Sara van den Berg

156 Chapter 6 Even though most research has focused on the large numerical expansions of CMV-specific CD8 + T-cells, also the CD4 + T-cell pool is significantly altered in CMV+ compared to CMV- individuals. We found that the expression of late-stage differentiation markers, as well as proliferation and anti-apoptotic markers was affected by CMV. Also, a trend towards a decreased CD4 + T EM/EMRA cell production rate was observed in CMV+ versus CMV- individuals in vivo by deuterium analysis. As we did not use peptide-MHC class II tetramers, we could not quantify to what extent CMV-specific CD4 + T-cells were responsible for these differences and to what extent CMV affected non-CMV-specific memory CD4 + T-cells. We speculate that the CD4 + T EM/EMRA cell population in CMV+ individuals mainly consists of CMV-specific cytotoxic T-cells. Cytotoxic CD4 + T-cells express a T EM/EMRA phenotype and are typically only present in small numbers in CMV- individuals [36, 37]. As CMV harnesses multiple mechanisms to escape neutralizing antibodies as well as the CD8 + cytotoxic T-cell response, CD4 + cytotoxic T-cells may be essential in the control of CMV [38]. CD8 + T-cells, which are known to have larger cytotoxic potential than CD4 + memory T-cells, have a longer expected lifespan than CD4 + T-cells [22], suggesting that cytotoxic function may be related to longevity. We also assessed if maintenance of CMV-specific CD8 + T-cells occurs due to the accumulation of long-lived cells, as was suggested by Wallace et al. [21]. In this deuterated glucose labelling study, tetramer+ CMV-specific T-cells were compared to kinetically heterogeneous tetramer- T-cell subsets (CD45RO + (i.e. T EM and T CM ) and CD45RO - (i.e. T N and T EMRA ) T-cells) within the same individual and were found to be longer-lived (lower in production and loss rate). In theory, our approach should be able to pick up even smaller differences by comparing to more strictly defined memory T-cell populations and also comparing to non-CMV-specific memory T-cell populations of CMV- individuals. Nevertheless, we did not find a significantly decreased production or loss rate of CMV-specific CD8 + T-cells in comparison with CD8 + T EM/ EMRA or T CM cells from CMV+ or CMV- individuals. In line with this, in a recent study of murine CMV, we found that the in vivo production rates of MCMV-specific T-cells were very similar to those of bulk memory T-cells (Baliu-Piqué et al. in preparation). We therefore believe that the maintenance of high numbers of CMV-specific CD8 + T-cells is not linked to substantially altered production or loss rates compared to other memory CD8 + T-cells, and that the large size of the CMV-specific T-cell pool is set shortly after CMV infection (Samson, Van den Berg et al. in preparation). The underlying mechanisms explaining these large CMV-specific expansions remain to be investigated. Despite the increased CD8 + T-cell numbers and their late stage differentiation phenotype, no significant differences in CD8 + T-cell proliferation and apoptosis markers or production and loss rates (as measured by in vivo deuterium labelling) were observed between CMV+ and CMV- individuals. It was previously proposed that chronic infection may lead to increased T-cell production rates due to the continuous recruitment of new antigen- specific T-cells from the naïve or central memory pool into the effector memory compartment [39]. Alternatively, it has been proposed that CMV-specific CD8 + T-cells might outcompete other pre-existing antigen-specific T-cells [40-42], and thereby affect their lifespan. In our