Multimodal mapping of cell types and projections in the central nucleus of the amygdala
- Janelia Research Campus, Howard Hughes Medical Institute, United States
- Friedrich Miescher Institute for Biomedical Research, Switzerland
- German Center for Neurodegenerative Diseases (DZNE), Germany
- Howard Hughes Medical Institute & Department of Neurosciences, University of California, San Diego, United States
Figures
Figure 1 with 3 supplements
Central nucleus of the amygdala (CEA) single-cell RNA-sequencing (scRNA-seq) data analysis.
(A) UMAP for molecularly defined neuron clusters in the CEA, with cell types color-coded by scRNA-seq clusters. A total of 1,643 cells were subjected to single-cell RNA sequencing, with an average depth of 6,611,566 ± 92,440 (mean ± SD) reads per cell. Among them, 1,393 CEA neurons were identified. (B) Heatmap of marker genes from scRNA-seq clusters. Colormap indicates z-score normalized expression. (C) Dendrogram representing transcriptional relationships of molecularly defined neuronal types from scRNA-seq. (D) Fraction of neurons in each scRNA-seq cluster. (E) Percent of correctly mapped neurons with increasing numbers of marker genes. The dotted vertical line indicates the cutoff for marker genes selected in this study. The top 100 most differentially expressed marker genes were included for this analysis. (F) Heatmap showing the fraction of neurons that were correctly assigned to their original scRNA-seq cluster using 29 selected marker genes.
Figure 1—figure supplement 1
Microdissection of central nucleus of the amygdala (CEA) tissue for single-cell RNA sequencing (scRNA-seq).
(A) Bilateral expression of AAV2/1-CAG-GFP in the parabrachial nucleus (PBN). (B) Afferent fibers from the PBN specifically label the CEA but not surrounding nuclei such as the basolateral amygdala (BLA), the medial amygdala (MEA), or the amygdalostriatal transition area (AST) of the ventral caudate putamen (CP). The CEA was selectively dissected and processed to obtain single neurons for sequencing. (C) Zoom-in view of tissue before and after CEA microdissection for scRNA-seq. (D) Mapping of all CEA microdissections for scRNA-seq. Data from seven animals. AAA: anterior amygdala area; BLA: basolateral amygdala; CEA: central amygdala; CP: caudate putamen; FS: fundus of the striatum; GPe: globus pallidus external; IA: intercalated amygdala nucleus; MEA: medial amygdala; SI: substantia innominata. Atlas schemes reproduced from the Allen Brain Reference Atlas (Allen Institute for Brain Science, 2011; http://atlas.brain-map.org/) with coronal sections shown. Reproduced with permission from The Allen Institute, copyright holder The Allen Institute, copyright year 2011. It is not covered by the CC-BY 4.0 license, and further reproduction of this panel would need permission from the copyright holder. (E) Silhouette scores (top) and Jaccard similarity index (bottom) of scRNA-seq clusters. (F–H) Example images showing differentially expressed genes identified in scRNA-seq cluster seq-c1 (F), seq-c4 (G), and seq-c12 (H). Scale bars: 840 µm. (F) Example images reproduced from the Allen Institute for Brain Science, Allen Reference Atlas – Mouse Brain Sagittal Atlas (http://atlas.brain-map.org/, specific dataset URLs: https://mouse.brain-map.org/experiment/show/69818338, https://mouse.brain-map.org/experiment/show/73573289, https://mouse.brain-map.org/experiment/show/100143572, https://mouse.brain-map.org/experiment/show/69734875, https://mouse.brain-map.org/experiment/show/69013272, Lein et al., 2007). (G) Example images reproduced from Allen Institute for Brain Science, Allen Reference Atlas – Mouse Brain Sagittal Atlas (http://atlas.brain-map.org/, specific dataset URLs: https://mouse.brain-map.org/experiment/show/77465048, https://mouse.brain-map.org/experiment/show/69817199, Lein et al., 2007). (H) Example images reproduced from Allen Institute for Brain Science, Allen Reference Atlas – Mouse Brain Coronal Atlas (http://atlas.brain-map.org/, specific dataset URLs: https://mouse.brain-map.org/experiment/show/509, https://mouse.brain-map.org/experiment/show/73592534, Lein et al., 2007).
© 2011, The Allen Institute. Atlas schemes in Figure 1 - figure supplement 1D reproduced from the Allen Brain Reference Atlas (Allen Institute for Brain Science, 2011) http://atlas.brain-map.org/, with coronal sections shown. It is not covered by the CC-BY 4.0 licence and further reproduction of this panel would need permission from the copyright holder.
