Corticostriatal neurons in auditory cortex drive decisions during auditory discrimination

Corticostriatal neurons in auditory cortex drive decisions during auditory discrimination Znamenskiy Zador

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

Corticostriatal neurons in auditory cortex drive decisions during auditory discrimination 2013 Nature 10.1038/nature12077 Znamenskiy2013-corticostriatal-neurons-in-auditory-cortex-d Znamenskiy2013-corticostriatal-neurons-in-auditory-cortex-d.txt, the project's staged copy af4b4a71854a2295150605101bf12e58acec8b62 not-redistributable neuroscience

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them whether the projection from auditory cortex to striatum carries the signals that drive choices during auditory discrimination Quotation withheld because this record is not cleared for redistribution. how auditory representations in cortex are transformed into motor commands Quotation withheld because this record is not cleared for redistribution. shows that activating corticostriatal neurons biases choices toward their preferred frequency and inactivating them biases choices away, establishing the pathway as both sufficient and necessary for the decision Quotation withheld because this record is not cleared for redistribution. mechanism Quotation withheld because this record is not cleared for redistribution. auditory perceptual decision-making Quotation withheld because this record is not cleared for redistribution. electrophysiology electrophysiology Quotation withheld because this record is not cleared for redistribution. mechanisms The auditory cortex-to-striatum projection carries frequency-specific acoustic information that is read out to select the motor response, so changing spiking in that projection changes the choice. Quotation withheld because this record is not cleared for redistribution. mechanisms Because exciting corticostriatal cells suppresses surrounding cortex rather than exciting it, the behavioural bias must be delivered by their long-range axons to the striatum, not by recruiting other cortical output pathways. Quotation withheld because this record is not cleared for redistribution. mechanisms A decision bottleneck: the small number of projection neurons that actually drive the choice provides finite bandwidth, which caps the effective pool size and hence psychophysical performance. Quotation withheld because this record is not cleared for redistribution. sufficiency and necessity are separated explicitly, and the direction of the necessity test is predicted before it is run Quotation withheld because this record is not cleared for redistribution. cell-type specificity is achieved by an intersection: a retrograde virus delivers Cre from the target region, and a Cre-dependent virus supplies the opsin in the source region Quotation withheld because this record is not cleared for redistribution. a failed non-specific manipulation is reported first, with three candidate explanations, and becomes the argument for targeting Quotation withheld because this record is not cleared for redistribution. full inactivation is deliberately avoided because compensatory plasticity would confound it, so only a patch of the tonotopic map is silenced Quotation withheld because this record is not cleared for redistribution. the same pathway is driven at two points - somata in cortex and axon terminals in striatum - so the result cannot depend on where the light went Quotation withheld because this record is not cleared for redistribution. recurrent cortical activity is pharmacologically removed while the axons are stimulated, which isolates the long-range output as the carrier of the effect Quotation withheld because this record is not cleared for redistribution. the behavioural effect is converted into stimulus units by fitting a logistic model, so a neural manipulation and an acoustic change become comparable Quotation withheld because this record is not cleared for redistribution. stimulation is restricted to a quarter of trials and still rewarded normally, to stop the animal adapting to the manipulation itself Quotation withheld because this record is not cleared for redistribution. tone intensity is randomised across trials specifically to block a loudness-based strategy Quotation withheld because this record is not cleared for redistribution. the effect size is tied to an anatomical count by combining fluorescence, a measured inactivation space constant, and cell density Quotation withheld because this record is not cleared for redistribution. uninjected animals receiving identical light serve as the control for both opsins separately Quotation withheld because this record is not cleared for redistribution. local suppression during corticostriatal excitation is used to argue the behavioural effect travels by the long-range output rather than through local circuitry Quotation withheld because this record is not cleared for redistribution. Non-specific cortical activation did not give tuning-predicted biases and lacked adequate control over the relevant projection neurons. what this limits: the non-specific activation experiments: THE METHOD OF NON-SPECIFIC ACTIVATION OF DIVERSE NEURONAL POPULATIONS USED IN THESE EXPERIMENTS DOES NOT PROVIDE SUFFICIENT control over the relevant populations Quotation withheld because this record is not cleared for redistribution. The stimulation biases were not statistically significant within individual sessions. what this