Herpes Simplex (false color) - Pasteur InstituteCytotoxic T lymphocytes (CTL) are important destroyers of virus, recognizing MHC class I on infected cells and killing the cell before virus replication is complete. Unsurprisingly, many viruses (especially, but not only, the large and complex viruses like herpesviruses and adenoviruses) have evolved mechanisms to block MHC class I antigen processing and, therefore, T cell recognition.

What has been surprising is how ineffective these viral immune evasion mechanisms seem to be. (I’ve talked about this before in a bit more detail.) In cells in the incubator (which is where most of the experiments with the antigen presentation immune evasion molecules have been done), the effects often look pretty dramatic; but when you look at the immune response in humans infected with human cytomegalovirus1 or herpes simplex virus2 (both of which have formidable immune evasion functions in tissue culture3) there are pretty hearty and effective CTL responses, suggesting that the immune evasion doesn’t work all that well in vivo.

Most of the viruses that have had immune evasion genes identified are human viruses; that’s where the clinical interest and the money are. One problem with this is that the large DNA viruses that are likely to have these immune evasion functions tend to be pretty species-specific. Human herpesviruses and adenoviruses don’t infect mice well, if at all, and even if they do infect lab animals it’s not really clear how accurately this infection mimics the natural course in humans. There are a couple of lab animal herpesviruses that may be useful models (though even here the effects of immune evasion molecules are quite small.4) Still, it would be nice to understand what’s going on with immune evasion in, say, herpes simplex infection.

Herpes simplex virus actually will infect mice (the alphaherpesviruses tend to be somewhat less host-restricted than the beta or gammaherpesviruses), and while mouse infection isn’t exactly like that in humans it does seem to be similar enough to offer useful insights. Not into immune evasion, though, because the immune evasion gene of herpes simplex (called “ICP47”) works very poorly in mice.5 Recently, though, Chris Wilson’s group took an ingenious approach to looking at the question, and a theme may be starting to emerge.

Blogging on Peer-Reviewed ResearchWilson reasoned that if ICP47 doesn’t work in mice,6 why not replace it with a functionally-similar gene that does? They made recombinant herpes simplex viruses that have immune evasion genes from mouse or human cytomegalovirus that block mouse MHC class I fairly well. These viruses are more neurovirulent in mice than the control viruses7 and this was because the CTL didn’t work as well in the recombinants. 8,9

Their latest paper10 shows something new and exciting with the recombinant virus: It reactivates better from neurons.

Trigeminal ganglion
Trigeminal ganglion

Remember that HSV establishes a latent infection in neurons and, in humans, periodically reactivates and may shed and infect new hosts this way. HSV latency is still a mysterious and little-understood field in spite of huge amounts of work over the years, but a host of recent findings have started to explain at least bits and pieces. It used to be felt that latent HSV didn’t express any proteins, and so should be invisible to the immune system. As I observed here, it’s now known that the immune system can detect HSV in neurons. That offers a mechanism for reactivation to be coupled with immune suppression. The Orr et al paper here offers more support for this concept, and suggests that one of the (most important?) functions for HSV immune evasion is to tip the scales a little further toward the virus being able to escape recognition in the neuron, and therefore being more easily able to reactivate and infect new hosts — something this ubiquitous virus does with extraordinary efficiency. Orr et al call this an “actively testing the waters” model, in which HSV constantly sticks its toes out into the stream of the immune response to see if it’s safe to reactivate.

What especially interesting about this is that we now have (that I know of) three models of MHC class I immune evasion in more or less authentic infections, and each case the immune evasion seems to be designed to support latency, reactivation, and transmission. Here we have HSV using immune evasion in latency and persistence. Ann Hill’s group has recently shown that mouse cytomegalovirus MHC class I immune evasion proteins help the virus persist in the salivary glands,11 so that (as with the HSV story) it is more able to infect new hosts. And in the third model I know of, mouse herpesvirus 68, the MHC class I immune evasion genes are important for, you guessed it, latency and persistence.12

Three data points aren’t very many, but it does suggest that this may be a common feature of the herpesviruses. Perhaps in acute infection the virus doesn’t really “want” to block immune responses all the effectively, because their lifestyle involves only a brief burst of acute infection followed by a long-term latency; and it’s that latency that allows infection of new hosts. The viruses “let” themselves be driven into apparent submission, but then use their immune evasion functions to mount intermittent stealth campaigns to recruit new hosts.