© 2011, The Allen Institute. Example images in Figure 1 - figure supplement 1E are reproduced with permission from The Allen Institute for Brain Science, Allen Reference Atlas - Mouse Brain Sagittal Atlas http://atlas.brain-map.org/, specific dataset URLs https://mouse.brain-map.org/experiment/show/69818338, https://mouse.brain-map.org/experiment/show/73573289, https://mouse.brain-map.org/experiment/show/100143572, https://mouse.brain-map.org/experiment/show/69734875, https://mouse.brain-map.org/experiment/show/69013272, Lein et al., 2007. It is not covered by the CC-BY 4.0 licence and further reproduction of this panel would need permission from the copyright holder.
© 2011, The Allen Institute. Example images in Figure 1 - figure supplement 1F are reproduced with permission from The Allen Institute for Brain Science, Allen Reference Atlas - Mouse Brain Sagittal Atlas http://atlas.brain-map.org/, specific dataset URLs https://mouse.brain-map.org/experiment/show/77465048, https://mouse.brain-map.org/experiment/show/69817199, Lein et al., 2007. It is not covered by the CC-BY 4.0 licence and further reproduction of this panel would need permission from the copyright holder.
© 2011, The Allen Institute. Example images in Figure 1 - figure supplement 1G are reproduced with permission from Allen Institute for Brain Science, Allen Reference Atlas - Mouse Brain Sagittal Atlas http://atlas.brain-map.org/, specific dataset URLs https://mouse.brain-map.org/experiment/show/509, https://mouse.brain-map.org/experiment/show/73592534, Lein et al., 2007. It is not covered by the CC-BY 4.0 licence and further reproduction of these panels would need permission from the copyright holder.
Figure 1—figure supplement 2
UMAP showing the expression of selected genes from single-cell RNA sequencing (scRNA-seq) data.
Color maps indicate normalized expression. UMI count is first normalized by the total counts in that cell, multiplied by 10,000 and then natural log transformed.
Figure 1—figure supplement 3
Integration of central nucleus of the amygdala (CEA) single-cell RNA-sequencing (scRNA-seq) data from two datasets.
(A–C) UMAP for molecularly defined neuron clusters in the CEA, with cells color-coded by their data source (A), original molecular cluster identity (B), and integrated new molecular cluster identity (C). (D) Jaccard similarity index between original CEA clusters and the integrated new CEA clusters. (E) Heatmap showing marker gene expression in integrated data. Cells are grouped by their new cluster identity. Colormap indicates z-score normalized expression. (F) Dendrogram showing the relationships between new scRNA-seq clusters. (G) Violin plots showing the number of genes detected and total RNA count across new clusters from the integrated data.
Figure 2 with 3 supplements
Method to combine expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH) with projection class mapping in the central amygdala.
(A) Schematics of the procedure. Fluorophore-labeled cholera toxin subunit B (CTbs) and fluorogold were used to retrogradely trace neurons in the central nucleus of the amygdala (CEA) that project to bed nucleus of the stria terminalis (BNST), lateral substantia nigra (SN), ventrolateral periaqueductal gray (vlPAG), parabrachial nucleus (PBN), and parvocellular reticular formation (PCRt). Confocal images were collected to identify projection classes in the CEA. Retrograde tracer fluorescence signals were then eliminated from tissue sections and the EASI-FISH procedure was performed as described in Wang et al., 2021. (B) Representative image showing the molecularly defined and projection-defined cell types in the CEA (left). Right: same as left, with projection types only. Dotted line: borders between the CEA and surrounding brain areas. AST: amygdalostriatal transition area; BLA: basolateral amygdala. Scale bars: 50 µm. (C) Zoom-in of the box in (B) showing registration between confocal image and EASI-FISH images (top). This allows the extraction of projection signals using EASI-FISH segmentation mask (bottom). Scale bars: 50 µm.
Figure 2—figure supplement 1
Retrograde tracer labeling in the central nucleus of the amygdala (CEA).
