Vm regulates mGlu5-induced Ca2+release
To investigate the effect of Vm on mGlu5 function and determine the optimal experimental conditions, we examined mGlu5-induced Ca2+ release from intracellular stores. mGlu5 stimulation is known to induce a transient increase of Ca2+ oscillations in various cell types, including cultured astrocytes , hippocampal neurons , midbrain neurons , as well as in heterologous expression systems such as Chinese Hamster Ovary cells and Human Embryonic Kidney 293T (HEK293T) cells . In line with these findings, in HEK293T cells expressing the GCaMP6s fluorescent Ca2+ sensor (Figure 1A ), mGlu5 agonist (RS)-3,5-dihydroxyphenylglycine (DHPG, 100 µM) induced a Ca2+ increase exclusively in cells expressing mGlu5, as evidenced by GCaMP6s fluorescence fluctuations measured in the cell population using a fluorimeter (Figure 1B ). The mGlu5 agonist-induced Ca2+ increase in mGlu5-expressing cells was inhibited by application of MPEP (10 µM), a mGlu5-specific negative allosteric modulator (NAM), confirming the requirement for mGlu5 receptor activation (Figure 1B ).
In order to determine the effect of Vm on mGlu5-induced Ca2+ increase, we treated cells with equimolar solutions containing increasing concentrations of KCl (from 3 mM, a physiological concentration referred to as “Vrest”, up to 100 mM) to induce a systematic, well-controlled, and long-lasting membrane depolarization (as previously demonstrated in neurons ). Interestingly, KCl 100 mM had no effect on basal intracellular Ca2+ concentrations (Figure Supp. 1A ), but significantly reduced mGlu5 agonist-induced Ca2+increase by 22.3 ± 7.1% compared to KCl 3 mM (Figure 1C ). We measured that KCl 100 mM application increased the membrane potential by 23.6 ± 0.9 mV (Figure 1D ), referred to as “Vdepol”. Whole cell patch clamp recordings indeed reported a reversible switch from Vrest = -28.4 ± 3.5 mV at physiological KCl concentration (3 mM) to Vdepol = -4.8 ± 0.9 mV with KCl 100 mM (Figure 1D ). The dose response curve of DHPG-induced Ca2+ release further revealed a reduction of the agonist efficacy at Vdepol compared to Vrest (Figure 1E ). Additionally, saturating concentrations of another mGlu5 agonist, quisqualate (10µM), triggered a Ca2+ increase at Vrest, which was reduced by 45 ± 9.5% at Vdepol (Figure 1F ). These initial findings indicate that the ability of the mGlu5 receptor to increase intracellular Ca2+ is dependent on Vm. To investigate whether the effect of Vm was specific to the mGlu5 receptor, we examined another Gq-coupled receptor, the angiotensin II type 1 receptor (AT1), and did not observe any influence of Vmon AT1 receptor-mediated Ca2+ release, thereby supporting a specific effect of Vdepol on mGlu5 signaling in this context (Figure 1G ).
To investigate the effect of Vm on mGlu5-induced Ca2+ increase at a cellular level, we performed live cell imaging of GCaMP6s fluorescence (Figure 2A andsupplemental Video1 ). Initially, large recordings in the entire field of view showed a global increase in Ca2+ induced by the mGlu5 agonist, quisqualate (10µM), at Vrest. This increase was significantly reduced at Vdepol (30 ± 10 % reduction compared to Vrest at the peak of fluorescence,Figure 2B and supplemental Video2 ). The global Ca2+ increase was an average of the single cell Ca2+ fluctuations, which were classified into 3 distinct populations of cells depending on the type of responses: silent cells, cells displaying 1 single Ca2+ spike, and Ca2+ oscillating cells (Figure 2C ), as reported previously . Vdepol modified the distribution of cells in each of these categories, with a global shift towards less oscillating populations (Figure 2C ), and a significant reduction in the proportion of cells with an oscillating Ca2+ pattern (10.6% reduction of the proportion of oscillating cells compared to Vrest, Figure 2D ). Subsequently, for each oscillating cell, we studied the global frequency of Ca2+ oscillations induced by mGlu5 agonist. Quisqualate induced a mean of 0.49 ± 0.02 oscillations/min per cell at Vrest, but only 0.39 ± 0.02 oscillations per cell at Vdepol. Thus, Vdepol induced a 20.3 ± 4.8% reduction in the frequency of Ca2+oscillations triggered by mGlu5 stimulation (Figure 2E ).
Given that depolarization in neurons can also occur following receptor stimulation, we investigated the impact of Vdepolsubsequent to mGlu5 activation. Specifically, we measured the effect of Vdepol compared to Vrest on oscillating cells that had undergone Ca2+ oscillations at resting membrane potential as a result of mGlu5 stimulation. To do so, we determined the instantaneous frequency, which represents the average inter-oscillation frequency between the first and last oscillations. Our findings consistently revealed a significant reduction in the instantaneous frequency, from 1.18 ± 0.11 at Vrest to 0.85 ± 0.04 oscillation/min at Vdepol, representing a 28 ± 4.7% decrease (Figure 2F ). These results suggest that a membrane depolarization of approximately 20-25mV, whether it occurs before or after mGlu5 activation, can diminish the receptor’s ability to induce Ca2+ release from intracellular stores, in few seconds. Specifically, Vdepol reduced the number of agonist-responsive cells and the efficiency of the responding cells, resembling the characteristics of a NAM .