Interrogation of a projection-defined CLA neuronal subpopulation.

(A) Schematic of the injection strategy. Three different CTB-AlexFluor conjugates were injected into separate rostrocaudal regions of RSP. (B) Representative 100 μm sections spanning the rostrocaudal axis of the brain in which the CLA can be seen retro-labeled by CTB from rostral (blue), intermediate (red), and caudal (green) injections into RSP (n = 3 mice). Scale bar = 1 mm. (C) Insets of B in which CLA neurons are labeled with CTB. Scale bar = 200 μm. (D, top to bottom) Intrinsic electrophysiological profiles, expanded spike waveforms (from top inset), proportion of neurons found to be CTB+ during experiments (cells for which CTB status was known), and example morphological reconstructions for each electrophysiological cell type. Scale bars (top to bottom): 20 mV/100 ms, 300 pA, 20 mV/10 ms, 100 μm. Dashed border around morphologies represents the average CLARSP area across slices from Supp. Fig. S2. Somas aligned to their approximate location in this region during patching.

Stereotaxic targets and coordinates. DV coordinates measured as depth from pia.

Intraclaustral connectivity is common and cross-modular

(A) Schematic of injection and patching strategy in the CLA. (B) Transmitted light images taken during patching. Magenta arrows indicate ChrimsonR-tdTomato-expressing neurons while green arrows indicate CTB-labeled neurons. Note that these populations only partially overlap. Scale bar = 50 μm. (C) Expression of AAV-FLEX-ChrimsonR-tdTomato and CTB in the CLA along the rostrocaudal axis. Notice the lack of tdTomato+ cell bodies in the far rostral and far caudal sections beyond the spread of the virus. Scale bar = 200 μm. (D) CLA neurons frequently displayed EPSPs in response to presynaptic CLA neuron photostimulation with 595 nm light (top). CLA neurons not labeled with CTB were found to be proportionally the most likely to respond to presynaptic photostimulation (n = 28/46 neurons, bottom). (E, top) Most electrophysiological types were represented among neurons responsive to CLA input (n is the number of neurons recorded in each group, total n = 32 responsive/46 neurons). No significant difference in response probability was found between excitatory and inhibitory cell types (p = 0.29, Fisher exact test), while a significant difference in responsivity was found between CTB+ and CTB-neurons (p = 0.009, Fisher Exact Test) (bottom). (F) Schematic of injection and patching strategy in the CLA. (G) Confocal image of CTB and ChrimsonR-tdTomato expression in the CLA from separate injections into PL and RSP. Scale bar = 200 μm. (H) Quantification of responses from experiments in F. CLA neurons that were labeled with CTB from PL were more likely to respond to CLARSP input than those that weren’t, but not by a significant margin (I, p = 0.19, Fischer Exact Test). (J) Model of the circuit investigated, showing that CLARSP neurons are more likely to form active synapses with non-CLARSP neurons.

Spatial distribution of afferent projections onto the CLA.

(A) Schematic of injection sites in the cortex. Individual Chrimson-tdTom (magenta) injection sites were combined with an injection of CTB (green) into the RSP. (B) Coronal sections of example injection sites. Scale bars = 1 mm. (C) Schematics of a representative section of the CLA at 1 mm rostral to bregma. (D) Histological sections from the representative section in c and each input cortical area in B. Scale bar = 200 μm. (E) Example image of CLARSP neurons overlaid with the average contour representing the CLARSP core from all mice (n = 21 mice). (F) Heatmaps of normalized fluorescence of corticoclaustral axons in the CLA. Scale bar = 200 μm, 15 μm/pixel. (G) Schematic of the patching strategy used to investigate single-cortex innervation of the CLA. (H) Individual response and CTB proportions of CLA neurons to each cortex investigated in these experiments. CLARSP neurons were found to be the most responsive to frontal-cortical input and less responsive to inputs from other regions, with the exception of VISam.

Dual-color optogenetics reveals integration among CLA neurons.

(A) Schematic of injection and patching strategy in the CLA. (B) Confocal images of AAV-Chrimson-tdTomato and AAV-Chronos-GFP expression in cortical axons in the CLA. Scale bars = 500 μm, 100 μm. (C) Example traces of different response outcomes for CLA neurons after sequential cortical axon photostimulation. From top to bottom: integrating responses indicated a neuron was responsive to both input cortices, Chrimson-responsive indicated a neuron responsive to only one cortex while Chronos-responsive indicated a neuron responsive to the other cortex. Finally, no response indicated no detectable synaptic connection. (D) Dual-color optogenetics response and CTB proportions for cortices examined in Fig. 3 using the strategy in C. (E) Expected vs. measured response probabilities for neurons in each dual-color optogenetics combination. All cortical combinations displayed a slightly higher response probability than expected by chance. (F) River plot displaying the proportion of neurons projected to by each cortical area, categorized by cell type. (G) Proportion of integrating cells within each electrophysiological cell type.