(A, B) Comparison of retrograde labeling in the CEA with AAV2/retro and cholera toxin subunit B (CTb). CTb-555 and AAV2/retro were mixed at 1:1 ratio and co-injected via the same glass pipette. Example injection sites are shown on the left and representative images in the CEA with AAV2/retro-GFP labeling (middle) and CTB-555 labeling (right) are shown. CTb labeled a larger population of parabrachial nucleus (PBN)-projecting (A) and periaqueductal gray (PAG)-projecting (B) neurons in the CEA compared to AAV2/retro. (C) Co-injection of CTB-555 and FluoroGold into the PBN. Note that there is a larger spread of FluoroGold at the injection site. Scale bar: 500 µm. (D) Retrogradely labeled neurons in the CEA. Most neurons were double-labeled with CTB-555 and FluoroGold (78.2%). In accordance with the larger spread of FluoroGold, a fraction of neurons was only labeled with FluoroGold (16.8%, white arrows), while a small proportion showed selective CTB-555 labeling (5.1%, yellow arrow). Data from n = 3 sections from N = 1 mouse. Scale bar: 10 µm.
Figure 2—figure supplement 2
Retrograde tracer labeling of central nucleus of the amygdala (CEA) neurons for five projection targets.
(A) Example images of injections (ANM #2) into the bed nucleus of the stria terminalis (BNST), the lateral part of the substantia nigra (lateral SN), the ventrolateral periaqueductal grey (vlPAG), the parabrachial nucleus (PBN), and the parvocellular reticular nucleus (PCRt). Scale bars: 500 µm. (B) Systematic mapping of all injection sites. Data from three animals. Atlas schemes reproduced from the Allen Brain Reference Atlas (Allen Institute for Brain Science, 2011; http://atlas.brain-map.org/), with coronal sections shown.
© 2011, The Allen Institute. Example images in Figure 2 - figure supplement 2B are reproduced with permission from Allen Institute for Brain Science, Allen Reference Atlas - Mouse Brain Sagittal Atlas http://atlas.brain-map.org/, specific dataset URLs https://mouse.brain-map.org/experiment/show/509, https://mouse.brain-map.org/experiment/show/73592534, Lein et al., 2007. It is not covered by the CC-BY 4.0 licence and further reproduction of these panels would need permission from the copyright holder.
Figure 2—figure supplement 3
Method validation.
(A) Representative images showing fluorescence tracer labeled neurons (CTB-555 in red, FluoroGold in green) before tissue clearing (left), after tissue clearing (middle), and after ExM (right). Scale bars: 20 µm. (B) Central nucleus of the amygdala (CEA) neurons labeled with fluorescence tracers were first detected after tissue clearing (Before ExM, left). Then the tissue was subjected to ExM procedures to remove fluorescence tracers and free up all fluorescence channels for expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH) (After ExM, right). Scale bars: 20 µm. (C) Tissue clearing does not compromise RNA quality. Ezr spot count in Prkcd+ neurons in the CEA, Igf1 spot count in Pmch+ neurons in the lateral hypothalamus with or without tissue clearing. Error bars indicate mean ± SD. (D) Workflow to register pre-ExM projection images with post-ExM EASI-FISH images. (E) Representative images showing registration between projection images and EASI-FISH images. DAPI-stained nuclei were used to register confocal images and round 0 EASI-FISH image. Fluorescence in situ hybridization with a ribosomal RNA probe was used to register round-0 EASI-FISH image (before DNase I) and cytoDAPI in subsequent rounds of EASI-FISH images (after DNase I treatment). (F) Boxplot showing summarized fraction of Oprk1, Scn4b, and Vipr2 spots detected in round 11/12 relative to round 1 across four samples. (G) Spot counts of Oprk1, Scn4b, and Vipr2 detected in round 1 and round 11 from the same neurons of one EASI-FISH sample. (H) Spot counts of Scn4b detected in round 1 and round 12 from the same neurons of one EASI-FISH sample. Gray lines in (G) and (H): linear regression fit (equation and R2 are shown on the plot).
Figure 3 with 2 supplements
Central nucleus of the amygdala (CEA) expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH) gene expression profiling.
(A) UMAP for molecularly defined EASI-FISH clusters in the CEA. (B) Heatmap of 29 FISH marker genes in EASI-FISH clusters. Colormap indicates z-score normalized spot count. (C) The proportion of neurons from EASI-FISH clusters assigned to scRNA-seq clusters based on cross-correlation of marker gene expression. (D, E) Based on FISH cluster assignment to scRNA-seq clusters and their spatial location, scRNA-seq clusters belonging to separate branches of the dendrogram (from Figure 1C) mapped to separate CEA subregions. (D) Dendrogram showing the gene-expression relationships of scRNA-seq clusters, same as Figure 1C, with mapped subregions colored (CeC and CeL: red, CeM: green). (E) Spatial distribution of EASI-FISH clusters mapped to scRNA-seq clusters on the separate branches of the dendrogram (class 1: red; class 2: green). (F) Morphological properties of somata in EASI-FISH clusters. Top: soma volume; middle: solidity; bottom: the ratio between the minor axis and the major axis of an ellipse fitting the cell outline. Dotted lines: population average. p-values with medium and high effect sizes (Cohen’s d>0.5, or rg>0.28) are shown. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
Figure 3—figure supplement 1
Central nucleus of the amygdala (CEA) expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH) data analysis.
