Back

Compartment-resolved in vivo phage display biopanning reveals constraint-driven peptide sequence landscapes

Okano, J.; Katagi, M.; Nakae, Y.; Furuhashi, K.; Fujino, K.; Shindo, A.; Furusho, Y.; Kojima, H.

2026-04-17 molecular biology
10.64898/2026.04.16.719084 bioRxiv
Show abstract

In vivo phage display has emerged as a powerful approach for identifying targeting ligands in living organisms. However, despite the extensive identification of organ-associated peptides, the principles governing peptide targeting remain poorly understood, in part because most studies have measured peptide enrichment at the level of whole organs without resolving the underlying microanatomical compartments. Here, we combined laser-capture microdissection with high-throughput sequencing to perform compartment-resolved in vivo phage display across multiple organs, enabling separate analysis of vascular and parenchymal peptide selection landscapes. Across the bulk, capillary, and parenchymal datasets, peptide selection showed reproducible non-random structure, indicating that peptide recognition is shaped by biological constraints that differ across tissue compartments. Motif-centered analyses identified distinct constraint regimes, ranging from bounded families of related sequences to near collapse onto a single sequence, thereby revealing a continuum of permissible solutions across tissue environments. Notably, vascular- and parenchyma-anchored motifs converged on a subset of interface-associated organs, suggesting coupling between vascular entry and tissue-level enrichment. As a proof of concept, we identified a brain-targeting peptide from the parenchymal dataset that was not recovered as a meaningful candidate in the bulk analysis. Together, these findings establish compartment-resolved in vivo phage display as a framework for revealing peptide targeting principles that are obscured by whole-organ measurements.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.