Neocortical neurons were isolated from both male and female P1-2 mouse pups as reported previously (Phillips et al., 2008). All animal procedures were approved by VAPORHCS IACUC in accordance with the U.S. Public Health Service Policy on Humane Care and Use of Laboratory Animals and the NIH Guide for the Care and Use of Laboratory Animals. Animals were deeply anesthetized with isoflurane before decapitation. The brain was extracted and cortices were dissected in cold Mg²⁺- and Ca²⁺-free Hanks’ balanced salt solution (Corning) plus 20% fetal bovine serum (FBS). Cortices were then incubated in trypsin (5 mg/mL) and DNAse (0.1 mg/mL) and then dissociated with a heat polished pipette in a high magnesium solution. Dispersed cells were cultured in minimal essential medium plus 5% FBS (MEM+) on glass coverslips that had been pre-coated with dilute Matrigel. ARAC (4 μM) was added 48–72 h after plating to limit glial division, then diluted two-fold with MEM + (2 μM final concentration) 72 h later. Cells were used after ≥ 14 days in culture.

The vast majority of experiments utilized cultures derived using a C57BL/6J and 129/SvJ mouse strain. We also used PLC-β1 null-mutant (PLC-β1^–/–) mice, kindly provided by Dr. Hee-Sup Shin (KIST). The PLC-β1^–/– mice were generated by a gene targeting method as described previously (Kim et al., 1997). Briefly, the PLC-β1 was disrupted by homologous recombination and transfected into embryonic stem cells. In our hands, the homozygous mutation was lethal and we were unable to generate any PLC-β1^–/– pups. This differed from other studies who reported a reduced life expectancy compared to wild-type due to seizures and sudden death (Kim et al., 1997). The cause for the higher lethality we observed was unidentified. In an attempt to improve survival, we changed the underlying strain by back-crossing heterozygotes (strain C57BL/6J) with wild-type (WT) C57BL/6J and 129/SvJ mice for seven generations. This back-cross substantially increased the survival of the PLC-β1^–/– mice. Then, mice heterozygous for the mutation\were crossed to produce animals for PLC-β1^–/– and WT (PLC-β1^+/+) cultures. The mutation consists of a PGK-neomycin cassette inserted into the exon encoding amino acids 50–82 of the protein for gene disruption. To genotype, DNA from mouse tail samples was extracted by treatment with 25 mM NaOH and 0.2 mM EDTA at 98°C for 30–50 min. Forty mM Tris buffer (20 mM final), pH 5.5, was added and the mixture centrifuged. The DNA-containing supernatant was used for polymerase chain reaction (PCR). PCR was performed using three primers: F1: 5′-CAAGTTAAGTCCGGCAAACACC-3′, B1: 5′-ACCTTGGGAGCTT-TGGCGTG-3′, and PGK22: 5′-CTGACTAGGGGAGGAGTAGAAG-3′. The PCR products were run on 1% agarose gels. Previous microarray experiments of 2 week old cultures using the GeneChip Clariom S Mouse Array (Affymetrix/Applied Biosystems) (Lindner et al., 2022) were analyzed using Affymetrix Expression Console software ver.1.4.1.46. The microarray data are available at NCBI GEO (GSE218028).