(A) UMAP for neurons and non-neurons in the CEA, measured with EASI-FISH. Cells were color-coded based on initial clustering analysis to split neurons and non-neurons from the CEA, as well as neurons from surrounding brain regions. (B) Violin plots showing the number of genes detected, total marker gene spot count, Gad1 spot count, and cell body volume across clusters. Note that cluster 1 has fewer genes detected, low total marker gene spot count, low Gad1 spot count, and smaller cell body volume. It is thus identified as the non-neuronal cluster and removed from subsequent analysis. (C) UMAP for molecularly defined cell types in the CEA, as in Figure 3A, with neurons separated by sample, showing the distribution of clusters in different samples from two animals. (D) Prkcd expression (spot count) in EASI-FISH clusters, shown in UMAP. (E) Summary table showing correspondence between EASI-FISH cluster and scRNA-seq clusters; also see Figure 3—figure supplement 1—source data 1. (F, G) Differentially expressed membrane-protein-encoding genes in seq-c2 (F) and seq-c3 (G) (corresponding to MC-1 and MC-9, respectively) that showed distinct soma size and shape. Colormaps indicate z-score normalized data.
-
Figure 3—figure supplement 1—source data 1
Table summarizing the correspondence between expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH) clusters and single-cell RNA-sequencing (scRNA-seq) clusters, related to Figure 3—figure supplement 1E.
- https://cdn.elifesciences.org/articles/84262/elife-84262-fig3-figsupp1-data1-v2.xlsx
Figure 3—figure supplement 2
UMAP showing marker gene expression as measured by expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH).
Colormaps indicate log10 transformed spot counts for each gene.
Figure 4 with 4 supplements
Expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH) spatial analysis.
(A) Representative diagrams detailing the anatomical parcellation procedure in the central nucleus of the amygdala (CEA). First, marker genes (Ppp1r1b, Penk, and Nefm) identified from principal component analysis (PCA) were used to classify neurons and create the anatomical parcellation using the probabilistic Gaussian process classification. BLA: basolateral amygdala; AST: amygdalostriatal transition area; PAL: pallidum; ITCs: the intercalated cells of the amygdala; MEA: medial amygdala. (B) Anatomical parcellations from anterior (top), middle (middle), and posterior (bottom) CEA from one animal (ANM #1). (C) EASI-FISH cluster enrichment in the parcellated subregions. Gray circles: fractions of neurons profiled that belong to selected clusters. Color bar: fraction of MC neurons in each subregion. (D) Spatial distribution of molecularly defined neuron types enriched in parcellated subregions. Colors represent cluster identity. All neurons are colored in light gray in the background. Scale bars in (A, B, D): 200 µm. M: medial, L: lateral, D: dorsal, V: ventral.
-
Figure 4—source data 1
Spatial distribution of central nucleus of the amygdala (CEA) clusters.
Each dot indicates the centroid position of a neuron. Each column is a sample from selected animal. Panels are maximum axial projections of the entire volumes.
- https://cdn.elifesciences.org/articles/84262/elife-84262-fig4-data1-v2.pdf
Figure 4—figure supplement 1
Central nucleus of the amygdala (CEA) spatial analysis.
(A, B) Principal component analysis (PCA) of the spatial distribution of marker genes in animal 1 (ANM #1) (A) and animal 2 (ANM #2) (B). Eigen images from the top 4 principal components (PCs) are shown, with explained variance from each PC shown on top and loadings of marker genes for each PC on the right. Scale bars: 200 µm. M: medial; L: lateral; D: dorsal; V: ventral.
Figure 4—figure supplement 2
Spatial parcellation in the central nucleus of the amygdala (CEA).
(A, C) Anatomical parcellations in anterior (left), middle (middle), and posterior (right) CEA from animal 1 (ANM #1) (A) and animal 2 (ANM #2) (C). Scale bars: 200 µm. M: medial; L: lateral; D: dorsal; V: ventral. (B, D) Expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH) cluster enrichment in the parcellated subregions in animal 1 (B) and animal 2 (D). Gray circles: fractions of neurons profiled that belong to selected clusters. Color bar: fraction of MC neurons in each subregion.
