Kiah Hardcastle
@kiahhardcastle.bsky.social
neuroscientist and new mom
currently Harvard postdoc, Stanford PhD
curious about how animals learn stuff
currently Harvard postdoc, Stanford PhD
curious about how animals learn stuff
Also, check out this cool write-up in Harvard Gazette on this work!
news.harvard.edu/gazette/stor...
news.harvard.edu/gazette/stor...
Possible clue into movement disorders like Parkinson’s, others — Harvard Gazette
Rodent study suggests different signaling “languages” in parts of brain for learned skills, natural behaviors.
news.harvard.edu
August 12, 2025 at 6:26 PM
Also, check out this cool write-up in Harvard Gazette on this work!
news.harvard.edu/gazette/stor...
news.harvard.edu/gazette/stor...
BIG shout-out to my co-first author Jesse Marshall, who got the ball rolling and collected tons of data, to my labmates who helped me keep the ball rolling, especially when the birth of my kid coincided with experiments, and to my mentor @olveczky.bsky.social, without whom none of this would exist!
a group of people hugging each other in a living room with the words group hug above them
ALT: a group of people hugging each other in a living room with the words group hug above them
media.tenor.com
August 12, 2025 at 6:26 PM
BIG shout-out to my co-first author Jesse Marshall, who got the ball rolling and collected tons of data, to my labmates who helped me keep the ball rolling, especially when the birth of my kid coincided with experiments, and to my mentor @olveczky.bsky.social, without whom none of this would exist!
Another question – what *exactly* triggers this change in the neural code? Does it happen quickly, at the start of a new context that requires learning, or does it emerge slowly as a new skill develops? TBD, but with any luck we will have a better idea soon
August 12, 2025 at 6:26 PM
Another question – what *exactly* triggers this change in the neural code? Does it happen quickly, at the start of a new context that requires learning, or does it emerge slowly as a new skill develops? TBD, but with any luck we will have a better idea soon
There are lots of interesting questions and ideas that come from this. For one, what the heck is DLS activity doing during free exploration?! Why is it so locked to motor output if it’s not required for it? We’re thinking about this now, but lmk if you have any ideas!
August 12, 2025 at 6:26 PM
There are lots of interesting questions and ideas that come from this. For one, what the heck is DLS activity doing during free exploration?! Why is it so locked to motor output if it’s not required for it? We’re thinking about this now, but lmk if you have any ideas!
But then, we compared mapping from neural activity to movement – and saw a big difference!
🤩
While neurons in both cases represented ongoing movement, they did so in radically different ways.
🤩
While neurons in both cases represented ongoing movement, they did so in radically different ways.
August 12, 2025 at 6:26 PM
But then, we compared mapping from neural activity to movement – and saw a big difference!
🤩
While neurons in both cases represented ongoing movement, they did so in radically different ways.
🤩
While neurons in both cases represented ongoing movement, they did so in radically different ways.
What could explain this change in function? We next turned to neural activity, recording from DLS neurons from animals in both contexts.
Maybe neurons were more active in the task, or activity was more locked to behavior?
At first, we saw a whole lotta nothing that screamed “big difference”.
🤨
Maybe neurons were more active in the task, or activity was more locked to behavior?
At first, we saw a whole lotta nothing that screamed “big difference”.
🤨
August 12, 2025 at 6:26 PM
What could explain this change in function? We next turned to neural activity, recording from DLS neurons from animals in both contexts.
Maybe neurons were more active in the task, or activity was more locked to behavior?
At first, we saw a whole lotta nothing that screamed “big difference”.
🤨
Maybe neurons were more active in the task, or activity was more locked to behavior?
At first, we saw a whole lotta nothing that screamed “big difference”.
🤨
But even with the latest and greatest tracking and behavioral quantification, we couldn’t find effects of lesioning DLS 😶🌫️. This is a big contrast to the huge effect in the task. This suggests DLS is not an essential part of the motor control machinery, but specializing in shaping learned behaviors.
August 12, 2025 at 6:26 PM
But even with the latest and greatest tracking and behavioral quantification, we couldn’t find effects of lesioning DLS 😶🌫️. This is a big contrast to the huge effect in the task. This suggests DLS is not an essential part of the motor control machinery, but specializing in shaping learned behaviors.
We first examined DLS function by taking it away (via lesion).
We already know DLS impacts learned movements – but what about innate ones?
Prior work has implicated DLS, but to know for sure, we needed to track these behaviors in detail. And thanks to advances in tracking, this is now easy-peasy!
We already know DLS impacts learned movements – but what about innate ones?
