HomeResearchComputational Neurobiology

Our Scientists

Computational Neurobiology

Research Summary

Terrence Sejnowski's goal is to discover the principles linking brain mechanisms and behavior. His laboratory uses both experimental and modeling techniques to study the biophysical properties of synapses and neurons and the population dynamics of large networks of neurons.

All of the visual information transmitted from the retina to the brain is encoded in the spike trains generated by the ganglion cells, which are relayed by the lateral geniculate nucleus (LGN) to the visual cortex. How is this information decoded by the visual cortex to create visual perceptions, and how does the brain select what should be further processed? We have used a combination of experimental and modeling techniques to probe the neural mechanisms that are involved.

Neural Codes
A spike train can carry information in both the average rate of spikes and the pattern of spikes. But to be useful, the information in the spike pattern must be decoded by neural circuits.

In recordings from the LGN of the cat in response to an irregularly fluctuating whole-field stimulus, neurons fired spikes at highly reliable spike times, with millisecond precision, across trials, across neurons, and across cats. Although this stimulus is unnatural, the results of the experiment show that neurons in the visual system can respond with great timing precision. Under more natural conditions, synchronous spikes are generated in the retina after fast eye movements, called saccades, which occur around three times a second, and in response to moving stimuli.

We have developed a detailed biophysical model of the spiny stellate neurons in the primary visual cortex that receive direct inputs from the LGN to test whether the cortex can decode the synchronous spikes generated by the LGN cells. The convergence of synchronous LGN inputs onto spiny stellate neurons can drive these neurons reliably, even though only 95 percent of the synapses on these neurons are from other cortical neurons. We estimated that the optimal number of synchronously firing inputs needed for reliable transmission of information was around 4–6 LGN fibers, each of which makes 2 to 10 synapses on the spiny stellate cell. This level of reliability was also obtained when the visual stimulus was a moving grating, which elicits much less precisely timed spikes in the LGN.

This spike-timing precision is preserved in the hierarchy of visual-processing cortical areas that receive inputs from the primary visual cortex. In the middle temporal area (area MT) of the monkey, where neurons respond selectively to the direction of motion of a visual stimulus, the spike-timing precision in responses to moving stimuli that fluctuated between the preferred and nonpreferred directions was less than 5 milliseconds. This result is surprising given that the visual inputs to neurons in area MT are at least eight synapses removed from photoreceptors.

There is a hierarchy of areas in the visual cortex. Natural visual stimuli such as a movie produce more precisely timed bursts of spikes, or "events," in a small number of cells along the cortical pathway. These sparse events may be involved in regulating the communication of information to higher stages of processing, just as the relative timing of spikes in LGN neurons could regulate which visual stimuli reliably reach the primary visual cortex.

Inhibitory Regulation of Spike Timing
In addition to the flow of information upward along the visual hierarchy, which might be called passive bottom-up processing, there is also active top-down anticipation and control of visual search.

Fast-spiking parvalbumin-positive (PV) inhibitory interneurons constitute 5 percent of the neurons in the cerebral cortex and make multiple synapses around the soma of pyramidal cells; for this reason, they are called "basket" cells. These interneurons receive strong excitatory inputs from the more numerous pyramidal cells, which they reciprocally inhibit, and are involved in generating fast oscillations in the gamma frequency (30–60 Hz) range. Using a biophysical model of basket cells, we have shown that they can be transiently synchronized through "synchrony by competition," in which inputs from higher cortical areas through feedback projections depolarize a subset of interneurons, increasing their firing rate and synchrony. This reduces the firing rate of the remaining interneurons and synchronizes the "winners," which could explain why the firing rates of a subset of inhibitory neurons increase when a monkey selectively attends to a stimulus.

When an interneuron network is in a synchronized oscillation, a postsynaptic pyramidal cell receives volleys of synchronized inhibitory inputs from the network, which could act as a gate. The pyramidal cell produces spikes that are precisely timed with respect to the inhibitory rhythm. Only excitatory inputs that arrive during the period when the inhibitory conductance is low can reliably elicit an output spike, which is locked to the local inhibitory rhythm. Using this mechanism, the specific path that feedforward information follows through cortical areas could be rapidly altered to achieve behavioral goals. Thus, inhibitory interneurons may have a dynamic role for regulating the flow of information in the cortex by internally controlling spike timing and synchrony of spikes in cortical circuits.

Schizophrenia
The pathophysiology of schizophrenia is complex and involves many different cortical and subcortical systems. The slow development of fast-spiking PV-interneurons, although essential for shaping neuronal circuits during postnatal brain development, increases their vulnerability to insults that can permanently affect their maturational process and lead to cortical dysfunction and perhaps schizophrenia in adulthood. Reduced expression of GAD67 in fast-spiking PV-interneurons (the main isoform synthesizing GABA in the brain) is one of the most replicated findings in postmortem brain studies of schizophrenia.

Fast-spiking PV-interneurons are also involved in regulating working memory and information transmission between cortical areas, and their dysfunction may account for the disruption in evoked gamma frequency oscillations as well as the cognitive deficits observed in schizophrenia. We are exploring an experimental rodent model of schizophrenia that can be induced by subanesthetic injections of NMDA receptor blockers such as ketamine or phencyclidine (PCP). In the adult mouse, this leads to the temporary down-regulation of GAD67, but we have shown that this is permanent if the injections occur during the first postnatal week. We have also developed cortical circuit models to explore the impact of down-regulating the PV-interneurons on the function of the cortex and have shown that a reduction in both GAD67 and fast-spiking PV expression may contribute to the decrease in gamma power. Parvalbumin is a calcium binding protein; when it is depleted, there is an increase in asynchronous GABA release, which stabilizes the disinhibited network but also dampens gamma oscillations.

Future Directions
One major conclusion is that we need to reconsider current views on the reliability of cortical neurons, since there is good experimental evidence that they are capable of highly reliable spike time responses with millisecond precision. The firing rate tells us what is being represented, and the relative timing of spikes in a population of neurons could be used as an internal variable in the cortex to represent the importance of a stimulus. Spike times could also have consequences for how memories are stored in the cortex, since the relative spike timing of cortical neurons also controls synaptic plasticity. Finally, higher cognitive functions could be disrupted when the ability of the cortex to organize information on a fine timescale is compromised. This may give us clues to how thought disorders like schizophrenia could someday be treated and perhaps even cured.

This research was supported in part by the National Institutes of Health and the National Science Foundation.

As of May 30, 2012

Scientist Profile

Investigator
Salk Institute for Biological Studies
Biophysics, Neuroscience