NJ025-701

Transcriptomic Profiling of Aging Reveals Coordinated Pathway Decline and a Candidate Reversal Compound 

Kyungmin Kim1, Bumho Yoo1, and Seong-Eui Hong2*

1HuGeX, Co., Ltd., 20, Songdogukje-daero 286beon-gil, Yeonsu-gu, Incheon, 22013, Republic of Korea. 

2 4th Floor, Dream Hall, Korea Polytechnic Colleage– Seongnam Campus 398, Sujeong-ro, Sujeong-gu, Seongnam-si, Gyeonggi-do, 13122 Republic of Korea.

Corresondence to: Seong-Eui Hong, sehong@kopo.ac.kr

Received: July 11, 2025; Revised: September 22, 2025; Accepted: September 27, 2025; Published: September 29, 2025

NATPRO J. 2025, 2, 25-33 

https://doi.org/10.23177/NJ025.701 

Copyright ©  The Asian Society of Natural Products

Abstract

Aging is a complex biological process marked by widespread and coordinated transcriptional changes across multiple functional pathways. To investigate these changes, we analyzed RNA-seq data from human breast primary skin fibroblast cultured over extended passages, comparing early (P6) and late (P36) stages to model replicative aging. Differential expression analysis revealed 1,378 genes significantly altered with aging. Among these, five major pathways were consistently affected: G2-M Checkpoint, Myc Targets V1, Cholesterol Homeostasis, Interferon Alpha Response, and Inflammatory Response. To explore therapeutic implications, we integrated these gene signatures with the LINCS L1000 Chem Pert up dataset. Bosutinib, a Src/Abl kinase inhibitor, emerged as a top candidate predicted to reverse aging-associated transcriptional decline across all five functional categories. These findings highlight the utility of transcriptomic profiling combined with perturbation-based analysis to uncover biological changes linked to aging and nominate pharmacological candidates with geroprotective potential.

Keywords

aging; transcriptome; pathway

Introduction

Aging is a progressive, multifactorial biological process marked by a gradual loss of cellular and systemic homeostasis. Rather than resulting from the dysregulation of a few isolated genes, aging reflects broad and coordinated alterations in gene regulatory networks, affecting pathways involved in proliferation, metabolism, immune response, and stress resistance [1]. Recent integrative multi-omics analyses further support this view, showing that aging involves coordinated transcriptional, epigenetic, and metabolic shifts that converge on key functional networks such as inflammation, mitochondrial activity, and proteostasis [2]. This systems-level nature of aging suggests that macro-scale transcriptomic profiling is essential for uncovering its underlying molecular signatures and these insights have laid the groundwork for biomarker discovery and pathway-targeted interventions.

Building on this foundation, we applied a straightforward model of cellular aging based on long-term passaging of human breast primary skin fibroblast. Previous studies have also sought to systematically characterize aging-related changes within cultured cells and to quantify cellular aging using biomarker panels. For instance, Hartmann et al., proposed a standardized set of biological age markers—including DNA damage, senescence-associated secretory phenotype (SASP) factors, and telomere dynamics—to estimate the biological age of cells and other fibroblast systems in vitro [3]. By profiling gene expression across a wide range of passages and contrasting early (P6) and late (P36) stages [4], we aimed to capture replicative senescence–associated transcriptional dynamics in vitro.

In parallel with mechanistic studies, there has been growing interest in identifying compounds that may reverse aging-associated transcriptomic changes [5]. Recent studies have demonstrated the utility of transcriptome-based drug screening in aging research. For example, Lee et al., systematically integrated transcriptomic profiles from multiple aging models and applied signature-matching approaches to identify compounds that reverse aging-associated gene expression changes [5]. This strategy successfully nominated pharmacological agents capable of restoring youthful transcriptional states, highlighting the feasibility of leveraging public transcriptomic databases for geroprotective drug discovery.

In this study, we integrated transcriptomic aging signatures with perturbation data to explore therapeutic candidates. Despite relying on relatively simple methodologies, our approach effectively prioritized known aging-associated pathways and identified bosutinib— a Src/Abl kinase inhibitor recently shown to restore mitochondrial function, reduce pro-inflammatory gene expression, and improve physical performance [6]. These findings support the utility of transcriptome-informed frameworks for elucidating aging mechanisms and guiding pharmacological intervention strategies.

