If variety is the spice of life, then photoactivation of neurons just got hotter. Recently, researchers described a photoactivatable cation channel from algae called channelrhodopsin-2 that depolarizes transfected neurons in response to blue light. Can mammalian receptors be photoactivated? Szobota et al. report glutamate receptors with attached ligands that bind in the presence of ultraviolet light in a recent article in Neuron.
Isomerization alters a molecule's geometry. The authors designed a glutamate analog on a linker that would change conformation in response to different wavelengths of light. The linker–ligand complex contained maleimide to attach the complex to a cysteine residue in the receptor, azobenzene for photoisomerization and a glutamate analog (MAG). The authors aimed to anchor MAG near the ligand-binding domain of the ionotropic kainate glutamate receptor. However, glutamate receptors do not contain cysteines in this region. Therefore, the authors mutated a leucine near the kainate receptor ligand-binding domain to a cysteine. Once covalently attached, MAG remained in the trans conformation under visible light (500 nm). When exposed to ultraviolet light (380 nm), MAG switched to the cis conformation, fitting snugly in the ligand-binding pocket and depolarizing the cell. To terminate activation, the authors re-exposed neurons to 500 nm light.
They transfected rat hippocampal slices with the light-activated glutamate receptor and incubated them with MAG. Brief exposure to 380 nm light induced action potentials, which were terminated by 500 nm light. In contrast, neurons in untransfected hippocampal slices exposed to MAG did not respond to light, suggesting that MAG is specific for the mutated kainate receptor. Brief pulses (1-5 ms) of ultraviolet light followed immediately by 500 nm light induced single action potentials. Longer ultraviolet illumination induced trains of action potentials that lasted for tens of minutes, prolonged activation that is possible because the half-life for spontaneous MAG isomerization is approximately 18 minutes. Either reduced light intensity or off-peak wavelength light diminished the amplitude of the neuronal response.
Like channelrhodopsin-2, light-activated glutamate receptors altered behavior in vivo. The authors generated zebrafish hemizygous for the light-activated glutamate receptor driven by the yeast upstream activating sequence. They mated these fish with zebrafish hemizygous for the upstream activating sequence binding-protein Gal4 driven by a promoter expressed in hindbrain, Rohon-Beard, trigeminal and vagal ganglion neurons, which are all important in the zebrafish touch-activated escape reflex. They incubated the progeny larvae in medium containing MAG. Although all of the progeny showed normal motor and visual behavior, 380 nm light blocked the escape reflex in approximately 28% of the fish, consistent with the proportion likely to be doubly transgenic (25%). Subsequent genotyping confirmed that nearly 93% of the light-sensitive zebrafish expressed the transgene.
Is one system of photoactivation better than the other? The answer depends on the situation. The light-activated glutamate receptor passes more current, terminates activation faster and can sustain longer activation than channelrhodopsin-2. However, channelrhodopsin-2 does not require exogenous ligands and may therefore be easier to use in vivo.