Summary: Researchers say resolving a key debate about how multiple items are held and evaluated in working memory is essential to understanding how the brain maintains its “sketchpad of conscious thought.”
Source: MIT
Researchers debate the mechanisms behind working memory—specifically whether the prefrontal cortex sustains information through continuous neural firing or through brief, coordinated bursts that interact with synaptic changes—in two papers published in the Aug. 8 issue of the Journal of Neuroscience.
Working memory enables us to hold and manipulate information for short periods—like remembering directions, keeping a waiter’s list of specials in mind, or juggling several ideas at once. Because working memory capacity correlates strongly with intelligence and because deficits in this cognitive system are linked to psychiatric conditions such as schizophrenia and autism, gaining a clear mechanistic understanding is a priority for both basic neuroscience and clinical research, says Mikael Lundqvist, a postdoctoral researcher at MIT’s Picower Institute for Learning and Memory and the lead author of one of the papers.
“Working memory deficits are associated with virtually every major psychiatric disorder,” says corresponding author and Picower Professor Earl Miller. “If we can figure out how working memory works, we can figure out how to fix it. Working memory is the sketchpad of consciousness—understanding it advances our knowledge of the conscious mind.”
The opposing “Dual Perspectives” piece is led by Christos Constantinidis of Wake Forest School of Medicine.
Core disagreement: persistent firing versus brief bursts
The central question is what neuronal activity in the prefrontal cortex looks like during the delay period: after you perceive something and must hold it in mind for later use. One view holds that neurons maintain memories through persistent, sustained firing—akin to an engine idling. The alternative view proposes that neurons fire in brief, coordinated bursts, while transient changes in synaptic strengths store the information between those bursts, a process more similar to mechanisms associated with longer-term memory.
In their essay, Lundqvist, Miller and Pawel Herman (KTH Royal Institute of Technology) argue for the burst-and-synapse model. They note that recent experiments that record many neurons simultaneously on a trial-by-trial basis reveal short, coordinated spiking events across neural populations. These bursts, coordinated with large-scale brain rhythms, can explain how the brain efficiently and flexibly controls multiple items in working memory with precise timing.
Importantly, Miller emphasizes that this view does not dismiss the role of spiking in working memory. Instead, it reframes how spiking contributes: spikes occur in brief bursts that drive temporary synaptic changes, which together maintain information. “We’re showing additional mechanisms by which spiking maintains working memory and gives volitional control,” Miller says. “Our work adds support to the idea that delay-period spiking plays a role, while revealing additional, more granular dynamics.”
Much of the disagreement traces back to differences in data collection and analysis. Earlier studies supporting persistent activity often averaged firing rates across trials and across small neuronal samples, smoothing over transient dynamics. By contrast, modern recordings that capture many neurons per trial reveal population-level, brief bursts of activity that are coordinated across the circuit.
Functionally, brief bursts coordinated with synaptic storage make strong computational sense. Such dynamics use less energy than continuous firing, help prevent interference when multiple items are held in mind (different items can be represented by bursts at different times), and make stored information more resilient to distraction by transferring transient representations into synaptic patterns between bursts.
“A hybrid strategy of spiking plus synaptic storage gives the brain flexibility,” Lundqvist explains. “It allows multiple memories to be activated or held temporarily without mutual interference, and synapses can maintain information while spiking processes other tasks—helping working memory withstand distractions.”
Where the field should go next
Both authors agree that new types of experiments are needed to settle the debate and refine models of working memory. Lundqvist says the field is at a productive point where emerging data and recording technologies can resolve outstanding questions.
The MIT team recommends several best practices for future experiments: record activity from large populations of individual neurons, analyze each trial separately rather than relying on long-term averages, design tasks that require animals to control multiple items simultaneously, and measure neural oscillations as well as spike timing. These approaches will clarify whether working memory is best characterized by persistent firing, brief coordinated bursts with synaptic storage, or a combination of mechanisms—and will advance our understanding of the neural basis of cognition.
Funding: The Miller lab’s research on working memory is funded by the National Institutes of Health, the Office of Naval Research, and the MIT Picower Institute Innovation Fund.
Source: David Orenstein, MIT. Publisher: Organized by NeuroscienceNews.com. Image Source: NeuroscienceNews.com image is in the public domain. Original Research: The study appears in the Journal of Neuroscience.