Figure 4—figure supplement 3
Molecularly defined neuron types enriched in central nucleus of the amygdala (CEA) subregions.
(A) Molecularly defined neuron type compositions in CEA subregions and surrounding brain regions. (B) Spatial overlap of molecularly defined neuron types.
Figure 4—figure supplement 4
Expression of selected neuromodulatory GPCRs in the central nucleus of the amygdala (CEA).
(A) Dot plot showing expression of selected GPCRs in the CEA expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH) molecular clusters. The size of each dot indicates the percentage of cells that expressed the selected GPCR; the colors represent the average spot count in each cluster, with darker color indicating higher expression levels. (B) Distribution of the selected GPCR expression in the CEA. Six samples from two representative animals are shown. The colors represent z-score normalized spot counts in each cell, with darker red color indicating higher expression levels. Scale bar: 200 µm. M: medial; L: lateral; D: dorsal; V: ventral.
Figure 5 with 1 supplement
Projection of central nucleus of the amygdala (CEA) molecular clusters to five downstream targets.
(A) Representative images showing neurons projecting to bed nucleus of the stria terminalis (BNST), lateral substantia nigra (SN), parabrachial nucleus (PBN), parvocellular reticular formation (PCRt), and ventrolateral periaqueductal gray (vlPAG). Note that neurons with collaterals (n = 281) are not shown to avoid confusions. Scale bar: 200 µm. (B) Molecularly defined CEA neuron types projecting to five downstream brain regions. Line thickness represents percent of neurons from selected MCs projecting to selected brain region. (C) Representative image showing the distribution of dominant PBN-projecting clusters, MC-3 and MC-16 and their marker gene expressions. Scale bars: 100 µm (left) and 50 µm (right). (D) Representative images showing a subset of Pdyn/Sst co-expressing neurons that project to the PBN (top). These neurons are Vipr2, Tac2, Nts, and Crh-negative (MC-3) and are localized more laterally (bottom). Scale bars: 50 µm.
-
Figure 5—source data 1
Spatial distribution of bed nucleus of the stria terminalis (BNST)-projecting central nucleus of the amygdala (CEA) clusters.
Each dot indicates the centroid position of a neuron. Each column is a sample from a selected animal. Panels are maximum axial projections of the entire volumes.
- https://cdn.elifesciences.org/articles/84262/elife-84262-fig5-data1-v2.pdf
-
Figure 5—source data 2
Spatial distribution of lateral SN-projecting central nucleus of the amygdala (CEA) clusters.
Each dot indicates the centroid position of a neuron. Each column is a sample from a selected animal. Panels are maximum axial projections of the entire volumes.
- https://cdn.elifesciences.org/articles/84262/elife-84262-fig5-data2-v2.pdf
-
Figure 5—source data 3
Spatial distribution of parabrachial nucleus (PBN)-projecting central nucleus of the amygdala (CEA) clusters.
Each dot indicates the centroid position of a neuron. Each column is a sample from a selected animal. Panels are maximum axial projections of the entire volumes.
- https://cdn.elifesciences.org/articles/84262/elife-84262-fig5-data3-v2.pdf
-
Figure 5—source data 4
Spatial distribution of parvocellular reticular formation (PCRt)-projecting central nucleus of the amygdala (CEA) clusters.
Each dot indicates the centroid position of a neuron. Each column is a sample from a selected animal. Panels are maximum axial projections of the entire vol umes.
- https://cdn.elifesciences.org/articles/84262/elife-84262-fig5-data4-v2.pdf
-
Figure 5—source data 5
Spatial distribution of ventrolateral periaqueductal gray (vlPAG)-projecting central nucleus of the amygdala (CEA) clusters.
Each dot indicates the centroid position of a neuron. Each column is a sample from a selected animal. Panels are maximum axial projections of the entire volumes.
- https://cdn.elifesciences.org/articles/84262/elife-84262-fig5-data5-v2.pdf
Figure 5—figure supplement 1
Molecularly defined neuron types project to different brain regions.