limits: any session-level claim: THE BIASES WERE NOT SIGNIFICANT ON INDIVIDUAL SESSIONS (p<0.05 ON 1/33 SESSIONS), IN PART BECAUSE THE NUMBER OF STIMULATED TRIALS PER SESSION WAS TYPICALLY SMALL (70+/-24 (s.d.) TRIALS) Quotation withheld because this record is not cleared for redistribution. The inactivation effect on chronometric functions did not reach significance. what this limits: the chronometric result specifically: A SIMILAR TREND WAS REFLECTED IN THE SUBJECTS' CHRONOMETRIC FUNCTIONS BUT DID NOT REACH SIGNIFICANCE Quotation withheld because this record is not cleared for redistribution. Cell counts could not be estimated for two fibres whose tracks were not found histologically. what this limits: the cell counts for two named fibres: FOR 2 FIBERS IN ONE OF THE RATS WE COULD NOT LOCATE THE TRACKS IN HISTOLOGICAL SECTIONS AND PRECISELY ESTIMATE CELL COUNTS Quotation withheld because this record is not cleared for redistribution. Other parallel pathways contributing to the decision are not excluded. what this limits: any claim that this pathway is the only one: OUR RESULTS DO NOT EXCLUDE THE PARTICIPATION OF OTHER PARALLEL PATHWAYS Quotation withheld because this record is not cleared for redistribution. other parallel pathways are not excluded what this limits: the same scope as limitations[4] - other parallel pathways remain possible Quotation withheld because this record is not cleared for redistribution. individual sessions were underpowered; the effect is visible in aggregate what this limits: the same scope as limitations[1] - individual sessions are underpowered at 70+/-24 trials Quotation withheld because this record is not cleared for redistribution. at sites with large biases, performance was also modestly affected, possibly a ceiling effect what this limits: the sites with large choice biases: STIMULATION ALSO HAD A MODEST EFFECT ON THE SUBJECTS' PERFORMANCE AT SITES WITH LARGE CHOICE BIASES, POSSIBLY DUE TO A CEILING EFFECT Quotation withheld because this record is not cleared for redistribution. the chronometric counterpart of the inactivation result trended but did not reach significance what this limits: the inactivation result: it BIASED SUBJECTS' CHOICES AWAY FROM THE REWARD PORT ASSOCIATED WITH THE FREQUENCY BAND OF THE INACTIVATION SITE BUT DID NOT AFFECT the chronometric measure Quotation withheld because this record is not cleared for redistribution. auditory striatal neurons had only ever been characterised in anaesthetised or passively listening animals before this what this limits: every PRIOR characterisation this paper builds on: AUDITORY STRIATAL NEURONES HAVE PREVIOUSLY BEEN CHARACTERIZED IN ANAESTHETISED AND PASSIVELY LISTENING ANIMALS, NOT DURING BEHAVIOUR Quotation withheld because this record is not cleared for redistribution. the function of sensory-recipient striatal regions was poorly established at the outset what this limits: sensory-recipient striatal regions as a class: THE FUNCTION OF STRIATAL REGIONS WHICH RECEIVE DIRECT INPUT FROM SENSORY CORTICAL AREAS IS LESS WELL ESTABLISHED Quotation withheld because this record is not cleared for redistribution. the random-dot-motion logic of a graded two-alternative discrimination, remade for frequency using clouds of tones Quotation withheld because this record is not cleared for redistribution. the ideal observer, applied to single neurons so that cortical and striatal selectivity can be compared on one scale Quotation withheld because this record is not cleared for redistribution. a retrogradely transported herpes virus carrying Cre, which turns an anatomical projection into a genetically addressable cell class Quotation withheld because this record is not cleared for redistribution. the fit re-run 100 times from RANDOM starting parameters, keeping the solution with the smallest mean square error eliminates: a local minimum as the reported fit - a single optimisation run reports wherever it happens to stop, and its parameters would be a property of the starting point · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. optogenetic activation - non-specific across a cortical population versus targeted to a projection-defined class; the first gives biases unrelated to tuning and degrades performance, the second gives biases predicted by tuning and spares it Quotation withheld because this record is not cleared for redistribution. sufficiency versus necessity - activation showing the pathway CAN drive choices, inactivation showing it normally DOES; the paper separates the two claims and tests them in that order Quotation withheld because this record is not cleared for redistribution. stimulating the pathway at the soma versus at the axon terminal; the same cells reached from either end, which is what rules out a location artefact Quotation withheld because this record is not cleared for redistribution. a behavioural effect measured as a choice bias versus as a shift in response time; the chronometric curve moves as if acoustic evidence had been added Quotation withheld because this record is not cleared for redistribution. the neuronal pool limiting psychophysical performance - as constrained by correlated firing versus by the bandwidth of the few projection neurons that drive decisions Quotation withheld because this record is not cleared for redistribution. striatal regions receiving motor and prefrontal input, whose functions are well studied, versus those receiving direct sensory cortical input, whose function was not established Quotation withheld because this record is not cleared for redistribution. claim_type: mechanism a second reading reaching the same value. Recorded: mechanism — note: a second reading reaching the same value from a different sentence. