Researchers have indicated that, similar to declarative memories, the short-term and long-term memories associated with motor skills develop in different areas of the brain. The cerebellum plays a vital role in the creation of long-term skill memories.
Do you recall your second-grade teacher’s name or what you had for lunch today? Although these memories could be decades apart, both are classified as long-term memories.
Over 50 years ago, neuroscientists discovered that injury to an area of the brain known as the medial temporal lobe (MTL) led to significant difficulties with long-term declarative memory — which involves explicit information like names and dates — while extremely short-term memory remained unaffected. Individuals with MTL damage could engage in a brief conversation but would quickly forget it just a minute or two later.
Interestingly, these individuals were capable of acquiring new motor skills and retaining them for an extended period, whether that was days, months, or even longer, demonstrating that MTL damage did not significantly impact motor skill memories.
This raises the question: which area of the brain is responsible for long-term memories related to motor skills, such as riding a bicycle? Are there specific regions that handle short- and long-term sensorimotor memory distinctly? Researchers have been exploring these inquiries for many years.
Recently, a team from the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) has revealed that, much like declarative memories, both short-term and long-term memories for motor skills are processed in different regions of the brain, with the cerebellum being essential for developing long-term skill memories.
The findings appear in the Proceedings of the National Academy of Sciences (PNAS).
“This research enhances our understanding of the cerebellum’s role in sensorimotor learning and suggests that the cerebellum serves as a gateway for forming stable memories related to sensorimotor skills, largely functioning independently of short-term memory systems,” explained Maurice Smith, Gordon McKay Professor of Bioengineering at SEAS and the study’s senior author.
While it has long been recognized that the cerebellum is crucial for motor learning, the specifics of its involvement in creating short- and long-term skill memories was less clear. To clarify this connection, Smith and lead author Alkis Hadjiosif, a postdoctoral fellow at SEAS and Massachusetts General Hospital, drew from an array of previous studies that suggested varying levels of motor learning difficulties in patients with cerebellar injuries.
Despite discovering that individuals with cerebellar damage consistently showed signs of hindered sensorimotor learning, the degree of these effects varied considerably across studies.
“These discrepancies could stem from differences in the extent or specific areas of brain damage or variations in the types of motor tasks used,” stated Smith.
However, Smith and Hadjiosif posited that minor differences in time intervals between trials — what they termed the memory window — might account for a significant portion of these observed discrepancies.
“This would hold true if long-term sensorimotor memory were specifically hampered by cerebellar damage since longer memory windows would necessitate greater reliance on the diminished long-term memory,” Hadjiosif noted.
The challenge was that the timing intervals were rarely detailed in prior research publications. Acting as part researchers and part detectives, Smith and Hadjiosif tracked down comprehensive raw data from two of these studies to analyze intertrial intervals across all participant sessions.
The researchers discovered that both studies had overall short intertrial intervals and reported only minimal impairments in learning for those with severe cerebellar issues compared to healthy individuals. This indicated that when participants repeated a task, such as five times with only a few seconds between repetitions, those suffering from cerebellar degeneration showed only slight decline in performance.
However, by delving deeper into the data, Smith and Hadjiosif stumbled upon an intriguing finding. Sometimes, between trials, there was increased time allowing either the research team to reset or the participants to briefly rest.
“When we analyzed these differences from trial to trial, we found that patients who performed nearly normally on short-interval tasks showed significant impairment on long-interval tasks within the same session. This finding was consistent in the data from both studies,” Hadjiosif explained.
The team then examined over a dozen additional studies where individuals with cerebellar degeneration engaged in motor tasks. They found that studies utilizing a greater variety of movement directions — which would increase the time between trials of the same direction that would rely on shared sensorimotor memory — exhibited considerable memory impairments compared to those tasks with fewer movement directions.
“These results emphasize the importance of timing in understanding memory loss among patients with cerebellar degeneration and resolve the puzzle of variability in the effects of cerebellar damage on sensorimotor learning abilities from trial to trial and study to study,” Smith concluded. “Although our research typically involves designing new experiments to generate innovative datasets that enhance our understanding of learning and memory, sometimes reevaluating existing data through a different perspective can be even more enlightening.”
The research was co-authored by Tricia Gibo.
