Single-cell transcriptomic and genomic changes in the aging human brain

Abstract

Aging brings dysregulation of various processes across organs and tissues, often stemming from stochastic damage to individual cells over time. Here, we used a combination of single-nucleus RNA-sequencing and single-cell whole-genome sequencing to identify transcriptomic and genomic changes in the prefrontal cortex of the human brain across life span, from infancy to centenarian. We identified infant-specific cell clusters enriched for the expression of neurodevelopmental genes, and a common down-regulation of cell-essential homeostatic genes that function in ribosomes, transport, and metabolism during aging across cell types. Conversely, expression of neuron-specific genes generally remains stable throughout life. We observed a decrease in specific DNA repair genes in aging, including genes implicated in generating brain somatic mutations as indicated by mutation signature analysis. Furthermore, we detected gene-length-specific somatic mutation rates that shape the transcriptomic landscape of the aged human brain. These findings elucidate critical aspects of human brain aging, shedding light on transcriptomic and genomics dynamics.

Extended Data Figure 1.. Cell type ratios do not change significantly with age.

(a) Neuron to glia ratio shows no significant change with age. (b) Excitatory to inhibitory neuron ratio shows no significant change with age. (c) Contribution of each cell type to each donor, shown as a percentage of total cells. Columns in S5 plus Fig. 2B sum to 100. (IN – inhibitory neurons, AST – astrocytes, Endo – endothelial cells).

Extended Data Figure 2 (above).. Infant-specific gene expression across ages in two independent data sets.

Heatmaps (top) plotting infant-specific gene expression ordered by age in the this study and Herring et al. showing higher expression in infant and gestational cases and lower expression in adults for L2/3 neurons, L4 neurons, and astrocytes. Box plots (bottom) showing mean expression of infant-specific genes and adult-specific genes in L2/3 neurons, L4 neurons, and astrocytes across ages in this study and Herring et al. Expression of infant-specific genes is significantly higher in donors ≤4 compared to donors ≥15. Expression of adult-specific genes is significantly lower in donors ≤4 compared to donors ≥15. (****, p<0.0001; Wilcoxon Rank Sum test).

Extended Data Figure 3 (above).. Fold change in elderly vs. adult brains for all ribosomal proteins.

Fold change of ribosomal proteins in aging for each cell type. Ext, excitatory neurons; Inb, inhibitory neurons; Oli, oligodendrocytes; OPC, oligodendrocyte precursor cells; Ast, astrocytes; Micro, microglia; Endo, endothelial. (*, p < 0.05; two-sided T-test).

Extended Data Figure 4.. Fold change in elderly vs. adult brains for all nuclear-encoded mitochondrial proteins.

Fold change of nuclear-encoded mitochondrial proteins in aging for each cell type. Ext, excitatory neurons; Inb, inhibitory neurons; Oli, oligodendrocytes; OPC, oligodendrocyte precursor cells; Ast, astrocytes; Micro, microglia; Endo, endothelial. (*, p < 0.05; two-sided T-test).

Extended Data Figure 5.. Gene class specific expression and correlation between immediate early gene expression and age (above).

(a) Expression of neuron-specific genes shows no significant fold change in aging. (b) Expression of neuronal and housekeeping genes in the relevant cell types. (c) Line represents linear model and shaded area shows a 95% confidence interval. Pearson’s R² and p-value shown.

Extended Data Figure 6.. Mutation spectrum of sSNVs in human neurons (above).

(a) Total mutation accumulation per neuron correlates significantly with age at a rate of 15.5 SNVs gained/year. (b) Mutation spectrum of sSNVs called in human neuron scWGS data. Each bar represents a specific mutation in a different trinucleotide context. (c) Cosine similarity of the two signatures, A1 and A2, derived de novo from the total mutation spectrum to each single-base substitution signature in the COSMIC data base. Signature A1 is most similar to SBS5. Signature A2 is most similar to SBS30. (d) Difference in somatic SNV spectra in housekeeping and neuron specific genes. Same plot as shown in figure 6i but with trinucleotide context specified below each bar.

Extended Data Figure 7.. Fold change in elderly vs. adult brains for DNA damage repair genes.

Fold change of polymerase genes (green), nucleotide excision repair (blue), and base excision repair (red) in aging for each cell type. Ext, excitatory neurons; Inb, inhibitory neurons; Oli, oligodendrocytes; OPC, oligodendrocyte precursor cells; Ast, astrocytes; Micro, microglia; Endo, endothelial. (*, p < 0.05; two-sided T-test).

