Material below is adapted from the SfN Short Course Local Circuit Analysis With Integrated Large-Scale Recording of Neuronal Activity and Optogenics, by György Buzsáki, PhD. Short Courses are daylong scientific trainings on emerging neuroscience topics and research techniques held the day before SfN’s annual meeting.
Understanding all the parts of a neuronal circuit will ultimately help explain how the circuit — and the brain — gives rise to behavior. To investigate how a circuit functions, different types of neurons spread across the brain need to be monitored simultaneously, and the contribution of each neuron determined. This requires hardware capable of recording many individual neurons, methods to separate spikes, and a way to identify different neuronal types.
When placed close enough to an individual neuron, electrodes can specifically record from that particular neuron. However, placing sharp electrodes so close to the cell’s body presents problems: the presence of the electrode can disrupt the cell’s firing, long-term recording is not practical as slight movements can disturb the electrode, and recording from multiple neurons (each with its own electrode) creates too much tissue damage.
On the other hand, monitoring neurons from farther away has problems as well. With only one electrode, different neurons at the same distance to the electrode may give identical signals, making it difficult to tell them apart. Using multiple recording sites (often four wires bunched together into a “tetrode”) can help separate neurons by triangulating the distances. But again, multiple electrodes create more damage.
An alternative to sharp electrodes is a micro-electro-mechanical system with silicon probes. Silicon probes are smaller than tetrodes, so they create less damage. Each probe has multiple recording sites in a known arrangement, allowing for the determination of various neurons’ locations and for simultaneous recording across different cortical layers. Silicon probes can also measure extracellular current, which (along with action potentials) can help calculate both inputs and outputs of a group of neurons.
Spikes can generally be sorted in two ways: either by their amplitude and waveform or by their location. But neither method is ideal. When based on amplitude and waveforms, one strongly activated neuron may appear to be multiple neurons. Alternatively, using triangulation to separate signals based on location assumes all signals from a neuron come from a single point, but signals may come from many places along the membrane. Again, because of this, a single neuron may appear as multiple neurons in different locations. Recording from multiple sites might help reduce these problems, but better, automatic methods and algorithms are needed to reduce human error.
Finally, coupling silicon probe recordings with optogenetics allows a way to classify the neurons. When analyzing a circuit, it’s important to know what types of neurons contribute and how. By integrating recording with optogenetics, neurons can be identified genetically, characterized physiologically, and tested for their roles within the circuit. Hybrid devices with both silicon probes and optical fibers reduce the light power required, minimizing secondary effects of stimulation and shrinking the size of the equipment needed, allowing for more freedom of movement in experimental animals.
Currently, optogenetics can only identify one or two neuron types in a single experiment. In the future, better understanding of the physiology of different cell types may lead to better identification of neurons without the need for optogenetics. In the meantime, the combination of extracellular recording and optogenetics enables better understanding of the roles of individual neurons within local circuits, and ultimately how perception, memory, and behavior are created.