(A) Fraction of neurons projecting to selected five brain regions in different samples from two animals. (B) Fraction of neurons from each molecular cluster that projects to selected brain regions. (C) Number of neurons from each molecular cluster that projects to selected brain regions, including collateral projections. (D, E) Projection patterns of molecularly defined cell types. Number (D) and fraction (E) of neurons from each molecularly defined cell type that projects to the five downstream brain regions. (F) Validation of projection targets for selected molecularly defined cell types in a third animal (ANM #3). From left to right: enrichment of bed nucleus of the stria terminalis (BNST) projecting neurons in MC-13, enrichment of parabrachial nucleus (PBN) projecting neurons in MC-16, enrichment of lateral substantia nigra (SN) projection in Drd1 expressing MC-7 compared to MC-5 and MC-20, the other AST-occupying, Drd2-expressing clusters. Statistics: chi-square test. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
Figure 6 with 1 supplement
Prediction of projection classes with marker genes.
(A) Schematic of the analysis. Expressions of 29 marker genes from each neuron were used to train a multiclass logistic regression model to predict its axonal projection target. (B) Model performance for predicting projection classes (bed nucleus of the stria terminalis [BNST], lateral substantia nigra [SN], parabrachial nucleus [PBN], parvocellular reticular formation [PCRt], ventrolateral periaqueductal gray [vlPAG], and unlabeled) with marker genes, was compared to performance scores generated with shuffled data. p-value was calculated with permutation test. (C) Normalized confusion matrix with true class labels in rows and predicted class labels in columns with the logistic regression model in (B). Data normalized as true positive (TP) over the total number of neurons from this class (TP + false negative [FN]). (D) F1 score for predicting projections to BNST, lateral SN, hindbrain (PBN, vlPAG, PCRt) regions as well as neurons that were unlabeled with marker genes compared to shuffled data. p-value was calculated with permutation test. (E) Normalized confusion matrix with true class labels in rows and predicted class labels in columns with the logistic regression model in (D). Data normalized as TP over the total number of neurons from this class (TP + FN). (F) Area under the receiver-operating characteristic curve (AUC-ROC) scores with sequentially selected features for each projection class. Recursive feature elimination with cross-validation was used to rank features and identify feature(s) that best predict projection targets. Dotted line: AUC-ROC=0.5 Statistics in (B, D): permutation test. **p<0.01.
Figure 6—figure supplement 1
Marker gene expressions predict projection classes.
(A) Area under the receiver-operating characteristic curve (AUC-ROC) scores of logistic regression model for all projection classes compared with shuffled data. (B) ROC curve showing optimal prediction of projection classes with gene expression. (C) AUC-ROC scores of logistic regression model for all projection classes, projections grouped by bed nucleus of the stria terminalis (BNST), lateral substantia nigra (SN), hindbrain regions (parabrachial nucleus [PBN], ventrolateral periaqueductal gray [vlPAG], parvocellular reticular formation [PCRt]) and unlabeled. (D) Feature coefficients of marker genes in predicting each projection class based on the logistic regression model. Dotted lines: feature coefficient=0. (E) AUC-ROC scores with sequentially selected features for each projection class. From the top to bottom, AUC-ROC scores of subsets of features after sequential feature elimination. Optimal combinations of features were selected and ranked by recursive feature elimination with cross-validation. (F) Model performance in predicting BNST, PBN, and vlPAG projections using selected features (Dlk1, Pdyn, and Crh, respectively) with data from a third animal (ANM #3). Dotted lines in (A, C, F): AUC-ROC=0.5. Statistics in (A, C, F): permutation test. **p<0.01.
Author response image 1
Comparison of gene expression in identified CEA clusters between Peters et al.
dataset (top dot plot) and our dataset (bottom dot plot).
Author response image 2
Candidate marker genes and their expression from seq-c1, c4 and c12.
(A) Sagittal sections from Allen ISH database showing expression patterns of selected seq-c1 marker genes at the central amygdala and surround area, as schematized from reference atlas-defined regions. (B) Violin plot showing expression of selected seq-c4 and seq-c12 marker genes in scRNA-Seq clusters. Gene expression is first normalized by the total counts in that cell, multiplied by 10,000 and then natural log transformed. (C) Left: Sagittal sections from Allen ISH database showing expression patterns of selected seq-c4 marker genes at the central amygdala and surrounding area, as schematized from reference atlas-defined regions. Right: Coronal sections from Allen ISH database showing expression patterns of selected seq-c12 marker genes at the central amygdala and surrounding area, as schematized from reference atlas-defined regions. Scale bars in A and C: 840 µm. (D) Fxyd6 expression in the central amygdala, as detected by FISH. Duplicated from O’Leary et al. study. (E) Volcano plots showing differentially expressed genes in seq-c4, seq-c12 and combined (seq-c4 & seq-c12) clusters. Genes with log2 fold change greater than 0.75 and adjusted p-value less than 1e-5 are labeled.
