Figures and data
The plCoA encodes innately-valenced odor identity using a population code.
(A) Schematic representation of virus injection and GRIN lens implantation into aplCoA or pplCoA for two-photon microscopy.
(B-C) Representative images (B) and traces (C) of fluorescence changes in individual neurons over an approximately 20-minute period that includes periods of odor stimulation.
(D) Schematic of odor exposure paradigm. Each trial presented 5 seconds of odor followed by a variable inter-trial interval (20-30s). Odors were present in blocks of 20 trials per odor, with 2 counterbalanced block schedules (1 & 2). Six odors were used: the appetitive odors 2-phenylethanol (2PE) and peanut oil (PEA), the neutral odors heptanol (HEP) and isoamyl acetate (IAA), and the aversive odors trimethylthiazoline (TMT) and 4-methylthiazoline (4MT).
(E) Heatmap of trial-averaged and Z-scored odor-evoked activity over time from pooled plCoA neurons. Responses are grouped by hierarchical clustering, with the dendrogram (right). Odor delivery marked by vertical red lines.
(F) Average of trial-averaged and Z-scored odor-evoked activity for each cluster concatenated. The order of color-coded blocks corresponds to the order of clusters in (E).
(G) Proportion of neurons responsive to different numbers of odors. Bars represent the mean across 13 animals and the error bars show SEM.
(H) Proportion responsive to each odor for aplCoA (red) or pplCoA (blue).
(I) Valence scores of individual neurons. White circles show the median of each distribution, whereas the gray rectangle shows the 25th-75th percentile range.
(J) Proportion of neurons with significant valence scores calculated as a function of trial number. Calculated with a 10-trial moving window. Top half shows those with significant positive valence scores, the bottom half shows those with significant negative valence scores.
(K) The percentage of neurons with responses (Z>2 for at least 5 frames) as a function of trial number for each odor.
(L) Left, MNR accuracies for all pooled plCoA neurons (data) and a control distribution where the training labels are shuffled (shuffled) in a violin plot. Right, proportion of neurons in each animal that have MNR accuracy greater than the 95th percentile of the shuffled MNRs.
(M) Cross-validated average accuracies of multinomial SVM’s plotted as a function of the number of neurons used for training during the odor period. Circles represent the mean across 100 iterations of random sampling of neurons and error bars show the standard deviation.
(N) Cross-validated accuracy of ecoc-SVM classifiers for a 6-odor classification task trained using 200 neurons as a function of time. Lines indicate means and shaded areas show the standard deviation across 100 random samplings of 200 neurons from the pooled data, and shuffled training controls where the label vectors are randomly shuffled.
(O) An example confusion matrix for a multinomial SVM trained with 200 neurons.
(Q) Comparison of inter-valence and intra-valence confusion across number of neurons used in training the classifiers. Filled circles show the average of the data across 100 iterations, open circles show shuffled controls.
(4) The normalized average distance between odor pairs that have different valence (inter) or same valence (intra).
Across panels, ns, not significant. Additional specific details of statistical tests can be found in Supplemental Table 1.
The plCoA has topographic organization capable of driving approach and avoidance behaviors.
(A) Schematic of plCoA domains divided into anterior (aplCoA), middle (mplCoA), and posterior (pplCoA) regions based on histology, positioning, and gradients observed in past observations[23].
(B) Strategy to activate anterior-posterior topographical ensembles via optogenetics.
(C) Representative histology and fiber/virus placement for aplCoA and pplCoA ChR2 animals.
(D) Schematic of four-quadrant open field behavioral assay with closed-loop photostimulation. (E-F) Linear-fit of change in performance index (E) or mean port distance (F) as a function of anterior-posterior position along plCoA for optical stimulation.
(G) Paths traveled during the stimulus period for a representative mouse (left) and baseline-normalized collective heatmaps (right) from both the ChR2- and eYFP-infected groups with aplCoA-localized fiber implants. Lower right stimulus quadrant indicated in blue.
(H-I) Mean effect of Photostimulation of aplCoA neurons on time spent in stimulated quadrant (performance index) (K) and distance from the corner (port distance) (I).
(J) Paths traveled during the treatment period for a representative mouse (left) and baseline-normalized collective heatmaps (right) from both the ChR2- and eYFP-infected groups with pplCoA-localized fiber implants. Lower right stimulus quadrant marked in blue.
(K-L) Mean effect of photostimulation of pplCoA neurons infected with ChR2, but not eYFP, is sufficient to increase time spent in the stimulation quadrant (K) and reduce its average distance from the stimulation port during the stimulation period (L).
