How is Memory Stored Within the Brain?

Ryan started out as an evolutionary geneticist and quickly became fascinated by the idea that the consolidation of long-term memories requires a new gene activation, while short-term memories don’t.

“This is why I became involved in exploring memory, but I’m now convinced gene expression patterns aren’t storing memories but, rather, are changing accessibility patterns and probably compensating for plasticity by contributing to homeostasis within the brain,” he says.

Recently, researchers gained the ability to “hitchhike onto particular components of memory engram cells, which are the physical substrates of how we think we’re encoding memory,” Ryan says.

Memory engrams refer to specific changes occurring within the brain that account for learned pieces of information.

“Your memory engrams must be able to last as long as you, because some memories—even mundane ones—can clearly last an entire lifetime,” Ryan points out. “So the mechanism of memory storage must be stable. But at the same time, we have this other side of memory always being destroyed, lost, or degraded. We forget stuff all the time and many memory disorders exist.”

Memory is often described as “fragile” or “long lasting,” which are contradictions, so how can we make sense of that?

Ryan has always been puzzled by memory being represented in both of those terms, but thinks part of the reason science represented them that way is because it was before the technology to really explore memory existed.

In 2012, the Tonegawa Lab at MIT, where Ryan was a postdoctoral researcher, developed memory engram labelling technology to allow researchers to tag and manipulate a mouse’s specific brain cells.

“By combining optogenetics—which uses laser light to control brain cells genetically engineered to respond to specific wavelengths—and immediate early gene labelling technology, we were able to switch on specific memory engrams in a mouse,” Ryan says. “In one experiment, we created memory loss, or amnesia, in a mouse. Then we stimulated engram cells and discovered we could retrieve lost memories from memory loss. This was a point of significant progress in my own thinking about memory.”

It showed Ryan that the two contradictory aspects of memory can be reconciled. “We could explain why memory is both fragile and lasts forever, because not all memory loss is about losing the memory,” he explains. “Memory loss can be about losing access to the memory, but the memory itself survives and you can stimulate it optogenetically, and there may be natural ways to reverse that access.”

Ryan’s lab began to reverse that access and realized accessibility to a memory may be differentially modulated by experience—which is quite different than the memory itself.

“Novel methodologies led to experiments that open doors to new ways of thinking about memory and forgetting as an adaptive process,” he says. “If it lasts your entire lifetime, it must be stored by a durable mechanism.”

He thinks it’s probably stored by changing the wiring of the brain itself, changing connectivity patterns, which enable long-term information storage. “This means we may be storing memory within microanatomical patterns rather than just the strength of synapses,” Ryan says. “If true, we can consider memory to be the same as instinct.”

Instincts are genetically wired and stored within the brain as hardwired anatomical pathways. If memory and instinct are stored in the same way, memory can influence the evolution of instinct.

“Because we can test drive different brain states within a population with different memories, if one particular brain state works—say a particular cultural idea—then it becomes common because everyone within the population learns it,” he elaborates. “Then it persists across generations and potentially we can create an environment where it’s more likely an instinct that mimics this memory would evolve and be selected for.”

In a sense, the brain may not know the difference between a memory and an instinct. “They may be essentially the same thing, just derived from different origins,” Ryan says. “This is a hypothesis my lab is currently doing experimental work on.”

By understanding the general components of how long-term memory and instincts (really long-term memories) are stored, “we get closer to the minimal essential components of information storage within the brain,” he adds.

Ryan’s lab is also exploring forgetting. “Forgetting, in general, may be functional and it may be a product of environmental experience,” he says. “It could be a purposely reversible process—the brain is a constantly developing machine that acquires and changes its beliefs about the world. And a lot of that involves differential learning and forgetting information and relearning.”

He likens the brain to an evolving sculpture, rather than a painting. “We shouldn’t think of the brain as a blank slate we paint information on,” he points out. “Rather, we should consider ourselves born with fairly developed sculptures that carry instincts and are in effect memories of ancestral generations of our species and closely related species, which change during our lifetime—it’s always an imperfect, messy, changing sculpture that is responding to the environment so we can survive.”

To understand memory, he adds, it’s important to consider the evolutionary, developmental history of the species and individual, but also the struggle of always being adaptive to our environment.

“Within the next 20 years, the neuroscience community will make a lot of advances in understanding how information is stored within the brain,” Ryan says. “We’re making progress in challenging previous ideas and starting to form new ones.”

He hopes his work will help reveal how evolution wired the brain so it can be reverse-engineered to improve human education and well-being. But he’s also concerned about the potential impact of human-created, brain-inspired artificial intelligence (AI). “I think the greatest output a neuroscientist can have is to inform how AI is created in the future. I’d really like to see my team’s work contribute to these things,” Ryan notes.