Deuterium‐Oxide Labelling Proteomics Illustrates the Importance of Sample Timing and the Limitations of Serial Biopsies

Background Measuring in vivo protein synthesis is important for testing how interventions, disease and ageing remodel the proteome. A limitation of approaches using short‐term infusions of labelled amino acids was that long‐term changes had to be extrapolated from short‐term measurements. Deuterium oxide (D 2 O) is now widely used because it can measure protein synthesis over periods of days to weeks. We previously showed that the duration of labelling impacts measured protein synthesis, but many still employ repeated tissue sampling during D 2 O provision. We hypothesized that the timing of tissue sampling alters which proteins contribute to the overall protein fractional synthesis rate (FSR), thereby invalidating comparisons between serial samples collected during continuous labelling. Methods To test this hypothesis, we used a five‐timepoint (0d, 4d, 15d, 30d and 45d) D 2 O labelling time course followed by liquid chromatography–tandem mass spectrometry proteomics to calculate individual protein synthesis rates and plateau values in mouse gastrocnemius muscle. Individual proteins were stratified into slow and fast groups, such that each group was responsible for the slowest or fastest 20% of total synthesis over the first 4 days of labelling, respectively. Results In a 4‐day period representing a baseline period, the synthesis rate of summed proteins was 2.09%/day. When the period was extended to 15, 30 or 45 days, the synthesis rates of summed proteins declined to 1.64%, 1.31% and 1.10%/day, respectively. The contribution of slow and fast proteins to these rates changed as a function of time. The fast and slow proteins accounted for 4.3% and 34.0% of measured proteome abundance, respectively. From Day 0 to Day 4, fast and slow proteins contributed 19.8% and 20.2% of the calculated FSR, respectively. By Day 45, these contributions shifted to 6.5% and 28.8%. Further, we demonstrated that the assumption that the fraction new of proteins linearly increases to 100% new is not true for the majority of proteins. Instead, most proteins plateaued at some point less than 100%, which stops their individual contribution to a summed protein synthesis rate. Conclusions In continuous long‐term labelling studies, biopsy timing changes the measured bulk FSR because the relative contributions of individual proteins to that measurement over time. Fast proteins that plateau early decrease their contribution as labelling time increases. These findings indicate that studies with different labelling periods, or multiple biopsies over time during continuous labelling of days to weeks cannot be compared to biopsies taken at other times in the labelling period.