Stereo-seq of the prefrontal cortex in aging and Alzheimer's disease

Abstract

Aging increases the risk for Alzheimer's disease (AD), driving pathological changes like amyloid-β (Aβ) buildup, inflammation, and oxidative stress, especially in the prefrontal cortex (PFC). We present the first subcellular-resolution spatial transcriptome atlas of the human prefrontal cortex (PFC), generated with Stereo-seq from six male AD cases at varying neuropathological stages and six age-matched male controls. Our analyses revealed distinct transcriptional alterations across PFC layers, highlighted disruptions in laminar structure, and exposed AD-related shifts in layer-to-layer and cell-cell interactions. Notably, we identified genes highly upregulated in stressed neurons and nearby glial cells, where AD diminished stress-response interactions that promote Aβ clearance. Further, cell-type-specific co-expression analysis highlighted three neuronal modules linked to neuroprotection, protein dephosphorylation, and Aβ regulation, with all modules downregulated as AD progresses. We identified ZNF460 as a transcription factor regulating these modules, offering a potential therapeutic target. In summary, this spatial transcriptome atlas provides valuable insight into AD's molecular mechanisms.

Fig. 1. Spatially resolved transcriptomic profiles of the multiple layers in human prefrontal cortex.

A The analysis pipeline of the study, with select figures created using BioRender (Created in BioRender. Gong, Y. (2024) https://BioRender.com/m56f734; Created in BioRender. Gong, Y. (2024) https://BioRender.com/l73a167; Created in BioRender. Gong, Y. (2024) https://BioRender.com/s92q855; Created in BioRender. Gong, Y. (2024) https://BioRender.com/c29s609). B The H&E staining, immunohistochemistry (IHC) staining for Aβ plaques (Aβ42; red) and nuclei (DAPI; blue), and layer clustering for NA (B01809A3), moderate (B01809C2), and severe (A02092E1) AD groups. The IHC experiment was conducted once. Scale bar for the upper panel of the IHC, 1,500 $\mu m$, while the scale bar for the lower panel of the IHC, 100 $\mu m$. C UMAP visualization of seven clusters across 12 samples, annotated by cortical layer I to VI (L1 to L6) and the white matter (WM). D Heatmap of the marker genes in each cortical layer and the WM. The x-axis represents six cortical layers and the WM. Colors represent the mean expression of the gene in each cortical layer and the WM. Violin and scatter plots illustrating the expression levels of previously identified (E) and unreported (F) layer marker genes in B01809A3. In the violin plot, the x-axis represents the layer specific markers, and the y-axis represents the cortical layers and the WM. The color represents the median gene expression in the cluster. For the scatter plots, six plots represent the distribution of the gene expression levels of the layer markers, respectively. The color represents the gene expression levels.

Fig. 2. Layer specific DEGs between NA, moderate AD, and severe AD groups at bin110 resolution.

A Proportion of the cortical layers in each group. The color represents cortical layers I-VI. B Bar plot of the layer proportion across NA, moderate, and severe AD groups. One-way ANOVA and two-way t-test was conducted to identify significant differences in proportion between NA, moderate, and severe AD groups for each layer. False discovery rate (FDR) adjustment was applied for the P-value adjustment. N.S. indicates non-significant (adjusted P-value > 0.05) and *** represents adjusted P-value < 0.01. The number above the N.S. and *** indicate the specific adjusted P-value for each comparison. C The expression level of the GFAP and NEAT1 in NA, moderate, and severe AD groups. Two-side t-test and FDR adjustment were applied for the adjusted P-value calculation. The adjusted P-value for the DGE analysis of GFAP between the NA and severe AD groups is 0.0, and between the moderate and severe AD groups is also 0.0. Similarly, the adjusted P-value for NEAT1 between the NA and severe AD groups is 0.0, and between the moderate and severe AD groups is 0.0. *** represents adjusted P-value < 0.01. D Number of the layer-specific up- and down-regulated genes of the pairwise DGE analysis for each layer and the WM across all three groups. E Significant layer-specific up-regulated genes in each group after the pairwise DGE analysis for the WM, layer I, II/III, and V among all three groups. Size of the spot indicates the proportion of the cells expressing the genes, and the color represents mean expression level of the genes in each group.

Fig. 3. Layer-layer interactions across six cortical layers and the WM in NA, moderate, and severe AD groups.

A Networks of layer-layer interactions across six cortical layers and the WM in NA, moderate, and severe AD groups. The colors of the dots and edges represent the specific layers and the outgoing signaling emanating from them. Number on the edge represents the number of the LR pairs. B Pair-wise comparison of the number of outgoing and incoming LR pairs among different cortical layers and the WM across the NA, moderate, and severe AD groups. The color of the squares in the heatmap represents the number of increased (red) or decreased (blue) signaling in the first group compared to the second one. The top color bar represents the total number of changes (increases or decreases) in incoming signaling, and the right color bar corresponds to changes in outgoing signaling, both comparing the first group to the second. C The communication strength comparison of the specific LR pairs of Glutamate (from layer V to other layers) and NRXN (from layer I to other layers) signaling pathways across control, moderate and severe AD groups. The x-axis represents the direction of the LR pairs, and the y-axis indicates the specific LR pairs. Permutation test and Bonferroni correction were used for adjusted P-value calculation. Dot size represents the adjusted P-value, and the color reflects the communication strength of the LR pairs.