In vitro measurements of CLA afferents in cortex uncover layer specificity.

(A, top) Schematic of injection and patching strategy for assessing cortex responses to photostimulation of CLA axons. (B, left) Representative images of opsin expression in CLA cell bodies. Scale bars = 1 mm (top), 200 μm (bottom). (B, right) Opsin expression in CLA axons innervating ACA (top) and RSP (bottom). Scale bars = 200 μm. (C) Normalized CLA axonal fluorescence in ACA and RSP. (D) Example image of biocytin-filled neurons in ACA. Scale bar = 200 μm. (E) Example average traces of IPSC (0 mV, blue trace) and EPSC (−70 mV, red trace) responses from a single neurons in both the ACA and RSP, aligned to light onset. (F) Pharmacological investigations of EPSC and IPSC responses to photo stimulation in normal ACSF (top) and during bath application of TTX and 4AP (middle) or TTX, 4AP, DNQX, and APV (bottom; n = 9 cells, 3 mice). (G) Quantification of normalized PSC magnitude and latency in pharmacological experiments. (H) Same as in G with neurons recorded solely in ACSF (n = 92 cells, 10 mice). (I) Expanded visualizations of currents in E demonstrating the differences in EPSC and IPSC latency in both ACA and RSP. (J) Quantification of EPSC and IPSC latency in ACA and RSP (ACA p = 0.0003, RSP p = 0.057, Cochran–Mantel–Haenszel test). (K) PSC probability in cortical neurons sorted by the layer in which neurons were patched.

In vivo responses of CLA axons in cortex to sensory stimulation.

(A, left) Schematic of injection strategy and window placement over bregma. (A, right) Schematic of in vivo recording strategy with symbols for stimuli (upper left: complex tone, upper right: white LED light, lower right: whisker stimulator). (B) Example FOV with CLA axons expressing GCaMP7b. Highlighted area and arrowhead indicate the axon from which the trace below was recorded. Scale bar = 50 μm, 10 seconds. Inset: GCaMP7b expression in CLA from approximately 0.0 mm bregma. Inset scale bar = 200 μm. (C) Passive stimulation protocol using three stimulus modalities. Stimuli and combinations thereof were presented 8-11 s apart (randomized) with a “blank” period where no stimulus was presented every ∼8 trials. (D) Average dF/F traces for each stimulus and combination across all axons responsive to that modality. Black line indicates the population mean, grey shaded regions indicate the 95% CI. (E, top left) Proportion of all recorded axons displaying significant responses to one or more trial types (n=4 mice, 1364 axons, 78 recordings). (E, top right) Proportion of all responsive axons displaying a uni-or multisensory response pattern. (E, bottom) Some axons were modulated only by multimodal trial types (left), both unimodal and multimodal trial types (center), or only unimodal trial types (right). (F) Dots represent the probability of observing a sensory evoked response to any trial type averaged across all axons, FOVs, and mice in an imaging session. Pearson’s r (0.21) indicates that response probability was not strongly correlated with experimental session. (G) Violin plots displaying the trial-by-trial response probability of axons to different trial types. (H) Violin plots displaying the trial-by-trial response magnitude of axons to different trial types.

CLA silencing reduces sensitivity to multimodal stimuli.

