Discussion
To our knowledge this is the first reported use of a label free technology to characterise the pharmacology of recombinantly expressed MRGPRD receptors. DMR is now a well established technique for studying GPCR function with examples including Muscarinic M1 (Peters et al., 2010), Muscarinic M2 (Peters et al., 2010, Lee et al., 2008), Muscarinic M3 (Dodgson et al., 2009), Dopamine D3 (Lee et al., 2008), corticotropin-releasing factor (CRF1) (Peterset al., 2010) and formyl peptide 1 (Christensen et al.,2017) receptors. GPCRs can couple through and activate multiple signaling pathways, which can be ligand or cell specific and may result in differences in pharmacological profile when using different assay formats, ligands and cell types. Biophysical technologies such as DMR provide an opportunity to measure a phenotypic, cellular, response in real time in the absence of labels, which can be prone to compound interference. Scott and Peters (2010), compared both cellular impedance (cellkey) and optically based (BIND and DMR) label free systems and noted that all three systems produced consistant responses that were dependent on the type of G-protein coupling of the receptor investigated. In general, GPCRs coupled via Gi pathways produced an increase in response, stimulation of Gqcaused a transient decrease followed by a larger increase, while Gs activation produced a decrease in the response. They also found that with optical biosensors both Gi and Gq responses were rapid in onset and peaked within around 5 minutes of agonist addition before declining, whilst Gs-coupled responses were slower and more sustained. Here we observed that the putative MRGPRD agonist, β-alanine produced a dose-dependent transient increase in the DMR response with a peak at approximately 4 minutes before declining, which is consistent with the observations of both Scott and Peters (2010) and Lee et al.(2008) for Gi-coupled receptor responses. The lack of response to β-alanine in untransfected control CHO cell lines confirms the specificity of the activation of MRGPRD receptors. In addition, we wanted to use an orthogonal method for confirming the activity observed in the DMR assay and so monitored intracelullar Ca2+flux using a FLIPR, a technique that has previously been used to study many GPCRs including MRGPRD receptors (Ajit et al., 2010; Zhanget al., 2007). We observed a rightward shift in pEC50 of approximately one order of magnitude when comparing DMR with FLIPR MRGPRD responses for all three agonists tested. One explanation for this discprenency may be differing G-protein coupling between the two assays, in addition to differences in the kinetics of the two readouts. MRGPRD receptors have been reported to be able to couple via both Gi and Gq G protiens (Shinohara et al., 2004; Crozier et al., 2007; Ajit et al., 2010; Uno et al., 2012), and we found that the Gq inhibitor, YM-254890 abolished the response to β-alanine in Ca2+ assays whereas incubating with the Gi inhibitor PTX only resulted in an approximate 50% loss in response, corresponding with the findings of Ajit el al , (2010). However, in the DMR assay, YM-254890 caused only minimal inhibition of the agonist response, with the most marked reduction in response due to PTX. These data indicate that in the DMR assay and, by extension, the gross cellular re-arrangements of mass that take place in response to MRGPRD receptor activation, were principally coupled via activation of Gi while in contrast the increase in intracellular Ca2+ was primarily via Gq. This difference in signalling is a likely explanation for the shift in agonist potencies observed between the two techniques and we speculate that MRGPRD when interacting with either Gi or Gq G proteins may adopt different conformers that alter the affinity of β-alanine. Dodgson et al.,(2009) compared DMR and FLIPR assays when investigating the muscarinic M3 receptor and observed similar shifts in agonist potency between the techniques leading them to suggest that the signal coupling, when measured proximal to the receptor, tends to be amplified leading to an increase in apparent agonist potency compared to DMR which measures a global change in cellular phenotype. They also comment on the time period being measured; in the FLIPR, responses are recorded within the first minute following activation, while in the DMR assay the peak measurement is taken after 4 minutes. Lee et al., (2014) also found discrepancies in the potency of antagonists of the urotensin receptor, which were less potent in the DMR assay compared to Ca2+ mobilisation.
The MRGPRD agonists GABA and glycine both produced a response with a similar DMR profile to β-alanine, but with reduced potency, that was consistent across both assay formats and consistent with the rank order of agonist potency described in the literature (Ajit et al.,2010). The differential coupling pathways for GABA and glycine in the two assays were also broadly consistent with those observed with β-alanine, although the weak DMR response to glycine was not fully inhibited by PTX while the β-alanine and GABA responses were not entirely abolished by co-administration of the Gi and Gq inhibitors, which may suggest the presence of additional signalling pathways.
Thioridazine hydrochloride, described by Ajit et al., (2010) as an MRGPRD antagonist, caused a response in DMR which was unaffected by the addition of PTX and which was also observed in the control cell line, indicating that its activity is not related to MRGPRD in our assay system; additionally, it only inhibited the β-alanine induced Ca2+-response in the FLIPR assay at the highest concentrations tested. We also show that thioridazine hydrochloride is cytotoxic at these higher concentrations (Fig 5E), bringing into question the specificity of this compound as an MRGPRD antagonist. Another putative MRGPRD inhibitor, PD123,319, first described as an AT2R angiotensin receptor antagonist and shown to inhibit the agonist effects of alamandine in CHO cells expressing MRGPRD (Lautner et al., 2013), failed to demonstrate any significant inhibition of the agonists used in either of the assay formats studied here. We also tested the effects of alamandine in both FLIPR and DMR assays but observed only minimal responses (supplementary data) and so could not characterise its sensitivity to antagonism. It is, therefore, possible that the agonist activity of alamandine (Lautner et al.,2013) is tissue or cell specific, similarly the effects of PD123,319, whereas we find no evidence, in a similar recombinant system, that PD123,319 affects β-alanine, GABA or glycine MRGPRD responses. Finally, we observed that MU-6840 (Uno et al., 2012) dose-dependently inhibited the response to all three agonists. However, analysis of the FLIPR Ca2+-data indicated that inhibition by MU-6840 although competitive may be cooperative as suggested by a Schild slope of 2, which is different to the results obtained in the DMR assay where the slope was not significantly different from 1 (pA2=6.25). This divergent pharmacology observed between two different assay techniques may, therefore, provide evidence for two distinct populations of receptor due to differential G protein coupling.
In conclusion, this study presents for the first time, a novel label free assay for studying MRGPRD receptors. Comparing the pharmacology of agonists and antagonists between DMR and a well characterised Ca2+ mobilisation assay has provided interesting insights into how MRGPRD receptors can differentially couple through separate G proteins in the same cell line. In addition, we find good evidence for MU-6840, but neither thioridazine hydrochloride nor PD123,319, to be a MRGPRD antagonist with an indication of cooperativity in its mechanism when coupled via Gq. This data argues for a holistic, unbiased approach to studying MRGPRD receptor pharmacology by employing multiple techniques that facilitate dissection of divergent on-target pharmacology and potentially off target effects, which will be particularly important in defining the activity of molecules identified in drug discovery programmes.