Compartment-specific eccDNA patterns reveal senescence associated biomarkers in hUC-MSCs

Abstract

The therapeutic potential of human umbilical cord-derived mesenchymal stem cells (hUC-MSCs) is significantly constrained by replicative senescence during in vitro expansion. To investigate this phenomenon, we established a long-term passaging model of hUC-MSCs and validated senescence-associated phenotypes. Through integrated transcriptomic and eccDNA profiling, we systematically analyzed three compartments—cell pellets (CP), cell culture media (CM), and extracellular vesicles (EVs)—at sequential timepoints (days 5, 22 and 43, post-seeding). Intriguingly, while eccDNA abundance in CP remained stable during senescence progression, CM and EVs exhibited higher eccDNA loads, respectively, compared to CP. Moreover, CM- and EVs-derived eccDNAs shared conserved size distribution patterns distinct from CP-associated eccDNAs. Furthermore, the compartment-specific eccDNA dynamics over time coincided with activation of growth-regulatory gene expression. We identified recurrent eccDNA species across all compartments, harboring genes linked to senescence-associated processes. Our study uncovers compartment-specific eccDNA dynamics during hUC-MSC aging and proposes their potential as biomarkers for senescence monitoring. These findings provide a foundation for developing strategies to mitigate senescence-related limitations in hUC-MSC clinical applications.

Supplementary Information: The online version contains supplementary material available at 10.1038/s41598-025-25377-7.

Keywords: Biomarkers; Circle-Seq; Extracellular vesicles; Extrachromosomal circular DNA; Intercellular communication; Mesenchymal stem cells; Replicative senescence.

Fig. 1

Characterization of replicative senescence features in hUC-MSCs. (A) The MSCs from the umbilical cords of nine puerperal volunteers were acquired and amplified in-vitro. Different methodologies are performed on three types of the cell contents. The cell pellets (CP), conditional cell culture media (CM) and extracellular vesicles (EVs) were sampled at passage 1 and passage 6, passage 12. Whole genome sequencing (WGS) was performed on CP samples from passage 1. RNA-Seq was performed on samples from all three timepoints of CP. And Circle-Seq was applied for eccDNA detection in samples of all three passages of CP, CM, and EVs. (B–D) The diameter, cell viability, and cumulative population doubling level of the nine MSCs cell lines during the long-term passage (****p < 0.0001, one-way ANOVA test).

Fig. 2

Identification of the extrachromosomal circular DNA features in cell pellets of MSCs. (A) The workflow of eccDNA enrichment through DNA extraction, linear DNA removal by exonuclease V, Phi-29 rolling circle amplification, sequencing, and analysis by Circle-Map pipeline. (B) The eccDNA per million mappable reads of CP samples of passage 1, 6 and 12 (Wilcoxon test). (C) The differential eccDNA gene abundance between P1, P6 and P12. The X-axis and Y-axis represent sample id and gene id respectively. 1382 eccGenes presented differently between 3 groups by the Wilcox test. (D) Normalized ratio of eccDNA in different genomic elements. (E) Normalized mapping ratio of eccDNA reads aligned to repeats.

Fig. 3

Extracellular circular DNA identification from EVs and CM samples. (A) Technical pipeline of EVs extraction from the cell culture media. (B,C) The size distribution and the density of eccDNA size of P1, P6, P12 samples in EV, and CM respectively. (D,E) The comparison of eccDNA number in EVs and in CM of P1, P6 and P12 (Wilcoxon test). (F) The overall eccDNA number of CP, CM, EV samples from all timepoints are compared (Wilcoxon test). (G) EPM normalization scores under different conditions (CP, CM, EV) across three time points (p1, p6, p12). Statistical analysis was performed using Wilcoxon test. (H) The median GC values under different conditions (CM, CP, EV) across three time points (P1, P6, P12). Statistical significance between groups was assessed using Wilcoxon test.

Fig. 4

The genomic function analysis of eccDNA in long term passage of hUC-MSCs. (A) Normalized ratio of eccDNA in different genomic elements. (B) Normalized mapping ratio of eccDNA reads aligned to repeats. (C) The number of genes associated with eccDNA across different genomic regions: Gene2KU, 5’UTR, Exon, Intron, 3’UTR, and Gene2KD. Positive bars (green) indicate enriched regions, while negative bars (brown) represent underrepresented ones. Values on each bar provide a quantitative comparison across these elements. (D) The distribution of genes associated with different genomic regions, including Gene2KU, 5’UTR, Exon, Intron, 3’UTR, and Gene2KD. The x-axis shows the p-value (− log10), and the y-axis represents the adjusted p-value (padj, − log10). Genes with significant expression changes are highlighted, with thresholds set at p-value = 0.01 and padj = 0.01 (marked by blue lines). Points are color-coded based on their genomic region, with labeled genes representing notable candidates within their respective categories. (E) The distribution of recurrent genes on eccDNAs across P1, P6, and P12. Each bar is segmented by sample count within the respective group, with colors representing the different time points (black for P1, brown for P6, and blue for P12). (F) The enriched biological pathways based on the gene set analysis of recurrent presented eccDNAs. Pathways include the regulation of the actin cytoskeleton, Ras signaling, axon guidance, growth hormone synthesis, secretion, and action, T cell receptor signaling, mRNA surveillance, Fc epsilon RI signaling, biosynthesis of nucleotide sugars, and D-amino acid metabolism. The x-axis indicates the count of genes associated with each pathway, showing the relative enrichment across these biological processes.

Fig. 5

Recurrent eccDNA is identified in CP, CM, EVs at P1, P6, P12. (A) The recurrently presented eccDNA in passage 1(P1), passage 6 (P6), passage 12 (P12) in CP, CM, EVs. The black bar represents eccDNA number in P1, blue bar represents eccDNA number in P6, and the yellow bar represents eccDNA number in P12. (B) The comparison of length distribution of eccDNA in P1, P6, P12. The brown line represents the interaction of CP and CM, blue line represents the interaction of CP and EV, and the yellow line represents the interaction of CP, EV and CM.