The retina, by Santiago Ramón y Cajal. Light enters from the left. The rods & cones on the right; ganglion cells in purple.
The retina, by Santiago Ramón y Cajal. Light enters from the left. The rods & cones on the right; ganglion cells in purple.

There’s a relay race in your eye that allows you to see. It begins when light hits the rod and cone cells in your retina, triggering a cascade of electrical signals. These pass through other types of cell in later layers and eventually to ganglion cells, which convey the signals to the brain.

In many eye diseases, the first stage of the relay—the light-sensing rods and cones—gradually die. The rest of the visual system is intact, but since it can’t respond to light, the world blackens.

Many scientists have tried to solve this problem by creating stand-ins for the lost light sensors. Two groups have developed implantable chips that do the same job: detect light and send out electrical currents. A US firm is using stem cells to regenerate the dying rods and cones, so they can be transplanted back into the eye. (More on these techniques here)

But Richard Kramer from the University of California, Berkeley, is working on a different solution: he is turning the surviving cells into light sensors. He is working with what’s there, rather than replacing what’s not.

In 2012, his team dosed retinal ganglion cells with a molecule called AAQ and found that they started responding to light. A promising start, but AAQ has two major drawbacks. It gets cleared from the eye within a day, so if it were ever used in humans, it would mean daily injections. And it needs ultraviolet light to work, which not only damages the eye but is filtered out by the lens.

Now, Kramer’s team, headed by Ivan Tochitsky, have found a better alternative. They screened a large library of chemicals for compounds that would do the same thing as AAQ, but without the drawbacks. They found one: a molecule called DENAQ.

When the team injected DENAQ into young mice, which had lost nearly all their rods and cones, it turned their ganglion cells into light detectors and allowed the totally blind mice to sense light again. They explored their arenas when they saw flashes of light. They could be trained to associate those flashes with an electric shock.

DENAQ’s targets are HCN channels—pore-like proteins that are found on the surfaces of ganglion cells and other neurons. DENAQ actually blocks these channels but it shifts when light shines upon it, allowing the channels to open and the ganglion cells to fire.

Best of all, DENAQ doesn’t do anything in normal working retinas. It only bestows light sensitivity on those that have lost rods and cones. In other words, it doesn’t try to fix what ain’t broke. That’s interesting in itself. It suggests that the loss of rods and cones changes the ganglion cells in ways that we still don’t really understand. Perhaps they start making more of the HCN channels? “Totally by accident, our DENAQ molecule is taking advantages of some of these changes,” says Kramer.

DENAQ ticks a lot of other boxes too. It lasts for days in the eye. The remodelled cells also work best with white light, and they need just 1 percent of the intensity that AAQ-transformed cells did. “It’s potent enough to be a really serious candidate for use in humans and we’re starting to do some preclinical research,” says Kramer. He plans on working up to rats, pigs or monkeys to see if the molecule is safe and effective.

He says that this approach has some advantages over alternatives like retinal implants. For a start, it involves no surgery and no foreign material being implanted in the delicate eye. You’d need regular injections, but some drugs for eye diseases are already administered in this way. The molecule’s fleeting nature might even be a good thing, compared to less reversible options like implants or gene therapy. “Until you know that the temporary change is going to work, diving off the cliff with something that will be permanent is pretty risky,” says Kramer.

They are lagging behind though. It is encouraging that scientists are pursuing many different options, but other techniques for restoring vision have already been tested in human clinical trials, and to some success. While Kramer’s blind mice can see light again, there’s no evidence that they can do more important tasks like recognise objects. By contrast, people who have received retinal implants can make out patches of white and dark—enough to read very large letters, see streetlamps, or catch a smile.

Reference: Tochitsky. Polosukhina, Degtyar, Gallerani, Smith, Friedman, Van Gelder, Trauner, Kaufer & Kramer. 2014. Restoring Visual Function to Blind Mice with a Photoswitch that Exploits Electrophysiological Remodeling of Retinal Ganglion Cells. Neuron