(A) Schematic of retrograde injection and patching strategy for assessing silencing of CLARSP terminals in cortex. (B) Example average voltage clamp electrophysiology from RSP neurons in mice injected with PBS and AAV-ChrimsonR-tdTomato (top, n = 33 cells) or AAV-FLEX-TetTox and ChrimsonR (bottom, n = 16 cells, 2 mice), aligned to light onset. (C, left) Quantification of normalized PSC magnitude in RSP neurons in Tet+ (red) and PBS (gray) conditions. (Right) Percent of neurons in each injection condition showing EPSCs in response to CLARSP axon photostimulation. (D) Experimental pipeline for behavioral experiments, beginning with injection and finishing with histological verification. Plots shaded blue correspond to the reversal learning task while plots shaded orange correspond to the multimodal conditioning task. (E) Average probability of poking the high reward probability port in the trials before and after the transition to a new block (trial 0). Dashed lines indicate chance performance and block transition (two way repeated measures ANOVA, effect of group F(1,25) = 0.413, p = 0.526). (F) Loading of logistic regression predictors based on data from the final five sessions (Welch’s t-test for previous choice p = 0.482, previous outcome p = 0.539, and interaction p = 0.581). In the multimodal conditioning task, two way repeated measures ANOVA was used to compare the latency from stimulus onset until the first poke, regardless of which port was poked, on hit (G; F(1,26) = 6.252, p = 0.019) and miss (H; F(1,26) = 3.527, p = 0.072) trials. (I) Percentage of trials classified as hits for each trial type (two way repeated measures ANOVA, effect of group F(1,26) = 0.056, p = 0.814). Dashed line indicates chance performance of 33%. (J) d’ values calculated separately for each stimulus type. Multimodal conditioning plots include data from experienced mice only (training sessions after day 3). Data from sham mice (n=12) are plotted in black and TetTox mice (n=16) are plotted in red. Error bars show the standard error of the mean. Symbols indicate an effect of treatment where p ≤ 0.1 (#), p ≤ 0.05 (*), or p ≤ 0.001 (***).

(A, left) Representative sections spanning the rostrocaudal axis of the brain in which CLA neurons were retrogradely labeled with CTB from three different injection locations along the RSP (blue: rostral RSP, red: intermediate RSP, green: caudal RSP). (A, right) Insets of CLA from sections in A, left, at 10X magnification. (B) Percent of counted neurons that project to different regions of RSP per section from experiments in A. (C) Venn diagram demonstrating the overlap between CLA cells projecting to various RSP locations. (D) Overall, caudal RSP injections yielded a higher penetrance of CLA neurons on a per slice basis. (E) Immunohistochemistry of a representative CLA section (AP = 0.0 mm bregma) against PV (magenta) and MBP (cyan) with CLARSP retro-labeled CLA neurons (green). Regional labels: CPu = caudoputamen, ec = external capsule, dCLA = dorsal CLA, vCLA = ventral CLA, DEn = dorsal endopiriform. Scale bar = 200 μm.

(A) Summary of processing steps for confocal images of the CLA (n = 3 mice). Slice images across mice were centroid-aligned and averaged for each matched slice. Active contours were initialized using an ellipse about the centroid and computed for each image. (B) 3D representation of contours of CLARSP neurons, rostral at left, colored by contour area. CLA contours are center-aligned and do not match anatomical position. (C) Probability density distributions of the euclidean distance between CLARSP neurons across the coronal axis in 50 μm sections across the rostrocaudal axis, colored by the cross-sectional area of CLARSP neurons. (D) Number of cells counted in the sections from C (n = 3 mice).

(A) Schematic of injection and patching strategy in the CLA. Both CTB+ and CTB-neurons were patched while gathering intrinsic electrophysiological profiles. (B) 40X transmitted light (top) and CTB (bottom) example images of a neuron during patching. Scale bar = 100 μm (C) Confocal microscopic images of a filled neuron post-patching. Scale bar = 50 μm. (D) Example standardized intrinsic electrophysiological trace overlaid with sections from which quality criteria values (right) are computed.

(A, left, center) K-means clustering and silhouette analysis of the dimensionally reduced excitatory neuron dataset. (A, right) Silhouette analysis found that two clusters best separated the data (k = 2, avg. silhouette score = 0.61). (B, left) Probability density of ADP voltage above spike offset for excitatory cell types. Inset shows example traces from cells with and without an ADP. (B, center) E2 cells displayed lower fAHP and higher ADP than E1 cells. (B, right) Spike amplitude for the first ten spikes, normalized to the first, for both excitatory cell types. Shaded areas represent the 95% confidence interval. ADP, fAHP, and spike amplitude adaptation reliably distinguished excitatory cell types. (C, left, center) K-means clustering and silhouette analysis of the dimensionally reduced inhibitory neuron dataset. (C, right) Silhouette analysis found that four clusters best separated the data (k = 4, avg. silhouette score = 0.85). (D) Comparison of separating features between interneuron types. (E) UMAP dimensionality reduction of all cells. (F) Proportion of all patched neurons of each putative cell type (n = 540).