Prior work has implicated DLS, but to know for sure, we needed to track these behaviors in detail. And thanks to advances in tracking, this is now easy-peasy!
August 12, 2025 at 6:26 PM
We first examined DLS function by taking it away (via lesion).
We already know DLS impacts learned movements – but what about innate ones?
Prior work has implicated DLS, but to know for sure, we needed to track these behaviors in detail. And thanks to advances in tracking, this is now easy-peasy!
We already know DLS impacts learned movements – but what about innate ones?
Prior work has implicated DLS, but to know for sure, we needed to track these behaviors in detail. And thanks to advances in tracking, this is now easy-peasy!
To examine this, we considered two behavioral contexts on each end of the learning spectrum:
(1) free exploration, where a rat goes about their business expressing innate behaviors like rearing
(2) task execution, where the rat expresses a sequence of arm movements it learned to solve a lever task
(1) free exploration, where a rat goes about their business expressing innate behaviors like rearing
(2) task execution, where the rat expresses a sequence of arm movements it learned to solve a lever task
August 12, 2025 at 6:26 PM
To examine this, we considered two behavioral contexts on each end of the learning spectrum:
(1) free exploration, where a rat goes about their business expressing innate behaviors like rearing
(2) task execution, where the rat expresses a sequence of arm movements it learned to solve a lever task
(1) free exploration, where a rat goes about their business expressing innate behaviors like rearing
(2) task execution, where the rat expresses a sequence of arm movements it learned to solve a lever task
But… is this just because rats *always* need their DLS for motor output, learned or otherwise?
... or ...
Does it play a special role for *learned* movements, only becoming involved when a movement is tailored to a task?
Basically – does DLS function change with learning?
... or ...
Does it play a special role for *learned* movements, only becoming involved when a movement is tailored to a task?
Basically – does DLS function change with learning?
a woman says quick q while holding a glass of water
ALT: a woman says quick q while holding a glass of water
media.tenor.com
August 12, 2025 at 6:26 PM
But… is this just because rats *always* need their DLS for motor output, learned or otherwise?
... or ...
Does it play a special role for *learned* movements, only becoming involved when a movement is tailored to a task?
Basically – does DLS function change with learning?
... or ...
Does it play a special role for *learned* movements, only becoming involved when a movement is tailored to a task?
Basically – does DLS function change with learning?
We focused on changes within the sensorimotor striatum – also called dorsolateral striatum (DLS) in rats – which is a big input region in the basal ganglia.
We know DLS is required for learned skills: if a rat learns a new motor skill and then you mess with DLS, they can no longer execute it.
We know DLS is required for learned skills: if a rat learns a new motor skill and then you mess with DLS, they can no longer execute it.
August 12, 2025 at 6:26 PM
We focused on changes within the sensorimotor striatum – also called dorsolateral striatum (DLS) in rats – which is a big input region in the basal ganglia.
We know DLS is required for learned skills: if a rat learns a new motor skill and then you mess with DLS, they can no longer execute it.
We know DLS is required for learned skills: if a rat learns a new motor skill and then you mess with DLS, they can no longer execute it.
Another question – what *exactly* triggers this change in the neural code? Does it happen quickly, at the start of a new context that requires learning, or does it emerge slowly as a new skill develops? TBD, but with any luck we will have a better idea soon :)
August 12, 2025 at 6:17 PM
Another question – what *exactly* triggers this change in the neural code? Does it happen quickly, at the start of a new context that requires learning, or does it emerge slowly as a new skill develops? TBD, but with any luck we will have a better idea soon :)
There are lots of interesting questions and ideas that come from this. For one, what the heck is DLS activity doing during free exploration?! Why is it so locked to motor output if it’s not required for it? We’re thinking about this now, but lmk if you have any ideas!
August 12, 2025 at 6:17 PM
There are lots of interesting questions and ideas that come from this. For one, what the heck is DLS activity doing during free exploration?! Why is it so locked to motor output if it’s not required for it? We’re thinking about this now, but lmk if you have any ideas!
But then, we compared mapping from neural activity to movement – and saw a big difference!
🤩
While neurons in both cases represented ongoing movement, they did so in radically different ways.
🤩
While neurons in both cases represented ongoing movement, they did so in radically different ways.
August 12, 2025 at 6:17 PM
But then, we compared mapping from neural activity to movement – and saw a big difference!
🤩
While neurons in both cases represented ongoing movement, they did so in radically different ways.
🤩
While neurons in both cases represented ongoing movement, they did so in radically different ways.
What could explain this change in function? We next turned to neural activity, recording from DLS neurons from rats in both contexts.
Maybe neurons were more active in the task, or activity was more locked to behavior?