Cells were visualized with an Olympus IX70 or Zeiss IM35 × 40 inverted microscope. Recordings were made in whole-cell voltage clamp mode using a HEKA 9 or 10/2 amplifier. Holding potential was −70 mV, corrected for experimentally-derived liquid-junction potentials and cells were perfused with extracellular solution containing (in mM) 150 NaCl, 4 KCl, 1.1 CaCl₂, 1.1 MgCl₂, 10 HEPES, 10 glucose, pH 7.35. In experiments where [Ca²⁺]_o was manipulated (Figures 1, 3), the concentration of [Mg²⁺]_o was decreased to 0.5 mM. Recordings of mEPSCs and hypertonic sucrose-induced currents were made in the presence of tetrodotoxin (TTX, 1 μM) and bicuculline (Bic, 10 μM) to block voltage-gated sodium channel currents and GABA-activated currents, respectively. Both potassium gluconate and cesium methane-sulfonate pipette solutions were used. Potassium gluconate solution consisted of (in mM) 113 K⁺ gluconate, 9 EGTA, 10 HEPES, 4 MgCl₂, 1 CaCl₂, 4 Na₂ATP, 0.3 diphosphate-GTP, 14 creatine phosphate, pH 7.2. Cesium methane-sulfonate solution consisted of (in mM) 108 Cs methanesulfonate, 9 EGTA, 10 HEPES, 4 MgCl₂, 1 CaCl₂, 4 Na₂ATP, 0.3 diphosphate-GTP, 14 creatine phosphate, pH 7.2. Patch electrodes had resistances of 2–7 MΩ. Miniature EPSCs were filtered at 1 kHz and digitized at 10 kHz. For recording of HS-induced currents (Figures 4–6), currents were filtered at 3 kHz and sampled at 20 kHz. Series resistance (Rs) was monitored, and recordings were discarded if Rs changed significantly during the course of a recording. Rs was compensated by up to 70%.

For recordings of mEPSCs, solutions were bath-applied constantly through a perfusion pipette placed ∼1 mm from the patch pipette tip. The perfusion pipette was connected to a manifold (Warner Instruments) and stopcocks used to switch solutions. Local solution equilibration occurred in substantially less than 10 s as measured by open-tip conductance changes. U73122 and U73343 stocks (5 mM) were prepared in DMSO and diluted 1:1000 in Tyrode’s for experiments which resulted in final drug concentrations of 5 μM and a final vehicle concentration of 0.1% in the recording solutions, which had no apparent effect in control experiments. Preparation of stocks required warming to 30°C in water bath then agitation using a sonicator (20 kHz for 10 × 10 s burst, FB150, Fisher Scientific, Hampton, NH, USA).

To measure the RRP, we evoked fusion of the vesicles with a hypertonic extracellular solution (Tyrode plus 500 mM sucrose) applied for 5 s using a custom-built Piezo-driven perfusion system described previously (Vyleta and Smith, 2008) and recorded the corresponding currents in the post-synaptic cell. Briefly, a piece of double-barreled theta glass was used to perfuse both control and hypertonic sucrose solutions simultaneously. A whole-cell recording was made from a neuron being bathed only by the control solution. Upon stimulation by a high-voltage stimulus isolator (World Precision Instruments, Sarasota, FL, USA, product # A365D), a Piezo bimorph (Piezo Systems, Inc., product # T234-A4CL-203X) moved the tip of the theta glass horizontally so that the recording was then bathed by extracellular solution containing hypertonic sucrose. Each barrel of the theta glass contained two small capillaries delivering solutions by gravity to the recording chamber.

Miniature EPSC frequency was measured in 10 s bins and normalized to the average mEPSC frequency during the preceding ≥ 100 s of recording in 1.1 mM [Ca²⁺]_o and 0.5 or 1.1 mM [Mg²⁺]_o (details in legends). The response to elevated [Ca²⁺]_o and other interventions reflects the average relative frequency measured over the last 80–100 s of the 6 mM calcium application, reflecting the steady state response. Analysis was performed using IgorPro (Wavemetrics, Lake Oswego, OR, USA) software. Single spontaneous release events were detected using a template matching method (Clements and Bekkers, 1997) using routines written by Dr. H. Taschenberger (López-Murcia et al., 2019). Data were acquired on a PIII computer and analyzed with IgorPro (Wavemetrics, Lake Oswego, OR, USA) software. Miniature EPSCs were identified using an 8 ms current template that comprised of a baseline (0.4 ms), a rising phase (tau 0.4 ms), and a decaying phase (tau 2.5 ms). The template was occasionally altered to better fit the shape of mEPSCs in an individual recording. IgorPro software analyzed the compliance of the raw current data to the template and gave a criterion score based on how well each segment of the data matched the template. A criterion threshold was set such that recording noise scored below the threshold, and obvious mEPSCs scored above threshold. All criterion scores at or above that value were identified as mEPSCs.