If this is true for herpesviruses (and of course it’s just a hypothesis) is it also true for other viruses that have MHC class I immune evasion functions? Maybe for some (I wonder about the adenoviruses in particular, since they’re also very capable of long-term persistence in spite of immune responses) but I doubt it’s universal — HIV, for example, doesn’t obviously fit into the same lifestyle pattern as the herpesviruses, and yet if has a protein (nef) that targets MHC class I. Still, if offers a framework for thinking about the problem, which is very valuable.


  1. Manley, T. J., Luy, L., Jones, T., Boeckh, M., Mutimer, H., and Riddell, S. R. (2004). Immune evasion proteins of human cytomegalovirus do not prevent a diverse CD8+ cytotoxic T-cell response in natural infection. Blood 104, 1075-1082.[]
  2. Posavad, C. M., Koelle, D. M., and Corey, L. (1996). High frequency of CD8+ cytotoxic T-lymphocyte precursors specific for herpes simplex viruses in persons with genital herpes. J. Virol. 70, 8165-8168.[]
  3. York, I. A., Roop, C., Andrews, D. W., Riddell, S. R., Graham, F. L., and Johnson, D. C. (1994). A cytosolic herpes simplex virus protein inhibits antigen presentation to CD8+ T lymphocytes. Cell 77, 525-535.
    and
    Jones, T. R., Hanson, L. K., Sun, L., Slater, J. S., Stenberg, R. M., and Campbell, A. E. (1995). Multiple independent loci within the human cytomegalovirus unique short region down-regulate expression of major histocompatibility complex class I heavy chains. J. Virol. 69, 4830-4841.[]
  4. For example, Munks, M. W., Pinto, A. K., Doom, C. M., and Hill, A. B. (2007). Viral interference with antigen presentation does not alter acute or chronic CD8 T cell immunodominance in murine cytomegalovirus infection. J Immunol 178, 7235-7241.[]
  5. Jugovic, P., Hill, A. M., Tomazin, R., Ploegh, H., and Johnson, D. C. (1998). Inhibition of major histocompatibility complex class I antigen presentation in pig and primate cells by herpes simplex virus type 1 and 2 ICP47. Journal of Virology 72, 5076-5084.[]
  6. To be strictly accurate, ICP47 does work in mice, just not very well — 100 times less than in humans, maybe, though what the unit of comparison is or should be isn’t clear[]
  7. I first wrote “than wild-type virus”, but I don’t think we know that — the process of recombination made the virus less virulent and the comparisons were done to this reduced-virulence control virus; I don’t see a direct comparison to wild-type virus.[]
  8. Orr, M. T., Edelmann, K. H., Vieira, J., Corey, L., Raulet, D. H., and Wilson, C. B. (2005). Inhibition of MHC class I is a virulence factor in herpes simplex virus infection of mice. PLoS Pathog 1, e7.[]
  9. This actually parallels and supports previous work showing that ICP47 does in fact have a small effect on neurovirulence in mice.[]
  10. ORR, M., MATHIS, M., LAGUNOFF, M., SACKS, J., WILSON, C. (2007). CD8 T Cell Control of HSV Reactivation from Latency Is Abrogated by Viral Inhibition of MHC Class I. Cell Host & Microbe, 2(3), 172-180. DOI: 10.1016/j.chom.2007.06.013 []
  11. Lu, X., Pinto, A. K., Kelly, A. M., Cho, K. S., and Hill, A. B. (2006). Murine cytomegalovirus interference with antigen presentation contributes to the inability of CD8 T cells to control virus in the salivary gland. J Virol 80, 4200-4202.[]
  12. Bennett, N. J., May, J. S., and Stevenson, P. G. (2005). Gamma-herpesvirus latency requires T cell evasion during episome maintenance. PLoS Biol 3, e120.[]