The data presented here indicate that contrary to previous reports [12
], the nominally selective agonist CHPG can activate mGluR1 with similar efficacy and potency as mGluR5. Consistent with the literature however, CHPG did not produce any detectable activation of mGluR2 or mGluR4. The effects of CHPG were examined using heterologous expression of each receptor in rat sympathetic neurons, an adult neuronal cell type with null-mGluR expression, and assayed using G protein mediated modulation of native ion channel currents as an assay for receptor signaling. Further, using M-current inhibition as an assay for receptor function, a pathway that depends only on Gq
signaling, revealed that in some cells, CHPG agonism of mGluR1 appeared to show some biased agonism. Specifically, while some mGluR1-expressing cells showed similar M-current inhibition using CHPG or Glu as an agonist, others were strongly inhibited by Glu, but only very weakly by CHPG. These data provide some contrast to those obtained using calcium current inhibition as an assay, which proceeds through a combination of Gq/11
and pertussis toxin sensitive Gβγ activation [11
]. Examination of calcium current inhibition by CHPG in mGluR1 expressing cells revealed that the Gq/11
-mediated, voltage independent inhibitory pathway was not as strongly activated by CHPG as Glu, supporting the hypothesis that CHPG is a poorer Gq/11
activator than Glu. However, the difference was not robust. Thus, while these data indicate that CHPG can effectively act as an agonist at mGluR1a and its splice variant mGluR1b, the overall balance of Gi/o
protein activation may be altered when CHPG rather than Glu is used as the agonist.
The apparent biased agonism of CHPG when applied to mGluR1-expressing cells was interesting, but the variability of the effect is difficult to explain. It is possible that the variability was related to mGluR1 expression level, but this conclusion is difficult to reconcile with the data. Due to the nature of these studies, receptor expression cannot be directly measured, but an estimate of expression can be made by assuming that expression level correlates with the magnitude of inhibition by Glu in each cell. For example, the M-current inhibition data in Figure
shows that in the 8 mGluR1a expressing cells examined, M-current inhibition ranged from about 35% to about 90%. While there may be some relationship between CHPG responses and Glu responses in these cells, the relationship is weak. For example, in the 4 cells that showed virtually no response to CHPG, the range of Glu responses was quite broad, including two cells with substantial inhibition by Glu (~ 60 and 80%). Further, examination of the voltage dependence of calcium current inhibition (Figure
) yields similar, ambiguous results (Figure
). If the Post/Pre calcium current inhibition ratio, a measure of the strength of Gq activation by CHPG, is plotted against the amplitude of total inhibition by Glu, an estimate of expression level, the effects are widely scattered and poorly correlated (Figure
). Thus, the data do not provide strong evidence that the degree of G protein activation bias by CHPG via mGluR1 is related to receptor expression levels. It should also be noted that while CHPG was less efficient than Glu at producing Gq
-mediated current inhibition in most of the mGluR1 expressing cells examined in the calcium current studies, CHPG still activated this pathway to a relatively strong degree (Figures
, ). Therefore, while some bias in G protein activation of CHPG via mGluR1 is detectable, it is unlikely that this effect will be meaningful in a physiological or in vivo context. However, the existence of a group I mGluR agonist with G protein bias is novel, and suggests that development of compounds with more significant biased properties is possible.
Figure 8 Relative strength of Gqactivation by CHPG is not well correlated with apparent mGluR1 expression level. Relative strength of the Gq inhibitory pathway, expressed as the Post/Pre inhibition ratio vs. apparent mGluR1 expression level, expressed as magnitude (more ...)
Given the effect of CHPG on mGluR1a, it was not surprising that the drug had similar effects on mGluR1b since both splice variants are identical in the N-terminal, ligand binding region. In fact, these proteins differ only in their extreme cytoplasmic C-termini, which is not expected to alter receptor pharmacology. Furthermore, there is no evidence that G protein activation differs in these splice variants, as both mGluR1a and 1b can produce qualitatively similar calcium current and M-current inhibition in SCG neurons [11
]. Effects of CHPG on both variants was tested primarily to confirm the rather surprising result of CHPG agonism on mGluR1, a result which has not been previously reported despite fairly widespread use of this drug for over a decade [12
]. Indeed, the mGluR1a data shown in Figure
B is combined data from SCG neurons expressing mGluR1a from two separate, but similar, plasmids. One is an untagged rat mGluR1a, and the other is an N-terminally myc-tagged mGluR1a, both in pCDNA3.1. The data were combined because CHPG acted identically on cells expressing both constructs (not shown). Both constructs (as well as mGluR1b, in pRK5) were tested for responses to CHPG, and sequence-verified. Finally, it should be noted that the results with the mGluR1 constructs were generated using CHPG from two separate sources (Tocris and Ascent, see Materials and Methods) with indistinguishable results (not shown).