Tracing and Tagging Memories: Science Fiction Becomes Science Fact
Toward the end of the February 10, 2016 episode of PBS’s NOVA, The Memory Hackers, Christine A. Denny, PhD, Assistant Professor of Clinical Neurobiology in the Department of Psychiatry at Columbia University Medical Center, places a small mouse gently into a lidless, transparent plastic box with a hard, white plastic floor, surrounded by bright lighting. The mouse huddles in a corner.
Dr. Denny then turns a key on a small box nearby sending a laser light through two narrow fiber-optic tubes attached to the mouse’s skull, aimed at a few hundred cells in its brain. The day before, these exact neurons had been tagged with the mouse’s experiences – as it was having them – of soft, earthy bedding, dark surroundings, and a plastic rock cave to hide under. The mouse’s perception in the plastic box is now rife with this specific stimulated memory. The mouse moves from the corner, sniffing and exploring. He grooms himself, a clear sign of feeling safe, and peruses the entire landscape of this moments-ago adverse environment. When Dr. Denny turns off the laser input, the mouse retreats to the corner. When she turns the memory back on, he ventures forth again. With control of the laser light, she can turn his particular memory on and off.
Dr. Denny compares this startling demonstration, including her own initial sense of surprise, to the science fiction of movies like Eternal Sunshine of the Spotless Mind or Inception. “The first time we did it,” she tells NOVA, “we didn’t believe it. But when you see inside of the brains of these mice,” she says, as NOVA shows images of the specifically tagged neurons in phosphorescent yellow-green on Dr. Denny’s computer screen, “and then to think that you are only manipulating those cells and changing the behavioral output of the animal, that’s science fiction.” Dr. Denny knows, however, this remarkable tagging of the neurons of a specific memory is very much science fact, and predicated on the work of more than 10 years at Columbia.
Memory Defining Questions
Dr. Denny is interested in the molecular mechanisms underlying learning and memory and she and her research team have developed a novel technique to label the cells that encode individual memories in the brains of mice. They are then able to indelibly tag these neurons using fluorescent molecules.
“In humans, we can look at brain regions of activity using fMRI or a number of methods; in an animal model you can look at activity based on gene expression,” she explains. “While this can tell us what cells are active when you learn or remember something, there wasn’t a tagging system available where you could permanently label one memory, then wait for the lifetime to occur or disease to set in, and then look at what happened to that memory. Therefore, we created these mice to allow for an indelible tag of one particular memory. We look at cells that are active during learning and cells that are active during remembering information learned. And then we look for overlap. And if that cell is active at both time points, we call that a memory trace or an engram.”
Dr. Denny’s lab was the first to inhibit the specific cells that are activated when something aversive is learned. “When we put the animals back into the aversive context, we could shut off the cells of very small subsets – a few hundred cells or less – and we could block retrieval of that fear memory.”
An article published in 2014 in Neuron, for which Dr. Denny is lead author, examines how the permanence of memory traces or engrams permits remarkably specific location of memory in the hippocampus (HPC) and begins to examine the vicissitudes of experience and stress on the development of the network of neurons involved in learning and memory. The study utilized the immediate early gene (IEG) Arc to understand how a memory trace is formed and retrieved in the HPC under a number of circumstances. The research team designed the ArcCreERT2 bacterial artificial chromosome (BAC) transgenic mice to test this hypothesis not just on a short timescale, but also indefinitely. “We were able to compare cells activated during the encoding of a memory to cells activated during the expression of that memory,” notes Dr. Denny.
Their hypothesis “that the cells reactivated during expression of memory are a component of memory traces” finds its evidence in the mouse dentate gyrus (DG) and CA3. The researchers showed that, “mice re-exposed to a fear-inducing context froze more and had greater percentage of reactivated cells in the dentate gyrus (DG) and CA3 than mice exposed to a novel context.” They also corroborated that the differences disappeared over time in keeping with the observation that memories become generalized. Additionally, silencing DG or CA3 cells that were recruited during encoding of a fear-inducing context prevented expression of the corresponding memory.
“Arc is definitely in humans,” Dr. Denny says, in connecting her research in mouse models to the clinic. “And it’s been implicated in a number of diseases, like Alzheimer’s and others.” Dr. Denny and her colleagues have created mouse models to investigate what happens to hippocampal memory traces in normal, aged, and Alzheimer’s diseased mice. “Is the problem that you just do not remember as well or is the memory there and not being accessed correctly? When you have a tagging system, you can answer those types of questions. Without the tagging system, you can basically only determine global brain activity at any given time point, but not whether the cells were active at an earlier time point.
“A model that is permanent, that lets you have a tag forever, can start to get at these questions that occur with disease onset,” adds Dr. Denny. “The nice thing about this mouse model is you can breed it with any other disease model to study how that disease impacts memory. This enables you to visualize what is going on before and after disease progression and then to manipulate it to answer very specific questions. We hope to identify the altered memory circuits and how to manipulate them in order to improve memory retrieval seen during cognitive aging.”
Most recently, Dr. Denny and her colleague, René Hen, PhD, Professor of Pharmacology (in Psychiatry) at Columbia University Medical Center and Director, Division of Integrative Neuroscience, New York State Psychiatric Institute, have received a substantial grant to perform whole-brain imaging. “Instead of looking at just individual parts of the brain to identify which cells are part of a memory trace, we can now image the entire brain and get a memory trace map across all circuits of the brain,” says Dr. Denny. “I’m a clinically inspired scientist. My philosophy is that you use mouse models to gain insight or information on disease processes on a level that wasn’t previously possible before. And then use that information to better design future studies that could impact disease progress or understand how it occurs in the first place. If you do that, you really can make an impact on bringing your work from pre-clinical models back into the clinic.”
Cazzulino AS, Martinez R, Tomm NK, Denny CA. Improved specificity of hippocampal memory trace labeling. Hippocampus. 2016 Jun;26(6):752-62.
Denny CA, Kheirbek MA, Alba EL, Tanaka KF, Brachman RA, Laughman KB, Tomm NK, Turi GF, Losonczy A, Hen R. Hippocampal memory traces are differentially modulated by experience, time, and adult neurogenesis. Neuron. 2014 Jul 2;83(1):189-201.