Fig. 4. Identity of the location of the dysfunctional neurons in the three groups at bin50 resolution.

A UMAP visualization of pseudo-spot clusters. B Expression levels of the gene markers in each cluster. The color of the violin plot reflects the median expression of the gene in each cluster. Based on the gene markers, we annotated the clusters as astrocytes (Ast), endothelial cells (End), excitatory neurons (Ex), inhibitory neurons (Inh), microglia (Mic), oligodendrocytes (Oli), and oligodendrocyte progenitor cells (Opc). C Distribution of different cell clusters on the captured area of Stereo-seq platform (B01809A3 in NA, B01809C2 in moderate, and B02008C6 in severe AD group), and the proportions of cell clusters in all samples from the NA, moderate, and severe AD groups, respectively. D The violin plot of the mitochondrial gene ratio of Ex in each sample. The x-axis represents 12 samples and the y-axis indicates the mitochondrial gene ratio. The mitochondrial gene ratio ranges are as follows: for the NA group, maxima 16.3–29.5%, 75th percentile 6.6–8.6%, median 5.5–7.4%, 25th percentile 4.4–6.5%, and minima 1.5–3.0%; for the Moderate AD group, maxima 13.3–25.7%, 75th percentile 5.5–8.7%, median 4.5–7.3%, 25th percentile 3.7–6.1%, and minima 0.9–2.2%; and for the Severe AD group, maxima 21.2–35.1%, 75th percentile 12.2–21.0%, median 10.6–18.4%, 25th percentile 9.2–16.0%, and minima 3.1–5.4%. The dashed line represents the threshold of the mitochondrial gene ratio (15%) to classify the Ex into highly stressed or normal ones. E Distribution of the highly stressed Ex in NA, moderate, and severe AD groups. The yellow dots represent the pseudo-spots covering the highly stressed Ex, and the blue dots are pseudo-spots containing the normal Ex and other types of the cells. F Proportion of the highly stressed Ex pseudo-spots among all Ex pseudo-spots in NA, moderate, and severe AD groups. G The significant up- and down-regulated genes of the highly stressed Ex pseudo-spots, compared to normal Ex pseudo-spots. Two-sided t-test and FDR adjustment were used for the adjusted P-value calculation. Number indicates the exact adjusted P-value and *** represents adjusted P-value < 0.01.

Fig. 5. Concentric analysis at the single-nucleus resolution.

A UMAP visualization of clusters of nuclei. B Heatmap of the gene marker expression in each cluster of nuclei. The color represents the expression levels of the gene in each cluster of nuclei. C The nuclei type distribution on B01809A3 (NA) sample. D Distribution of nuclei in B01809C2 (moderate) across levels I, II, and III in the concentric circle analysis. The red dots represent the selected nuclei at each level, and gray dots indicate other nuclei. E The nucleus-type specific DGE analysis between the nuclei in level I and III across NA, moderate, and severe AD groups. The color represents the mean expression of the genes in each level. F The nucleus-type specific DGE analysis between the nuclei in NA, moderate, and severe AD groups at level I.

Fig. 6. Differences in the cell-cell communication between the nuclei in level I and III.

A The total number (top) and the strength (bottom) of the LR pairs in level I and III groups across control, moderate, and severe AD groups, respectively. B Networks of cell-cell interactions show the number of LR pairs (edges) between nuclei clusters (nodes) within level III and I across NA, moderate, and severe AD groups, respectively. The colors of the dots and edges represent the specific types of nuclei and the outgoing signaling emanating from them. The number on the edge indicates the number of the outgoing signals. The size of the spot are proportional to the number of nucleus within each level of specific AD group. C The communication strength of the specific LR pairs of the nuclei in level I and III in the NA, moderate, and severe AD groups, respectively. Nuclei type in each row indicates the nuclei expressing the ligands and the column indicates the one expressing the receptors. Permutation test and Bonferroni correction were used for adjusted P-value calculation. The x-axis represents the level I and III across NA, moderate, and severe AD groups in each nucleus type. The color of the dots reflects the communication strength, and the size of the spot represents the adjusted P-value. D, E The chord diagram of the PSAP-GPR37L1 and APP-CD74 LR pairs of the nuclei in level I and III in the NA, moderate, and severe AD groups, respectively. Each segment of the circle in different color represents different types of nuclei. The chords connecting the segments represent interactions between two types of nuclei, and the thickness of each chord reflecting the strength of the interactions.

Fig. 7. Gene co-expression networks in excitatory and inhibitory neurons.

A The gene co-expression modules based on the Ex (Ex1) and Inh (Inh1 and Inh2) from level I to III across NA, moderate, and severe AD group. The nodes represent the hub genes, while the edge connecting two nodes indicates the co-expression of those genes. B-D GO Enrichment analysis on the top 50 hub genes in the Ex1, Inh1, and Inh2 modules. The length of the bar indicates the gene numbers enriched in the GO term and the color represents the adjusted P-values for enrichment analysis. Adjusted P-values were calculated through Fisher exact test and FDR adjustment. E Enrichment analysis to identify the Motif overlaps with Ex and Inh modules. Fisher exact test and FDR adjustment were applied for the adjusted P-value calculation. The color indicates the adjusted P-value.