Hypertonic sucrose-induced charge transfer, representing the RRP, was calculated as the integral of the total currents induced over 5 s of sucrose application. The duration of sucrose application was clearly identified by stimulus artifacts in the recordings (Figures 5, 6). In some experiments, currents did not show clear transient and steady-state phases. Thus total current elicited by HS was integrated in all experiments, without subtracting off a steady-state component, and used to estimate the RRP (Stevens and Sullivan, 1998). Curve fitting was performed using IgorPro (Wavemetrics, Lake Oswego, OR, USA).

Statistical significance was determined using parametric or non-parametric testing using ANOVA or Kruskal–Wallis tests with Dunn’s multiple comparisons tests or Student’s t-test or Mann-Whitney test as appropriate (Microsoft EXCEL, Richmond, WA; Graphpad Prism). Non-Gaussian distribution data are reported displaying the median. Normally distributed data were reported as mean ± SEM (text and legends). P-values < 0.05 were considered significant and cut-offs were denoted in Figures with the following symbols (p-values < 0.05, 0.01, 0.001, and 0.0001 were indicated with *, ^**, ^***, and ^****, respectively and n.s. used to identify non-significant p-values).

A number of GPCRs have been implicated in the regulation of spontaneous release by [Ca²⁺]_o (Kubo et al., 1998; Tabata et al., 2002; Yamasaki et al., 2006; Vyleta and Smith, 2011). We tested if PLC, downstream of GPCRs, was contributing to the enhancement of spontaneous glutamate release by elevated [Ca²⁺]_o using the PLC inhibitor, U73122 (Smith et al., 1990; Horowitz et al., 2005). Whole-cell patch clamp recordings were made from neocortical neurons voltage-clamped at −70 mV in the presence of tetrodotoxin (TTX, 1 μM) and bicuculline methiodide (10 μM). We measured mEPSC frequency at physiological [Ca²⁺]_o (Tyrode with 1.1 mM [Ca²⁺]_o, T₁.₁) and then following a step to 6.0 mM [Ca²⁺]_o (Tyrode with 6 mM [Ca²⁺]_o, T₆) in untreated controls, or following a 20- to 50-min incubation with U73122 (5 μM), the vehicle DMSO, or the inactive analog, U73343 (5 μM; Figure 1). Exemplar traces show that the mEPSC frequency increased robustly in response to T₆ but that this increase was attenuated by U73122 (Figure 1A). Similarly, the time course of the response to T₆, which occurred over multiple 10 s of seconds consistent with earlier reports (Vyleta and Smith, 2011), was absent following U73122 treatment (Figure 1B). Key controls included testing if the response to T₆ was affected by the vehicle, DMSO, or the inactive analog, U73343. In a comparison of the normalized response to T₆, the medians were different (Figures 1C; P = 0.0002 by Kruskal–Wallis test). The control group increased to 3.02 of baseline which was substantially higher than the 0.88 times baseline observed after U73122 (5 μM) treatment (P = 0.0001 by Dunn’s multiple comparisons). The control responses to T₆ were unaffected by treatment with DMSO (median 2.3 times baseline, n = 14, P > 0.9999) or U73343 (median 1.6 times baseline, n = 13, P = 0.2058, Dunn’s; Figure 1C). Taken together these data indicate that [Ca²⁺]_o stimulation of spontaneous release of glutamate is dependent on the activation of PLC.

Ca²⁺-sensing receptor detects changes in [Ca²⁺]_o and is estimated to regulate ∼30% of basal spontaneous glutamate release in neocortical neurons (Vyleta and Smith, 2011). We hypothesized that PLC may also determine the mEPSC frequency in the presence of T₁.₁. However, basal mEPSC frequency was not impacted by U73122, U73343, or DMSO (p = 0.33, Kruskal–Wallis; Figure 1D). The large population variance in basal mEPSC frequency increases the likelihood that differences may not be detected (Figure 1D). Consequently, we examined the time-course of U73122 and U73343 effects on basal mEPSC frequency within single cells. Using this approach, we observed that the response to U73122 was biphasic and consisted of an initial increase then subsequent decrease in the average normalized mEPSC frequency (Figures 2A, B). At steady state, U73122 decreased mEPSC frequency by 40 ± 28% (n = 8, red line Figure 2B). The response to U73443 was monophasic with spontaneous glutamate release increasing by 246 ± 44% (n = 5) following U73443 application. The median steady state mEPSC frequency changed to 74% and 355% of baseline for U73122 and U73343, respectively, (P = 0.0016, Mann–Whitney, Figures 2B, C). This difference probably indicates that a fraction of basal spontaneous glutamate release is PLC-dependent. However, the similarity of the initial increase in mEPSC frequency caused by both U73122 and U73343 indicates these agents presumably have off-target effects.