(A, top) Raw confocal images of filled spiny, sparsely spiny, and aspiny neurons used for morphological reconstruction and analysis. Scale bars = 50 μm. (A, bottom) Insets of A (top) showing the presence/absence of dendritic spines (red arrows). Scale bars = 10 μm. (B) Percent of spiny (spiny + sparsely spiny) vs. aspiny neurons in the morphological dataset for which image quality was sufficient to count spines (72/134 morphologies). (C) Percent of spiny vs. aspiny neurons sorted by electrophysiological type. (D) Number of dendritic spines counted per 100 μm of dendrite across cell types. (E, left) UMAP dimensionality reduction of the morphological dataset alone, overlaid with electrophysiological cell type labels. (E, right) UMAP on the combined electrophysiological and morphological datasets. (F) Average polar histograms of CLA neurons by type centered on the soma.

All CTB+ interneurons encountered during in vitro whole-cell patch-clamp experiments, arranged by type. In two experiments, we recorded from Nkx2.1-Cre;Ai9+ animals, finding two CTB+ FS interneurons co-labeled with tdTomato.

(A) Schematic of retrograde labeling and immunohistochemistry strategy to identify projection inhibitory neurons. Mice (n = 4) were injected with either retroAAV-Cre or CTB and fixed slices were stained against Cre and GABA. (B) Immunohistochemistry of GABA against retro-Cre and CTB retrograde markers of CLA. (C) Quantification of neurons labeled by GABA in each experiment that were co-labeled with one of the two retrograde markers (n = 4 mice, 175 counted cells, p = 0.0004 Fischer Exact Test).

(A) Normalized EPSP magnitude and onset latency for intraclaustral-responsive neurons. Neurons directly expressing opsin had larger light-driven responses compared to monosynaptically connected neurons. Shaded region indicates window of monosynaptic latencies. (B) Example stimulus-evoked responses from neurons corresponding to the datapoints highlighted with dashed boxes in (A). (C) Among responsive neurons of all types, CTB+ neurons made up a small proportion. (D) Response probability for n = 46 neurons based on their dorsoventral location relative to CLARSP during patching. Neurons were less likely to respond if they occupied the CLARSP core.

(A) Schematic of VSDI experiments in horizontal sections of a rat brain. (B) Nissl stained horizontal section containing the CLA demonstrating the region of stimulation in C. (C) Voltage indicator images of the CLA in response to electrical stimulation of the rostral CLA 0, 5, 10, and 15 ms after stimulation onset. Colored arrows indicate locations used for analysis in D. (D) Traces from the pixels in c demonstrating the delay in time to peak from rostral to caudal regions of the CLA (top). DNQX + APV application to slices during experiments extinguished this response (bottom). (E) Quantification of the time to peak signal for each pixel in c and D (top). Caudal pixels show a significant delay in ACSF experiments but no response when glutamatergic signaling is blocked in DNQX + APV experiments. (F) Voltage indicator images of the CLA in response to electrical stimulation of the caudal CLA 0, 5, 10, and 15 ms after stimulation onset. Colored arrows indicate locations used for analysis in G. (G) Traces from pixels in F demonstrating the delay in time to peak from caudal to rostral regions of the CLA. (H) Quantification of the time to peak signal for each pixel in F and G (top). Scale bars = 100 μm.

Histology from posterior ACA (ACAp) injections relative to anterior ACA (ACAa; n = 3 mice per injection site).

(A) Overlay of image masks (dorsal, core, ventral) used in subsequent analysis. Scale bar = 200 μm. (B-I, top) Average heat maps and histograms of fluorescence normalized to maximum signal along the dorsoventral and mediolateral axes for each injected cortical region (n = 3 mice per cortical area; dashed line is the average contour of CLARSP neurons from all experiments). (B-I, middle) Comparison of fluorescence in each mask region for each cortical area. (B-I, bottom) Response probabilities of CLA neurons to input from cortical axons after a brief (4 ms) photostimulation of presynaptic terminals, split by soma location relative to CLARSP during patching. (J) On the basis that a cortical area primarily displayed fluorescence in one masked CLA region, “core” and “shell” labels were applied to areas that projected to the core and dorsal/ventral CLA, respectively. Heat maps from within-label areas were then averaged and subtracted from heat maps from the other label. Displayed here is an index of similarity of pixels in the CLA to the “core” or “shell.”

