Dendritic computation for abstract rule-based flexible decision-making.
Saved in:
| Title: | Dendritic computation for abstract rule-based flexible decision-making. |
|---|---|
| Authors: | Zhang, Yuan (AUTHOR), Liu, Yanhe (AUTHOR), Hu, Yachuang (AUTHOR), Zhang, Zhenkun (AUTHOR), Du, Jie (AUTHOR), Cui, Lele (AUTHOR), Liu, Qi (AUTHOR), Zhong, Lin (AUTHOR), Xin, Yu (AUTHOR), Chai, Ruiming (AUTHOR), Deng, Li (AUTHOR), Pan, Jingwei (AUTHOR), Schiller, Jackie (AUTHOR), Xu, Ning-long (AUTHOR) |
| Source: | Science. 10/15/2026, Vol. 394 Issue 6821, p1-12. 12p. |
| Subjects: | Dendrites, Cognitive flexibility, Pyramidal neurons, Cognitive computing, Optogenetics, Sensorimotor integration |
| Abstract: | Intelligent behavior involves flexible integration of sensory input and abstract rule knowledge. Dendritic computation has been shown to contribute to top-down modulation of sensory processing; however, whether and how it is involved in flexible cognitive processing remains to be elucidated. Combining a rule-based flexible decision-making task with in vivo two-photon imaging and optogenetics in mice, we found that within single layer 5 extratelencephalic neurons in the auditory cortex, apical dendrites preferentially encoded the inferred abstract rule and somata preferentially represented sensory stimuli, whereas choice signals were present in both compartments. Optogenetic dendritic inhibition impaired rule-dependent somatic choice coding, supporting a task-relevant integration across the dendritic and somatic compartments. These results indicate that dendritic computation in cortical neurons mediates flexible behavior. Editor's summary: Different neuronal compartments receive different inputs but at the same time add their own local computational features. To identify the contribution of single compartments requires methods to monitor their activity. Zhang et al. recorded from the apical dendrites and somata of auditory cortex layer 5 neurons while mice performed a context-dependent auditory conditioning task. The authors demonstrated compartmentalization of rule encoding (in apical dendrites) and sensory processing (in the somata) to drive flexible choices. The results suggest that integration of dendritic and somatic inputs is required for flexible decision-making. —Mattia Maroso INTRODUCTION: Intelligent behavior requires flexibility: The same sensory input can call for different actions depending on the current rule. A ringing phone means "answer it" at home but "silence it" in a meeting. How individual neurons carry out this flexible mapping remains poorly understood. Cortical pyramidal neurons extend elaborate, branching apical dendrites into the upper layers of the cortex, where they collect a wide range of contextual inputs. Theory has long held that these dendrites integrate distinct streams of information within a single cell, yet direct evidence linking dendritic processing to flexible cognition has been lacking. Most studies of decision-making treat each neuron as a single point, overlooking the computations that dendrites may perform. RATIONALE: To probe dendritic computation during flexible behavior, we trained mice on a rule-switching flexible categorization task. Mice heard tones and reported whether each was "high" or "low" relative to a category boundary, which randomly changed without warning. The animal had to continually infer the hidden task rule from reward feedback to flexibly guide its decisions. A computational model of this behavior allowed us to estimate the animal's inferred rule on each trial. We then used two-photon imaging with subcellular resolution to record activity separately from the apical dendrites and the somata of layer 5 extratelencephalic (L5-ET) neurons in the auditory cortex to measure dendrite- and soma-specific encoding of cognitive and task variables and used optogenetics to selectively suppress dendritic activity to test the contribution of dendritic activity to neuronal computation. RESULTS: We found that within single L5-ET neurons, the two compartments represented different information. The apical dendrites preferentially encoded the inferred abstract rule, whereas the somata preferentially represented the sensory stimulus; signals related to the animal's choice appeared in both. Imaging dendrites and somata of the same neuron at the same time confirmed this division directly: Rule information carried by the dendrites was largely absent at the soma. Optogenetic suppression of dendritic activity by activating inhibitory somatostatin (SST) interneurons that target distal dendrites selectively impaired rule-dependent choice coding at the soma. This impairment was strongest in trials where the correct choice depended on the rule, indicating that dendritic rule information must be integrated with somatic sensory input to guide flexible choices. CONCLUSION: These findings indicate that a single cortical neuron can perform a cognitive computation for flexible behavior. Its apical dendrites carry an internally inferred rule that combines with sensory information arriving at the soma to produce rule-dependent choices that neither compartment conveys alone. Dendritic computation thus provides a cellular substrate for cognitive flexibility, extending the classical two-compartment integration model from sensory and motor processing into abstract, rule-based cognition. More broadly, this form of within-neuron computation expands the capacity available to cortical networks during flexible behavior and may inform the design of more-efficient artificial systems. Cortical dendrites implement a computation underlying rule-dependent auditory decision-making.: (Left) Head-fixed mice performed a rule-based flexible categorization task by inferring an abstract rule indicating the current categorization boundary for mapping tone stimuli to left or right choices. Correct choices for tone stimuli falling between the two alternative boundaries could only be made by applying the current rule. (Right) Compartmentalized apical dendritic signals carry rule information, whereas compartmentalized somatic signals carry sensory information. Rule-dependent choice signals are generated following the integration of rule and sensory information across compartments within the same neuron. Optogenetic inhibition of dendritic signals disrupted the generation of correct choice signals. [ABSTRACT FROM AUTHOR] |
