Mathematical models have been developed that describe possible sources of the free amino acids (AA) present in the red blood cells (RBC). These models reflect three hypotheses, namely (i) direct transport between plasma and RBC, (ii) enhanced direct transport by deformation of the RBC during passage through capillaries and (iii) direct exchange between tissues and RBC. The latter implies that RBC samples from appropriate blood vessels might offer an alternative to tissue biopsies in protein turnover studies.

These compartmental models were tested against data from ovine studies. For hypothesis (i), a single dose of [U-13C] AA was added to blood in vitro. The rate of disappearance of label from plasma was low (kP=0·0009–0·004 per min), indicative of slow direct transport between plasma and RBC. For hypotheses (ii) and (iii), a single dose of [1-13C] AA was administered intravenously to sheep. Label disappeared from plasma quicker (kP=0·33–0·53 per min) than in vitro, with the appearance of labelled AA in the RBC higher than could be explained from direct transport alone. The model reflecting enhanced exchange of free AA between plasma and RBC due to capillary action over-predicted RBC enrichments and pool sizes by up to five-fold. The third hypothesis, namely that AA in the RBC are predominantly derived from exchange with tissues, gave better agreement. This model suggested that at least 0·63 of the AA in the RBC might come from exchange with tissue.

To investigate whether peripheral tissues (skin and muscle) might be involved in such exchange, an additional experiment was conducted in five sheep prepared with arterio-venous catheters across the hindquarters. [1-13C,15N] leucine was continuously infused, with both blood samples and tissue biopsies taken. The ratio of [1-13C] leucine to [1-13C,15N] leucine was 1·8–2·9 fold greater in muscle and skin compared with arterial plasma, indicative of intracellular transamination. If free leucine exchanged directly between tissues and RBC then the ratio should be higher in RBC isolated from venal caval compared with arterial blood. In practice, the ratios from the RBC in aorta and vena cava were similar (P=0·30). These findings suggest that, for leucine, RBC enrichments do not form an alternative to tissue biopsies to estimate the enrichment of the precursor pool in protein turnover studies of the hindquarter. Instead, other tissues must play a more important role in the exchange of AA with the RBC.