Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Oct 7;8(40):eabo3932.
doi: 10.1126/sciadv.abo3932. Epub 2022 Oct 5.

Aging compromises human islet beta cell function and identity by decreasing transcription factor activity and inducing ER stress

Shristi Shrestha  1 , Galina Erikson  2 , James Lyon  3 , Aliya F Spigelman  3 , Austin Bautista  3 , Jocelyn E Manning Fox  3 , Cristiane Dos Santos  4 , Maxim Shokhirev  2 , Jean-Philippe Cartailler  1 , Martin W Hetzer  5 , Patrick E MacDonald  3 , Rafael Arrojo E Drigo  4
Affiliations

Aging compromises human islet beta cell function and identity by decreasing transcription factor activity and inducing ER stress

Shristi Shrestha et al. Sci Adv. .

Abstract

Pancreatic islet beta cells are essential for maintaining glucose homeostasis. To understand the impact of aging on beta cells, we performed meta-analysis of single-cell RNA sequencing datasets, transcription factor (TF) regulon analysis, high-resolution confocal microscopy, and measured insulin secretion from nondiabetic donors spanning most of the human life span. This revealed the range of molecular and functional changes that occur during beta cell aging, including the transcriptional deregulation that associates with cellular immaturity and reorganization of beta cell TF networks, increased gene transcription rates, and reduced glucose-stimulated insulin release. These alterations associate with activation of endoplasmic reticulum (ER) stress and autophagy pathways. We propose that a chronic state of ER stress undermines old beta cell structure function to increase the risk of beta cell failure and type 2 diabetes onset as humans age.

