All animal procedures were reviewed and approved by the University of Florida Institutional Animal Care and Use Committee (IACUC). Juvenile male Sprague–Dawley rats (P16-28) were used for this study. Acute brain slices were prepared as previously described (Harden and Frazier, 2016). Briefly, rats were administered a ketamine/xylazine cocktail (100 mg/kg ketamine, 10 mg/kg xylazine, i.p.) and after adequate anesthesia was achieved (no response to hind-paw pinch) rats were rapidly decapitated and brains were rapidly extracted and submerged in ice-cold sucrose-laden artificial cerebrospinal solution (ACSF) containing (in mM) 205 sucrose, 10 dextrose, 1 MgSO4, 2 KCl, 1.25 NaH2PO4, 1 CaCl2, and 25 NaHCO3, oxygenated with carbogen (95% O2/5% CO2). Brains were coronally sectioned at 300 microns using a Leica VT1200S vibratome, then transferred to an incubation solution containing (in mM) 124 NaCl, 10 dextrose, 3 MgSO₄, 2.5 KCl, 1.23 NaH₂PO₄, 1 CaCl₂, 25 NaHCO₃, oxygenated with carbogen and maintained at 35°C for 30 min. Brain slices were then permitted to passively equilibrate to room temperature for at least 30 min prior to recording. Brain slices were transferred to a perfusion chamber and submersed in ACSF containing (in mM) 126 NaCl, 11 dextrose, 1.5 MgSO₄, 3 KCl, 1.2 NaH₂PO₄, 2.4 CaCl₂, 25 NaHCO₃, continuously oxygenated with carbogen, perfused at 2 mL/min, and maintained at 28°C. ACSF for all experiments was supplemented with glutamate- and GABA receptor antagonists (20 μM DNQX, 40 μM DL-AP5, 100 μM PTX, and 10 μM CGP-55845) to minimize influence of synaptic signaling (on AMPA, NMDA, GABA_A, and GABA_B receptors, respectively) and enhance measurement of isolated somatic currents.

Brain slices were visualized using an Olympus BX-WI upright stereomicroscope equipped for infrared differential interference contrast microscopy, a 12-bit CCD camera (QImaging Rolera-XR), and micro-manager software (Edelstein et al., 2014). Patch-clamp experiments were performed using a Multiclamp 700B patch-clamp amplifier with a CV-7B headstage, a DigiData 1440A digitizer (Molecular Devices), and pClamp 11 software. Data were obtained at 20 kHz and voltage-clamp data were filtered using a 7-pole Bessel low-pass filter with a 2 kHz − 3 dB cutoff. On-line analysis was performed with custom software written using Python 3.11 and the pyABF package. Off-line analysis was performed with OriginPro (OriginLab, Northampton, MA) using custom software written in OriginC by CJF.

Patch pipettes were pulled from borosilicate glass (Sutter Instrument BF150-86-10) using a Flaming/Brown pipette puller (Sutter Instrument SU-P97) and had an open-tip resistance of 4–6 MΩ when filled with a K-gluconate based internal solution containing (in mM) 115 K-gluconate, 10 di-tris phosphocreatine, 10 HEPES, 0.5 EGTA, 2 MgCl₂, 4 Na₂ATP, 0.4 Na₃GTP, 5 KCl. This solution was adjusted to pH 7.25 and 295 mOsm. In some experiments K-gluconate was replaced with an equimolar concentration of Cs-gluconate. For experiments requiring intracellular delivery of BAPTA, a modified internal solution was created substituting 10 mM of the primary cation with BAPTA. The liquid junction potential (LJP) was calculated according to the stationary Nernst–Planck equation (Marino et al., 2014) using LJPcalc (RRID: SCR_025044) and estimated to be 14.8 mV and 15.3 mV for the potassium and cesium based internal solutions, respectively. Results presented are based on calculations that used raw data uncorrected for the LJP.

Following establishment of a whole-cell recording, passive membrane properties were evaluated in all neurons using a membrane test protocol that delivered a 200 ms 10 mV hyperpolarizing pulse from a holding potential of −70 mV. Principal neurons were identified by their large soma and regular firing pattern with a maximum firing frequency of <40 Hz, consistent with previous studies (Hernandez et al., 2019; Ryan et al., 2018). Time course experiments in which access resistance drifted >30% were also excluded from this study. Whole-cell capacitance was measured using a bidirectional ramp protocol described by Golowasch et al. (2009). Experiments in which membrane test properties were compared before vs. after CCh report mean values from 2.5–4.5 and 8–10 min after CCh application relative to the mean value over a 5-min baseline period before CCh application. Current-clamp experiments evaluating the AHP and/or ADP were performed using a protocol that applied a constant current to hold cells near −70 mV, and a 500 ms depolarizing pulse to cause robust action potential firing (at > = 66% of the maximum observed rate). AHP and/or ADP measurements were obtained over a 2 s window beginning 100 ms after termination of the that depolarizing current pulse and are reported as area under the curve (AUC) relative to the mean voltage over a 1 s period immediately before the depolarizing step. Voltage-clamp experiments evaluating the currents underlying the ADP and/or AHP used a protocol in which neurons were continuously held at −60 mV and stepped to +50 mV for 500 ms. Variations of this protocol recorded multiple sweeps with stimulation duration (0–500 ms in 50 ms steps) or post-stimulation voltage commands of −90 to +20 mV (in 10 mV steps).

