Whole-cell patch clamp recordings

For all recordings, slices were submerged in a heated, 32–34°C chamber, and perfused with bubbled artificial cerebral spinal fluid, which was perfused at 1–2 ml/min with artificial cerebral spinal fluid containing: 125 mM NaCl, 3 mM KCl, 1.25 mM NaH₂PO₄, 25 mM NaHCO₃, 1 mM MgCl₂, 2 mM CaCl₂, 10 mM dextrose, and 3 mM pyruvate at pH ∼7.4. Slices were visualized on an Axioskop 2 (Carl Zeiss Microscopy) with differential interference contrast optics and an infrared video camera (DAGE-MTI). Healthy pyramidal neurons in the middle of the proximal-distal axis of CA1 were targeted. Data were acquired with a Dagan BVC-700A amplifier with a 0.1 N headstage (Dagan Corp.), and digitized with an ITC-18 (HEKA Instruments Inc.). Data in this study were acquired at 10–20 kHz and filtered at 3–10 kHz. The pipette capacitance was compensated and the bridge was balanced throughout all recordings. Series resistance was monitored through recordings, and ranged from 8–30 MΩ for somatic recordings and 13–35 MΩ for dendritic recordings. The liquid junction potential, estimated to be ∼12 mV, was not corrected.

Borosilicate capillary glass 1.65-mm external diameter (World Precision Instruments) was pulled with a Flaming/Brown micropipette puller (model P-97, Sutter Instruments). Electrodes used for somatic recordings were pulled to have a resistance of 4–6 MΩ. For dendritic recordings electrodes had a resistance of 6–9 MΩ and were wrapped with Parafilm to reduce the capacitance of the electrode. Electrodes were filled with a solution containing: 120 mM potassium gluconate, 8 mM NaCl, 16 mM KCl, and 11 mM HEPES, 4 mM Mg-ATP, 0.3 mM Na-GTP, 7 mM 2K-phosphocreatine, and 0.2% neurobiotin; pH 7.37. For dendritic recordings, 16 µM Alexa Fluor 594 (Thermo Fisher Scientific) was included to determine recording location.

For all experiments, external AFSF saline included 20 µM 6,7 dinitroquinoxaline-2,3-dione (DNQX), 25 µM D-22-amino-5 phosphonovaleric acid (D-APV), which were obtained from Alomone Labs, and 2 µM SR-95531 (gabazine; Abcam). In some experiments, 10 µM ZD7288, 10 µM XE991, 2 µM CGP-55845, or 0.5 µM TTX was included in the ACSF (Abcam). In addition, for some experiments, 2 mM NiCl₂ or 50 µM BaCl₂ (Sigma-Aldrich) was included in the ACSF. ZD7288 was only introduced through the bath transiently, for 3–4 min. This prevented a nonspecific depolarization that occurs with continuous bath application, yet provided a stable block for ∼30 min (Kim and Johnston, 2015).

Data were acquired and analyzed with a custom written software in Igor Pro (Wavemetrics). To measure the input resistance and rebound slope, a family of 800-ms-long current injections from –150–50 pA was delivered to the cell. Voltage traces were excluded from the analysis if action potentials were generated. To calculate the input resistance, the change in voltage was plotted against the current amplitude (–70–10 pA), and the slope of a linear fit was reported. The amplitude of the rebound depolarization was plotted against the membrane potential at the end of the step, and the slope of a linear fit was referred to as the rebound slope. To calculate the peak resonance frequency, a ±50-pA sinusoidal current that increased in frequency from 0 to 15 Hz over 15 s was injected to the cell. The current and the voltage response were then transformed from a times series into the frequency domain with a fast Fourier transform. A ratio of the real portion of the transformed voltage and current was used to calculate the impedance amplitude response. This relationship was fit with a polynomial function, and the frequency at which the impedance was at its maximum was defined as the peak resonance frequency.

To measure the firing intensity a family of 800-ms-long current injections from 100 to 500 pA were delivered to the cell. The action potentials generated were measured at –20 mV. From traces that had 8–11 action potential, features of action potential shape were measured. The threshold was defined as the membrane potential at which the first derivative exceeded 20 mV/ms. The maximum of the first derivative was measured. The membrane potential at the peak of each action potential was subtracted from rest to calculate the amplitude. The fast afterhyperpolarization (fAHP) was detected by finding location within 1.5 ms of each spike where the derivative crossed zero for the second time. The membrane potential at this time was then subtracted from the threshold membrane potential to calculate the fAHP amplitude. The spike frequency accommodation (SFA) was reported as a ratio of the sixth spike to the first.