Materials and methods

RNA-seq Dataset and Sample Selection

Transcriptomic data were obtained from the publicly available GEO dataset GSE179848 [4], which includes RNA-seq profiles of human breast primary skin fibroblast subjected to long-term passaging. A total of 22 samples spanning passages 6 to 36 were analyzed, with passage 6 (P6) designated as the "young" group and passage 36 (P36) as the "old" group.

Quality Control and Read Preprocessing

Raw RNA-seq data (FASTQ files) were subjected to initial quality control using fastp (v0.20.1) to assess base quality distribution, GC content, and adapter contamination [7]. Adapter trimming and read filtering were performed using cutadapt (v3.4) with standard settings to remove sequencing artifacts and low-quality bases [8].

Read Alignment and Quantification

Cleaned reads were aligned to the human reference genome (GRCh38/hg38) using STAR aligner (v2.7.9a) with default parameters [9]. Aligned reads were then processed with featureCounts (Subread v2.0.1) to generate gene-level raw counts [10]. To assess consistency in sequencing depth across samples, raw counts were normalized to counts per million (CPM), and their distribution was visualized using boxplots, which confirmed a relatively uniform distribution of read counts among samples.

Data Normalization and DEG Analysis

Raw counts were normalized using DESeq2 (v1.36.0) in R (v4.2.0) to compute differentially expressed genes (DEGs) between P6 and P36 groups [11]. DEGs were defined using a threshold of |log2 fold change| ≥ 1 and adjusted p-value (padj) < 0.05. Transcripts per million (TPM) were also computed using RSEM [12] to enable cross-sample comparisons and downstream visualizations. These TPM values were subsequently used for heatmap visualization and sample clustering via Uniform Manifold Approximation and Projection (UMAP) to explore global transcriptomic patterns [13].

Functional Enrichment Analysis

Preranked gene set enrichment analysis (GSEA) was performed using the MSigDB Hallmark 2020 gene sets [14–16], where genes were ranked by log2 fold change values. To minimize gene overlap and ensure functional independence among the enriched pathways, pairwise Jaccard similarity was computed based on leading edge gene sets. The Jaccard similarity index between two gene sets A and B was calculated using the formula:

J(A, B) = |A ∩ B| / |A ∪ B|

where |A ∩ B| is the number of genes common to both sets, and |A ∪ B| is the total number of unique genes across both sets. Because Jaccard similarity provides a pairwise measure of gene overlap between pathways rather than an intrinsic score for each pathway, it cannot be used to rank or prioritize individual terms. Therefore, one representative pathway was selected from each major functional cluster based on prior knowledge and its well-established relevance to aging biology. This approach enabled the identification of functionally distinct clusters, from which five representative biological categories were defined through hierarchical clustering. 

To determine representative genes for each of these categories, we intersected the leading edge genes identified by GSEA with the list of differentially expressed genes (DEGs), selecting only those that met the DEG criteria (|log2 fold change| ≥ 1 and padj < 0.05).

Drug Perturbation and Candidate Prioritization

Given that all enriched pathways were downregulated during aging, we aimed to identify compounds capable of reversing these transcriptomic changes. For this purpose, the LINCS L1000 Chem Pert up dataset—representing genes upregulated in response to compound treatment—was queried via the Enrichr platform [17, 18]. The enrichment analysis was performed using Fisher's exact test, which evaluates the overlap between input genes and drug-induced gene signatures. Enrichment significance (adjusted p < 0.05) and breadth of gene regulation across functional categories were used to prioritize candidates

Visualization and Computational Tools

All analyses were conducted using R (v4.2.0) and Python (v3.10). Visualization tools included seaborn for heatmaps and hierarchical clustering, and plotly for Sankey diagram generation.

