Peptide-tiling screens of cancer drivers reveal oncogenic protein domains and associated peptide inhibitors

Reprogramming Adeno-Associated Virus Tropism Via Displayed Peptides Tiling Receptor-Ligands

“…First, we sought to interrogate how peptides derived from specific ligand protein regions alter AAV tropism. This was motivated by our previous observation that tiled peptide enrichment patterns in a screen can yield insight into functional domains of the proteins from which they were derived ( 37 ). To do this, we interrogated tissue-specific AAV variants identified in the screen, focusing first on a lung-specific variant (termed as AAV9.DKK1) that was derived via display of a peptide fragment from the DKK1 gene on the AAV9-Loop2 scaffold.…”

“…Second, a defined library allows for robust quantification of the fitness of each AAV variant, enabling facile stratification of AAV variants by infectivity across organs of interest. While we primarily applied this methodology to engineer AAVs, mining nature for functional biomolecules has applications in a wide range of protein engineering challenges, such as engineering orthogonal viral (including lentiviruses) and non-viral (including lipid nanoparticles ( 74 )) vectors or identifying biologic inhibitors of critical protein/protein interactions, as we have shown recently ( 37 ).…”

“…sgRNA sequences targeting LRP6 or non-targeting controls were identified using CRISPick ( 108, 109 ) and cloned into the lentiCRISPR v2 plasmid backbone ( 110 ). Lentivirus was then produced as described in ( 37 ). Briefly, HEK293T cells were seeded at ∼40% confluency the day before transfection.…”

“…With a goal to develop a rational and potentially programmable strategy for enabling tissue targeting, we hypothesized that receptor-ligand interactions which mediate the spectrum of naturally occurring cell-protein and cell-cell interactions could be repurposed for engineering AAV tropism. However, as neither the interaction interfaces nor the cell-type specificity profiles of receptor-ligand interactions are fully mapped, to comprehensively interrogate as well as engineer these, we thus leveraged our recently developed PepTile approach for mapping and targeting putative protein interaction domains ( 37 ). Specifically, we fragmented all annotated protein ligands known to bind human receptors into tiling short peptides and displayed these onto surface loops of AAV capsids, an approach we term AAV-PepTile .…”

“…Short peptides grafted into stabilizing molecular scaffolds can recapitulate local protein domain structure ( 38 ), and as protein-protein interface sites are typically 1200-2000 Å 2 ( 39 ), with specific peptide hot loops ranging from 4-8 amino acids (AAs) contributing maximally to the binding energy of protein-protein interactions ( 40 ), we hypothesized that 20 amino acid peptide insertions into the AAV capsid could drastically alter its binding ability, and therefore, its transduction profile in vivo. While insertion of still longer peptides could in principle more faithfully recapitulate local ligand structures ( 37, 41 ), AAVs do not generally tolerate larger than 20-25 amino acid insertions without severely compromising capsid packaging efficiency ( 42 ). Based on this, we proceeded with a 20AA peptide + flanking 2AA linker insertions, all together constructing a library of >1 million AAV variants by inserting corresponding oligonucleotide pool synthesized gene fragments coding for potential receptor-ligands and cell membrane permeable proteins into one of two surface loops on AAV5 and AAV9.…”