Extrasynaptic NMDARs Enhance Inhibitory Neuron Excitability
Lulu Yao, Yi Rong, Xiaoyan Ma, Haifu Li, Di Deng, et al.
(see pages 3066–3079)
Activation of synaptic NMDA receptors (NMDARs) in excitatory neurons has a well established role in synaptic plasticity. Activation of extrasynaptic NMDARs in excitatory neurons can lead to synapse removal and even cell death. Considerably less is known about the roles of synaptic and extrasynaptic NMDARs in inhibitory neurons, despite evidence that hypofunction of these receptors contributes to schizophrenia. To clarify the effect of NMDARs on inhibitory neuron activity, Yao et al. treated cortical slices with M-8324, a positive allosteric modulator of NMDARs that was previously shown to act selectively on inhibitory interneurons, likely because of higher ambient glutamate levels at these synapses.
M-8324 increased spontaneous spiking in inhibitory neurons, as well as reducing the amount of depolarizing current needed to evoke a spike. In addition, more spikes were elicited by EPSC-like current pulses in GABAergic neurons in slices treated with M-8324 than in control slices. These effects were attributable partly to a hyperpolarizing shift in the spike threshold after M-8324 treatment. M-8324 did not affect the resting membrane potential, input resistance, or spike shape, however. Notably, the effects of M-8324 were mimicked by a selective enhancer of GluN2C/2D-containing receptors and by EU1794-4, a selective modulator of extrasynaptic NMDARs. Moreover, the effects of both M-8324 and EU1794-4 were abolished by a selective antagonist of GluN2C/2D NMDARs, but not a selective antagonist of GluN2B-containing NMDARs. Finally, M-8324 increased spiking in presumptive inhibitory neurons in prefrontal cortex in vivo, and this was associated with a decrease in spiking in presumptive excitatory neurons.
These results suggest that activation of extrasynaptic GluN2C/2D-containing NMDARs in GABAergic cortical neurons increases the excitability of these neurons, in part by lowering the spike threshold. This likely leads to increased inhibition of excitatory neurons, resulting in reduced glutamate release when ambient glutamate levels are elevated. Disruption of NMDAR function in cortical inhibitory neurons would therefore lead to disinhibition of excitatory neurons, which may contribute to psychotic symptoms in schizophrenia. Future work should determine the molecular pathways through which the activation of extrasynaptic NMDARs lowers spike threshold and any other effects of NMDAR activation in inhibitory neurons.
Application (at time indicated by black vertical line) of hNR1 (green) causes a significantly greater increase in spiking of cultured cortical neurons than application of control antibody (gray) or a nonsaturating dose of NMDAR antagonist (yellow). See Andrzejak et al. for details.
Anti-NMDAR Antibodies Reduce Cortical Inhibition
Ewa Andrzejak, Eshed Rabinovitch, Jakob Kreye, Harald Prüss, Christian Rosenmund, et al.
(see pages 3253–3270)
Certain kinds of peripheral tumors express NMDA receptors (NMDARs), leading to the production of anti-NMDAR antibodies. If these antibodies enter the brain, they can cause anti-NMDAR encephalitis, characterized by psychosis, memory loss, seizures, and autonomic dysfunction. Studies in mice treated with patient-derived antibodies suggest that memory loss results partly from internalization of NMDARs in hippocampal pyramidal neurons, resulting in a reduction in NMDAR currents. Loss of NMDARs on excitatory neurons would not be expected to cause seizures, however. Instead, seizures may result from loss of NMDARs on inhibitory neurons and subsequent disinhibition of excitatory neurons. Intriguingly, studies of schizophrenia suggest that loss of NMDAR function in inhibitory neurons may also underlie psychotic symptoms. Andrzejak et al. confirmed a predominant effect of a patient-derived anti-NMDAR antibody (hNR1) on inhibitory neurons in mouse cortical cultures.
Microarray recordings revealed that overall neuronal activity increased after cortical cultures were treated with hNR1. This hyperactivity included an increase in burst firing similar to what occurs when inhibition is removed. Consistent with loss of inhibition, hNR1 decreased the frequency of miniature IPSCs and occluded the disinhibitory effects of a GABAA receptor antagonist in excitatory neurons. Remarkably, hNR1 decreased NMDAR currents and synaptic output only in inhibitory neurons, not in excitatory neurons. Furthermore, immunostaining indicated that hNR1 bound to a larger proportion of glutamatergic synapses on inhibitory neurons than on excitatory neurons. In contrast, hNR1 labeled ∼30% of inhibitory synapses on excitatory neurons, but only ∼11% of inhibitory synapses on inhibitory neurons. Finally, hNR1 decreased the intensity of staining of both the vesicular GABA transporter and the GABA-synthesizing enzyme GAD65 selectively at inhibitory synapses onto excitatory neurons, not at synapses onto inhibitory neurons.
These results suggest that hNR1 selectively reduces inhibitory drive to excitatory neurons in cortical cultures and this occurs in part via a reduction in GABA synthesis and packaging in presynaptic terminals. Notably, this differs from the predominant effect of hNR1 on excitatory neurons in hippocampus. Future work should explore the reasons for this difference by determining whether GluN2 subunits affect binding of hNR1 to NMDARs, where the affected receptors are expressed (presynaptically, postsynaptically, or extrasynaptically), and whether hNR1 affects the excitability of cortical GABAergic neurons.
Footnotes
This Week in The Journal was written by Teresa Esch, Ph.D.