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. method: rats performing a two-alternative cloud-of-tones frequency discrimination, with tetrode recording and projection-targeted channelrhodopsin and archaerhodopsin in corticostriatal neurons a second reading DIFFERING from the recorded value. Recorded: electrophysiology — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. phenomenon: perceptual decision-making - converting an auditory representation into a motor choice a second reading DIFFERING from the recorded value. Recorded: auditory perceptual decision-making — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. builds-on → perceptual decision experiments in primate area MT (ref. 16) Quotation withheld because this record is not cleared for redistribution. extends → prior characterisations of auditory striatal neurones in anaesthetised and passively listening animals (refs. 14, 15) Quotation withheld because this record is not cleared for redistribution. replicates → previous results on the laminar location of corticostriatal neurones (ref. 22) Quotation withheld because this record is not cleared for redistribution. replicates → previously reported response-time increases on difficult trials in rats (ref. 18) Quotation withheld because this record is not cleared for redistribution. builds-on → the known topography of corticostriatal projections (ref. 23) Quotation withheld because this record is not cleared for redistribution. extends → work on striatal regions receiving motor and prefrontal input, by asking what the sensory-recipient striatum does Quotation withheld because this record is not cleared for redistribution. critiques → non-specific optogenetic activation of mixed cortical populations as a way of testing a pathway's role Quotation withheld because this record is not cleared for redistribution. critiques → complete-inactivation lesion designs, whose interpretation compensatory plasticity can obscure differs in: THE AXIS IS THE EXTENT OF THE INACTIVATION, AND THE REASON IS NAMED. 'Rather than attempting to inactivate all corticostriatal cells - a technically difficult feat Quotation withheld - we sought instead to silence corticostriatal neurones WITHIN A RESTRICTED REGION of the tonotopic map.' The critique of whole-population lesions is not that they fail but that the preparation answers back. Quotation withheld because this record is not cleared for redistribution. builds-on → primate area MT microstimulation experiments, whose logic the task design is modelled on Quotation withheld because this record is not cleared for redistribution. extends → estimates of how few neurons animals can be trained to detect, from artificial detection to normal perception Quotation withheld because this record is not cleared for redistribution. supports → the account in which correlated firing limits the effective size of the neuronal pool underlying psychophysical performance Quotation withheld because this record is not cleared for redistribution. EVIDENCE CORRECTED (RE-GROUNDED) · corticostriatal —causes→ choice biases corticostriatal choice biases EVIDENCE CORRECTED: RE-GROUNDED. The old span named neither endpoint. This clause names both and states the dependence - the bias scales with how many corticostriatal neurones were silenced. — strength of the relation: 0.95 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span writes 'biased subjects' choices', not 'choice biases'. Quotation withheld because this record is not cleared for redistribution. Channelrhodopsin-2 —contrasted-with→ archaerhodopsin-3 Channelrhodopsin-2 archaerhodopsin-3 differs in: THE AXIS IS THE SIGN OF THE MANIPULATION, AND THE BIAS FOLLOWS IT. 'TARGETED CHANNELRHODOPSIN-2 (ChR2)-MEDIATED STIMULATION of corticostriatal neurones during the task biased decisions in the direction predicted by the frequency tuning of the stimulated neurones, WHEREAS ARCHAERHODOPSIN-3 (ARCH)-mediated inactivation biased decisions in T[HE OPPOSITE DIRECTION].' Two opsins are not being compared as tools - they are the two signs of one experiment, and the double dissociation is the evidence. · strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. EVIDENCE CORRECTED (RE-GROUNDED) · corticostriatal —part-of→ auditory cortex corticostriatal auditory cortex EVIDENCE CORRECTED: RE-GROUNDED. The old span named neither endpoint. This sentence names both and places one inside the other - corticostriatal neurones IN the auditory cortex. — strength of the relation: 0.9 Quotation withheld because this record is not cleared for redistribution. choice biases —operationalised-as→ logistic regression choice biases logistic regression strength of the relation: 0.95 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span writes 'subjects' choices' / 'behavioural effect', not 'choice biases'. Quotation withheld because this record is not cleared for redistribution. striatum —causes→ superior colliculus striatum superior colliculus strength of the relation: 0.75 Quotation withheld because this record is not cleared for redistribution. corticostriatal —causes→ choice biases corticostriatal choice biases strength of the relation: 0.95 · note: RE-GROUNDS another row of this record: the relation was always stated, the endpoint