The human APOBEC3G (A3G) protein is a DNA deaminase that can inhibit HIV replication. In most forms of HIV though, the deaminase activity is limited by the viral protein VIF. While recent work has shed light on the three dimensional structure of A3G, little is known about its complete structure and the domains that mediate self-interaction. For instance, inside some types of cells human, A3G forms much larger aggregates with itself, RNA and other proteins, but it is not known how these aggregates are put together. In order to fully understand how A3G interacts with polynucleotides (single-stranded DNA and RNA), how it is inhibited by VIF and how it is regulated within the cell, it is necessary to know how two A3G proteins dimerize. I hypothesized that a distinct site within A3G nucleates or mediates dimerization. To test this hypothesis, a large library of full-length A3G mutants was created. These mutants were then tested in a well-known protein-protein interaction assay called the yeast two-hybrid system. Mutant A3G proteins were pitted against the C-terminal domain of A3G to screen for those that had lost the capacity to dimerize. Candidate dimerization mutants were then sequenced to determine the changes in the nucleotide bases that were responsible. I currently have over ten non-interacting candidates. In the coming months, theses candidates will be tested for their ability to deaminate DNA, dimerize in an independent protein-protein interaction assay and restrict HIV replication in a model cell-based system. My data will help reveal whether the functional unit of A3G is a monomer, a dimer or some more complex polymer.