Results

Statistical overview of gene expression in aging transcriptomes

To capture key aging-related genes and their functional roles, we selected RNA-seq samples from the publicly available GSE179848 dataset [4], which provides transcriptomic profiles of human breast primary skin fibroblast subjected to long-term passaging. Passage 6 (P6) samples were designated as the "young" group, representing early-stage culture, while passage 36 (P36) samples were defined as the "old" group, reflecting advanced replicative aging through extensive subculture. For this analysis, we focused on four high-quality RNA-seq samples representing two biological replicates from each group: P6 (SRR15095913, SRR15095924) and P36 (SRR15095923, SRR15095934).

The number of reads after filtering ranged from 71.2 million to 84.2 million, indicating robust sequencing depth across all samples. Duplication rates were relatively consistent, ranging between 18.9% and 20.0%, which is within the expected threshold for high-quality RNA-seq libraries. A summary of sample-level quality metrics is presented in Table 1. These results confirm that the dataset provides sufficient quality and coverage for downstream differential expression and functional enrichment analyses across passage-associated aging conditions.

Following quality control, RNA-seq reads were aligned to the reference genome (GRCh38/hg38) using the STAR aligner [9], and gene-level raw counts were generated using featureCounts [10]. To normalize for library size differences, raw counts were converted to counts per million (CPM). Distribution of CPM values across samples was then examined to assess global expression patterns and detect any sample-specific bias or anomalies prior to differential expression analysis. As shown in Figure 1, the log2-transformed CPM distributions were highly consistent across replicates and passage groups, suggesting comparable global gene expression levels among the samples.

Key genes and their functional characteristics

To identify transcriptomic changes associated with aging, we performed differential expression analysis between the young (P6) and old (P36) groups. Gene-level counts were obtained using featureCounts, and differential expression was assessed using the DESeq2 package. Genes with a log2 fold change (log2FC) ≥ 1 and an adjusted p-value (padj) < 0.05 were considered significantly differentially expressed. A total of 1,378 genes met this threshold, with 635 genes upregulated and 743 genes downregulated in the old (P36) group compared to the young (P6) group. Figure 2 summarizes the DEG landscape. The volcano plot (Figure 2A) illustrates the distribution of all genes based on log2FC and adjusted p-values. Among the top 50 most significant DEGs, the majority were clearly downregulated, reflecting a transcriptional decline associated with replicative aging. The accompanying heatmap (Figure 2B) shows distinct expression profiles of selected DEGs across the four RNA-seq samples, with consistent patterns observed between replicates. KEGG enrichment analysis of the 635 upregulated genes revealed significant enrichment of the p53 signaling pathway and lysosome function. In contrast, downregulated genes were enriched in pathways related to cell cycle progression (e.g., Cell cycle, DNA replication, Oocyte meiosis) and DNA repair (e.g., base excision repair, mismatch repair, homologous recombination). These results indicate that aging-associated transcriptional changes include both the activation of stress-response pathways and the suppression of proliferative and genome maintenance programs (Figure 2C). The enrichment of the p53 signaling pathway among upregulated genes highlights its central role in enforcing cell cycle arrest and maintaining the senescent state through p21-mediated checkpoint control [19–21]. Concurrently, the enrichment of lysosome-related functions reflects the metabolic remodeling and secretory activity characteristic of senescent cells, including increased autophagic flux and senescence-associated β-galactosidase activity [22, 23]. Together, these pathways underscore the dual contribution of stress-response signaling and metabolic adaptation to the establishment of cellular senescence. To evaluate how these expression changes evolve over intermediate passages, we extended our analysis to include additional samples across a range of passage numbers. Transcript abundances were quantified as TPM (transcripts per million), enabling cross-sample comparisons. A UMAP projection based on TPM values across all samples revealed a gradual and continuous transition in global gene expression, supporting the progressive nature of aging-associated transcriptomic changes (Figure 2D). This progression reflects the accumulative molecular changes associated with cellular senescence and supports the dynamic nature of gene expression reprogramming across time in culture. 

Enriched biological functions associated with aging 

To investigate the biological processes associated with transcriptomic changes during aging, we conducted gene set enrichment analysis (GSEA) using the “MSigDB Hallmark 2020” gene sets. A preranked GSEA approach was applied, in which genes were ranked by their log2 fold change derived from the differential expression analysis between the young (P6) and old (P36) groups. GSEA results with a false discovery rate (FDR) below 0.1 were considered significant, and the selected pathways are listed in Table 2.

