A team lead by neurobiologist Richard H. Kramer, UC Berkeley professor of molecular and cell biology, and Dirk Trauner, assistant professor of chemistry, inserted a light-activated switch into brain cells normally insensitive to light, enabling the researchers to turn the cells on with green light and turn them off with ultraviolet light.
This trick could potentially help those who have lost the light-sensitive rods and cones in their eyes because of nerve damage or diseases such as retinitis pigmentosa or age-related macular degeneration. In these cases, the photoreceptor cells are dead, but other nerve cells downstream of the photoreceptors are still alive. In particular, retinal ganglion cells, which are the third cell in the path from photoreceptor to brain, could take over some of the functions of the photoreceptors if they could be genetically engineered to respond to light.
Kramer envisions a device, reminiscent of the eyepiece worn by the blind Geordi La Forge in "Star Trek - The Next Generation," that would provide some semblance of the real world.
"We may be able to use laser scanning to trace on and off patterns on the retina and allow people to see visual patterns," Kramer said. "Sometimes I'm not sure where the science ends and the fantasy begins, but I think we can make it work."
"With this technique, you also could confer light sensitivity on organisms that normally don't have vision, such as the nematode worm C. elegans," Trauner said. "Taking this from a chemical novelty to showing that it works in a biological system is a real breakthrough."
Kramer, Trauner and their colleagues will report their results on Nov. 21 in a paper published online in the journal Nature Neuroscience.
The idea of genetically engineering surviving retinal cells to be sensitive to light has various advantages over the most common approach to creating a bionic eye -- inserting electrodes into the optic nerve to simulate the cell firings a visual scene normally would excite. Though this technique works fairly well in the ear -- witness the success of cochlear implants -- the eye is a much more complicated place, Kramer said.
"This is a more organic, less invasive approach than electrodes," Kramer said, noting that insertion of electrodes can cause problems with biocompatibility. Electrodes also are large and tend to stimulate an entire bank of cells at once, which would limit the resolution.
"How well electrodes would work depends on the density of the electrode array and how well you can marry the electrodes with the neural elements underneath," he said. "Our approach is not a mere chip on the retina - it may allow us to cover the entire retina with light sensitive cells. If each nerve responds individually, you could do a very fine scan of the retinal field and create much, much better spatial resolution."
Current, admittedly early attempts at restoring sight with electrodes in the retinal ganglion cells, whose axons bundle together to form the optic nerve entering the brain, allow the patient to see little more than patches of light and dark, Kramer noted.
Kramer, a researcher with UC Berkeley's Helen Wills Neuroscience Research Institute and a member of the campus's Health Sciences Initiative, studies ion channels -- protein valves that regulate the flow of charged atoms in and out of cells. Spanning the membranes of nerve cells, sodium and potassium channels, in particular, facilitate the transmission of electrical signals along the length of the cell.
Trauner, on the other hand, specializes in synthesizing large, complex molecules. Together, the two scientists conceived the idea of modifying an ion channel to turn it into a remote-controlled switch that could be used to turn nerve cells on and off.
They decided to concentrate on the potassium channel, which opens when a voltage difference develops between the inside and outside of a nerve cell. The open channel lets positive potassium ions flow out of the cell, equalizing the voltage and turning the cell off.
Trauner, Kramer and their team designed a way to re-engineer the potassium channel to respond to light rather than voltage. To create this man-made channel and insert it into living cells, they took a two-step approach. First, they mutated the gene for the ion channel -- using as their starting material the potassium channel found in fruit flies -- so that, when expressed in a cell, the channel is broken and always stays open. They also added an extra molecule -- the amino acid cysteine -- to the channel protein so that, once the protein gets in place in the cell membrane, this molecule dangles off the outer surface of the cell like a fish hook.
They then inserted the mutated potassium channel gene into cells from the hippocampus of a rat -- cells that are found inside the brain and never see the light of day. To achieve this in their cell culture experiment, they flooded the culture with the mutated gene inside a circular piece of DNA called a plasmid, which cells readily take up. They checked to see how many of the hippocampal cells took up the gene by also washing the cells with a plasmid containing a gene for green fluorescent protein, which glows green when hit with UV light. Cells taking up one plasmid usually take up other plasmids, and nearly all the cells glowed green.
The second step was to wash the cells with a chemically synthesized switch that gloms onto the cysteine hook. The photoswitch - an azobenzene compound - was built like a drain plug on a rigid tether, so that when the end of the tether binds to cysteine, the plug fits snugly into the potassium channel.
The chemical was also designed to be sensitive to light -- when hit with long-wavelength ultraviolet light (390 nanometer wavelength), the tether kinks and shortens, pulling the plug and letting potassium out of the cell. Green light (500 nanometer wavelength), on the other hand, makes the chemical tether straight again, replugging the channel pore. They refer to the altered channel as a synthetic photoisomerizable azobenzene-regulated K (SPARK) channel, where K is the chemical signal for potassium.
Apart from possible therapeutic applications, or tricks such as giving sight to sightless organisms, the technique allows neuroscientists to ask more basic questions, Trauner said.
"Once we insert this artificial light-sensitive channel in a nerve cell, it opens an extra potassium channel that we can manipulate remotely to hyperpolarize the cell and silence it," Trauner said. "By selectively silencing neurons in a complex network of neurons, all of them talking to one another, we can try to figure out who talks to whom."
These potassium channels also can be made sensitive to molecules instead of light, so that a nerve cell could be turned on or off by DNA or heavy metals, for example. Kramer and Trauner are most excited about the possibility of artificial vision, however.
"We created a method for making light-regulated channels that are stably light sensitive, responding rapidly and reliably for hours," Kramer said. "Now, we're trying it in eyeballs."
To achieve the same trick in a living eye, Kramer will use a virus, such as the adeno-associated virus that is commonly used for experimental gene therapy, to carry the mutated channel genes into retinal ganglion cells. The viruses are injected directly into the vitreous or liquid center of the eye, where they have easy access to ganglion cells.
Kramer noted several problems with the approach, but possible fixes, too. For one, not all retinal ganglion cells are alike. Some are "on" cells that turn on when the eye is hit with light, while others - "off" cells - turn off. This is part of the eye's analysis circuitry, which helps pick out significant features of the visual field, such as edges and motion, even before the signals reach the brain. Inserting the same switch in all retinal ganglion cells could result in a visual muddle.
"Your brain would be confused, like feeling hot and cold at the same time," he said. "Electrodes would have this problem, too, indiscriminately stimulating on and off cells."
One solution, Kramer said, is to re-engineer a sodium channel to function just the opposite of the mutated potassium channel, then target the engineered sodium channel to "on" cells and the engineered potassium channel to "off" cells, using cell specific promoters.
"If you're using electrical stimulation, there is no way to selectively deliver information to two different channels," he said. "But with genetics, we can do something that electrical stimulation can never do."
"We haven't cured blindness yet," Kramer added, "but that's our main motivation in this work."
Coauthors with Kramer and Trauner are graduate students Matthew Banghart and Katharine Borges, and Ehud Isacoff, UC Berkeley professor of molecular and cell biology.
The work was supported by a grant from Fight-for-Sight and an award from Lawrence Berkeley National Laboratory.