Feature Story Fall 2014 Issue 33

In Search of Smell’s Meaning

Laboratory technician Nora Benavidez trains a mouse to associate an odor with a reward. Photo: Justin Knight
Laboratory technician Nora Benavidez trains a mouse to associate an odor with a reward. Photo: Justin Knight

A few years ago, as Gloria Choi and her husband were finishing their postdoctoral fellowships at Columbia University—she in neuroscience, he in immunology—they decided to try an experiment together. Their idea was to explore the possibility, suggested by observations in humans, that an inflammatory immune response during pregnancy might be a risk factor for autism in later life.

The project was a diversion from Choi’s primary expertise in olfaction, but her postdoctoral work was near completion and she had plans to join MIT as a McGovern Investigator and assistant professor of brain and cognitive sciences in 2013. With little to lose, Choi thought, “Why not try it?”

Now, in her own lab, she is studying mice as a model for understanding complex human social behaviors, including autism. She continues to focus on olfaction, since mouse behavior is strongly driven by the sense of smell. But her larger goal is to understand the brain changes that underlie learning, and in particular to understand how an initially neutral sensory stimulus can acquire meaning, including social meaning, that guides behavior.

Building a Bench

When Choi first arrived at MIT in 2013, her laboratory space was still under construction, and she began work using borrowed bench space and a few items of equipment that she had brought with her from Columbia University. She was accompanied by Nora Benavidez, a technician with whom she had worked in the laboratory of Nobel laureate Richard Axel. As soon as the renovations were completed, Choi and Benavidez set about transforming the new space into a working laboratory, filling it with boxes containing everything from animal cages to a coffee machine. “It was so intense, but fun, actually, to see what it takes to put a lab together from the ground up,” says Benavidez.

Circuit Breaker

A mouse inside a box where it learns to associate an odor with a water reward. Photo: Justin Knight

A mouse inside a box where it learns to associate an odor with a water reward. Photo: Justin Knight

Inside the lab, Benavidez, who plans eventually to go to graduate school, is training a mouse to associate an odor with a reward. The training session is part of her larger project to map the brain circuits that control olfactory learning. Some odors have innate meaning for mice—for instance, they instinctively avoid the smell of fox feces—but like humans, they can also form associative memories for odors that have no intrinsic meaning, such as orange, anise, ethyl acetate (nail polish remover), and octanol, a liquid that is Benavidez’s favorite.

Benavidez places the mouse inside a shoebox-sized plexiglass box mounted to the wall, and puffs in the odor—in this case, citronella—while simultaneously providing a reward of water. The pairing lasts a few seconds until an exhaust tube removes the odor. The next puff brings a different smell—this time, it’s nail polish remover—but without the water. Over time, the mouse will learn to seek water when presented with citronella.

To understand how the brain learns these associations, Benavidez is using a technique called optogenetics that allows her to stimulate or silence different parts of the brain using light. She is especially interested in a structure called the piriform cortex that is thought to be the brain’s olfactory processing center. As a postdoc, Choi had shown that stimulating this structure could mimic the effect of odors to produce learned behaviors, and Benavidez is trying to understand how this happens. She flips a switch, activating the lasers that will silence some neurons while stimulating others, and waits to observe the effect. The mouse shows some difficulty in responding, which could be an interesting result. But Benavidez suspects there may be limits to her optogenetic technique, so she and Choi have decided to try a different method, in which they will use a drug to silence the piriform cortex. “I’ll be doing the first drug injection tomorrow,” says Benavidez, her excitement visible, even though it will be just the first of many trials to come.

Building a Platform

Around the corner from Benavidez, postdoc Han Kyoung Choe works quietly. Like Choi, he is from Korea, and he was the second person to join her lab, arriving at MIT in early 2014 after completing his graduate studies at Seoul National University. “Han grounded us,” says Benavidez. “His arrival was a signal to us. This is really happening.”

