Back

Noncanonical roles of chemokine regions in CCR9 activation revealed by structural modeling and mutational mapping

Pinheiro, I. D. M.; Dawson, J. R. D.; Calo, N. V.; Paolini-Bertrand, M.; Akondi, K. B.; Tan, G.; Handel, T. M.; Kufareva, I.; Hartley, O.

2024-06-04 molecular biology
10.1101/2024.06.04.596985 bioRxiv
Show abstract

The G protein-coupled CC chemokine receptor 9 (CCR9) plays a major role in inflammatory bowel disease and has been implicated in cancer. Despite its importance as a drug target, there is limited mechanistic understanding of how CCR9 engages and is activated by its endogenous chemokine agonist CCL25. Here, by combining structural modeling with multimodal pharmacological assessment of receptor mutants, we generated a functional map of the CCR9-CCL25 interaction interface and delineated key determinants of binding, agonism, constitutive activity, and G protein vs arrestin signaling. In contrast to all complexes studied to date, where chemokines drive receptor activation through their N-termini, we determined that CCL25 activates CCR9 via a distinct region, its 30s loop. In support of this non-canonical mechanism, CCR9 signaling is tolerant to alanine mutations in the N-terminus of CCL25 but strongly affected by modifications to the 30s loop. Through molecular evolution of the CCL25 N-terminus, we identified chemokine analogs with enhanced binding properties. However, in contrast to other receptor-chemokine systems, these analogs remained full agonists, consistent with the localization of CCL25 signaling determinants outside of the N-terminus. The non-canonical signature of CCR9 activation provides new insights to aid CCR9 drug discovery and may also inform structure-based design of drugs targeting other chemokine receptors.

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.