Phospholipase C modulates signaling by increasing IP₃ and DAG (Berridge and Irvine, 1984). Stimulation of spontaneous release by increasing [Ca²⁺]_o was reduced by buffering intracellular [Ca²⁺] ([Ca²⁺]_i) at inhibitory but not excitatory synapses (Vyleta and Smith, 2011; Williams et al., 2012). This finding indicates that IP₃-induced rises in [Ca²⁺]_i were an unlikely cause for the change in mEPSC frequency to T₆ (Figure 1). Therefore, we next tested if PLC activation by presynaptic GPCRs might enhance spontaneous release rate by increasing levels of DAG. Phorbol esters are synthetic DAG analogs that enhance spontaneous release in central neurons (Lou et al., 2008). Consistent with this, phorbol 12,13-dibutyrate (PDBu, 1 μM) produced a robust enhancement of spontaneous release in the presence of T₁.₁ following a 200 s application (4.16 ± 0.45 times baseline, n = 12, P < 0.001; Figures 3A, B). This result is consistent with DAG increasing spontaneous release of glutamate. Additionally, there was a modest increase of mEPSC amplitude after PDBu treatment (mean 2.11 ± 0.54 times baseline, but similar median values, P = 0.039 by Wilcoxon test; Figures 3A, B), which is not consistent with previous data (Lou et al., 2008). If elevation of [Ca²⁺]_o enhances the rate of spontaneous neurotransmission by producing DAG, then exogenously applied DAG analogs may impair the effect of elevation of [Ca²⁺]_o on spontaneous release. To examine the interaction between a high intracellular concentration of phorbol ester and cell’s ability to respond to the addition of elevated [Ca²⁺]_o, we applied PDBu and subsequently added T₆. T₆ increased normalized mEPSC frequency to 3.52 ± 1.03 of baseline and PDBu increased it to 5.49 ± 1.12 and 12.64 ± 3.68 of baseline in T1.1 and T6, respectively (Figure 3C, n = 6). The median ratio of mEPSC frequency in T₆ to T₁.₁ decreased significantly from 2.8 to 1.6 following PDBu treatment (P = 0.031) indicating that the phorbol ester reduced the sensitivity of spontaneous release to increases in [Ca²⁺]_o.

Some studies indicated that PLC-β1 is the most abundant PLC isoform in the neocortex (Ross et al., 1989; Lim et al., 2022). Therefore, we tested if PLC-β1 mediated [Ca²⁺]_o-dependent spontaneous release of glutamate by comparing the sensitivity of mEPSC frequency between neurons deficient of PLC-β1 (PLC-β1^–/–) with littermate control neurons (PLC-β1^+/+) and the impact of U73122 in PLC-β1^–/– cultures (Figure 4). These genotypes were confirmed using PCR (Figure 4B). There were significant differences between the response to T₆ across the three groups (Kruskal-Wallis, P < 0.0001). In PLC-β1^–/– neurons and littermate control neurons (PLC-β1^+/+), mEPSC frequency increased similarly following application of T₆ (median = 2.0 and 2.2 times baseline, respectively, n = 14 for both groups, p = 0.94, Dunn’s; Figures 4A, C). However, the response to T₆ was completely suppressed in U73122 pre-treated PLC-β1^–/– neurons (median = 0.8 times baseline, p < 0.0001, Dunn’s; Figures 3A, C). These data indicate that PLC-β1 may not be contributing to the [Ca²⁺]_o-dependent spontaneous release of glutamate in wild-type neurons or there may be upregulation of compensatory pathways (Namkung et al., 1998). Alternatively, more recent data has shown that other PLC isoforms may have a substantial presence in the neocortex (Lindner et al., 2022) possibly facilitating these [Ca²⁺]_o-dependent changes in spontaneous release in PLC-β1^–/– neurons.

