Using light to control neuronal activity

A group of researchers led by Ed Boyden, an assistant professor at the MIT Media Lab and leader of the Neuroengineering and Neuromedia Laboratory, has developed a technique by which pulses of light are used to inhibit the activity of specific individual neurons on a millisecond-by-millisecond timescale. The findings were published last week in the open access journal PLoS One.

The method uses a protein called halorhodopsin, which was recently isolated from Natronomas pharaonis, an extremophile archaebacterium that thrives in highly salty conditions. Halorhodopsin is a light-activated chloride channel, which opens in response to green-yellow light, allowing an influx of chloride ions. Boyden and his colleagues fused the halorhodopsin gene to the gene encoding green fluorescent protein (GFP). The construct was under the control of the CAMKII promotor, a regulatory DNA element which drives gene expression in neurons of the mammalian forebrain. Using a lentivirus vector, Boyden’s team shuttled the construct into cultured rat hippocampal neurons; cells expressing the construct could be easily visualized because they fluoresced green.

Microelectrodes were used to inject ionic solutions into the cells; the electrical currents evoked trains of up to 20 action potentials at a frequency of 5 per second. At the same time, pulses of light were delivered to the cells, at specific phases in the trains of action potentials, to inhibit the activity of the cells. The light activates the halorhodopsin molecules, which then pump negatively charged chloride ions into the cell. This hyperpolarizes the membrane of the nerve cell, i.e. increases the difference in voltage between the inside and outside of the membrane, so that the cell is less likely to generate a nervous impulse. The induced hyperpolarizations occurred within 15 milliseconds of the light pulse, and were reversible – the currents were deactivated within 15 milliseconds of cessation of the light pulse. This temporal resolution is so high that the light pulses could abolish single spikes in the train; a pulse timed to coincide with the 17th action potential in the train eliminated that spike from the train, but left spikes 16 and 18 unaffected.

Boyden’s team assayed the effects of the construct on cultured cells. It was found that, in the absence of light, the electrical properties and basal activity of the transfected cells appeared to be no different from wild-type (normal) cells. Expression of the halorhodopsin-GFP construct remained stable for up to one week and did not lead to cell death.

In previous work by the same group, a similar technique was used to activate neurons. In this case, the protein used was channelrhodopsin-2 (ChR2), a light-gated ion channel from the green alga Chlamydomonas rheinhardtii. In transfected rat hippocampal cells, ChR2 is also expressed robustly, and is observed to be localized at the cell membrane for weeks after transfection. Activation of ChR2 by pulses of blue light generates inward currents of protons (hydrogen ions). These currents are induced even more rapidly than those mediated by halorhodopsin – within 1 millisecond of onset of the light pulses. The currents depolarize the cells, i. e. they bring the membrane voltage nearer to zero, inducing action potentials; rapid light pulses delivered in quick succession evoke realistic trains of action potentials.

In the PLoS One paper, Boyden’s group describe combining the two systems by transfecting cells with both constructs, so that they simultaneously expressed halorhodopsion and ChR2. In these cells, alternate pulses of yellow and blue light induce hyperpolarizations and depolarizations of targeted neurons, respectively, providing a method for ultrafast, rapidly reversible and extremely precise control of neuronal activity. The high spatiotemporal resolution of the technique provides the fastest control yet of neuronal membrane voltage, enabling the electrical activity of the neurons to be controlled without disturbing the normal firing patterns of the cells.

Optical inhibition and/ or activation of neurons will prove to be a powerful tool for researchers. It will, for example, enable them to introduce temporary functional lesions in specific cells or groups of cells. By silencing specific subsets of cells for periods of less than a second during the performance of a task, the role of the timing of action potentials in neural computation can be investigated. Advances in in vivo imaging techniques, such as two-photon fluorescence microscopy, should enable researchers to photoactivate cells within intact tissues. Eventually, the technique could also lead to the development of optical neural prostheses for controlling the aberrant electrical activity associated with conditions such as Parkinson’s Disease and epilepsy. Boyden and his colleagues plan to start experimenting with such devices in mice later this year.

References:

Han, X. & Boyden, E. S. (2007). Multiple-color activation, silencing and desynchronization of neural activity, with single-spike resolution. PLoS One 2: doi: 10.1371/journal.pone.0000299. [Full text]

Boyden, E. S., et al. (2005). Millisecond-timescale, genetically targeted optical control of neural activity. Nat. Neurosci. 8: 1263–1268. [Full text]