In silico structural and docking models of dipteran FXPRLamide neuropeptides support ligand-receptor coevolution and suggest mechanisms for ligand bias.

Pyrokinin (PK) neuropeptides are characterized by a conserved C-terminal FXPRLamide motif and modulate a range of physiological functions and behaviors in species spanning the Eumetazoa. The insect pyrokinin gene pban encodes a taxon-specific number of PKs including the eponymous pheromone biosynthesis activating neuropeptide (PBAN). The pban gene of basal Diptera resembles that of other insect orders while in more derived Diptera (where it is referred to as the hugin gene) the PBAN peptide coding sequence appears to be absent. In the present study, in silico structural models and docking simulations of the Drosophila melanogaster PK hugin and its receptor PK2-R1 are compared with those of the PBAN and PK2-3 neuropeptides (the latter the likely homolog of hugin) and their receptor PK2/PBAN-R belonging to the more basal species Aedes aegypti. The binding pockets for all three peptides overlap extensively as do individual amino acid contacts; these interactions also concur with data for PBAN and NMU binding of their cognate receptors in the silk moth Bombyx mori and humans respectively. C-terminal FXPRL core sequences of all peptides preferentially bind conserved residues in the transmembrane regions (TMs) of the receptor while the variable N-termini interact with amino acids in the extracellular loops (ECLs) that differ between the two species. The A. aeg PBAN peptide forms rigid secondary and tertiary structures with its long N-terminus that uniquely interact with non-conserved amino acids in the extended ECLs of PK2/PBAN-R, providing a basis for functional differentiation from binding of the short and flexible PK2-3 peptide to the same receptor, perhaps via a mechanism such as ligand bias. D. melanogaster hugin is similar in structure to PK2-3 but interacts with amino acids in areas of PK2-R1 that contact PBAN but not PK2-3 in A. aegypti, even though amino acids at those receptor sites are non-synonymous substitutions compared with PK2/PBAN-R. The ECLs of PK2-R1 are also shorter than those of PK2/PBAN-R, reflecting the loss of selection for contacts with the much longer PBAN peptide. Taken together these changes suggest that loss of PBAN impacted ligand-receptor coevolution in the higher Diptera.