Acquired immunity tends to mean that secondary virus infections are much less severe than the first infection. An exception is secondary infection with dengue virus, which is associated with the deadly dengue haemorrhagic fever (DHF). Why the second hit should be worse that the first is a mystery. One hypothesis is that DHF might be caused by the overproduction of cytokines by virus-specific T cells — but we know almost nothing about T-cell responses to the virus. Now, researchers in the UK and in Thailand have developed dengue-virus-specific MHC class 1 tetramers, which give the first insights into the virus-specific T-cell responses and their possible role in DHF.

Mongkolsapaya et al. identified a new dengue virus T-cell epitope presented by an MHC class I molecule that is prevalent in South East Asian populations (HLA-A*11). From this, they generated HLA-A*11–epitope tetramers and used them to track dengue-virus-specific CD8+ T cells in Thai patients with secondary dengue-virus infection. Early in infection, they found that the epitope-specific T cells were present in low numbers and could not produce the antiviral cytokine interferon-γ (IFN-γ) — a state the authors refer to as “stunned”. But, at later stages, the number of specific T cells increased markedly, peaking at 2 weeks, and they acquired the capacity to secrete IFN-γ. Notably, the patients with the most severe DHF had the highest peak levels of specific T cells (more than 2% of peripheral blood T cells).

Why are the numbers of virus-specific T cells initially so low? Although the epitope-specific T cells in the blood seemed to be proliferating vigorously during acute infection, further investigation indicated that most were apoptotic. So, activation-induced cell death might account for the lack of antigen-specific T cells during the crucial early stages of infection.

But the most intriguing finding of this study concerns the serotype specificity of the T cells. There are four dengue-virus serotypes, and tetramers were made with epitope variants representative of each serotype. In many cases of secondary infection, a large number of T cells reacted preferentially with tetramers of a serotype that was different to that of the current infection. This hints that populations of memory T cells that are specific for a primary dengue-virus infection might be preferentially expanded and activated in a subsequent infection with a different serotype. The potential problem lies in the fact that the reactivated memory cells probably have a lower affinity for the secondary virus and will be less effective at controlling it. So, a dengue-virus vaccine might have to induce immunity to all four serotypes if it is to prevent DMF.