Fig 1.
Clustering analysis and t-SNE plots of transcriptome data obtained from 6-month-old (young) and 29-month-old (old) mice.
(A) SVF cell populations from young (cell number, n = 271) and old (cell number, n = 1015) mice were classified into eleven numbered groups (Groups 0 to 10) using the package Seurat version 2.4. A heat map for the genes that were differentially expressed in each of the 11 groups is shown. (B) The t-SNE plot of SVFs from young and old mice is colored according to the group number (left) and age (right).
Fig 2.
ASC marker gene expression in SVFs.
(A) The gene expression levels of Gsn, Cxcl1, Col1a2, Col6a1, Mmp2, and Mmp14 in Groups 1, 3, and 5 combined (as indicated in X) (cell number, n = 461) and eight other groups (as indicated in Y) (cell number, n = 825) are shown. (B) Heat maps of Gsn, Cxcl1, Col1a2, Col6a1, Mmp2, and Mmp14 in t-SNE plots are shown. Values are presented as the means ± standard error of the means (SEMs). The statistical analysis was performed using the two-tailed Welch’s t-test. *p<0.001.
Fig 3.
Age-associated changes in ASCs expressing ASC marker genes.
The percentages of cells positive for the expression of Gsn, Cxcl1, Col1a2, Col6a1, Mmp2, and Mmp14 in Groups 1, 3, and 5 from young (blue) and old (orange) mice are shown. G1: Group 1 (cell number, n = 191), G3: Group 3 (cell number, n = 136), G5: Group 5 (cell number, n = 134).
Fig 4.
Preadipocyte and white adipocyte marker gene expression among Groups 1, 3, and 5.
The gene expression levels of Lpl, Egfr, Efemp1, Igfbp4, Pparγ, Fabp4, and C/ebpα in Groups 1 (purple), 3 (red), and 5 (magenta) are shown. G1: Group 1 (cell number, n = 191), G3: Group 3 (cell number, n = 136), G5: Group 5 (cell number, n = 134). Values are presented as the means ± SEMs. The statistical analysis was performed using one-way analysis of variance followed by the Tukey-Kramer test. *p<0.05, **p<0.01.
Fig 5.
Pseudotime analysis using Monocle 3.
(A-D) The data of Group 3 (cell number, n = 136) and Group 5 (cell number, n = 134) from young and old mice were reanalyzed using Monocle 3. Black lines in the plot show the paths of pseudotime. UMAP plots are colored according to three clusters (A), age (B), groups (C), and pseudotime (D). (E-H) Heat maps showing the Log10 expression levels of Lpl, Egfr, Efemp1, and Igfbp4 in the UMAP plot. Gray dotted lines indicate the Group 5 area. (I) The heat maps of the hierarchical cluster analysis and expression of modules between Group 3 (G3) and Group 5 (G5) are shown. The modules consist of genes clustered according to the correlation of expression levels. Genes whose expression changed significantly over pseudotime were extracted using Moran’s I test for spatial autocorrelation (q-value < 0.05). (J) GO terms of module 6 genes searched using DAVID (p-value < 0.05). (K) Heat maps showing the Log10 expression levels of Adamts7, Snai2, and Tgfbr1 in the UMAP plot. Gray dotted lines indicate the Group 3 area. (L) Violin plots showing the distribution of Adamts7, Snai2, and Tgfbr1 expression in Group 3 (red) and Group 5 (magenta). (M) Kinetic plots showing the expression of Adamts7, Snai2, and Tgfbr1 over pseudotime are shown.
Fig 6.
Age-associated changes in gene expression patterns.
Kinetic plots showing the expression of Adamts7, Snai2, Tgfbr1, Lpl, Efemp1, Egfr, and Igfbp4 over pseudotime are shown. The orange line indicates young cells (cell number, n = 61), and the blue line indicates old cells (cell number, n = 209).
Fig 7.
GSEA and PPI analysis between young and old ASCs.
GSEA results showing the enrichment score (green line) and gene distributions (black line) among 12,729 genes rearranged according to the relative expression levels in Group 3 between young and old mice. Bars colored according to the gene distribution between the young (blue) and old (red) groups are shown. (A-C), (E-G) Results obtained using the GO public database (c5.all). (D) Results obtained using a curated public database (c2.all). All figures’ nominal p-values are lower than 0.05. (A–G) PPI among core enrichment genes of each gene set were analyzed using STRING. The confidence score cutoff for showing edges was set to ‘high’ (confidence score > 7.0). Edge colors indicated the type of evidence such as (i) co-expression (black), (ii) text-mining (yellow), (iii) biochemical/genetic data (“experiments”) (magenta), and (iv) previously curated pathway and protein-complex knowledge (“databases”) (cyan) of STRING.