WHAT'S HOT IN... BIOLOGY, Jan./Feb. 2008

Most viral infections the immune system deals with swiftly and effectively. Some elude destruction and go on to establish a chronic infection. The paper at #10 identifies the main difference between the two types of infection and suggests a therapeutic approach that could help the body to clear persistent viral infections such as hepatitis B, hepatitis C, and HIV.

Rafi Ahmed and his colleagues at Emory University in Atlanta, Georgia, and elsewhere made use of lymphocytic choriomeningitis virus (LCMV), which can go either way, acute or chronic, in mice. One viral strain, the Armstrong strain, is cleared within a week. Another, known as clone 13, differs from Armstrong at only two amino acids across its entire genome, and yet results in a chronic infection.

Ahmed showed almost a decade ago that the primary difference is that in mice with a chronic infection the CD8 T cells no longer have the ability to kill infected cells. In mice that have recovered from an acute infection the CD8 T cells "remember" the virus and are a vital part of acquired immunity. This paper identifies the basis of T-cell ineffectiveness and delivers a pick-me-up to exhausted CD8 T cells.

The differences between Armstrong strain and clone 13 do not affect any of the regions that stimulate T-cell recognition of LCMV; the main difference is in the end result. Harvesting T-cells specific to LCMV from chronic and acute cases, Ahmed's group used a genome-wide microarray analysis to look for differences between the effective T memory cells and the exhausted T cells.

As an aside, it's worth noting that the use of once-miraculous DNA-chip technology (see, for example, Science Watch, 12[4]: 8, July/August 2001) now merits a single sentence. Exhausted cells, it turned out, expressed vastly more of a protein called PD-1, for programmed death 1. Functional memory T cells did not express detectable levels of PD-1.

Looking at the time course of infection in more detail, the researchers noted that both strains elicited the expression of PD-1 early on, but in the acute infection this was rapidly down-regulated, while in the chronic infection expression remained high. PD-1 recognizes two ligands, PD-L1 and PD-L2. In chronically infected mice, PD-L1 is also expressed at high levels, especially on virally infected cells. All of which suggests that PD-1 and its ligand PD-L1 are involved in regulating CD8 T-cell function.

An antibody that blocked PD-1's access to PD-L1 resulted in a greater number of T cells specific to LCMV, and those cells were more capable of killing infected host cells. Treated mice cleared the virus from their systems, remained healthy, and showed no signs of the disease. CD8 killer cells normally derive a certain amount of support from CD4 helper cells, but even in mice without CD4 helper cells, blocking PD-L1 promoted the growth of CD8 cells and restored their potency.

Just to be sure, Ahmed and his colleagues also gave chronically infected mice an antibody to PD-1 itself. This worked too, but not as effectively as blocking PD-1's target. It seems that the interaction between PD-1 and PD-L1 is what suppresses the CD8 cell's ability to proliferate and target infected cells. Interrupt the interaction, even relatively briefly, and you reinvigorate the exhausted CD8 cells for several weeks.

As a final test, Ahmed and his colleagues infected mice that lack the PD-L1 gene. They responded just like wild-type mice when infected with the Armstrong strain, producing a normal CD8 T cell response that cleared the acute infection and resulted in memory T cells specific to the virus. Mice infected with strain 13, however, died as a result of immunopathologic damage.

This suggests that down-regulating PD-1 actually protects the animal from its own immune response, for which there is other evidence too. Some viruses and viral strains have evolved to take advantage of this and establish a chronic infection.

Of course the big news in this paper is not the nitty-gritty detail but the discovery that blocking PD-L1 revives exhausted CD8 cells. This raises two big questions. First, will it work in important chronic human viral infections? Yes, at least for HIV, where there is also apparently a difference in PD-1 between patients in whom the disease progresses normally and those who are "long-term nonprogressors." (See G.J. Freeman, et al., J. Exp. Med., 2006, DOI:10.1084/jem.20061800; and J. Zhang, et al., Blood, 2007, DOI:10.1182/blood-2006-09-044826.) Secondly, vaccines do nothing to help patients deal with chronic infections, probably because their T cells are exhausted. Could blocking PD-L1 enhance the potential of therapeutic vaccines? No word on that. Yet.

Dr. Jeremy Cherfas is Science Writer at Bioversity International, Rome, Italy.