names were not in the span. Quotation withheld because this record is not cleared for redistribution. logistic regression logistic regression Quotation withheld because this record is not cleared for redistribution. Wilcoxon signed-rank test Wilcoxon signed-rank test Quotation withheld because this record is not cleared for redistribution. linear regression (ordinary least squares) linear regression (ordinary least squares) Quotation withheld because this record is not cleared for redistribution. bootstrap bootstrap Quotation withheld because this record is not cleared for redistribution. cross-validation cross-validation Quotation withheld because this record is not cleared for redistribution. confidence interval confidence interval Quotation withheld because this record is not cleared for redistribution. p-value p-value Quotation withheld because this record is not cleared for redistribution. standard error standard error Quotation withheld because this record is not cleared for redistribution. mean squared error minimisation mean squared error minimisation Quotation withheld because this record is not cleared for redistribution. logistic regression model logistic regression model Quotation withheld because this record is not cleared for redistribution. signed-rank test signed-rank test Quotation withheld because this record is not cleared for redistribution. ideal observer ideal observer Quotation withheld because this record is not cleared for redistribution. slope of the psychometric curve slope of the psychometric curve Quotation withheld because this record is not cleared for redistribution. Optogenetic activation of auditory corticostriatal neurons shifted rats' choices toward the reward port associated with the frequency the stimulated neurons preferred. Quotation withheld because this record is not cleared for redistribution. Arch-mediated silencing of corticostriatal neurons pushed choices away from the port matching the inactivated site's frequency band while leaving psychometric slope (sensitivity) unchanged. Quotation withheld because this record is not cleared for redistribution. Non-specific (untargeted) activation of auditory cortex produced biases whose direction did not follow the frequency tuning of neurons near the fiber. Quotation withheld because this record is not cleared for redistribution. Non-specific cortical activation also impaired task performance, flattening the psychometric curve. Quotation withheld because this record is not cleared for redistribution. Stimulating corticostriatal axons still biased choices after the auditory cortex was pharmacologically silenced, showing the effect travels through the long-range striatal projection rather than through recurrent cortical circuitry. Quotation withheld because this record is not cleared for redistribution. Silencing only about one third to eight tenths of one percent of the local neurons (roughly 1,000-2,400 Arch-expressing cells) was enough to produce robust choice biases. Quotation withheld because this record is not cleared for redistribution. Corticostriatal stimulation also shifted response-time (chronometric) curves exactly as adding real acoustic evidence for the site's preferred frequency would. Quotation withheld because this record is not cleared for redistribution. Ideal-observer analysis showed that cortical and striatal neurons had similar stimulus selectivity. Quotation withheld because this record is not cleared for redistribution. activating corticostriatal neurons biased choices toward the port associated with the stimulation site's preferred frequency Quotation withheld because this record is not cleared for redistribution. inactivating them biased choices away from that port, as predicted Quotation withheld because this record is not cleared for redistribution. non-specific cortical activation produced biases whose direction the tuning near the fiber did not predict Quotation withheld because this record is not cleared for redistribution. stimulating corticostriatal axons in the striatum reproduced the effect from the other end of the pathway Quotation withheld because this record is not cleared for redistribution. the bias survived pharmacological silencing of the cortex, so it is carried by the long-range output rather than recurrent cortical activity Quotation withheld because this record is not cleared for redistribution. stimulation shifted chronometric curves as if acoustic evidence favouring that choice had been added Quotation withheld because this record is not cleared for redistribution. 84% of targeted neurons were in layer V Quotation withheld because this record is not cleared for redistribution. robust biases appeared at sites with 1040-2444 Arch-expressing cells, 0.35-0.83% of the local population Quotation withheld because this record is not cleared for redistribution. silencing that minute fraction changed behaviour, suggesting a privileged role for these cells Quotation withheld because this record is not cleared for redistribution. inactivation efficiency decayed with a space constant of about 564 micrometres Quotation withheld because this record is not cleared for redistribution. cortical and striatal neurons showed similar stimulus selectivity, so the striatum preserves the information needed for the task Quotation withheld because this record is not cleared for redistribution. targeted stimulation left psychophysical sensitivity intact, unlike non-specific stimulation Quotation withheld