Sara van den Berg

105 4 CMV on influenza infection CMV- individuals. Furthermore, these levels did not seem to be associated with the influenza virus specific T-cell response at any time point. Of course, we cannot rule out the possibility that other immune modulatory factors may be influenced by CMV infection and may act on the influenza-specific T-cell response. Most studies claiming a negative effect of CMV on a immune response to a heterologous challenge have been performed in mice infected with murine CMV (MCMV). Several mouse studies have shown that only lifelong infection with MCMV leads to decreased immunity against heterologous infections [58-60]. The magnitude of the effect of CMV infection might thus be linked to the duration of CMV infection and the experienced amount of viral reactivation in an infected host. Here, we found no evidence for decreased immunity against a heterologous infection in humans infected with CMV. Also, high CMV-specific antibody levels were not associated with the height of the influenza virus-specific T-cell response, as was previously suggested for influenza antibody vaccine responses [37, 61]. One of the reasons for the observed differences between mice and men, could be the order of infections by CMV and influenza. In mouse studies, mice are typically first infected with MCMV long before they are challenged with a heterologous acute infection. Many humans, in contrast, may have undergone their first influenza virus infection before they were infected with CMV, which may lead to the presence of influenza-specific memory T-cells before the CMV-specific immune response is established. A potential harmful effect of CMV might be less pronounced in a host who already has a proper immune response against influenza. Furthermore, as mouse models of CMV are almost exclusively done in specific pathogen free mice, and humans are exposed to dozens of infections and triggers during life, it might be that the small effect of CMV is magnified in mice. Even if CMV can modulate other immune responses, there is no substantial evidence that CMV impacts the function of the immune system by hampering immune responses against heterologous infections in humans. The increased CD8+ T-cell response to influenza virus infection that we observed in CMV+ older individuals remains partially unexplained. Previously, enhanced influenza vaccine responses in humans and mice were explained by an increase in IFN γ in serum [1]. As we did not observe a difference in IFN γ levels in serum between CMV+ and CMV- individuals, this could not explain the small positive association with CMV infection in our study. We cannot exclude the possibility that CMV infection may only affect CD4+ T-cells, as this effect was observed in IFN γ ELIspot assays, in which CD4+ T-cells may contribute as well. Another explanation for the difference in the early influenza virus-specific T-cell response between CMV- and CMV+ individuals could lie in the migratory capacity of the responding T-cells. We observed a trend towards enhanced CXCR3 expression of influenza-virus specific CD8+ T-cells in CMV- individuals, which may lead to early migration towards tissues such as the lung. Since the presence and function of tissue-resident T-cells are crucial in respiratory diseases such as influenza, the effect of CMV on influenza virus-specific T-cell responses at the site of infection instead of the blood would be of great interest, and requires further research.