Empirical assessment of model validity

The metabolism of amino acids is far more complicated than a 1- to 2-pool model. Yet, these simple models have been extensively used with many different isotopically labeled tracers to study protein metabolism. A tracer of leucine and measurement of leucine kinetics has been a favorite choice for following protein metabolism. However, administering a leucine tracer and following it in blood will not adequately reflect the complex multi-pool nature of the leucine system. Using the tracer enrichment of the ketoacid metabolite of leucine, alpha-ketoisocaproate (KIC), to reflect intracellular events of leucine was an important improvement. Whether this approach is adequate to follow accurately leucine metabolism in vivo or not has not been tested. From data obtained using simultaneous administration of leucine and KIC tracers, we developed a 10-pool model of the in vivo leucine-KIC and bicarbonate kinetic system. Data from this model were compared with conventional measurements of leucine kinetics. The results from the 10-pool model agreed best with the simplified approach using a leucine tracer and measurement of KIC enrichment