Signature morphoelectric properties of diverse GABAergic interneurons in the human neocortex

Abstract

Human cortex transcriptomic studies have revealed a hierarchical organization of γ-aminobutyric acid-producing (GABAergic) neurons from subclasses to a high diversity of more granular types. Rapid GABAergic neuron viral genetic labeling plus Patch-seq (patch-clamp electrophysiology plus single-cell RNA sequencing) sampling in human brain slices was used to reliably target and analyze GABAergic neuron subclasses and individual transcriptomic types. This characterization elucidated transitions between PVALB and SST subclasses, revealed morphological heterogeneity within an abundant transcriptomic type, identified multiple spatially distinct types of the primate-specialized double bouquet cells (DBCs), and shed light on cellular differences between homologous mouse and human neocortical GABAergic neuron types. These results highlight the importance of multimodal phenotypic characterization for refinement of emerging transcriptomic cell type taxonomies and for understanding conserved and specialized cellular properties of human brain cell types.

Fig. 1.. Slice culture paradigm and multimodal characterization of human cortical GABAergic neurons.

(A) Tissue processing schematic for acute and culture paradigm, Patch-seq targeting guided by brightfield or fluorescence, and subsequent multimodal analysis or characterization. Scale bars: human brain specimen, 500 μm; brightfield of patched neuron, 10 μm; fluorescent image, 50 μm. (B) Box plot represents the difference in the number of genes detected between the two paradigms. Asterisks indicate significant pairwise comparisons (****P < 0.0001, FDR-corrected Mann-Whitney test). (C) Scatter plot of the average gene expression in acute versus culture for the top 25 differentially expressed genes for the LAMP5/PAX6, VIP, SST, and PVALB subclasses. Red line is regression line in each plot. (D) Box plot representing the difference in NMS score between the two paradigms. There were no significant pairwise comparisons (FDR-corrected Mann-Whitney test). (E) UMAP representation of the PVALB-subclass transcriptomic space with PVALB WFDC2 acute and culture neurons highlighted. (F) Cortical depth–matched PVALB WFDC2 morphologies from acute and culture shown aligned to an average cortical template, with corresponding voltage responses to a 1 s–long current step of −90 pA and rheobase +80 pA. (G) Box plots showing select morphology features for cortical depth–matched PVALB WFDC2 neurons from the acute and culture paradigm. There were no significant pairwise comparisons (FDR-corrected Mann-Whitney test). (H) UMAP representation of morphology space with PVALB WFDC2 acute and culture neurons highlighted. (I) UMAP representation of electrophysiology space with PVALB WFDC2 acute and culture neurons highlighted. (J) Overlaid single action potential sweeps from acute and culture PVALB WFDC2 neurons. Black lines represent the mean and are overlaid to the right for direct comparison. (K) Overlaid and normalized voltage response to a −90 pA hyperpolarizing current step from acute and culture PVALB WFDC2 neurons. Black lines represent the mean of the group. (L) Box plots showing select distinguishing features from PVALB WFDC2 neurons from the acute and culture paradigm. Asterisks indicate significant pairwise comparisons (****P < 0.0001, FDR-corrected Mann-Whitney test). ISI, interspike interval; FI curve slope, frequency-current curve slope.

Fig. 2.. Human neocortical GABAergic neuron subclass characterization.

(A) Integrated snRNA-seq and Patch-seq transcriptomic UMAP of GABAergic neurons. UMAP representations of (B) electrophysiology and (C) morphology space. (D) Representative morphologies from GABAergic subclasses with each transcriptomic type (t-type) represented by different colors and the corresponding t-type gene names (primary t-type gene name representing subclass removed for clarity) displayed in gray at bottom. The morphologies are aligned to an average cortical template with corresponding voltage responses to a 1 s–long current step to a −90 pA and rheobase +80 pA, shown below. (E) Box plots representing distinguishing morphology features by subclass (Kruskal-Wallis ANOVA on ranks; P < 0.05, FDR corrected. Post-hoc Dunn’s test; *P < 0.05, **P < 0.01, FDR corrected). (F) Overlaid single action potential sweeps from each GABAergic subclass. Black lines represent the mean and are overlaid to the right for direct comparison. (G) Overlaid and normalized voltage response to a 1-s −90-pA hyperpolarizing current step from each GABAergic subclass. Black lines represent the mean of the group. (H) Box plots representing key distinguishing electrophysiological features by subclass (Kruskal-Wallis ANOVA on ranks; P < 0.05, FDR corrected. Post-hoc Dunn’s test; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, FDR corrected).

Fig. 3.. Phenotypic alignment of SST FRZB neurons with PVALB subclass.

