Down With Inhibition: Modulating GABAergic Neocortical Control
24 July 2008. There is evidence that local cortical circuits are disturbed in the brains of people with schizophrenia. In particular, complementary hypotheses have arisen concerning deficits in parvalbumin-containing, fast-spiking interneurons of prefrontal cortex (see SRF related news story) and abnormalities of cortical gamma band synchronization (see SRF Current Hypothesis paper by Woo and colleagues). Thus, basic science findings on control of cortical activity are of special interest to scientists investigating schizophrenia.
In a study published in the June 26 issue of Neuron, Illya Kruglikov and Bernardo Rudy of New York University report that neuromodulators, including acetylcholine, serotonin, and adenosine, regulate GABA release from fast-spiking interneurons onto neocortical excitatory cells. Specifically, they control the perisomatic release of GABA—i.e., onto the cell bodies and initial axon and dendrite segments of the pyramidal cells. The researchers were also able to show that this modulation regulates the integration of thalamocortical signaling.
Modulators of inhibition
Neuromodulators such as acetylcholine, serotonin, and norepinephrine modify cortical circuits, changing the flow of information. To do this, their effects include critical actions on neurons employing GABA, the brainís main inhibitory neurotransmitter. However, the direct effects of neuromodulators on the interneurons that release GABA are not well understood.
Kruglikov and Rudy studied the effects of activating a number of different neuromodulator receptors on fast-spiking basket cells, the primary interneurons of the mammalian neocortex. These cells provide a major source of inhibition to neocortical pyramidal cells. A subset of parvalbumin-containing, fast-spiking interneurons are of special interest to research groups that have identified deficits in these cells in people with schizophrenia (see interview with David Lewis).
The investigators initially screened a series of neuromodulators and drugs that act on neuromodulator receptors, assessing their ability to modify inhibitory post-synaptic currents (IPSCs) in cultured mouse brain slices of somatosensory cortex. They recorded from multiple pyramidal cells in the same slices, driving electrical activity of GABAergic interneurons with a nearby stimulating electrode.
Agents that blocked the inhibitory effects of the stimulation included muscarine, serotonin, adenosine, and baclofen. Muscarine stimulates the muscarinic acetylcholine receptor and baclofen stimulates the GABAB receptor. Compounds targeting metabotropic glutamate, adrenergic, histaminergic, cannabinoid, somatostatin, or cholecystokinin receptors did not modulate the IPSCs.
The researchers wanted to confirm that the IPSCs that they observed were indeed due to the fast-spiking basket cells, particularly since a variety of neuronal cell types (especially Martinotti interneurons) could have been activated by the stimulating electrode. Several pieces of evidence implicated the fast-spiking basket cells.
For one, they comprise the largest population of interneurons in layer five of somatosensory cortex, and consistent with their numbers, fast-spiking cells provided the greatest contribution to the IPSCs provoked by the stimulation. This was determined by comparing inhibitory recordings from Martinotti cell-pyramidal cell pairs versus fast-spiking cell-pyramidal cell pairs. In addition, blocking P/Q calcium channels, which are specific to the terminals of fast-spiking cells, also reduced the IPSCs. Blocking N-calcium channels, which are found in terminals of other types of interneurons, had no effect. And most importantly, according to the authors, these modulations were also observed on the IPSCs generated on a pyramidal cell when a single, synaptically connected fast-spiking interneuron was stimulated.
Kruglikov and Rudy also assessed whether the effects of the four neuromodulators on inhibitory activity were dependent on the intracellular signaling of protein kinases. Serotonin did not block the IPSCs evoked by stimulation of the slices when the protein kinase inhibitor staurosporine was present, indicating that the effects of serotonin were largely protein kinase-dependent. The inhibitory effects of the other three neuromodulators were not affected by staurosporine. Interestingly, serotonin also had a slow time course of blocking the IPSCs measured in the slices, suggesting second-messenger involvement, whereas the other three agents acted more quickly, which the authors take to indicate that these neuromodulators have effects only at the membrane, via calcium channels.
A bigger slice of cortical activity
The researchers then moved to a slice preparation that included both neocortex and its major input source, the thalamus, in order to study the interaction of neuromodulators and excitatory thalamic input. Afferents from thalamus to cortex contact both excitatory and inhibitory neurons, exciting them via glutamate release. There is a brief span of time between the excitation of the primary excitatory cells via the thalamic input and the feed-forward inhibition of the interneurons onto these cells. This "temporal integration window" is believed to be important for processing sensory inputs to neocortex.
Because muscarine was particularly effective at inhibiting GABA release from neocortical fast-spiking interneurons, and because of the powerful effect acetylcholine is known to have on cortex, Kruglikov and Rudy focused on the role that cholinergic receptors might play in reducing inhibition in the cortex. While stimulating the somatosensory region of the thalamus, they measured the effects of muscarine on both inhibitory and excitatory post-synaptic currents (EPSCs) in somatosensory pyramidal neurons. The researchers report that feed-forward IPSCs decreased, producing an increase in the window for temporal integration. Baclofen, the GABAB receptor agonist, had the same effect, but completely eliminated the feed-forward IPSCs. Nicotine, which stimulates the nicotinic subtype of the acetylcholine receptor, increased EPSCs in excitatory cells, but not in inhibitory cells.
The authors conclude that cholinergic activity increases thalamic excitation of neocortex, mainly by reducing the feed-forward inhibition produced by release of GABA from fast-spiking interneurons. This effect is mediated via muscarinic receptors and is complemented by the excitation of the excitatory neurons by acetylcholine via nicotinic receptors.
There is some sparse evidence that communication between thalamus and cortex is disturbed in schizophrenic patients (for review, see Sim et al., 2006), but the most relevant impact of this research is likely to be in the elucidation of mechanisms involved in cortical integration, suggesting neuronal circuitry of interest for researchers studying the disease. Future research may also address whether inhibitory circuits in hippocampus, another region of prime interest for schizophrenia researchers, are modulated in the same fashion.—Alisa Woods and Hakon Heimer.
Kruglikov I, Rudy B. Perisomatic GABA release and thalamocortical integration onto neocortical excitatory cells are regulated by neuromodulators. Neuron. 2008 Jun 26;58(6):911-24. Abstract