1. The properties of receptors for amino acid neurotransmitters expressed by developing cortical neurons were studied with the use of whole-cell recording in the intact cerebral cortex of embryonic turtles in vitro. The inhibitory agonist-gamma-aminobutyric acid (GABA) and the excitatory agonist glutamate were focally applied to single cells under voltage clamp, and the ionic dependence, voltage dependence, and pharmacological sensitivity of the responses were characterized. The responses mediated by a glutamate receptor subtype, the N-methyl-D-aspartate (NMDA) receptor, produced by glutamate and by evoked release of an endogenous excitatory agonist, were compared further. Fluctuation analysis was used to characterize the properties of the NMDA channels and the mechanism of action of receptor antagonists. 2. When postmitotic neurons first appeared at stage 15, all neurons tested responded to GABA with a current that reversed at the equilibrium potential for chloride ions and that was sensitive to the GABA(A) receptor antagonist bicuculline methiodide (BMI). As development proceeded, an increasing proportion of neurons also responded with a BMI-insensitive current that reversed near the equilibrium potential for potassium ions. This current was blocked by the GABA(B) receptor antagonist 3-amino-2-propyl phosphonic acid (phaclofen). The GABA(B) agonist baclofen, however, failed to produce a detectable postsynaptic current. 3. Neurons at stage 15 showed a biphasic response to glutamate that reversed at the equilibrium potential for cations. All neurons tested showed a slow, sustained response associated with an increase in current variance compared with background, and, as development proceeded, an increasing proportion also exhibited a fast, transient response. Both fast and slow responses varied linearly with voltage in the absence of Mg2+ ions, but the addition of Mg2+ ions to the bathing medium attenuated the slow response at hyperpolarized potentials. As a result, the current-voltage relation of the slow response in the presence of Mg2+ ions exhibited a region of negative slope conductance, like that of currents mediated by NMDA receptors. 4. The fast and slow responses to glutamate differed in their pharmacological sensitivity. The fast responses were sensitive to the non-NMDA receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), whereas the slow responses were sensitive to the NMDA receptor antagonist D(-)-2-amino-5-phosphonovalerate (D-APV). 5. When cells were held at -70 mV, glutamate evoked a fluctuating current consisting of channel currents with a mean open time, tau, of 4.42 +/- 0.47 (SE) ms in early postmitotic neurons at stage 15 and 4.99 +/- 0.38 ms at stages 17-20. Addition of the NMDA receptor antagonist D-APV reversibly blocked the current and associated fluctuations in a dose-dependent way without changing the mean open time, consistent with a competitive antagonist mechanism of action. 6. The glutamate-evoked current and associated fluctuations were also attenuated at hyperpolarized potentials by Mg2+ ions. Fluctuation analysis of responses in Mg2+ revealed addition of a second Lorentzian component in power spectra of neurons that, in the absence of Mg2+, had only a single component (tau in Mg2+ = 5.04 +/- 0.76 and 0.94 +/- 0.27 ms). This shift in open times is consistent with a noncompetitive antagonist mechanism of action. 7. In addition to attenuating responses to focally applied glutamate, the presence of the antagonists D-APV and Mg2+ in the bathing solution caused a parallel decrease in a background current. In young animals, this current was composed of unitary events with a conductance (50 pS) resembling that of NMDA receptor currents in previous studies. Glutamate-activated channels in outside-out patches from juvenile neurons had a similar conductance of 47 +/- 1 pS. In more mature animals, the background current consisted of a net inward current. 8. Mg2+ blockade of background current is exerted directly on channels of the postsynaptic cell, rather than indirectly by preventing presynaptic release of excitatory neurotransmitter. When NMDA receptor-dependent current noise was recorded in the presence of spontaneous excitatory postsynaptic currents (EPSCs), the current noise was blocked in Mg2+ while the EPSCs persisted. These EPSCs, caused by release of presynaptic excitatory neurotransmitter, were blocked by the non-NMDA receptor antagonist CNQX. 9. NMDA receptor activation by endogenous agonist did not require sodium-dependent action potential activity, because the current continued in the presence of tetrodotoxin (TTX). However, action potential activity could increase the current: stimulation of specific afferents to the cerebral cortex in early development caused activation of channels with the same open time and conductance as the endogenously activated NMDA channels. 10. Cortical neurons in situ respond to putative excitatory and inhibitory neurotransmitters soon after their generation and before the time synaptic currents are observed. The appearance of receptor subtypes occurs in a characteristic sequence, with GABA(A) and NMDA receptors detectable first, followed by GABA(B) and non-NMDA receptors. When first expressed, these receptors exhibit many of the pharmacological and ionic mechanisms of their counterparts in mature brain. NMDA receptors expressed by the earliest differentiating neurons in turtle, when activated by endogenous agonist, exhibit properties similar to those of channels activated by glutamate and similar to previously reported mature channels. Neurotransmitter receptors with the same basic properties may play a role in shaping cellular differentiation at early developmental stages and may subsequently mediate synaptic communication in the mature brain.