Choe is studying how the brain learns to associate odors with social, and in particular, sexual cues. He places a male mouse in the central compartment of amaze with three chambers. On one side, separated by a barrier, is a female in heat, accompanied by the smell of octanol. On the other side, Choe presents a different odor. Over time, males in this maze will learn to associate the smell of octanol with the possibility of a new mate.

Graduate student Michael Reed adjusts a plexiglass mouse maze as postdoc Han Kyoung Choe looks on. Photo: Justin Knight

Graduate student Michael Reed adjusts a plexiglass mouse maze as postdoc Han Kyoung Choe looks on. Photo: Justin Knight

The design of the maze emerged by trial and error. The original prototype was broken during shipping to MIT, so Choe had to construct a new one. He used black material to prevent the male from seeing the female, but then he also couldn’t see the mice. With the help of Choi’s first graduate student, Michael Reed, Choe built a new maze, this time using transparent red plexiglass, since mice do not see well in red light. Then some males started jumping over the walls in search of the female that they could smell but not see. “They can be very determined,” says Choe. “We had to raise the walls.”

Once the maze was working as intended, Choe and Reed added a computerized system to automate their experiments, instead of running them manually. The system tracks the male mouse’s movements by video, controls the release of odors, and vacuums them away at the end of every trial. It can even control the mouse’s brain activity, via optogenetic lasers. “The first time I watched it, I was so happy,” says Reed. “It just makes everything so easy.”

Reed is now beginning to design his own experiments, and turns frequently to Choe for advice. Choe in turn is learning to be a mentor, but he is quick to point out that Reed also contributes many ideas to the project. It’s a relationship built on mutual respect, common goals, and also kindness, something endemic in Choi’s lab.

For Choi, their interaction represents a huge milestone for her laboratory. “I’m not alone,” she says. “People are bringing in new techniques and new ideas. They are teaching me and teaching each other. I hope this collaborative spirit will permeate the lab for years to come.”

Building a Community

Choi says that a mentor once told her that a principal investigator must fulfill two roles: a leader with scientific vision, and a manager who builds a lab community. Choi deftly blends both through a weekly ritual. Every Monday, she meets with each of her laboratory members individually to review progress. The meetings can be long and demanding, but afterwards, the entire team goes out together for lunch. “We all trek across the road and get huge burgers and come back here and eat them,” says Benavidez. “We joke that it’s a form of reward conditioning.” For Choi, sharing meals as a group comes naturally. “Families make time to share meals,” she says, “and the lab is almost like extended family.”

The Choi lab enjoys lunch together after one of their weekly lab meetings. Photo: Justin Knight

The Choi lab enjoys lunch together after one of their weekly lab meetings. Photo: Justin Knight

Meanwhile, Choi has her own family to juggle. Her husband, now starting his own lab at the University of Massachusetts Medical Center in Worcester, has a long daily commute. Her daughter, a toddler, comes with her to MIT for daycare most days, and is often seen playing in Choi’s office as the workday winds down. The long days can be taxing, but Choi takes satisfaction from seeing her research agenda building momentum. She recently hired a second postdoc, Yeong Shin Yim, who is picking up on the possible connection between inflammation during pregnancy and autism. Yim is studying similar effects in mice, in collaboration with Choi’s husband Jun Huh and his former postdoctoral mentor, Dan Littman at New York University. “Our goal is to understand how infl ammation and immune signaling affects the developing brain,” says Choi. “Knowing that could help us to figure out which brain regions we need to target for future therapeutics.”

Choi knows this is still a distant goal, but she describes her attitude as one of cautious optimism. As lab head she must instill enthusiasm in her younger colleagues while tempering it with the knowledge that research is a slow business and that it can take years for a project to bear fruit. As the day’s experiments wind down, Choi retreats to her office. There is grant writing to be done, and results to be analyzed. Later on, there might be a break for conversation, or for few minutes of practice on the cello that she keeps near her desk. Then, if everyone is working late, they may order take-out together, perhaps with a glass of wine, the aroma calming, and a subtle reminder of their aspirations.

Share This Print This