Tables
Key resources table
| Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
|---|---|---|---|---|
| Peptide, recombinant protein | Cholera Toxin Subunit B (Alexa Fluor-488) | Thermo Fisher | Cat. # C34775 | |
| Peptide, recombinant protein | Cholera Toxin Subunit B (Alexa Fluor-555) | Thermo Fisher | Cat. # C34776 | |
| Peptide, recombinant protein | Cholera Toxin Subunit B (Alexa Fluor-594) | Thermo Fisher | Cat. # C34777 | |
| Peptide, recombinant protein | Cholera Toxin Subunit B (Alexa Fluor-647) | Thermo Fisher | Cat. # C34778 | |
| Chemical compound, drug | FluoroGold | Fluorochrome | ||
| Chemical compound, drug | Melphalan | Cayman Chemicals | Cat. # 16665 | |
| Chemical compound, drug | Acryloyl-X, SE | Thermo Fisher | Cat. # A20770 | |
| Commercial assay or kit | RNase-Free DNase Set | QIAGEN | Cat. # 79254 | |
| Commercial assay or kit | Proteinase K | NEB | Cat. # P8107S | |
| Chemical compound, drug | DAPI | Sigma | Cat. # D9542 | |
| Chemical compound, drug | Janelia Fluor 669, SE | Tocris | Cat. # 6420 | |
| Chemical compound, drug | N,N,N′,N′-Tetramethyl ethylenediamine | Sigma | Cat. # T22500 | |
| Chemical compound, drug | Ammonium persulfate | Sigma | Cat. # A3678 | |
| Chemical compound, drug | Acrylamide solution | Sigma | Cat. # A4058 | |
| Chemical compound, drug | 4-Hydroxy-TEMPO | Sigma | Cat. # 176141 | |
| Chemical compound, drug | N, N'-Methylenebisacrylamide | Sigma | Cat. # M7279 | |
| Chemical compound, drug | Acrylamide | Sigma | Cat. # A9099 | |
| Chemical compound, drug | Acrylic acid | Sigma | Cat. # 147230 | |
| Chemical compound, drug | DMSO | Sigma | Cat. # 570672 | |
| Chemical compound, drug | MOPS buffer | Sigma | Cat. # M1254 | |
| Chemical compound, drug | 20× SSC | Thermo Fisher | Cat. # AM9763 | |
| Chemical compound, drug | Nuclease-free water | Thermo Fisher | Cat. # AM9932 | |
| Chemical compound, drug | NaOH | Fisher scientific | Cat. # SS267 | |
| Chemical compound, drug | Poly-L-lysine | Pelco | Cat. # 18026 | |
| Chemical compound, drug | Dextran sulfate 50%, 100ML | Sigma | Cat. # S4030 | |
| Chemical compound, drug | Formamide | Fisher Scientific | Cat. # BP227-100 | |
| Chemical compound, drug | PBS | Fisher Scientific | Cat. # BP24384 | |
| Chemical compound, drug | RNase away/DNase away | Fisher Scientific | Cat. # 10328011 | |
| Chemical compound, drug | Photo-Flo 200 | EMS | Cat. # 74257 | |
| Commercial assay or kit | QIAquick Nucleotide Removal Kit | QIAGEN | Cat. # 28304 | |
| Strain, strain background (mouse, male) | C57Bl/6 | Jackson Laboratory | JAX stock #000664 | |
| Sequence-based reagent | HCR probes | Molecular Instrument | N/A | |
| Sequence-based reagent | HCR Amplifier B1 | Molecular Instrument | N/A | |
| Sequence-based reagent | HCR Amplifier B2 | Molecular Instrument | N/A | |
| Sequence-based reagent | HCR Amplifier B3 | Molecular Instrument | N/A | |
| Sequence-based reagent | HCR Amplifier B4 | Molecular Instrument | N/A | |
| Sequence-based reagent | HCR Amplifier B5 | Molecular Instrument | N/A | |
| Sequence-based reagent | Custom-DNA probe | This study | Ribosomal RNA probes | Sequence: gcgggtcgccacgtctgatctgaggtcgcg/3ATTO550N/ |
| Software, algorithm | EASI-FISH pipeline | Wang et al., 2021 | https://github.com/JaneliaSciComp/multifish | |
| Software, algorithm | Seurat 4.0.1 | Stuart et al., 2019 | RRID:SCR_016341 | https://satijalab.org/ |
| Software, algorithm | Fiji | ImageJ | https://imagej.net/software/fiji/ | |
| Software, algorithm | Python v3.7 | RRID:SCR 008394 | https://www.python.org/ | |
| Software, algorithm | n5-viewer | Saalfeld et al., 2022; Saalfeld lab | https://github.com/saalfeldlab/n5-viewer | |