Abbreviations: aplCoA, anterior zone of posterolateral cortical amygdala; mplCoA, middle zone of posterolateral cortical amygdala; pplCoA, posterior zone of posterolateral cortical amygdala. Across panels, ns, not significant; * p < 0.05; ** p < 0.01; *** p < 0.001. Specific details of statistical tests can be found in Supplemental Table 1.
Transcriptomic heterogeneity of plCoA molecular cell types.
(A) Schematic of freeze and-re-pool strategy for snRNA-seq.
(B) Two-dimensional UMAP (n = 47,132 nuclei, see also Figure S4.3), colored by broad cellular identity assigned by graph-based clustering of neuronal and non-neuronal nuclei.
(C) Cell-type-specific expression of canonical marker genes indicating broad cellular identity in the brain. Dot size is proportional to percentage of nuclei expressing the marker, with color scale representing normalized expression level.
(D) Total proportion of cells of each identified type in each domain of plCoA.
(E) Two-dimensional UMAP of glutamatergic neurons, colored by molecular cell type.
(F) Clustered heatmap showing Euclidean distance between averages of each subtype positioned based on hierarchical clustering (left), and dot plot of marker genes for all glutamatergic subtypes (right).
(G) Two-dimensional UMAP of GABAergic neurons, colored by molecular cell type, like in (E).
(H) Clustered heatmap showing Euclidean distance between averages of each subtype positioned based on hierarchical clustering (left), and dot plot of marker genes for all GABAergic subtypes (right), like in (F).
Glutamatergic neurons subtypes in plCoA are spatially distributed along an anteroposterior molecular gradient.
(A) UMAP of all plCoA nuclei colored by zone of origin, with dotted outlines and labels denoting the major cell types.
(B) Relative proportion of nuclei from each domain within each broad identity class. Dotted line indicates chance level for all plCoA nuclei.
(C) Top, abundance of domain-specific DEGs for each major cell type, either enriched in aplCoA nuclei (top) or pplCoA nuclei (bottom). Bottom, volcano plots for domain-specific DEGs for glutamatergic (left) and GABAergic neurons (right), the two cell types with the greatest degree of domain specific gene expression, where negative log-fold changes indicate enrichment in pplCoA and positive log-fold changes indicate enrichment in aplCoA.
(D) UMAP of plCoA glutamatergic neurons colored by domain of origin, with dotted outlines and labels denoting the subtypes on the graph. Groups of glutamatergic neuron types identified previously via Euclidean distance and hierarchical clustering are overlaid on top of the neuron types of interest.
(E) Relative proportion of molecular subtypes from each domain within glutamatergic neurons, where relevant subtypes are outlined according to their glutamatergic neuron group. Dotted line indicates chance level for plCoA glutamatergic neuron nuclei.
(F) UMAP of all glutamatergic neuron nuclei, colored by expression levels of VGluT2 (top) or VGluT1 (bottom).
(G) Left, representative images of in situ RNAscope labeling of VGluT2 RNA (red) and VGluT1 RNA (green) across plCoA domains. Right, proportions of glutamatergic neurons expressing VGluT2, VGluT1, or both. Scale bars, 500 µm (main image), 50 µm (inset).
(H) UMAP of all plCoA-overlapping Visium capture spots, colored by cluster. Broad spatial position of groups of clusters are overlaid on top of the capture spots of interest.
(I) UMAP of all plCoA-overlapping Visium capture spots, colored by expression levels of VGluT2 (top) or VGluT1 (bottom).
(J) Representative plCoA-overlapping region of one section on a Visium slide capture area, with capture spots colored by cluster.
(K) Representative plCoA-overlapping region of one section on a Visium slide capture area, with capture spots colored by expression levels of VGluT2 (top) or VGluT1 (bottom).
(L) Prediction scores for representative glutamatergic neuron subtypes within Group 1 (left) and Group 2 (right), shown on a UMAP of all plCoA-overlapping capture spots across all sections (top) and on a representative plCoA-overlapping region of one section (bottom).
(M) Prediction scores for a representative GABAergic neuron subtype, shown on a UMAP of all plCoA-overlapping capture spots across all sections (top) and on a representative plCoA-overlapping region of one section (bottom).
Across panels: * p < 0.05; ** p < 0.01; *** p < 0.001; ns, not significant. Additional specific details of statistical tests can be found in Supplemental Table 1.