Earlier experiments indicated that basal spontaneous release of glutamate was inhibited by U73122, indicating that PLC signaling may contribute to vesicle turnover and availability. The size of the RRP is one determinant of the probability of neurotransmission. To test if PLC regulates RRP size we elicited exocytosis by applying HS solution and recorded excitatory postsynaptic currents (EPSCs) in voltage-clamp (Rosenmund and Stevens, 1996). Hypertonic sucrose (Tyrode plus 500 mM sucrose) evoked large inward currents that were inhibited by a 20-min application of U73122 (Figure 5A). On average, the total HS-induced charge transfer was 772 ± 299 pC before and 225 ± 93 pC after U73122 (n = 7, P = 0.048, paired t-test; Figure 5B). This inhibition was not reversible during our experiments, as reported by others (Jin et al., 1994). We next confirmed this inhibition of synaptic vesicle fusion was specific to PLC blockade by measuring the effect of U73343 on sucrose-induced currents. A 20-min application of U73343 produced an apparent increase in sucrose-induced currents that was reversible (Figures 5C, D), however, this trend failed to reach statistical significance. On average, total sucrose induced charge transfer was 645 ± 165 pC before and 769 ± 198 pC after U73343 (n = 7, P = 0.061, paired t-test; Figure 5D). Taken together, these results indicate inhibition of PLC decreases the size of the synaptic vesicle pool available to fuse in response to hypertonic solution, and further suggests that at baseline, PLC regulates the size of the RRP in excitatory neocortical neurons.

Phorbol esters increase both the size and recovery rate of the RRP of synaptic vesicles evoked with hypertonic sucrose solution (Stevens and Sullivan, 1998). We hypothesized that if U73122 decreases RRP size by decreasing DAG levels in the presynaptic membrane, then the inhibition of sucrose-induced currents by PLC blockade should be rescued by exogenous phorbol ester. We evoked hypertonic sucrose-induced currents every 100 s while simultaneously applying U73122 and PDBu. Representative traces from one recording of sucrose-induced EPSCs at 300 and 600 s after application of U73122 illustrate U73122 still inhibits neurotransmitter release in the presence of PDBu (Figure 6A). The average normalized diary plots of total sucrose-induced charge transfer versus time suggest that phorbol ester application may slow the depletion of the RRP seen under PLC blockade (Figure 6B). A comparison of total sucrose-induced charge transfer between 500 and 600 s after onset of U73122 application indicates that inhibition of RRP is less complete in the presence of PDBu. Average total charge transfer was 0.14 ± 0.03 and 0.28 ± 0.02 of control for U73122 alone and U73122 + PDBu (n = 14 and 6 data points, respectively, P = 0.02), however, the two curves converge by 700 s of U73122 application (Figure 6B). Average time constants for inhibition of sucrose-induced currents 220 ± 44 and 349 ± 86 s for U73122 alone and U73122 + PDBu respectively, suggesting that PDBu slows RRP inhibition, however this difference was not statistically significant (n = 8 and 3, respectively, p = 0.18 test). Taken together, these results suggest that inhibition of neurotransmitter release by PLC blockade may be in part due to a consequent decrease in DAG levels in the presynaptic membrane.

U73343 does not inhibit PLC but its stimulatory action on mEPSC frequency shows that it is not inactive (Figures 2B, C). U73122 also increased spontaneous neurotransmission transiently, supporting the idea that this stimulation is not PLC-mediated but reflects an off-target effect. We attribute differences between the two agents to inhibition of PLC but, as mentioned, use an appropriate incubation period such that the off-target effects of U73122 should no longer be in effect and PLC inhibition by U73122 should be complete.

Our data indicate that PLC activity is necessary to enhance the spontaneous release of glutamate in response to elevated [Ca²⁺]_o in neocortical neurons. We propose that this occurs, at least in part, due to an increase in the size of the RRP which increases the overall likelihood of spontaneous release.