| Software, algorithm | Napari | Napari contributors, 2019 | https://napari.org/ | |
| Other | Zeiss Lightsheet Z.1 microscope | Zeiss | https://www.zeiss.com/microscopy/us/products/imaging-systems/light-sheet-microscope-for-lsfm-imaging-of-live-and-cleared-samples-lightsheet-7.html | |
| Commercial assay or kit | Press-to-Seal Silicone Isolator with Adhesive | Thermo Fisher | Cat. # P24743 | |
| Commercial assay or kit | 8 mm glass coverslip | Harvard Apparatus | Cat. # BS4 64-0701 | |
| Other | Zeiss Lightsheet Z.1 imaging holder | Svoboda Lab and Janelia Experimental Technology | The design of the imaging holder can be found at https://www.janelia.org/open-science/zeiss-lightsheet-z1-sample-holder | |
| Other | CEA scRNA-Seq | This study | GEO: GSE213828 | Sequencing data included in this study is available through GEO: https://www.ncbi.nlm.nih.gov/geo/ |
| Other | CEA EASI-FISH data | This study | FISH and projection data included in this study are publicly available through the following links: https://doi.org/10.25378/janelia.21171373; http://multifish-data.janelia.org/ |
Author response table 1
ScRNA-Seq cluster correspondence between the Peters et al.
study and this study.
| Peters, et al. (9 CeA clusters) | This study | Notes |
|---|---|---|
| CeL_Nts.Tac2 | seq-c7 (Sst/Pdyn/Crh) | |
| CeL_Sst | seq-c10 (Sst/Pdyn/Crh--) | |
| CeM_Vdr | seq-c9 (Vdr) | |
| CeM_Il1rapl2 | seq-c12(Fxyd6/Gnas) | may have included cells from seq-c2, seq-c3 and seq-c4 |
| CeM_Drd2.Rai14 | seq-c5 (Drd2/Scn4b) | matched based on Drd2 expression |
| CeM_Tac1.Sst | seq-c11 (Sst/Tac1/Sema3c) | |
| CeM_Dlk1 | seq-c13 (Cyp26b1/Dlk1) | |
| CeL_Prkcd | seq-c8 (Prkcd/Cartpt) | |
| CeC_Cdh9.Calcrl | seq-c6 (Cyp26b1/Crym) |
Additional files
-
Supplementary file 1
Differentially expressed genes in molecularly defined clusters identified from scRNA-seq data.
Gene names are listed in the first column, and molecular clusters in the last column. avg_logFC: log fold-chage of the average expression between cells in the cluster of interest and all other cells; pct.1: the percentage of cells where the feature is detected in the cluster of interest; pct.2: the percentage of cells where the feature is detected in all other cells; p_val_adj: adjusted p-value is based on the Bonferroni correction.
- https://cdn.elifesciences.org/articles/84262/elife-84262-supp1-v2.xlsx
-
Supplementary file 2
Marker-genes used for expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH) experiment in the central nucleus of the amygdala (CEA).
- https://cdn.elifesciences.org/articles/84262/elife-84262-supp2-v2.docx
-
Supplementary file 3
Summary of cells analyzed in expansion-assisted iterative fluorescence in situ hybridization (EASI-FISH).
- https://cdn.elifesciences.org/articles/84262/elife-84262-supp3-v2.docx
-
Supplementary file 4
Co-expression relationships of previously studied marker-genes in the CEA subregions based on EASI-FISH.
Percent of neurons that co-express selected marker genes (Prkcd, Sst, Crh, Tac2, Nts, Pdyn, Drd1, Drd2, Penk, Tac1, Cartpt, Cyp26b1, Nefm) are shown, with darker red color representing a higher percent of co-expression.
- https://cdn.elifesciences.org/articles/84262/elife-84262-supp4-v2.xlsx
-
Supplementary file 5
Summary of statistical analyses.
- https://cdn.elifesciences.org/articles/84262/elife-84262-supp5-v2.xlsx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/84262/elife-84262-mdarchecklist1-v2.pdf
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Multimodal mapping of cell types and projections in the central nucleus of the amygdala
eLife 12:e84262.
https://doi.org/10.7554/eLife.84262