(A) Schematics of control experiments of single-opsin injections into PL (n = 6 mice total, three each of Chrimson and Chronos injections). (B) Example traces of CLA neuron responses to cortical axon stimulation (C) Magnitude pre/post-stim comparisons of CLA EPSPs during sequential stimulation. Cross-talk in Chrimson-only experiments was effectively eliminated by extended exposure to 500 ms of 595 nm light. Responses to blue light photostimulation in Chronos-only experiments did not show cross-talk. (D) Magnitude responses to 470 nm and 595 nm light for all recorded cells (n = 259). Integrators displayed responses to both wavelengths (595: p = 6.1e-9, 470: p = 1.7e-10, Mann Whitney U test). (E) Latency of ESPS in integrating postsynaptic neurons was within the typical monosynaptic range. Onset of Chronos EPSPs was typically faster than Chrimson.

(A) Representative high and low magnification (inset) images of sections spanning the rostrocaudal axis of the brain from mice (M1–9) used for calcium imaging. Sections were stained for parvalbumin (magenta) and GFP (yellow). Red rectangle indicates animals excluded from further analysis. Scale bars 200um. (B) Percentage of all recorded axons from each mouse in the unimodal data set displaying significant responses to one or more of the stimuli. (c) Percentage of all significantly activated axons from each mouse in the unimodal data set responsive to each stimulus type. (D) Percentage of all recorded axons from each mouse in the multimodal data set displaying significant responses to one or more of the stimuli. (E) Percentage of all significantly activated axons from each mouse in the multimodal data set responsive to each stimulus type. (F) Percentage of responsive axons from each mouse in the multimodal data set tuned to multimodal stimuli only, multi-or unimodal stimuli, and unimodal stimuli only.

(A) Passive stimulation protocol using three stimulus modalities. Stimuli were presented 8-11 s apart (randomized) with a “blank” period where no stimulus was presented every ∼4 trials. (B) Average dF/F traces for each stimulus across all axons responsive to that modality. (C) Percentage of all recorded axons displaying significant responses to one or more of the stimuli (n=4 mice, 1342 axons, 117 recordings). (D, left) Percentage of all significantly activated axons responsive to each stimulus type in the unimodal stimulus condition. (D, right) Percentage of all significantly activated axons responsive to each stimulus type in the multimodal stimulus condition. Light gray bars show the percentage exclusively responsive to each stimulus type. (E) Axonal response probability across sessions for all mice.

Immunofluorescent staining of TetTox injected (CLA) brain tissue showing parvalbumin (PV, magenta, top), myelin basic protein (MBP, cyan, top center), retro-iCre-mCherry (CRE, red, center), FLAG (yellow, bottom center), and overlay (bottom). Staining shows the full slice (A), a zoomed in view of the CLA (B), and a zoomed in view of the dorsal cortex (C). Scale bars are 1mm, 250um and 100um.

Immunofluorescent staining of TetTox injected brain tissue (A) and PBS injected brain tissue (B) showing glial fibrillary acidic protein (GFAP; top), myelin basic protein (MBP; middle), and overlay (bottom). (C) Graph showing CLA GFAP fluorescence intensity normalized against whole slice fluorescence intensity for TetTox and PBS injected brain tissue.

No change due to claustrum silencing in 24/7 activity levels nor ethologically motivated anxiety assays.

Two way repeated measures ANOVAs were used to test for a main effect or interaction of treatment group and time. (A top) Daily distance traveled in the light phase (left; effect of group F(1,10) = 0.759, p = 0.404) and dark phase (right; effect of group F(1,10) = 0.352, p = 0.587). (A bottom) Animal activity index in the light phase (left; effect of group F(1,10) = 0.045, p =0.835) and dark phase (right; effect of group F(1,10) = 1.215, p = 0.296). (B) Frequency weighted histograms showing the distribution of activity bouts by duration (seconds) in the light cycle (left) and dark cycle (right) during the baseline (top) and expression periods (bottom). Two way repeated measures ANOVA of the baseline-subtracted distribution (not shown) found no effect of group in the light cycle (F(1, 10) = 0.901, p = 0.365) or dark cycle (F(1, 10) = 0.030, p = 0.866) (C) Mean duration of activity bouts recorded in the light (left; effect of group F(1, 10) = 1.027, p = 0.335) and dark cycles (right; effect of group F(1, 10) = 0.367, p = 0.558). (D) Mean number of activity bouts recorded in the light (left; effect of group F(1, 10) = 0.533, p = 0.482) and dark cycles (right; effect of group F(1, 10) = 0.292, p = 0.601). (E) RDI from the light (left; effect of group F(1, 10) = 0.018, p = 0.896) and dark cycles (right; effect of group F(1, 10) = 0.380, p = 0.552). Mean inter-daily stability (ISm; F; effect of group F(1,10) = 1.064, p = 0.327), mean intra-daily variability (IVm; G; effect of group F(1,10) = 0.161, p = 0.697) and relative amplitude (RA; H; effect of group F(1,10) = 3.601, p = 0.087). Welch’s t-tests were used to compare the time spent in the open arms of the elevated plus maze (I; p = 0.216), time spent in the centre zone during the open field test (J; p = 0.640), and in the light chamber during the light dark box test (K; p = 0.143). Except in panel B, data from the baseline period is plotted with open bars and data from expression time period is plotted with shaded bars. Yellow and blue backgrounds indicate data collected in the light and dark cycles. Data points for animals in the TetTox group are plotted in red while data points for animals in the sham group are plotted in black. Error bars show the standard error of the mean (n = 6 per group).