Among the significant pathways, substantial gene overlap was observed, highlighting the need to distinguish functionally independent categories. To address this, we computed pairwise Jaccard similarity between the gene sets of the top hallmark pathways, enabling the identification of overlapping gene content and reduction of redundancy among the enrichment results. As shown in Figure 3A, hierarchical clustering of the Jaccard index matrix revealed a distinct cluster composed of cell cycle–associated gene sets such as Spermatogenesis, Mitotic Spindle, G2M Checkpoint, and E2F Targets. Although Myc Targets V1 and V2 also contain cell proliferation–related genes, they were not grouped within this primary cluster, suggesting a partially distinct regulatory program. Additionally, inflammatory and interferon-related pathways also showed somewhat independent. Notably, Cholesterol Homeostasis and mTORC1 Signaling share partial gene content that may reflect coordinated metabolic adaptation during aging [24, 25].

Based on the clustering and functional overlap, we categorized the enriched pathways into five representative themes: G2-M Checkpoint (cell cycle), Cholesterol Homeostasis (lipid metabolism), Myc Targets V1 (proliferation-associated), Interferon Alpha Response, and Inflammatory Response (immune-related). This categorization facilitates a clearer interpretation of aging-associated biological changes (Figure 3B).

To further interpret the biological roles of each category, we defined representative genes by intersecting the leading edge genes of each hallmark pathway with the list of DEGs. Genes that belonged to both the GSEA lead gene sets and showed significant expression changes (log2FC ≥ 1 or ≤ -1, padj < 0.05) were considered as key contributors within their respective categories. These representative DEGs were used to highlight category-specific molecular signatures associated with aging. For example, UBE2C and CDC20 were prominent in the G2-M Checkpoint category [26–28], while PPARG and SCD were identified in the Cholesterol Homeostasis group [29, 30]. This approach allowed for more focused downstream interpretation and potential biomarker identification within each functional class.

To assess the temporal expression dynamics of representative DEGs in each category, we examined their transcript levels across a series of passages. As shown in Figure 4, genes associated with the G2-M Checkpoint, Myc Targets V1, and Cholesterol Homeostasis categories exhibited a gradual decrease in expression as passage number increased, reflecting reduced proliferative and metabolic activity during cellular aging. In contrast, genes in the Interferon Alpha Response and Inflammatory Response categories also showed a decreasing trend, but the pattern was less pronounced compared to the other categories, indicating a more moderate transcriptional modulation of immune-related pathways during aging. Through this series of analyses, we identified biological functions that progressively declined with aging. In particular, genes associated with cell division, lipid metabolism, and immune response consistently showed reduced expression across the aging continuum, highlighting key functional pathways affected during the aging process.

Potential Biomarker Candidates and Drug Associations

To identify potential pharmacological modulators of aging-associated transcriptomic decline, we investigated the relationship between downregulated genes from the five representative functional categories and known compound-induced gene expression profiles. Given that key aging-related genes across categories were predominantly downregulated, we utilized the "LINCS L1000 Chem Pert up" gene set via Enrichr to identify compounds that could potentially reverse these transcriptomic patterns. Instead of focusing on narrow biological functions, we prioritized compounds that influence gene expression across multiple aging-associated biological categories. This strategy favors agents with the potential to restore multiple biological programs affected by aging. As a result of this enrichment analysis, a total of eight compounds with statistically significant effects (padj < 0.05) were identified, as shown in Figure 5A. Among these, bosutinib emerged as the most prominent compound, exhibiting the broadest upregulation of genes within the major functional pathways. To further highlight the functional relevance of these compounds, we identified representative target genes regulated by bosutinib. This compound targets UBE2S, CKS2, and KPNA2 (G2-M Checkpoint); SCD, HMGCS1, SQLE, IDI1, DHCR7, LDLR, and FDFT1 (Cholesterol Homeostasis); and IFITM1 and NFKBIA (Interferon Alpha and Inflammatory Responses) (Figure 5B). 

