News & views author

For every immune challenge there is an opposing immune response. But scientists are using immune reactions to one challenge to treat a completely separate challenge. In the paper on page 682, scientists look at how one particluar response — a ligand–receptor pair that most probably evolved to prevent tissue damage following an immune response — could be manipulated to clear a different chronic infection. In a related News and Views (page 669), Michael Bevan, an immunologist at the University of Washington in Seattle, looks into the implications of the balancing act between immune challenge and response. Nature talked to Bevan about his take on the paper's broader implications for immunology and clinical practice.

In your News and Views, you talk about a “delicate balance”. What is this balance and what challenges and opportunities does it present?

The immune system is good. It protects us from lots of things. But it also does a lot of harm, when it gets too widespread with so many T cells going after things and killing them and releasing lots of cytokines. The response can be worse than the virus itself. In mice that don't have a functional immune system, you can infect them with this virus and they'll just retain the virus for life. But if you inject them with T cells and give them an immune system, they would die as a result of the inflammatory response.

So what does this tell us about treating humans?

That remains to be seen. It's hopeful for certain cases of chronic infections. But this is just a mouse model.

Is the reactivation of T cells a short- or long-term strategy in treating chronic viral infections?

We're afraid to call this a long-term strategy because of the threat of autoimmune reactions; things that would ordinarily be held in check for a short period could become full blown. The paper didn't show that, but that's the worry.

What's the next step?

In the paper that was published, the virus wasn't actually cleared. But it showed that the experiment can be carried out repeatedly. It would be interesting to do this with a virus that is actually cleared.

How does this resonate or mesh with your own work?

My lab doesn't work on chronic infections; we work mostly on acute affections, and with the same virus in the same mice.