At first, we saw a whole lotta nothing that screamed “big difference”.
🤨
Maybe neurons were more active in the task, or activity was more locked to behavior?
At first, we saw a whole lotta nothing that screamed “big difference”.
🤨
August 12, 2025 at 6:17 PM
What could explain this change in function? We next turned to neural activity, recording from DLS neurons from rats in both contexts.
Maybe neurons were more active in the task, or activity was more locked to behavior?
At first, we saw a whole lotta nothing that screamed “big difference”.
🤨
Maybe neurons were more active in the task, or activity was more locked to behavior?
At first, we saw a whole lotta nothing that screamed “big difference”.
🤨
But even with the latest and greatest tracking and behavioral quantification, we couldn’t find effects of lesioning DLS 😶🌫️. This is a big contrast to the huge effect in the task. This suggests DLS is not an essential part of the motor control machinery, but specializing in shaping learned behaviors.
August 12, 2025 at 6:17 PM
But even with the latest and greatest tracking and behavioral quantification, we couldn’t find effects of lesioning DLS 😶🌫️. This is a big contrast to the huge effect in the task. This suggests DLS is not an essential part of the motor control machinery, but specializing in shaping learned behaviors.
We first examined DLS function by taking it away (via lesion).
We already know DLS impacts learned movements – but what about innate ones?
Prior work has implicated DLS, but to know for sure, we needed to track these behaviors in detail. And thanks to advances in tracking, this is now easy-peasy!
We already know DLS impacts learned movements – but what about innate ones?
Prior work has implicated DLS, but to know for sure, we needed to track these behaviors in detail. And thanks to advances in tracking, this is now easy-peasy!
August 12, 2025 at 6:17 PM
We first examined DLS function by taking it away (via lesion).
We already know DLS impacts learned movements – but what about innate ones?
Prior work has implicated DLS, but to know for sure, we needed to track these behaviors in detail. And thanks to advances in tracking, this is now easy-peasy!
We already know DLS impacts learned movements – but what about innate ones?
Prior work has implicated DLS, but to know for sure, we needed to track these behaviors in detail. And thanks to advances in tracking, this is now easy-peasy!
We considered two behavioral contexts on each end of the learning spectrum:
(1) free exploration, where a rat goes about their business expressing innate behaviors like rearing and grooming
(2) task execution, where the rat expresses a sequence of arm movements it learned to solve a lever task
(1) free exploration, where a rat goes about their business expressing innate behaviors like rearing and grooming
(2) task execution, where the rat expresses a sequence of arm movements it learned to solve a lever task
August 12, 2025 at 6:17 PM
We considered two behavioral contexts on each end of the learning spectrum:
(1) free exploration, where a rat goes about their business expressing innate behaviors like rearing and grooming
(2) task execution, where the rat expresses a sequence of arm movements it learned to solve a lever task
(1) free exploration, where a rat goes about their business expressing innate behaviors like rearing and grooming
(2) task execution, where the rat expresses a sequence of arm movements it learned to solve a lever task
But… is this just because rats *always* need their DLS for motor output, learned or otherwise?
... or ...
Does it play a special role for *learned* movements, only becoming involved when a movement is tailored to a task?
Basically – does DLS function change with learning?
... or ...
Does it play a special role for *learned* movements, only becoming involved when a movement is tailored to a task?
Basically – does DLS function change with learning?
a woman says quick q while holding a glass of water
ALT: a woman says quick q while holding a glass of water
media.tenor.com
August 12, 2025 at 6:17 PM
But… is this just because rats *always* need their DLS for motor output, learned or otherwise?
... or ...
Does it play a special role for *learned* movements, only becoming involved when a movement is tailored to a task?
Basically – does DLS function change with learning?
... or ...
Does it play a special role for *learned* movements, only becoming involved when a movement is tailored to a task?
Basically – does DLS function change with learning?
We focused on changes within the sensorimotor striatum – also called dorsolateral striatum (DLS) in rats – which is a big input region in the basal ganglia.
We know DLS is required for learned skills: if a rat learns a new motor skill and then you mess with DLS, they can no longer execute it
We know DLS is required for learned skills: if a rat learns a new motor skill and then you mess with DLS, they can no longer execute it
August 12, 2025 at 6:17 PM
We focused on changes within the sensorimotor striatum – also called dorsolateral striatum (DLS) in rats – which is a big input region in the basal ganglia.
We know DLS is required for learned skills: if a rat learns a new motor skill and then you mess with DLS, they can no longer execute it
We know DLS is required for learned skills: if a rat learns a new motor skill and then you mess with DLS, they can no longer execute it