SUMMARY Long-term expression of broadly neutralizing antibodies (bNAbs) has the potential to suppress an established HIV-1 infection. However, current methods for maintaining high bNAb concentrations necessary for this control are inadequate. Passive infusion of bNAbs is prohibitively expensive and requires HIV-1 positive individuals to receive infusions on a weekly or monthly basis. Delivery of bNAbs by gene-therapy vectors almost invariably raises anti-drug antibodies (ADA) against expressed bNAbs, which are immunogenic due to their extensive hypermutation. Most importantly, no single set of antibodies can adequately suppress the range of viruses in the population, in large part because current antibody delivery systems fail to do what an immune system does well: adapt to a diverse and evolving pathogen. Here we describe a series of technical advances that allow us to introduce bNAb heavy- and light-chain genes into their native loci in primary B cells. These technologies enable in vivo improvement of bNAbs through affinity maturation in mice and primates, using the acquired wisdom of the humoral response to rapidly increase antibody potency, breadth, and bioavailability. They also allow us to test the core hypothesis of this proposal that B-cell delivered bNAbs can permanently suppress an established infection in the absence of anti-retroviral therapy (ART). The chief technical advance that enables these studies is the development of an efficient double-editing technique for simultaneously replacing the variable heavy and light chain segments of B cell receptors. This is made possible through use of a newly characterized Cas12a ortholog and a unique homology-directed repair template design capable of efficiently replacing nearly any endogenous BCR variable region. The net consequence is that, unlike related B-cell editing approaches, the full regulatory apparatus of the B cell is left intact, facilitating robust B-cell development and efficient affinity maturation of the B-cell receptor. The project is divided into three aims. Aim 1 will increase the breadth and potency of three well characterized bNAbs through affinity maturation in vivo. Aim 2 will extend CRISPR editing to the Fc domain, introducing a recently described set of mutations into the IgG1 Fc domain that facilitate antibody transfer across the blood- brain barrier. Finally, Aim 3 tests the ability of primary B cells expressing the bNAbs improved in Aim 1 to control a SHIV infection in rhesus macaques. A series of structured treatment interrupts will be performed to drive CAR B proliferation and generate an individualized response to virus that emerges from the reservoir. After these structured interruptions, ART will be permanently withdrawn to determine if CAR B cells alone can control an established infection.