Membrane properties underlying patterns of GABA-dependent action potentials in developing mouse hypothalamic neurons

Spikes may play an important role in modulating a number of aspects of brain development. In early hypothalamic development, GABA can either evoke action potentials, or it can shunt other excitatory activity. In both slices and cultures of the mouse hypothalamus, we observed a heterogeneity of spike patterns and frequency in response to GABA. To examine the mechanisms underlying patterns and frequency of GABA-evoked spikes, we used conventional whole cell and gramicidin perforation recordings of neurons (n = 282) in slices and cultures of developing mouse hypothalamus. Recorded with gramicidin pipettes, GABA application evoked action potentials in hypothalamic neurons in brain slices of postnatal day 2-9 (P2-9) mice. With conventional patch pipettes (containing 29 mM Cl-), action potentials were also elicited by GABA from neurons of 2-13 days in vitro (2-13 DIV) embryonic hypothalamic cultures. Depolarizing responses to GABA could be generally classified into three types: depolarization with no spike, a single spike, or complex patterns of multiple spikes. In parallel experiments in slices, electrical stimulation of GABAergic mediobasal hypothalamic neurons in the presence of glutamate receptor antagonists [10 microM 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), 100 microM 2-amino-5-phosphonopentanoic acid (AP5)] resulted in the occurrence of spikes that were blocked by bicuculline (20 microM). Blocking ionotropic glutamate receptors with AP5 and CNQX did not block GABA-mediated multiple spikes. Similarly, when synaptic transmission was blocked with Cd(2+) (200 microM) and Ni(2+) (300 microM), GABA still induced multiple spikes, suggesting that the multiple spikes can be an intrinsic membrane property of GABA excitation and were not based on local interneurons. When the pipette [Cl-] was 29 or 45 mM, GABA evoked multiple spikes. In contrast, spikes were not detected with 2 or 10 mM intracellular [Cl-]. With gramicidin pipettes, we found that the mean reversal potential of GABA-evoked current (E(GABA)) was positive to the resting membrane potential, suggesting a high intracellular [Cl-] in developing mouse neurons. Varying the holding potential from -80 to 0 mV revealed an inverted U-shaped effect on spike probability. Blocking voltage-dependent Na+ channels with tetrodotoxin eliminated GABA-evoked spikes, but not the GABA-evoked depolarization. Removing Ca(2+) from the extracellular solution did not block spikes, indicating GABA-evoked Na+ -based spikes. Although E(GABA) was more positive within 2-5 days in culture, the probability of GABA-evoked spikes was greater in 6- to 9-day cells. Mechanistically, this appears to be due to a greater Na+ current found in the older cells during a period when the E(GABA) is still positive to the resting membrane potential. GABA evoked similar spike patterns in HEPES and bicarbonate buffers, suggesting that Cl-, not bicarbonate, was primarily responsible for generating multiple spikes. GABA evoked either single or multiple spikes; neurons with multiple spikes had a greater Na+ current, a lower conductance, a more negative spike threshold, and a greater difference between the peak of depolarization and the spike threshold. Taken together, the present results indicate that the patterns of multiple action potentials evoked by GABA are an inherent property of the developing hypothalamic neuron.