I first learned how our eyes work in a college neuroscience class in the fall of 2002. My textbook showed colorful cartoons of the retina, lined with two types of cells that convert light waves into the electrical currency of the brain. There were rod cells, which are useful in low-light situations, and there were cone cells, which allow us to see colors. Neat and tidy.

Unbeknownst to me (because the information hadn’t made it into textbooks yet) neuroscientists had just discovered a third class of light-sensitive eye cells, called, rather uncreatively, ‘intrinsically photosensitive retinal ganglion cells’ (ipRGCs). These cells have nothing to do with vision. They absorb light in order to properly set our circadian clock.

In the decade since then, many studies have suggested that ipRGCs do even more than clock-setting. The latest of these reports, out this week in the Proceedings of the National Academy of Sciences, finds that these cells — and the light-sensitive protein they contain, melanopsin — are involved in short-term memory.

“To date it’s the most convincing data for a role of melanopsin in cognition,” says lead investigator Gilles Vandewalle, FNRS research associate at the University of Liège, in Belgium.

Others aren’t so sure about that, though, largely because of the many mysteries surrounding how melanopsin works.

Opsins are proteins that change their shape in response to light of certain wavelengths. Rhodopsin, found in rod cells, responds most strongly to light in the blue-green spectrum. Three different opsins reside in the cone cells, and preferentially respond to blue, green, and red light, respectively.

Melanopsin responds to blue light, but its activity is more complicated than that. It has two biochemical states. When activated by blue light, the protein switches into a configuration that allows it to produce an electrical signal. After that, orange-red light switches it back into a state that’s sensitive to blue.

Several years ago, Vandewalle’s team performed several brain-imaging experiments supporting the idea that melanopsin helps drive cognition. The researchers scanned volunteers’ brains while they saw either blue light or green light during a working memory task. Blue light drove more brain activity in memory-related regions, suggesting that melanopsin and the ipRGCs were responsible. Still, it was indirect evidence because there was no way to determine how much of the response was coming from the rods and cones.

The new study tries to tease apart the role of different cells by taking advantage of melanopsin’s two states. The researchers scanned the brains of 16 people while they performed an auditory working memory task (they would hear strings of letters and were asked to say whether the most recent letter matched the one that came three letters before). While in the scanner, participants saw 10 minutes of either blue, green, or orange light. Then they rested in darkness outside of the scanner for 70 minutes. Then they went back into the scanner and performed the same memory test while looking at green light.

The participants who were first exposed to orange light had stronger brain responses later in the prefrontal cortex and the pulvinar, which is part of the thalamus, compared with those who were first exposed to blue light, the study found. Because melanopsin is primed by orange light, these results strongly suggest that melanopsin is driving the increased cognitive response, the researchers say.

“We know that melanopsin and rods and cones, all three types of photoreceptors, probably participate in this enhancement of cognitive abilities,” says Howard Cooper, head of chronobiology at INSERM, who collaborated with Vandewalle on the study. “But this study shows that melanopsin is a predominant player.”

Other experts, though, aren’t quite ready to accept this explanation, because it’s unclear what happened to participants’ cells during the 70 minutes of downtime. “What is happening to the melanopsin when it is activated by light, and then the whole system is plunged into darkness?” asks David Berson, a professor of neuroscience at Brown University whose team discovered, back in 2002, that ipRGCs are sensitive to light. It could be that after an hour in darkness, the amount of melanopsin ready to respond to light is the same whether the subjects were previously exposed to orange or blue light, he says.

Unfortunately this type of mechanistic question can’t be investigated in people. That’s why most work on these cells over the past decade has been limited to animal models. “You have to admire what these guys have done because they’re trying to bring this story into the realm of human health, disease, and cognitive function,” Berson says. “I think that’s wonderful. It’s just that when you move into a human study, it’s harder to be confident of what exactly is going on.”

All of these experts agree on one thing: light, and particularly natural light, is good for our waking lives. “Light changes throughout the day, and as it changes, it changes us,” Cooper says. Most people prefer a sunny day to a gray one, and some people even experience seasonal depression in the winter.

Indoor lights, however, are typically yellow or white. “We’re all indoors so much, with artificial lighting that may not contain enough blue light,” Vandewalle says. This is especially important for school and office settings, he says. “We’re depriving our cells from the natural blue light we would be exposed to if we were outside as human beings used to be.”

Vandewalle uses blue-light therapy in the mornings and his office lights are always on. “It always makes my colleagues laugh,” he says. “They say it’s only the chronobiologists who keep the lights on during the day.”

Update, 3/13/14, 10:13am: Two errors have been corrected in this post related to the location of the pulvinar and the nature of the memory task.