Shining light on human emotion

Two billion years ago, on an Earth thin in oxygen and soaked in sunlight, a mitochondrial ancestor developed a protein to channel light into energy. When the Earth eventually filled its atmosphere with oxygen, this mitochondrial protein, known as cytochrome c oxidase, found oxygen atoms to be a better currency for making energy than photons. Today, mitochondria exist as supportive energy-producing organelles in every animal cell, including human brain cells, but they never lost their ability to respond to light.

Now, three scientists are championing the use of light to change, understand, and mimic the human brain. Each in their own way uses photonics technologies to connect us more deeply to our emotional lives.

In the late aughts, for example, Dan Iosifescu, an associate professor of psychiatry at New York University, noticed a problem with the mitochondria in his depressed patients' brain cells. Their brain cells showed abnormal levels of energy-rich molecules called phosphates that mitochondria use instead of sunlight to create adenosine triphosphate, the building block of energy that all cells rely on to live.

Iosifescu didn't know what to do with this information in terms helping his patients, but he knew what healing from a mood disorder looks like in the brain: Neurons build bridges to connect with other neurons, an ATP-intensive process. Could it be that his patients were stuck with depression because their mitochondria weren't providing them with the energy needed to change moods, he wondered?

When Iosifescu read in 2010 that other researchers had successfully induced mouse brain cells in petri dishes to grow and build neuron bridges by exposing them to beams of near-infrared (near-IR) light, he had an idea: Deliver near-IR light into the brains of people with what looks like a modified bicycle helmet, a form of therapy now known as photobiomodulation.

Photobiomodulation delivers IR light to the brain, a possible treatment for depression. Credit: Dan Iosifescu

Unlike other wavelengths of light, IR light can pass through living tissue, and therefore through human skin, blood, and bone, reaching the black box of the brain and its light-sensitive mitochondria.

"You're just using near-infrared radiation to hack into this old system that is not in use anymore," Iosifescu says. Like a programmer's old code, "it's just been left there by evolution."

In Iosifescu's ongoing experiment to treat depression with light, a patient puts on polarized sunglasses to protect their eyes from stray beams of infrared, and then they sit in an armchair next to a console the size of a small briefcase. The console generates the infrared light.

Next, his team equips the patient with a helmet connected to another diode-studded console which transmits the IR light as laser beams. The diodes are wired to probes in the helmet, which Iosifescu's team places on the patient's forehead. Doing so targets the frontal lobe, a brain site associated with emotion regulation. When the team turns the central console on, the patient feels their scalp warm up as the contraption shines infrared light through their forehead to the surface of the brain.

"We're essentially taking these insights about what we think is a problem in the brain when it has to do emotion processing in a very chronic stress condition," Iosifescu says of photobiomodulation therapy. Even though depression is complex and may appear for different reasons in people it affects, he believes delivering light to the brain has potential to treat the disorder. "If we can improve [the patient's] energy production and mitochondria functioning, maybe we can help them get better, even if we cannot address the original problem."

But Iosifescu cautions that the mitochondrial hypothesis of depression is unproven. "There are situations in medicine and biology where there are some wonderful ideas which just don't work. So, I mean, it's very cool, but we haven't proven this yet." Although his initial studies showed mixed results compared to placebo, by adjusting his equipment's parameters, such as light intensity and length of treatment, he has started to see a variety of antidepressant effects in more recent studies.

The real attraction of photobiomodulation, however, lies in its accessibility and lack of side-effects, making it safe and easy to use.

"Because lasers are not a very novel technology, this is something that could be scaled quite rapidly if we really believe that this is working," Iosifescu says. "Should we really prove that this is something that helps, there's a very good chance of this eventually becoming something people can do mostly at home." The fact that patients would be able to self-administer an antidepressant treatment at home by wearing a bicycle helmet for half an hour a few times a week would be a victory for psychiatry.

"Ultimately the field is a bit skeptical about these technologies," Iosifescu adds. "I think the only way we're going to answer to this deserved skepticism is to do high-quality studies."

Susan Perlman, an associate professor of psychiatry at Washington University in St. Louis, is doing high-quality studies with near-IR light to uncover the secrets of how our emotions evolve with age. Like Iosifescu, she uses a portable wearable cap, but instead of changing the brain, she only wants to understand how it works.

While studying three-to-six-year-old children in 2013, Perlman began testing out a new method of brain imaging called functional near-infrared spectroscopy (fNIRS). Similar to Iosifescu's photobiomodulation therapy, NIRS equipment features beams of near-infrared light delivered to the brain through a wearable cap, but the cap also includes detectors that catch near-IR light bouncing back from the brain. The ‘functional' in fNIRS refers to the substance that signals brain function: blood flow, specifically hemoglobin, which absorbs near-IR light. fNIRS is "telling you where the blood flow is by sending light in and measuring what comes back out," Perlman says.

In her lab, a room decorated in a science theme for children to feel comfortable, she had an idea. She developed a virtual game where the player had to click and grab digital bones away from a dog before the dog got them first. However, she rigged the game so that when the player proudly accumulated a large hoard of bones, the dog got faster, making the game harder to win. She invited children to participate in an experiment.

While the NIRS cap recorded the brain activity of children experiencing what it's like to lose a game they thought they were really good at, their parents recorded how frustrated their kid seemed in everyday life in a questionnaire. Perlman also timestamped whether a child was winning or losing the game at any given moment. Later, Perlman took the kids' NIRS data and paired it with the parents' questionnaire about their child's frustration. The method allowed her to look in the kids' brain data for a biological signature of their emotional moments.