Reversal learning task.

Two way repeated measures ANOVAs were used to test for a main effect or interaction of treatment group and training day. Mean number of trials completed (A; effect of group F(1,25) = 0.007, p = 0.936), rewards earned (B; effect of group F(1,25) = 0.001, p =0.981) and blocks completed (C; effect of group F(1,25) = 0.201, p = 0.658) per day of the reversal learning task. Mean latency from trial onset until mouse poked a choice port (D; effect of group F(1,25) = 0.002, p = 0.962) and from poking a choice port until poking the reward port (E; effect of group F(1,25) = 0.309, p = 0.583). Insets show a zoomed in version of the latency data from day 4 onwards.Mean length of a block in trials (F; effect of group (mixed effects model) F(1, 25) = 0.163, p = 0.689). Mean probability of choosing the same port after a rewarded trial (G; effect of group F(1,25) = 0.360, p = 0.554) or an unrewarded trial (H; effect of group F(1,25) = 0.975, p = 0.333). Data from sham mice (n=12) are plotted in black and TetTox mice (n=15) are plotted in red. Error bars show the standard error of the mean.

Multimodal conditioning task.

Two way repeated measures ANOVAs were used to compare the number of trials with each outcome across all trial types (A; effect of group F(1,26) = 1.926, p = 0.177), audio trials only (B; effect of group F(1,26) = 0.373, p = 0.547), visual trials only (C; effect of group F(1,26) = 0.805, p = 0.377), and audiovisual trials only (D; effect of group F(1,26) =1.692, p = 0.205). The number of audiovisual miss trials on which the animal chose the auditory and visual ports (E; effect of group F(1,26) = 2.145, p = 0.155. Tendency to choose the auditory port (positive values) or the visual port (negative values) on audiovisual miss trials (F; Welch’s t-test p = 0.812). Absolute value of the bias in F showing the strength of bias regardless of the preferred unimodal port (G; Welch’s t-test p = 0.951). Two way repeated measures ANOVAs were used to compare the frequency of different outcomes incorporated into the d’ for audio (H; F(1,26) = 1.351, p = 0.256), visual (I), and audiovisual stimuli (J). Boxes show the quartiles while the whiskers show the full range of the data. Note, the effect of treatment is consistent across H, I, and J. Plots include data from experienced mice only (training sessions after day 3). Data from sham mice (n=12) are plotted in black and TetTox mice (n=16) are plotted in red. Error bars show the standard error of the mean.

Select electrophysiological properties of CLA neurons. All values reported here as the mean ± the standard deviation. Abbreviations: Rin-input resistance, RMP-resting membrane potential, Thre.-spike threshold, Rheo.-rheobase current, fAHP-fast afterhyperpolarization potential, ADP-afterdepolarization potential, AP1/2-action potential half-width, Freq.-maximum recorded spike frequency, APmax-maximum spike height at rheobase, Spkdelay-delay to spike onset.

Select morphological properties of CLA neurons. All values reported here as the mean ± the standard deviation. Abbreviations: #Den.-number of primary dendrites, #Nodes-number of dendritic nodes, <Den.>-average dendritic length, Max Den.order-the highest order dendrite, <Seg.>-average dendritic segment length, ND-node density, Den.max-longest dendrite, Len.max-the polar bin of the longest dendrite.

Response probability of CLA electrophysiological cell types to inputs from individual cortices during optogenetic stimulation. Values here are reported as the number of responsive neurons of a given type divided by the total number that type recorded in all optogenetics experiments involving inputs from a given cortex. Entries shaded by gray bars indicate experiments in which < 4 neurons of that cell type were recorded.