- Imbalanced deoxyribonucleoside triphosphate (dNTP) pools are mutagenic for DNA synthesis in both intact cells and cell-free replication systems. Almost certainly, such mutagenesis involves competition between correctly and incorrectly base-paired precursors at replication sites. However, there are certain differences between the intact cell and cell-free systems that do not always allow direct comparisons to be made with regards to mutagenesis stimulated by dNTP pool imbalances. For example, even though dNTP concentrations can be carefully controlled in vitro, cell-free replication systems almost certainly do not contain all of the components found at replication forks in vivo. In contrast, intracellular dNTP concentrations, and base changes in DNA, have not always been measured in parallel with conditions thought to cause imbalanced dNTP pools and mutagenesis in vivo. I describe in this dissertation the further development of a procaryotic system, namely bacteriophage T4, as a model in vivo system for studying dNTP pool imbalances and mutagenesis. While these investigations focus on two enzymes in T4 deoxyribonucleotide metabolism, namely T4 ribonucleotide reductase and dCMP deaminase, the effects of other phage mutations on dNTP pools and mutagenesis were also investigated. Loss of T4 dCMP deaminase activity during phage infections resulted in hyperexpanded hmdCTP pools, decreased dTTP pools, and expanded dGTP pools. There was a concomitant increase in AT-to-GC mutagenesis, which was confirmed by nucleic acid sequencing of amber⁺ phage revertants. To some degree, the aberrant dNTP pools and AT-to- GC mutation rates could be returned to normal by the addition of thymidine to infections by the dCMP deaminase deletion mutant, pseTΔ4, but not during infections by the dCMP deaminase missense mutant, cdN16. Both AT-to-GC and GC-to-AT mutations were weakly stimulated during T4 ribonucleotide reductase mutant, nrdBamB55, infections of ED8689/ pPS2, a host cell overproducing Escherichia coli (E. coli) ribonucleotide reductase. During T4 nrd⁺ infections, BrUdR mutagenesis preferentially stimulates GC-to-AT mutagenesis; while in uninfected E. coli, BrUdR stimulates the opposite pathway, namely AT-to-GC transitions. However, during nrdBamB55 infections of ED8689/pPS2, BrUdR mutagenesis now resembles that of uninfected E. coli, namely, AT-to-GC transitions prevail. I discuss a model based on these results that explains how BrUdR stimulates opposite mutagenic pathways in T4-infected and uninfected E. coli.