But her work doesn't only use near-IR light to illuminate what emotion looks like in the brain. Like other researchers who use older brain imaging modalities, she uses the findings from her fNIRS equipment to help predict who is going to develop depression or psychosis later in life.

Perlman and researchers before her who imaged the brains of children and tracked them over their lifetime found that kids who experience high levels of frustration-throwing tantrums, having a hard time being flexible when things don't go their way-are at a greater risk for later mental health disorders like depression.

"We are finding, using techniques like fNIRS, that we can see an early brain basis of that," Perlman says. "We can see why they're so irritable in the brain, we can see circuitry that's functioning differently in some children than others."

While scientists like Perlman are using infrared light to divine the emotional futures of children, photonics entrepreneurs like Sam Guilaume are creating devices to usher in a future inspired by the biology of the human brain. He is furthering the quest to connect us more deeply with our emotional lives through photonics technologies that mimic the anatomy of smell.

Unlike most sensory neurons in the body which connect to the brain through the spinal cord, the human nose has neurons that connect directly to the brain. Smell is a uniquely powerful sense. Guilaume says that for many years in the biosensor community, "quite a few of us have been thinking that the best way to convey emotion is through olfaction." This has biological power behind it: unlike the brain's other sensory hubs, the olfactory bulb connects straight to the limbic system, the emotional center of the brain."

"Very often we get exposed to a specific context or a specific situation and the first thing that speaks is ‘ah I remember the smell. It reminds me when I was a kid, and I was drinking my cocoa; or I was by the ocean; or it really smells like my wife or my kid or my mother.' This is what really triggers emotions." Smells can elicit a sequence of powerful memories, like going through a treasured photo album in your mind.

But while vision has the camera to make memories shareable and objective, smell doesn't; or at least didn't, until researchers like Guilaume and his team at Aryballe dreamed up a digital nose. Aryballe's NeOse Advance uses silicon chips that capture and then detect the optical index of olfactive molecules to make fingerprints of scents.

In other words, the digital nose takes snapshots of the odors that make up our emotional memories, creating for odor what family photo albums are for vision.

We each have about 400 types of olfactory receptors in our nose, all of which combine to allow us to smell at least 10,000 unique odors. These receptors are connected directly to the olfactory bulb in our brains. When this bulb receives signals from the olfactory receptors, our brain builds an image of that unique smell. Our brain then compares this unique smell with whatever it's smelled in the past, taking note of its differences to determine its newness.

"What your brain does is that it will compare this [odor] signature with what you have recorded in the past," Guilaume says. "Ah! That's my grandmother's garden. Or that's cheese. Or I don't know what that is, but it smells woody and leathery. So essentially your brain is looking for a match into your olfactive memory."

Guilaume's NeOse Advance is 3 millimeters wide and has 64 biosensors inspired by the olfactive biosensors in our nose. "We combine these different biosensors into one picture," Guilaume says.

The human nose's smell receptors can detect volatile olfactive compounds (VOCs), odorous compounds emitted by common household objects. The NeOse uses an array of silicon interferometers that contain peptides mimicking the smell receptors in our nose. VOCs that pass by peptides bind to them, and each VOC-peptide bond is unique in its strength. The interferometers that contain the VOC-bound peptides shine a light that is unique to the strength of the bond, and a tiny camera within the NeOse captures the whole array as a pattern of 64 bright and dim dots. This pattern of dots is an odor's signature.

A NeOse Advance ‘smelling' a variety of scent samples in small containers. Credit: Aryballe

"And then we compare this to the olfactive library" of what the NeOse has smelled before, Guilaume says. When hooked up to a digital device like a computer, the NeOse can show the user if they have a match to what it has smelled previously. If it has no specific match, it can even relay to the user its proximity to similar scents it's picked up, such as if the smell is woody or leathery.

For some people, leathery smells are emotionally important; so important, in fact, that it's a crucial cog in a multibillion-dollar business. As certain industries like the automotive industry move green, they're losing smells that customers emotionally associate with their products.

"I'm BMW, for instance, and I want my car to smell BMW, right?" Guilaume says. However, with every car model's signature mesh of plastic, petroleum, metal, leather, and cloth, there are so many materials that coalesce to exude its familiar yet hard-to-articulate smell. This challenge makes it hard to please customers reliably. For instance, according to Guilaume, Europeans might be more likely to enjoy a leathery scent in their BMW, whereas in Asia, they might dislike a leathery scent. This poses a challenge for BMW workers who are tasked to create a car smell people will like. "It's a freaking nightmare for these guys because we have too many requirements for a specific smell," he says.

But the NeOse Advance's odor snapshot gives them an objective datapoint with which car scent testers can determine if they're getting the smell that their audience requests.

That satisfaction for the audience is what is most important. Guilaume says the NeOse Advance is "not about the content of the molecules, but it's more about the feeling it will generate into us humans."Photonics technologies that heal, probe, or are inspired by the brain can touch their creators, too.

"I'm not afraid today to associate two worlds, metrology and olfaction," Guilaume says. "And for a while, I thought it was just a fantasy. Now I understand that we can really claim that Aryballe provides a device that can do olfactory metrology. And I think that's a really significant breakthrough."

"We're cracking the nut open, eventually, at last."

Andrew Meissen is a science writer in New York City.