Autophagy and apoptosis are catabolic pathways essential for homeostasis in all eukaryotes. While autophagy usually promotes cell-survival by enabling degradation and recycling of damaged macromolecules, apoptosis is the canonical programmed cell death pathway. Dysfunction of autophagy and apoptosis is implicated in diseases like cancers, cardiac disease, and infectious disease. Beclin homologs are key for autophagy as they are core components of class III phosphatidylinositol 3-kinase autophagosome nucleation and maturation complexes. Anti-apoptotic Bcl-2s inhibit apoptosis through binding and antagonizing pro-apoptotic Bcl-2s. Anti-apoptotic Bcl-2s also down-regulate autophagy by binding to the Bcl-2 homology domain 3 (BH3D) of Beclin homologs, thereby enabling crosstalk between apoptosis and autophagy. The central goal of my doctoral research is to understand the structure-based mechanism of selected Bcl-2s and Beclin homologs in the regulation of autophagy and apoptosis. -Herpesviruses are common human pathogens that encode viral Bcl-2s to facilitate viral reactivation and oncogenic transformation. M11, a viral Bcl-2, and Bcl-XL, a cellular Bcl-2, bind with comparable affinities to the Beclin 1 BH3D. By assessing the impact of different Beclin 1 BH3D mutations on binding to M11 and Bcl-XL, we developed a cell-permeable inhibitory peptide that targets M11, but not Bcl-XL. The mechanism by which this peptide specifically binds to M11 was elucidated by determining the X-ray crystal structure of the peptide:M11 complex. Our attempts to investigate the role of other Bcl-2s in these pathways were unsuccessful. In one project, we were unable to express and purify BALF1, another viral Bcl-2. In another, no interaction was detected between purified samples of the anti-apoptotic Bcl-2 homolog Mcl-1, and the autophagy protein Atg12, that had previously been shown to bind. We also delineated the domain architecture of Beclin 2, a novel Beclin homolog and attempted to express and purify different Beclin 2 constructs for structural studies. We successfully purified and solved the X-ray crystal structure of the Beclin 2 coiled-coil domain (CCD), showing it is a curved, anti-parallel, meta-stable coiled-coil homodimer with seven pairs of non-ideal packing residues. In general, mutating the non-ideal packing residues to leucines enhanced Beclin 2 CCD homodimerization, but also weakened its binding to Atg14 CCD.