Sara van den Berg

141 6 T-cell dynamics in CMV tetramer, either with HLA-B*0702/TPRVTGGGAM-PE, or HLA-A*0201/NLVPMVATV-PE or -APC, or both HLA-A*0101/VTEHDTLLY-APC and HLA-B*0801/ELRRKMMYM-PE (all from Immudex), for 30 minutes at room temperature. An overview of CMV-tetramers used per donor can be found in Supplementary Table 1 . Cells were measured on a Fortessa™ flow cytometer (BD Biosciences) and data was analysed with FlowJo V10 software. T-SNE analysis was performed with of every donor 10.000 CD8 + T-cells with the senescence panel for CD8 + T-cells (cytobank.org ). In vivo deuterated water labelling protocol Ten participants (five CMV+ and five CMV-) were included in a longitudinal heavy water labelling study to quantify the underlying dynamics, i.e. production and loss rates, of different sorted T-cell subsets. In short, 2 H is incorporated via de novo DNA synthesis by cells undergoing cell division and lost when cells die, differentiate, or migrate to a different body compartment [25]. Participants received a ramp-up dose of 2 H 2 O on the first day, after which they drank a daily dose of 2 H 2 O for five weeks (except for participant E30 who drank 2 H 2 O for only three and a half weeks for logistical reasons). The procedures on the first day and the follow-up schedule were largely as described before [22], though we decreased the length of labelling from nine to five weeks. Urine was collected at thirteen time points during the up- and down-labelling phases to correct for 2 H 2 O-availability in the body. Blood was collected to sort T-cell subsets at eight time points, of which four during up-labelling and four during down-labelling, with an average follow-up time of 448 days (range 378 to 511 days). Sorting of T-cell subpopulations by flow cytometry Directly after Ficoll-Paque isolation, between 5∙10 7 and 20∙10 7 PBMCs were stained with CD95-FITC, CD4-APC-eF780 (both eBioscience), CD3-PerCP, CCR7-BV421, CD8a-BV510, and CD45RO-PE-Cy7 (all from BioLegend), and either CD56-APC, CD56-PE, or CD56-PE/ Dazzle-594 (the first purchased from BD Biosciences and the latter two from BioLegend). T-cell subpopulations were defined as follows: truly naive, T TN (CCR7 + CD45RO - CD27 + CD95 + ), central memory, T CM (CD45RO + CD27 + ), effector memory, T EM (CD27 - CD45RO + ) and effector memory re-expressing RA, T EMRA (CD27 - CD45RO - ). The full gating strategy can be found in Supplementary Figure 1 . Cells were sorted on a FACSAria™ II or FACSAria™ III sorter (BD Biosciences), and data was analysed with the BD FACSDiva™ v8.0.1 and FlowJo V10 software. DNA isolation and measurement of deuterium enrichment by GC/MS After sorting of the different T-cell populations, DNA was isolated after overnight storage of the cell pellets at 4⁰C using the ReliaPrep™ DNA isolation kit (Promega). DNA samples were stored in 200 μ L DNAse free water at -20⁰C until further processing. Next, we followed the protocols described previously [22, 25], with minor modifications, to obtain the percentage deuterium enrichment in body water and immune cell populations. Briefly, DNA samples from sorted lymphocytes were enzymatically hydrolysed into deoxyribonucleotides and conjugated to pentafluorotriacetate (PFTA). The PFTA derivative of deoxyadenosine was analysed by