Glutamatergic plCoAVGluT2+ and plCoAVGluT1+ neurons contribute to innate approach and avoidance behaviors.
(A) Schematic for selective photostimulation of distinct glutamatergic cell type. VGluT2::Cre and VGluT1::Cre animals were injected with Cre-dependent viral vectors into mplCoA with a fiber optic implant placed just above the injection site.
(B) Representative histology from ChR2 viral injection and fiber implantation site in an VGluT2::Cre animal (top) and an VGluT1::Cre animal (bottom).
(C) Baseline-normalized collective heatmaps from both the ChR2- and eYFP-infected groups in VGluT2::Cre and VGluT1::Cre animals with plCoA-localized fiber implants. Lower right stimulus quadrant marked in blue.
(D-G) Effect of photostimulation of plCoAVGluT2+ neurons (D-E) or plCoAVGluT1+ neurons (F-G) on time spent in the stimulation quadrant (D, F) and distance from the corner (E, G).
(H) Behavioral paradigm to assess innate valence responses to odor. Left, schematic of four-quadrant open field behavioral assay for spatially-specific odor delivery. Upper right, within-trial timeline. Lower right, odors delivered and their associated innate valence.3
(I) Schematic of strategy for selective chemoinhibition of molecularly defined glutamatergic plCoA neurons.
(J-M) Effect of chemoinhibition of plCoA VGluT1+ neurons on time spent in the odor quadrant (J, L) or decrease in mean port distance (K, M) in response to 2PE (J-K) or TMT (L-M).
(N-Q) Effect of chemoinhibition of plCoA VGluT2+ neurons on time spent in the odor quadrant (N, P) or decrease in mean port distance (O, Q) in response to 2PE (N-O) or TMT (P-Q).
Across panels, ns, not significant; * p < 0.05; ** p < 0.01; *** p < 0.001, **** p < 0.0001. Additional specific details of statistical tests can be found in Supplemental Table 1.
Projections to MeA and NAc from plCoA are Topographically Organized
(A) Left, whole-hemisphere view at AP = 0.98 mm from bregma. Scale bar, 500 µm. Right, Magnified images of the areas highlighted inside white dashed lines. Scale bar, 200 µm.
(B) Left, whole-hemisphere view at AP = -1.06 mm from bregma. Scale bar, 500 µm. Right, Magnified images of the areas highlighted inside white dashed lines. Scale bar, 200 µm.
(C) Other plCoA projections not found in cross-sections of the brain found in (A) and (B). Scale bar, 200 µm.
(D) Magnitude of anterograde synaptophysin-eYFP fluorescence in primary downstream targets of plCoA projection neurons ordered by total output strength, colored based on each region’s function.
(E) Schematic for topographic retrograde mapping strategy from MeA and NAc into plCoA. Red retrobeads are injected into MeA or NAc and topographical projection bias is examined along the anterior-posterior axis.
(F) Representative images (top) for injection into MeA (left) or NAc (right) and number of neurons labeled along the anterior-posterior axis as distance (mm) from bregma (bottom). Gray lines denote individual replicates, where colored lines indicate mean ± s.e.m.
(G) Proportion of retrobead-labeled neurons projecting to MeA or NAc for each 100 µm segment as a function of distance from bregma. Dashed line indicates overall balance of all retrobead-labeled neurons across entire plCoA.
(H) Proportion of retrobead-labeled neurons from either target within each plCoA zone. MeA-labeled neurons are significantly enriched in aplCoA compared to NAc-labeled neurons, while NAc-labeled neurons are significantly enriched in pplCoA compared to those labeled from MeA.
(I) Representative histological images for the injection sites in aplCoA (left) and pplCoA (right) from a representative animal. Scale bar, 200 µm.
(J) Representative histological images for MeA from the animal in (J). Scale bar, 200 µm.
(K). Representative histological images for NAc from the animal in (J). Scale bar, 200 µm.
(L) Output strength as a proportion of total fluorescence from aplCoA and pplCoA to MeA and NAc.
(M) Representative histological images for the injection site in plCoA from a representative VGluT1::Cre and VGluT2::Cre animal. Scale bar, 200 µm.
(N) Representative histological images from MeA and NAc from a representative animal of either genotype. Scale bar, 200 µm.
(O) Left, output strength as a proportion of total fluorescence from plCoAVGluT2+ and plCoAVGluT1+ neurons to MeA and NAc. Right, comparison of same data, but by target region within genotype.
(P) Same data as (O), but by target region within genotype.
(Q-V) Mapping collateral projections from NAc- and MeA projecting neurons.