Extended Data Figure 8 (above).. Density plots of gene length for all cell types.

Comparison of expressed gene (black), up-regulated gene (green) and down-regulated gene (yellow) size density in (a) L2/3 neurons, (b) L4 neurons, (c) L5/6 neurons, (d) inhibitory-SST neurons, (e) inhibitory-PV neurons, (f) inhibitory-VIP neurons, (g) inhibitory-SV2C neurons, (h) Inhibitory neurons, (i) microglia, (j) oligodendrocyte precursor cells, (k) astrocytes. L5/6-CC not shown due to low number of differentially expressed genes.

Extended Data Figure 9 (above).. Expression ratio of housekeeping genes by size decile for all cell types.

Comparison of elderly to adult expression of housekeeping genes by size decile in (a) L2/3 neurons, (b) L4 neurons, (c) L5/6 neurons, (d) inhibitory-SST neurons, (e) inhibitory-PV neurons, (f) inhibitory-VIP neurons, (g) inhibitory-SV2C neurons, (h) inhibitory neurons, (i) Microglia, (j) oligodendrocytes, (k) oligodendrocyte precursor cells, (l) astrocytes, and (m) endothelial cells. Median fold change shown under each box plot. L5/6-CC neurons not shown due to low number of cells in the cluster. All box plots depict median, and first and third quartile. Whiskers show one and a half times the interquartile range beyond the first and third quartiles.

Extended Data Figure 10 (above).. Expression ratio of neuron-specific genes by size decile for all cell types.

Comparison of elderly to adult expression of neuron-specific genes by size decile in (a) L2/3 Neurons, (b) L4 neurons, (c) L5/6 neurons, (d) inhibitory-SST neurons, (e) inhibitory-PV neurons, (f) inhibitory-VIP neurons, (g) inhibitory-SV2C neurons, and (h) inhibitory neurons. Median fold change shown under each box plot. L5/6-CC neurons not shown due to low number of cells in the cluster. All box plots depict median, and first and third quartile. Whiskers show one and a half times the interquartile range beyond the first and third quartiles.

Fig 1.. Droplet-based snRNA-seq of human PFC and cell-type classification.

(a) Experimental schematic of droplet-based snRNA-seq. (b) Dimensional reduction and clustering of all cells post-filtration yielded multiple clusters for each cell type (Oli, oligodendrocyte; AST, astrocyte; Endo, endothelial). (c) Percentage of cells in each cluster of our data that correspond to the annotated reference cluster. (d) Correlation of gene expression profiles for each sub-cluster within a cell type correspond most closely to the cells of the same lineage based on Pearson’s correlation coefficient. Bar plot above heatmap shows the number of genes expressed in each cluster.

Fig. 2.. Transcriptionally unique excitatory neurons and astrocytes detected in infant PFC.

(a) Clusters plotted by donor contribution as a percentage of total cells in the cluster. L2/3–2, L4–4, and AST-2 are infant-specific. (b) GO terms derived from differentially expressed genes up-regulated in infant-specific clusters plotted as general categories (see Table S5 for full term list and category designation). Development related terms are most common. (c) Contribution of OPCs (top) and oligodendrocytes (bottom) to the total cells in each donor, plotted as a percentage. OPCs are most abundant in infants and their population declines later in life. Oligodendrocytes are most abundant in adulthood and nearly absent in infancy.

Fig. 3.. Common down-regulation of genes across cell types.

(a) The number of down (blue) and up (red) -regulated genes for each cell type. (b) Heatmap showing Log₂ fold-change of elderly vs. adult differentially expressed genes across cell types.(c) Log₂ fold change of elderly vs. adult ribosomal protein genes from both the small and large subunit. (d) Log₂ fold change of elderly vs. adult nuclear encoded mitochondrial electron transport chain genes from all five complexes. (*, p < 0.05, two-sided T-test).

Fig. 4.. Aging-related down-regulation of essential housekeeping functions in neurons.