because this record is not cleared for redistribution. no biases appeared in uninjected animals given identical light Quotation withheld because this record is not cleared for redistribution. exciting corticostriatal neurons suppressed local cortical activity overall Quotation withheld because this record is not cleared for redistribution. cloud of tones stimulus cloud of tones stimulus Quotation withheld because this record is not cleared for redistribution. stimulus strength r stimulus strength r Quotation withheld because this record is not cleared for redistribution. preferred frequency preferred frequency Quotation withheld because this record is not cleared for redistribution. anti-bias anti-bias Quotation withheld because this record is not cleared for redistribution. cloud of tones cloud of tones Quotation withheld because this record is not cleared for redistribution. anti-bias anti-bias Quotation withheld because this record is not cleared for redistribution. two alternative choice two alternative choice Quotation withheld because this record is not cleared for redistribution. motor component of the response time motor component of the response time Quotation withheld because this record is not cleared for redistribution. magnitude of modulation of response times magnitude of modulation of response times Quotation withheld because this record is not cleared for redistribution. light activation light activation Quotation withheld because this record is not cleared for redistribution. target octave target octave Quotation withheld because this record is not cleared for redistribution. choice bias choice bias Quotation withheld because this record is not cleared for redistribution. response time response time Quotation withheld because this record is not cleared for redistribution. sound intensity sound intensity Quotation withheld because this record is not cleared for redistribution. mask LED mask LED Quotation withheld because this record is not cleared for redistribution. number of Arch-expressing neurones number of Arch-expressing neurones Quotation withheld because this record is not cleared for redistribution. stimulus strength stimulus strength Quotation withheld because this record is not cleared for redistribution. choice biases choice biases Quotation withheld because this record is not cleared for redistribution. preferred frequency preferred frequency Quotation withheld because this record is not cleared for redistribution. response times response times Quotation withheld because this record is not cleared for redistribution. 84% — share of ChR2-expressing neurons that lay in cortical layer V after the HSV1-Cre/AAV targeting Quotation withheld because this record is not cleared for redistribution. 0.35–0.83% — fraction of neurons in the illuminated volume that had to be silenced to yield robust choice biases Quotation withheld because this record is not cleared for redistribution. ~564 μ m — space constant of the decay of inactivation efficiency with distance from the fiber Quotation withheld because this record is not cleared for redistribution. 36/57 sites, p = 0.015 — proportion of non-specific cortical stimulation sites producing contralateral biases, and its significance Quotation withheld because this record is not cleared for redistribution. p = 0.016 — significance of the choice bias from corticostriatal axon stimulation while the cortex was pharmacologically inactivated Quotation withheld because this record is not cleared for redistribution. 20–50 ms — size of the response-time increase on difficult trials Quotation withheld because this record is not cleared for redistribution. 70±24 (s.d.) trials — typical number of stimulated trials per session, explaining why single-session biases rarely reached significance Quotation withheld because this record is not cleared for redistribution. 4 of 201 — cells that responded to blue-light pulses on more than half of trials, i.e. the identified light-driven corticostriatal population Quotation withheld because this record is not cleared for redistribution. 25% — proportion of trials on which photostimulation was delivered, capped to limit behavioural adaptation Quotation withheld because this record is not cleared for redistribution. 5-40 kHz over three octaves — the stimulus range, split into low and high target octaves Quotation withheld because this record is not cleared for redistribution. 84% — targeted ChR2-expressing neurons located in layer V Quotation withheld because this record is not cleared for redistribution. 1040-2444 neurones, 0.35-0.83% — Arch-expressing cells within 1 mm producing robust biases, as a fraction of the local population Quotation withheld because this record is not cleared for redistribution. ~564 μ m — space constant of inactivation efficiency Quotation withheld because this record is not cleared for redistribution. 25% — share of trials carrying photostimulation, capped to prevent adaptation Quotation withheld because this record is not cleared for redistribution. 70±24 trials — stimulated trials per session, the reason single sessions were underpowered Quotation withheld because this record is not cleared for redistribution. 45-75 dB (SPL) — the randomised intensity range blocking a loudness strategy Quotation withheld because this record is not cleared for redistribution. 30 ms tones at 100 Hz — the cloud-of-tones stimulus timing Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://doi.org/10.1038/nature12077.