PubMed Disclaimer

Figures

Fig. 1.
Fig. 1.. Gene transcription landscape of aging human beta cells.
(A) Integration of human islet single-cell RNA sequencing (scRNA-seq) and analysis for cell type identity, pseudotime trajectory, TF activity, and GRN analysis. Glucose-stimulated insulin release assays (GSIS) from isolated human islets quantified beta cell function, and confocal microscopy quantified human beta cell TF expression in situ. (B) Uniform Manifold Approximation and Projection (UMAP) of our integrated scRNA-seq dataset with 68 nondiabetic (ND) donors, 35 diseased donors, and 48,737 cells in total. Major islet cell types are shown; non-endocrine cells are in gray. (C) Pearson correlation matrix and hierarchical clustering analysis (HCA) of human beta cell transcriptomes shows high similarity between beta cells from younger (0 to 29 years old) and older (>60 years old) samples, or between middle age samples (30 to 59 years old). (D) Dot plot with HCA of beta cell TFs and beta cell–enriched genes in each decade of life. (E) Dot plot with HCA of beta cell ion channel, insulin secretion, vesicle organization, protein homeostasis, stress, autophagy, and amino acid activation genes. (F) GSIS of isolated islets from 268 ND donors incubated with 1 or 16.7 mM glucose concentrations and split across different age ranges. Data are shown as a fraction of the total beta cell insulin content. Each point represents one individual donor. Statically significant differences were determined by one-way analysis of variance (ANOVA) with Tukey multiple comparisons posttest.
Fig. 2.
Fig. 2.. Beta cell TFs, ER stress, and autophagy markers in aging beta cells.
Immunohistochemistry and confocal microscopy of human pancreas formalin-fixed paraffin-embedded samples from adults (22 to 35 years old) and old adults (61 to 79 years old). Human islets were stained with (A) insulin, PDX1, NKX6-1, and/or NKX2-2; (B) insulin and HSPA5 or XBP1; and (C) insulin, LC3A/B, and LAMP1. (D) Quantification of LC3-LAMP1 colocalization, area of beta cell cytosol covered by LAMP1+ granules, and LAMP1+ granule size. Each dot represents data from a single region of ~140 μm2 of human islets. Here, 1 region per islet, 10 regions per donor were quantified. In (A) and (B), the islet region is demarcated by the dotted yellow line. Scale bars, 50 μm (A and B), 20 μm (C), and 5 μm (inset of C). CI, confidence interval.
Fig. 3.
Fig. 3.. Effect of aging on the prevalence of distinct beta cell transcriptional states.
(A) UMAP of 11,279 human ND beta cells and transcriptionally different human beta cells. (B) Same as in (A); however, each beta cell is colored by the age decade of their respective donor. (C) Heatmap of differentially expressed genes per cluster. Rows on the left and top columns provide the identity of each beta cell cluster. Bars on the right show the total number of cells in each cluster. (D) Heatmap of differentially expressed genes in different decades of life. Rows on the left indicate decades of age, the top column identifies each beta cell cluster, the bottom column identifies differentially expressed genes found in each cluster, and bars on the right show the total number of cells in each age decade of life. (E) Illustration of changes in beta cell state composition associated with transcriptional heterogeneity phenotypes identified from our clustering approach. (F) Illustration highlighting the effect of aging on the expression dynamics of beta cell development, stress, function, and identity genes. In (A), cell colors match the colors assigned to each beta cell cluster from (A).
Fig. 4.
Fig. 4.. Identification of beta cell GRNs.
(A) Illustration of SCENIC analysis. (B) Heatmap and hierarchical clustering analysis (HCA) of TF activity patterns. TFs classified as “ON” are shown in black, while TFs classified as “OFF” are in white. Top rows show the distribution of cells within the heatmap subdivided by cell type, age group, percent of ribosomal genes, sex, and study origin. Column dendrogram represents HCA clades within each cell type using “Euclidean” distances. Rows are TFs identified using SCENIC. Beta cell–enriched TFs are in (C). (D and E) GRNs formed by TFs identified using SCENIC in the human pancreas and in beta cells, respectively. TFs are shown as pink nodes, while target protein-coding genes are shown in light gray in (D) and white in (E). Black lines connecting the nodes represent the importance metric (IM) between TF-gene pairs, where thicker edges indicate stronger TF-target relationship. Node size represents the “betweenness centrality” measurements that report on the influence of a given TF within the network. (F) Pearson correlation matrix of 609 human beta cell TFs. Correlation index scale is on the right. Bounding boxes highlight clusters of TFs with high degree of correlation.
Fig. 5.
Fig. 5.. Changes in the TF-GRN landscape of human beta cells in young, adult, and old adult humans.
(A) Schematics of the approach used to generate age group–specific GRNs. (B) Network graphs illustrating the GRNs formed by TFs identified using SCENIC in human beta cells from young (<6 years old), adult (20 to 59 years old), or old adult (>60 years old) donors. TFs are shown as pink nodes, while target protein-coding genes are shown in light gray. Black lines connecting the nodes represent the IM between TF-gene pairs, with thicker edges indicating stronger relation of TF-target. The size of each node represents the betweenness centrality measurements (often used, as a measure of “hub status” due to its influence in flow of information) of a given TF within the network. (C) Clustering analysis identified by TCseq resulted in six different families of TFs based on the number of targets gained or lost over increasing age. Identity of select TFs belonging to each family is shown at the bottom of each graph. (D) Same as in (B); however, these GRNs are focused on ER stress–related TFs to highlight age-dependent differences in the organization and scope of ER stress GRNs in human beta cells.
Fig. 6.
Fig. 6.. Pseudotime ordering of human beta cells classified by TF activity at the single-cell level.
(A and B) Force-directed layout to project the Leiden clusters of human beta cells from ND donors analyzed using SCENIC into two dimensions. Distinct Leiden clusters representing distinct beta cell TF states (A) and decades of age for each cell (B) are shown. (C) Trajectory analysis performed with partition-based graph abstraction (PAGA) on SCENIC binary output of ND beta cells’ regulons. Trajectory starts at node number 4 (containing mostly infants), and each following node represents a significant point in the trajectory characterized by a distinct TF signature. (D) Three pseudotime trajectories are shown, namely, transition, mature, and stress, and their respective cell frequency. (E) TF activity identified as differentially modulated across the three different pseudotime trajectories.
Fig. 7.
Fig. 7.. Working model of ER stress activation in old human beta cells.
Up-regulation of the beta cell transcriptional output increases protein translation and folding demands. To meet this increased demand, beta cells increase amino acid transport and metabolism. This is also associated with activation of the amino acid starvation response that stimulates autophagy. This process is impaired and leads to accumulation of lipofuscin bodies. Higher protein synthesis overloads the ER and triggers ER stress response pathways, which are compromised. This likely contributes to activation of other beta cell aging–dependent pathways, including senescence. Unresolved ER stress phenotype, combined with compromised expression of identity TFs, creates a chronic ER stress state that undermines beta cell structure function that could lead toward secretory failure and/or death.
See this image and copyright information in PMC

References

    1. López-Otín C., Blasco M. A., Partridge L., Serrano M., Kroemer G., The hallmarks of aging. Cell 153, 1194–1217 (2013). - PMC - PubMed
    1. Taylor R. C., Dillin A., Aging as an event of proteostasis collapse. Cold Spring Harb. Perspect. Biol. 3, a004440 (2011). - PMC - PubMed
    1. Petersen K. F., Befroy D., Dufour S., Dziura J., Ariyan C., Rothman D. L., DiPietro L., Cline G. W., Shulman G. I., Mitochondrial dysfunction in the elderly: Possible role in insulin resistance. Science 300, 1140–1142 (2003). - PMC - PubMed
    1. Kalyani R. R., Golden S. H., Cefalu W. T., Diabetes and aging: Unique considerations and goals of care. Diabetes Care 40, 440–443 (2017). - PMC - PubMed
    1. Basu R., Breda E., Oberg A. L., Powell C. C., Man C. D., Basu A., Vittone J. L., Klee G. G., Arora P., Jensen M. D., Toffolo G., Cobelli C., Rizza R. A., Mechanisms of the age-associated deterioration in glucose tolerance: Contribution of alterations in insulin secretion, action, and clearance. Diabetes 52, 1738–1748 (2003). - PubMed