CNS*2020 Online has ended
Welcome to the Sched instance for CNS*2020 Online! Please read the instruction document on detailed information on CNS*2020.
Back To Schedule
Monday, July 20 • 9:00pm - 10:00pm
P151: Generalisation of stimulus representation across somatosensory cortex areas in a cellular-resolution photostimulus detection task

Log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
Thijs L van der Plas*, James M Rowland*, Robert M Lees, Adam M Packer

Virtual poster presentationhttps://meet.google.com/zkk-vwjq-jbm

Introduction video & poster attached below!


Mice use whiskers to explore their environment. Whisker stimulation elicits a neural response in primary (S1) and secondary (S2) somatosensory cortex, two highly interconnected and hierarchically organised brain regions. Their interaction has been related to stimulus detection [1], although its precise functional role remains unclear [2]. Here, we aim to assign a function to this circuit for a stimulus detection task, by assessing how S1-S2 interactions facilitate stimulus perception.

We have conditioned mice to detect 2-photon optogenetic stimulation of random ensembles of S1 cells. This allows us to control the number of stimulated cells on a trial by trial basis, and to separate the initial stimulus representation from the ensuing network response. Simultaneously, we record the calcium activity of both stimulated and unstimulated cells in S1 and S2, rendering an all-optical approach to study neural dynamics [3]. In short, we are able to directly stimulate S1 neurons, hence defining the initial stimulus in S1, while recording the subsequent S1 and S2 neural response.

Mice were conditioned to report the photostimulus by licking a water spout. The task was divided into Go trials, where a varying number (5 - 150) of cells were stimulated, and Catch trials without stimulation. Behavioural accuracy increased as more cells were stimulated (Fig A), indicating that our task operated in the regime of perception.

We observed strongly elevated, sustained neural population activity in both S1 and S2 on successful Go trials (Hits), compared to both unsuccessful Go trials (Misses) and to licking behaviour in the absence of a stimulus (False Positives). This suggests that S1 and S2 encode information during Hit trials that is different from both passive stimulus-induced activity and neural signals driven by movement and reward.

To confirm whether neurons indeed encoded stimulus information, we performed a stimulus decoding analysis on S1 and S2 neural activity separately. We only consider trials where mice licked (i.e. Hits and False Positives), to avoid a behavioural bias. Here, we observe a significant difference between S1 (where the stimulus occurred) and S2 (Fig B): Stimulus information could only be decoded from S2 after a considerable time post-stimulus, while S1 could be decoded directly post-stimulus. Hence, stimulus information has propagated (directly or indirectly) from S1 to S2.

Furthermore, we find a striking dynamic property of information coding in the S1-S2 circuit. Directly post-stimulus at 1s, decoding accuracy in S1 depends on the stimulus strength, the number of stimulated cells (Fig C). However, after a delay of 3s, we find that accuracy has increased, and has become independent of the original stimulus strength (Fig C). S2 decoding accuracy increased equivalently (not shown), even though S2 decoding performs at chance level directly post-stimulus (Fig B).

The stimulus detection task design requires the animals to elicit the same response, independent of stimulus strength. Our results show that the S1-S2 circuit dynamically performs this computation: by propagating stimulus information between S1 and S2, the neural code becomes independent of the original stimulus strength. Hence, we uncover a putative mechanism of how interregional communication can transform stimulus information to facilitate stimulus detection.


[1] Kwon et al., 2016, Nat Neur

[2] Ni & Chen, 2017, EJN

[3] Packer et al., 2015, Nat Meth

avatar for Thijs L van der Plas

Thijs L van der Plas

PhD student, ​Department of Physiology, Anatomy, and Genetics, University of Oxford

Monday July 20, 2020 9:00pm - 10:00pm CEST
Slot 04