Population values are presented as mean ± the SEM. Electrophysiological changes before vs. after application of a drug were evaluated for significance in using paired t-tests on raw data, or one-sample t-tests on baseline subtracted data (null hypothesis mean = 0). Curve fitting was performed with OriginPro (Originlab, Northampton, MA).

Somatic effects of CCh were evaluated in BLA PNs targeted for whole-cell recording in acute brain slices according to their unique morphological and electrophysiological features (see Methods). Passive membrane properties evaluated in control conditions (membrane resistance: 78.71 ± 4.64 MΩ, whole-cell capacitance: 192.31 ± 7.97 pF, n = 34) were consistent with previous reports (Hernandez et al., 2019). Holding current at −70 mV was continuously measured as CCh was delivered using a syringe pump to achieve a bath concentration of 10 μM. We found that CCh reliably produced an excitatory shift in holding current (−45.65 ± 3.90 pA, n = 34, p = 2.59e-13) that plateaued approximately 8 min after perfusion began (Figure 1A).

To gain insight into which mAChR subtype mediates the response to CCh, experiments were repeated in the presence of M1/M4-selective mAChR antagonist pirenzepine (PZP, 1 μM). We found that PZP prevented CCh from causing an excitatory shift in holding current (−5.72 ± 4.03 pA, n = 10, p = 0.19, Figure 1B) suggesting that the CCh-induced excitatory current was mediated primarily by M1/M4 mAChRs. When the intracellular solution was supplemented with 10 mM BAPTA (a strong calcium chelator), CCh still produced a negative shift in holding current (−21.29 ± 7.65 pA, n = 10, p = 0.021, Figure 1C) but it was significantly smaller than the effect observed in control conditions (p = 5.35e-03), suggesting that excitation by CCh is at least partially dependent on an effector that is modulated by intracellular calcium. Cesium-based pipette solutions are commonly used to block some types of potassium channels, but when experiments were repeated using a cesium-based intracellular solution we found CCh still produced a large excitatory shift in holding current (−124.5 ± 20.1 pA, n = 6, p = 3.41e-3, Figure 1D). In fact, this effect was even larger than in observed with a potassium based internal (p = 0.02), likely due to the significantly higher basal membrane resistance observed with a Cs-based internal solution (107.8 ± 13.0 MΩ, p = 0.02 vs. control conditions).

Interestingly, despite observing an expected strong net excitatory effect of CCh on BLA PNs, we found that CCh reliably produced a biphasic change in membrane resistance (Figure 2A). Specifically, an early decrease in membrane resistance was observed (−9.4 ± 1.1%, p = 4.7e-6, n = 34, measured 2.5–4.5 min after CCh application), followed by an increase in Rm which was apparent by 8–10 min after the onset of CCh application (10.3 ± 2.4%, p = 1.8e-4, n = 34). This finding suggests that CCh may activate at least two distinct types of effectors: one that opens ion channels to lower membrane resistance, and another that closes ion channels to raise it. Consistent with our findings that PZP prevents CCh from changing holding current, we also found that PZP effectively prevents both the observed CCh-mediated decrease and later CCh-mediated increase in Rm (early: −0.66 ± 1.0%, p = 0.56, n = 10, late: −1.4 ± 2.0%, p = 0.48, n = 10, Figure 2B).

Next, we noted that when 10 mM BAPTA was included in the internal solution, CCh increased membrane resistance at the early time point (by 7.02 ± 3.20%, p = 0.057, n = 10), and continued to do so at the later time point (16.5 ± 4.3%, p = 0.004, n = 10, Figure 2C). In contrast, when a cesium-based internal solution was used we found that membrane resistance changed similarly to control conditions at both early and late time points (early: −5.59 ± 1.68%, n = 6, p = 0.039, late: +13.9 ± 3.7%, n = 6, p = 0.013, Figure 2D), with no statistically significant difference between cesium and control groups at either time point (p = 0.66 early, 0.54 late, vs. control conditions). Figures 2E,F provide summary data across all conditions at both the early and late time points, respectively. Collectively, these data in combination with those presented in Figure 1, suggest that CCh drives depolarization of BLA PNs from near the resting potential in part by activating a cesium-insensitive and calcium-dependent current that drives the membrane potential positive of rest, and also by inhibiting a calcium-independent and cesium-insensitive current that drives the membrane potential negative of rest.