(below) (a) GO terms of genes downregulated in aging plotted as general categories (see Table S8 for full GO results). Housekeeping functions (shades of blue) are most commonly downregulated in neurons and microglia. Only cell types with significantly enriched GO terms shown. (b) Housekeeping genes, which are significantly downregulated in elderly brains relative to adults in 7/8 neuron types. Boxes show median, first and third quartiles. Whiskers show one and a half times the interquartile range beyond the first and third quartiles. (c) Boxplots showing the mean gene effect score for all the down and up -regulated genes in the DepMap database from the Broad Institute. The down-regulated genes for both neurons (left plot) and microglia (right plot) are more essential than the up-regulated genes (two-sided T-test). Boxes and whiskers formatted as in b. All data points shown. Points beyond whiskers are outliers. (d) Expression of immediate early genes ARC, BDNF, EGR1, and NPAS4, in excitatory neurons decreases with age. (e) Number of DEGs linked to age-associated neurodegenerative disease by subtype. Red proportions depict up DEGs, blue denotes down DEGs. Microglial DEGs were enriched for neurodegeneration GWAS hits (p = 0.0068, χ² test), and proportionally more glial DEGs were neurodegeneration GWAS hits than neuronal DEGs (7.2% vs. 2.5%, p = 0.0099, two sided t-test). (f) Heatmap showing log₂(FC) of neurodegeneration GWAS hits that were DEGs in at least one subtype. Gene names are colored by disease. Dots near gene names reflect genes with >1 association. (*, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001) Boxes around cells in heatmap denote significant DEGs.

Fig. 5.. scWGS reveals sSNV mutation signatures linked to expression.

(below) (a) De novo mutation signature analysis of somatic SNVs (sSNVs) in human neurons revealed two signatures, A1 dominated by T>C mutations and A2 dominated by C>T mutations. Trinucleotide contexts are the same as shown in fig.S13. (b) Number of Signature A1 sSNVs in each neuron plotted by age. Signature A1 strongly correlates with age (R²=0.94) with an extrapolated mutation rate of 13.8 SNVs/year. (c) sSNV enrichment of Signature A1 in coding regions plotted by expression quantile (left) and genic vs. intergenic regions (right). Signature A1 is enriched in the highest expressed genes and genic regions. (d) Percent of total sSNVs derived from the transcribed strand broken down by expression quantile. T>C and C>T strand bias increases with expression. (e) Number of Signature A2 sSNVs in each neuron plotted by age. Signature A2 correlates with age (R²=0.31) with an extrapolated mutation rate of 1.8 SNVs/year. (f) sSNV enrichment of Signature A2 in coding regions plotted by expression quantile (left) and genic vs. intergenic regions (right). Signature A2 is depleted in the highest expressed genes and enriched in the lowest expressed genes as well as intergenic regions. (g) Log₂ fold change of elderly/adult expression data from snRNA-seq showing reduced expression in key neuron DNA damage repair pathway proteins (two-sided T-test). Polymerases, green; nucleotide excision repair, blue; base excision repair, red. (*, p < 0.05; ****, p < 0.0001).

Fig. 6.. Size-specific down-regulation of housekeeping genes and preservation of cell identity genes.

(below) (a) Bar graph showing the size of down-regulated genes by cell type relative to the median gene size for each cell type. (Wilcoxon Rank Sum Test). (b) Density plots of lengths of all expressed genes (black), upregulated genes (green), and downregulated genes (yellow) in three cell types. (c) Expression in topoisomerase complex genes across cell types. Stars denote significant difference between neurons and non-neurons in the percent of cells expressing the gene (Wilcoxon test). (d) Housekeeping genes are significantly shorter than neuron-specific genes. (n = number of genes in each category). (e) Elderly/adult fold change of housekeeping genes by size decile in excitatory neurons (R² = 0.41). The shortest housekeeping genes have the lowest expression in elderly cells. (f) Elderly/adult fold change of neuron-specific genes by size decile in excitatory neurons (R² = 0.09). (g, h) Somatic SNV rate per base pair in housekeeping genes (g) negatively correlates with gene size, plotted by decile, while (h) neuron-specific genes do not show a significant relationship with gene size. 95% confidence interval shown in gray. (i) Difference in single-base substitution frequency between housekeeping and neuron-specific sSNV mutation spectra. C>N substitutions are significantly enriched in housekeeping genes, with particular enrichment in C>G substitutions (exact binomial test). Trinucleotide contexts are the same as shown in Extended Data figure 6. (*, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001). All box plots depict median, and first and third quartile. Whiskers show one and a half times the interquartile range beyond the first and third quartiles.