viral evolution has taken advantage of the miRNA pathway to generate effectors that enhance the probability of successful infection."
The paper's authors wrote that, to their knowledge, SV40 miRNAs "are … the first miRNAs shown to have a function in virus biology. By down-regulating the accumulation of unnecessary T antigen, the [miRNAs] reduce [cytotoxic T lymphocyte] susceptibility and local cytokine release," they noted. "Thus, although this down-regulation is dispensable for viral growth in culture, it is likely to be of considerable importance in vivo.
As for how widespread viral miRNAs might be, Sullivan said that, based on work coming out of his and other people's labs, it appears that the answer might be 'very.'
"A lot of the herpes viruses have been proven [to encode microRNAs] and it's possible [that this is the case for] all the herpes viruses, although Tuschl's work would contradict some of that," he said.
Additionally, Sullivan said that unpublished work from his lab indicates that if computational predictions are run on "all of these small DNA tumor viruses that are lumped together — the polyoma family, the adeno, and the papilloma — they could all possibly encode microRNAs." He added that these findings, however, don't "mean much without confirmation."
Sullivan noted that a big question that remains to be answered is how cytoplasmic viruses get processed by Drosha, which appears to be predominantly nuclear. "Certainly, there's no proof that [many viruses don't make microRNAs], but we're going to have to evoke certain models where the transcripts get to the nucleus or you don't need Drosha to make them," he said.
Additionally, once viral microRNAs are identified, their roles need to be elucidated.
Going forward, Sullivan said he and his colleagues in Ganem's lab are now looking into the role of microRNAs in Kaposi's sarcoma.
"Two groups — Tuschl's group and Bryan Cullen's group [at Duke University] — have published [papers showing] KS makes microRNA," Sullivan said. "They used the traditional cloning approach, and we feel our computational approach is a nice complement to theirs.
"We're really pursing the KS side of things, not only finding new microRNAs that they may have missed, but also figuring out the function of them," he said. "We're also screening some other small viruses that are of interest" to individual members of the lab.
"A lot of the DNA viruses, I think we're going to find, make" microRNAs, he said. "Then the real question is going to be, 'Can we figure out what these guys do?'."
— Doug Macron ([email protected])