The CCh has been reported to act on calcium- and voltage-sensitive ion channels to influence excitability of BLA PNs in response to strong firing (Unal et al., 2015). To evaluate how mAChR activation influences the AHP and ADP in BLA PNs, a 500 ms current pulse was delivered from a membrane potential near −70 mV to produce action potential firing at > = 66% of the maximum observed rate. The AHP or ADP was then measured during a 2 s window shortly after the depolarizing current pulse (see Methods). In control conditions, strong firing was followed by an AHP in most cells tested (11 of 14, Figure 3A), with a population mean of −1.42 ± 0.49 mV * sec (p = 0.012, n = 14). Following 10 min of exposure to 10 μM CCh, strong firing resulted in a large ADP in all cells tested (13.9 ± 0.24 mV * sec, p = 2.99e-4, n = 10, Figure 3B), which was significantly more positive than the pre-stimulus baseline (p = 1.22e-04). In a subpopulation of cells (2 of 10) the ADP was large enough to reach suprathreshold potentials and cause continuous firing (Figure 3C), a phenomenon that was never observed in the absence of CCh.

Current-clamp experiments evaluating the CCh-induced enhancement of ADP were repeated using modified extracellular or intracellular solutions to gain insight into the signaling pathways that underlie this behavior. When experiments were repeated using a pipette solution containing 10 mM BAPTA, CCh still produced a small ADP (2.55 ± 0.73 mV, p = 0.018, n = 6, Figure 3D) but it was significantly smaller than the ADP observed without BAPTA (p = 2.87e-04) suggesting that much of the ADP is the result of a calcium-dependent intracellular signaling pathway. Similarly, experiments repeated in the presence of I_CAN antagonist flufenamic acid (FFA, 100 μM) demonstrated that CCh produces an ADP (4.20 ± 1.12 mV, p = 3.75e-3, n = 6, Figure 3E) significantly smaller than the ADP observed without FFA (p = 4.20e-04). Together these findings suggest that CCh acts on BLA PNs activate an FFA sensitive and calcium-dependent non-selective cation channel, which is primarily responsible for driving the ADP.

The excitatory current underlying the ADP observed in the presence of CCh (I_ADP) was further characterized in voltage-clamp configuration using a 500 ms step to +50 mV to simulate strong firing from a −60 mV subthreshold potential. I_ADP was measured as area under the curve (AUC) of the voltage-clamp trace for a 3 s period after the depolarizing voltage step relative to the current before the step (see Methods). In the absence of CCh, this protocol allowed the current underlying the AHP (Figure 3A, dashed line) to be measured as an outward current following stimulation (97.4 ± 27.8 pA * sec, p = 0.07, n = 3, Figures 4A,B). Following 10 min of exposure to 10 μM CCh, the depolarizing voltage step resulted in a large inward current, the I_ADP (−47.6 ± 11.4 pA * sec, p = 0.05, n = 3, Figures 4A,B). In order to better isolate and evaluate the I_ADP, these experiments were repeated with a cesium-based pipette solution, which blocked the potassium channels underlying the AHP that was observed in the absence of CCh. Using this pipette solution, we found a small inward rather than outward current was present in baseline conditions, and the size of this current was significantly increased by bath application of 10 μM CCh (see Figure 4C for representative example). The I_ADP activated by CCh can then be isolated by calculating the difference between the CCh trace and baseline trace, as illustrated in Figure 4D.

Using this method, we proceeded to evaluate the time and voltage dependence of the isolated CCh-activated I_ADP. Specifically, experiments that varied the stimulus duration revealed that the maximum I_ADP could be produced with a stimulus duration of ~250 ms (Figure 4E). The mean amplitude of the maximum I_ADP was 35.3 pA, and the time constant of the activation curve was 52.3 ms. To evaluate voltage dependence, we used a constant stimulus duration of 500 ms, but varied the post-stimulation voltage command. This experiment revealed that the I_ADP demonstrated little voltage dependence at subthreshold potentials but had a linear slope at suprathreshold potentials with a reversal potential near +10 mV (Figure 4F).

To more clearly summarize data across these experiments, in Figure 4G we present area under the curve of poststimulation currents (recorded at −60 mV after a brief step to +50 mV), both in control conditions and in the presence of CCh. Specifically, the data obtained with a K-gluconate internal (left side of Figure 4G) further emphasizes that CCh converts the post-depolarization current from outward to inward (n = 3, p = 0.017), consistent with the current clamp data in Figure 3 that revealed a CCh-induced shift from an AHP to an ADP. By contrast, in cells filled with Cs-gluconate (right side of Figure 4G), we observed a post-stimulus inward current in control conditions that was significantly larger when observed in the presence of CCh (n = 9, 10 for control vs. CCh, p = 5.58e-4). Finally, still in cells with Cs-gluconate internal, we noted that if both CCh and FFA were present in the bath, then the post-stimulus current was no larger than observed in control conditions (n = 9, 10, respectively, p = 0.86), and was also significantly smaller than observed in the presence of CCh without FFA (n = 10,10, respectively, p = 4.74e-5). These data qualitatively mirror the effects of FFA on post-stimulus responses in current clamp, but importantly represent measurements of a much better isolated cholinergic response. We expect this method to be useful in future studies for making detailed measurements of cholinergic function in a variety of pathological conditions.