Ligand-dependent temporal patterns of ERK activation and re-sensitisation of Ca2+ signalling at the type 2 neuromedin U (NMU2) receptor

O Bahattab1, K Al-Hosaini2, RAJ Challiss1, GB Willars1. 1University of Leicester, Leicester, UK, 2University of King Saud, Riyadh, Saudi Arabia, 3University of Leicester, Leicester, UK

The family A G protein-coupled receptors NMU1 and NMU2 couple to Gαq/11 and Gαi/o and mediate the physiological effects of the gut-brain neuropeptide, neuromedin U (NmU). More recently, neuromedin S (NmS), which has an identical N-terminal heptapeptide to NmU, has been identified as an alternate, endogenous ligand at these receptors. The distribution of NmS within the CNS suggests that it may be the ligand primarily responsible for the inhibitory effects of NMU2 on feeding. We have previously reported the essentially irreversible binding of NmU to its receptors and the subsequent internalisation of both ligand and receptor (Brighton et al., 2004). This suggests that the processing of internalised ligand by, for example, endosomally-located endothelin-converting enzyme 1 (ECE1; Roosterman et al., 2007), may regulate the life-time of the ligand-receptor complex and potentially regulate receptor recycling, re-sensitisation and G protein-independent signalling. Here, we have assessed the effects of human NmU (NmU-25) and NmS (NmS-33) on ERK activation and the desensitisation and re-sensitisation of Ca2+ signalling at the human NMU2 recombinantly expressed in HEK293 cells. We have also explored the role of ECE1 in these processes.

Using a plate-reader to determine changes in fluorescence of fluo-4-loaded cells, we showed that NmU-25 and NmS-33 evoke equipotent elevations of [Ca2+]i with pEC50 values of 9.05±0.05 and 8.93+0.08, respectively (all data are mean±s.e.m., n>3). These ligands also caused time- and concentration-dependent desensitisations of Ca2+ signalling. Recovery from desensitisation was reduced (p<0.05, Bonferroni’s terst) by inhibition of either dynamin-dependent internalisation (dynasore, 80 µM) or endosomal acidification (monensin, 50 µM). Recovery from desensitisation was slower following exposure of cells to NmS-33 (~6 h) compared to NmU-25 (~3 h). Furthermore, recovery following NmU-25, but not NmS-33 was reduced by an inhibitor of ECE1, SM-19712 (10 µM), whether this agent was added prior to, or immediately after the period of agonist stimulation. Both NmU-25 and NmS-33 caused ERK activation as assessed by immunoblotting of phospho-ERK (pEC50 8.82±0.08 and 10.03±0.24, respectively at 5 min). Stimulation with NmU-25 (30 nM, 5 min) followed by removal of ligand resulted in ERK activation that had returned to basal by 90 min. In contrast, following similar stimulation and removal of NmS-33, ERK activation was sustained for at least 180 min. Under these conditions, inhibition of ECE1 increased the ERK response to NmU-25 stimulation throughout the time-course (5-180 min, p<0.05, Bonferroni’s test), but had little effect on the response to NmS-33.

These data suggest that although G protein-dependent Ca2+ signalling by NMU2 is not different between the two peptide ligands NmU-25 and NmS-33, there are differences in longer-term effects, including re-sensitisation of Ca2+ signalling and the sustained phase of ERK activation. These differences may arise through differential intracellular processing of the ligands, with NmU-25 degradation occurring through the activity of ECE-1, most likely within an endosomal compartment. Such differences may have important implications for the design of peptide-mimetics, or synthetic small molecule agonists and for the therapeutic targeting of peptidergic receptors.

Brighton PJ et al., Mol Pharmacol 66:1544, 2004

Roosterman D et al., Proc Natl Acad Sci USA 104:11838, 2007