(Q) Representative histological images for the injection site in plCoA from a representative animal receiving retrograde virus into MeA or NAc. Scale bar, 200 µm.
(R) Representative histological images of NAc and MeA retro-Cre targeting (red) and outputs (green).
(S) Comparison of absolute integrated fluorescence intensities in MeA and NAc when retroAAV was injected into NAc (top) or MeA (bottom).
(T) Quantification of fluorescence in selected downstream brain regions from plCoA originating from plCoA-NAc neurons proportional to eYFP fluorescence in NAc (top) or MeA (bottom). Abbreviations: NAc, nucleus accumbens; BNST, bed nucleus of stria terminalis; MeA, medial amygdala; Pir, piriform cortex; BLA, basolateral amygdala; Ahi, amygdalo-hippocampal transition area; pmCoA, posteromedial cortical amygdala; Str, striatum; OT, olfactory tubercle; EA, extended amygdala; IPAC, inferior peduncle of the anterior commissure; AA, anterior amygdala; LA, lateral amygdala; HDB, horizontal limb of the diagonal band; VP, ventral pallidum; AIC, anterior insular cortex; mfb, medial forebrain bundle; MO, medial orbitofrontal cortex; LOT, lateral olfactory tract; ACo, anterior cortical amygdala; AOA, anterior olfactory area; DG, dentate gyrus; Rt, reticular nucleus; LPO, lateral preoptic area; VMH, ventromedial hypothalamus; DEn, dorsal endopiriform claustrum; LH, lateral hypothalamus; IL, infralimbic cortex; DP, dorsal peduncular cortex; LS, lateral septum; CxA, cortex-amygdala transition area; sox, supraoptic decussation; StHy, striohypothalamic nucleus; GP, globus pallidus; PLH, perirhinal cortex; ZI, zona incerta.
Across panels, ns, not significant; * p < 0.05; ** p < 0.01; **** p < 0.0001. Additional specific details of statistical tests can be found in Supplemental Table 1.
Projections from plCoA to NAc and MeA control innate olfactory attraction and aversion.
(A) Schematic for optogenetic MeA terminal stimulation in plCoA neurons. Strategy to activate MeA-projecting plCoA neuron terminals via optogenetics (top) and representative histology from ChR2 viral injection and fiber implantation site (bottom).
(B) Baseline-normalized collective heatmaps from both the ChR2- and eYFP-infected plCoA groups with MeA- and NAc-localized fiber implants. Lower right stimulus quadrant marked in blue.
(C-D) Optogenetic MeA terminal stimulation of plCoA neurons infected with ChR2, but not eYFP, is sufficient to reduce time spent in the stimulation quadrant (C) and increase its average distance from the port (D) during the stimulation period.
(E-F) Optogenetic NAc terminal stimulation of plCoA neurons infected with ChR2, but not eYFP is sufficient to increase time spent in the stimulation quadrant (E) and decrease its average distance from the stimulation port (F) during the stimulation period.
(G) Viral strategy for selective retrograde chemoinhibition of projection-defined plCoA neurons.
(H) Schematic for selective retrograde chemoinhibition of projection-defined plCoA neurons.
(I-L) Chemoinhibition of NAc-projecting plCoA neurons significantly eliminates the 2PE-evoked increase in time spent in the odor quadrant (I) and decreases in mean port distance (J). The response to TMT is unaffected in time spent in odor quadrant (K) or port distance (L).
(M-P) Chemoinhibition of MeA-projecting plCoA neurons does not affect 2PE-evoked increase in time spent in the odor quadrant (M) or decrease in mean port distance (N) significantly decreases the TMT-evoked reduction in time spent in the odor quadrant (O) or increase in mean port distance (P).
Across panels, ns, not significant; * p < 0.05; ** p < 0.01; *** p < 0.001. Additional specific details of statistical tests can be found in Supplemental Table 1.
Models that could support innate odor evoked attraction and avoidance.
Two potential models that could support valence responses to odor with a population code imposed on divergent circuitry. (A) Balance of activation between MeA- and NAc-projecting neurons determines the valence. In this model an odor may activate a different proportion of these projection defined neurons, and the valence is determined by the balance. For example activation of more NAc-projecting neurons should cause attraction, and activation of more MeA-projecting neurons should cause aversion.
(B) Dynamic activity evolves overtime due to recurrent processing or integration of behavioral state variables, in an attractor-like network. In this case the activity may evolve from an initional broad population code (T0) towards preferential activation of one output population over time (Tn).