(A) Heat map of snRNA-seq data with the top 20 differentially expressed genes for SST and PVALB subclass with SST FRZB highlighted in blue. Genes in subpanel are key genes for the major subclasses and classes. (B) Transcriptomic UMAP highlighting the PVALB and SST subclasses and SST FRZB in blue. (C) UMAP representation of electrophysiology space highlighting SST and PVLAB subclasses and SST FRZB in blue. (D) Overlaid single action potential sweeps from SST FRZB and PVALB subclass; black lines represent the mean and are overlaid to the right for direct comparison. (E) UMAP representation of electrophysiology space color-coded by upstroke/downstroke ratio and action potential width. Box plots to the right show the distribution of data of the SST and PVALB subclasses with SST FRZB highlighted in blue. (Kruskal-Wallis ANOVA on ranks; P < 0.05, FDR corrected. Post-hoc Dunn’s test; ****P < 0.0001, FDR corrected). (F) UMAP representation of morphology space with highlighting SST and PVLAB subclasses and SST FRZB in blue. (G) Morphology-feature box plots for SST, SST FRZB, and PVALB. EMD, earth movers distance (Kruskal-Wallis ANOVA on ranks; P < 0.05, FDR corrected. Post-hoc Dunn’s test; **P < 0.01, FDR corrected). (H) Morphology hierarchical coclustering of SST- and PVALB-subclass neurons where values represent the number of neurons found in each morphology type. (I) Representative morphologies from SST FRZB and PVLAB subclasses shown aligned to an average cortical template with associated voltage responses to a 1 s–long current step of −90 pA and rheobase +80 pA. Morphology types from (H) are shown above each reconstruction.

Fig. 4.. Morphological heterogeneity within SST CALB1 transcriptomic type.

(A) Morphologies from SST CALB1 t-type categorized by qualitative morphology type shown aligned to an average cortical template, with histograms to the right of the morphologies displaying average dendrite (darker color) and axon (lighter color) branch length by cortical depth (shading shows +/− 1 SD about mean; soma locations are represented by black circles). Voltage responses to a 1 s–long current step to a −90 pA and rheobase +80 pA are shown below. Box plots representing key morphological (B) and electrophysiological (C) features by morphology type (Kruskal-Wallis ANOVA on ranks; P < 0.05, FDR corrected. Post-hoc Dunn’s test; *P < 0.05, **P < 0.01, FDR corrected). (D) UMAP of all SST CALB1 neurons based on 253 genes differentially expressed between 10 DBC, MC, and sparse SST neurons. Neurons cluster into three main groups corresponding to morphology types (colored markers). Neurons with unknown morphologies are represented by gray markers.

Fig. 5.. Transcriptomic-type identity, spatial distribution, and multimodal properties of human DBCs.

(A) Putative DBC morphologies mapped to transcriptomic type SST CALB1 or SST ADGRG6 and shown aligned to an average cortical template. Histograms to the right of the morphologies display average dendrite (darker color) and axon (lighter color) branch length by cortical depth (shading shows +/− 1 SD about mean; soma locations represented by black circles). (Bottom) Accompanying each reconstruction is the voltage response to a 1 s–long current step to a −90 pA and rheobase +80 pA for each respective DBC. (B) Spatial distribution for SST CALB1 and SST ADGRG6 revealed by MERFISH. (C) Four specific electrophysiology features shown as box or scatter plots for the SST CALB1 and SST ADGRG6 t-types with putative DBCs highlighted in green. (D) UMAP representation of transcriptomics (isolated to the SST subclass), electrophysiology, and morphology. SST CALB1 and SST ADGRG6 t-types are colored orange and brown, respectively, and qualitatively defined putative DBCs are highlighted in green. (E) 63× MIP inset corresponding to Fig. 4A showing axon horse-tail bundles and boutons.

Fig. 6.. Species differences in morphoelectric properties for homologous types Pvalb 2 and Sst 5.

(A) Human and mouse representative morphologies, aligned to average cortical templates, from homologous types Pvalb 2 and Sst 5 with each t-type represented by different colors. (B) Box plots of select morphological features from humans (pink) and mice (gray) (FDR-corrected Mann-Whitney test; *P < 0.05, **P < 0.01, ***P <0.001) (C) Electrophysiology UMAP for mouse and human SST/PVALB neurons. (Left) Homologous type Pvalb 2 and (Right) homologous type Sst 5 with pink markers representing human and gray representing mouse. (D) Overlaid single action potential sweeps from human and mouse homologous types Pvalb 2 and Sst 5. Black lines represent the mean and are overlaid to the right for direct comparison. Overlaid and normalized voltage response to a 1-s −90 pA hyperpolarizing current step. Black lines represent the mean of the group and are overlaid to the right for direct comparison. Voltage responses to depolarizing current at rheobase and rheobase +110 pA from human and mouse homologous types Pvalb 2 and Sst 5. (E) Box plots of select distinguishing electrophysiological features from homologous types Pvalb 2 and Sst 5 in mice and humans. Asterisks indicate significant pairwise comparisons; ****P < 0.0001, FDR-corrected Mann-Whitney test.