A stochastic model of
protein synthesis was modified by including the process of dissociating
peptidyl-tRNA from ribosomes. To simulate ribosome editing, the probability of dissociation was assumed to be high if the
peptidyl-tRNA was erroneous; that is, if it resulted from transfer of a
peptide to an aminoacyl-
tRNA that was inappropriate relative to the
mRNA codon. The effects of
amino acid starvation on
protein synthesis were simulated both by increasing the probability of such erring at and by reducing the conditional probability of elongation at "hungry"
codons, those whose correct
amino acid was in short supply. These probabilities were varied systematically to simulate
tryptophan limitation during synthesis of coat
protein from bacteriophage MS2. Significant reduction, during
starvation, in the synthesis of complete coat
protein required large reductions in the probability of elongation at hungry
codons but only small increases in the probability of erring. Enhanced dissociation of
peptidyl-tRNA during
starvation, followed rapidly by dissociation of ribosomes from
mRNA, led to reductions in mean polysome size, a result that had been interpreted by others as due to some effect of
starvation on the initiation of
protein synthesis. Results from experiments by Goldman (1982) on the cell-free synthesis of MS2 coat
protein during
tryptophan starvation could be mimicked in detail by the computer simulations. A simple competition between correct and erroneous
amino acids was sufficient to explain the
tryptophan dependence of complete coat
protein and internal
peptide syntheses. Values for the Michaelis constants were derived from the computer simulations.