Dependence of the amplification factor on the body water deuterium enrichment

Heavy water (D₂O) labeling is the state-of-the-art technique to track the dynamics of circulating cells in vivo. D₂O labels dividing cells through incorporation of deuterium into newly synthesized DNA, which is measured using GC/MS. The labeling rate depends on (1) the level of body water enrichment, (2) cell kinetics, and (3) an amplification factor quantifying the deoxyribose enrichment relative to the body water enrichment. This amplification factor is typically estimated using a reference population undergoing rapid turnover (such as granulocytes), and is larger than one because deoxyribose contains seven hydrogens that can be replaced by deuterium. In a meta-analysis, we found that individuals differ markedly in this amplification factor. Since the amplification factor also depends on the level of body water enrichment, we use conventional binomial expressions to describe the fractions of deoxyribose incorporating zero, one, two, or more deuterium atoms. We extend this classic binomial model with a new parameter, 0<γ<1, describing the relative contribution of hydrogens from body water during deoxyribose synthesis. While for most studies, our 'novel Binomial' model reasonably explains the slope with which the amplification factor declines with the level of body water enrichment, we find that some individual amplification factors differ considerably from their expected values Re-fitting deuterium labeling data of granulocytes with the Binomial model reveals that the actual decrease is steeper than expected. We speculate that this residual variation depends on differences in diet, metabolism, and/or life style, which apparently correlate with daily fluid intake.