| Copyright of Science is the property of American Association for the Advancement of Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Psychology and Behavioral Sciences Collection |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | Intelligent behavior involves flexible integration of sensory input and abstract rule knowledge. Dendritic computation has been shown to contribute to top-down modulation of sensory processing; however, whether and how it is involved in flexible cognitive processing remains to be elucidated. Combining a rule-based flexible decision-making task with in vivo two-photon imaging and optogenetics in mice, we found that within single layer 5 extratelencephalic neurons in the auditory cortex, apical dendrites preferentially encoded the inferred abstract rule and somata preferentially represented sensory stimuli, whereas choice signals were present in both compartments. Optogenetic dendritic inhibition impaired rule-dependent somatic choice coding, supporting a task-relevant integration across the dendritic and somatic compartments. These results indicate that dendritic computation in cortical neurons mediates flexible behavior. Editor's summary: Different neuronal compartments receive different inputs but at the same time add their own local computational features. To identify the contribution of single compartments requires methods to monitor their activity. Zhang et al. recorded from the apical dendrites and somata of auditory cortex layer 5 neurons while mice performed a context-dependent auditory conditioning task. The authors demonstrated compartmentalization of rule encoding (in apical dendrites) and sensory processing (in the somata) to drive flexible choices. The results suggest that integration of dendritic and somatic inputs is required for flexible decision-making. —Mattia Maroso INTRODUCTION: Intelligent behavior requires flexibility: The same sensory input can call for different actions depending on the current rule. A ringing phone means "answer it" at home but "silence it" in a meeting. How individual neurons carry out this flexible mapping remains poorly understood. Cortical pyramidal neurons extend elaborate, branching apical dendrites into the upper layers of the cortex, where they collect a wide range of contextual inputs. Theory has long held that these dendrites integrate distinct streams of information within a single cell, yet direct evidence linking dendritic processing to flexible cognition has been lacking. Most studies of decision-making treat each neuron as a single point, overlooking the computations that dendrites may perform. RATIONALE: To probe dendritic computation during flexible behavior, we trained mice on a rule-switching flexible categorization task. Mice heard tones and reported whether each was "high" or "low" relative to a category boundary, which randomly changed without warning. The animal had to continually infer the hidden task rule from reward feedback to flexibly guide its decisions. A computational model of this behavior allowed us to estimate the animal's inferred rule on each trial. We then used two-photon imaging with subcellular resolution to record activity separately from the apical dendrites and the somata of layer 5 extratelencephalic (L5-ET) neurons in the auditory cortex to measure dendrite- and soma-specific encoding of cognitive and task variables and used optogenetics to selectively suppress dendritic activity to test the contribution of dendritic activity to neuronal computation. RESULTS: We found that within single L5-ET neurons, the two compartments represented different information. The apical dendrites preferentially encoded the inferred abstract rule, whereas the somata preferentially represented the sensory stimulus; signals related to the animal's choice appeared in both. Imaging dendrites and somata of the same neuron at the same time confirmed this division directly: Rule information carried by the dendrites was largely absent at the soma. Optogenetic suppression of dendritic activity by activating inhibitory somatostatin (SST) interneurons that target distal dendrites selectively impaired rule-dependent choice coding at the soma. This impairment was strongest in trials where the correct choice depended on the rule, indicating that dendritic rule information must be integrated with somatic sensory input to guide flexible choices. CONCLUSION: These findings indicate that a single cortical neuron can perform a cognitive computation for flexible behavior. Its apical dendrites carry an internally inferred rule that combines with sensory information arriving at the soma to produce rule-dependent choices that neither compartment conveys alone. Dendritic computation thus provides a cellular substrate for cognitive flexibility, extending the classical two-compartment integration model from sensory and motor processing into abstract, rule-based cognition. More broadly, this form of within-neuron computation expands the capacity available to cortical networks during flexible behavior and may inform the design of more-efficient artificial systems. Cortical dendrites implement a computation underlying rule-dependent auditory decision-making.: (Left) Head-fixed mice performed a rule-based flexible categorization task by inferring an abstract rule indicating the current categorization boundary for mapping tone stimuli to left or right choices. Correct choices for tone stimuli falling between the two alternative boundaries could only be made by applying the current rule. (Right) Compartmentalized apical dendritic signals carry rule information, whereas compartmentalized somatic signals carry sensory information. Rule-dependent choice signals are generated following the integration of rule and sensory information across compartments within the same neuron. Optogenetic inhibition of dendritic signals disrupted the generation of correct choice signals. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 00368075 |
| DOI: | 10.1126/science.adz5323 |