Together, these findings demonstrate that transcriptome-based analysis can uncover biological functions that decline with aging and enable the screening of candidate compounds capable of restoring the expression of genes involved in these key functional pathways.

Discussion

This study systematically characterized transcriptomic changes associated with cellular aging using long-term passaging of human breast primary skin fibroblast, identifying five major biological pathways—G2-M Checkpoint, Myc Targets V1, Cholesterol Homeostasis, Interferon Alpha Response, and Inflammatory Response—that exhibit consistent downregulation over time and nominating bosutinib as a top candidate compound with the potential to reverse these gene expression changes. While the study design was relatively simple, it provided valuable insights into aging at the transcriptomic level; nevertheless, deeper mechanistic understanding will require more complex and refined experimental approaches. For example, although these genes generally showed a gradual decline, a notable increase in expression was observed near passage 15. This suggests that transcriptomic alterations during aging may be nonlinear, highlighting the importance of more granular sampling and analysis at intermediate stages of the aging trajectory. To further clarify and strengthen these findings, several aspects of the current study warrant additional refinement and investigation.

We primarily focused on biological functions that exhibited progressive transcriptional decline; however, despite the presence of 635 upregulated genes, GSEA did not reveal any significantly enriched biological functions among them. However, KEGG-based enrichment analysis revealed that these upregulated genes were associated with the p53 signaling pathway and lysosomal function, indicating potential involvement in cellular stress responses and homeostasis mechanisms during aging [31]. The activation of the p53 signaling pathway in our dataset is consistent with its well-established role as a gatekeeper of the senescent phenotype, ensuring that damaged or stressed cells undergo permanent cell cycle arrest rather than uncontrolled proliferation [32–34]. Beyond enforcing arrest, p53 also modulates metabolic and stress-response programs, linking genome surveillance to the broader remodeling observed in aging cells. In parallel, the observed enrichment of lysosome-related pathways highlights their complex role in senescence. On one hand, increased lysosomal content and activity are hallmarks of aging cells, exemplified by elevated senescence-associated β-galactosidase activity and impaired autophagic flux [35]. On the other hand, lysosomal pathways are not merely passive markers but active regulators of cell fate, capable of both promoting and restraining senescence depending on context. For instance, enhanced lysosomal degradation may facilitate the clearance of damaged organelles and mitigate stress, whereas chronic lysosomal dysfunction can exacerbate metabolic imbalance and reinforce the senescent state [36]. Our transcriptomic data revealed an upregulation of lysosomal genes during late passages. This finding is consistent with the view that, in response to senescence-inducing stress, the lysosomal processing system (LYPAS) can be adaptively expanded through mechanisms such as mTORC1 regulation, Ca²⁺ signaling, and cGAS–STING activation to enhance autophagic flux, acidification, and exocytosis [37]. However, under conditions of persistent or severe stress, these same pathways may paradoxically contribute to the establishment of the senescent phenotype by promoting SASP expression and lysosomal exocytosis, underscoring the bidirectional role of lysosomes in cellular aging. Thus, lysosomal upregulation in our system may reflect an adaptive attempt to counterbalance accumulating damage, underscoring the complex and context-dependent role of lysosomes in cellular aging. A deeper exploration of these upregulated processes may reveal compensatory mechanisms or stress responses that emerge as cells undergo aging-related changes. 

To identify pharmacological modulators capable of reversing aging-associated gene expression changes, we applied a relatively straightforward approach by leveraging the LINCS L1000 Chem Pert up dataset [27]. Despite the simplicity of this method, it enabled the effective screening of potential compounds and yielded intriguing results, including the identification of bosutinib as a promising candidate. However, a limitation of this approach is that the Enrichr “LINCS L1000 Chem Pert up” library provides only qualitative lists of up-regulated genes for each compound, without quantitative fold-change information. As a result, while we could robustly assess enrichment of aging-associated DEGs in the bosutinib-responsive gene set, we could not directly visualize the magnitude of transcriptional reversal. In addition, although our analysis highlighted genes and pathways potentially modulated by bosutinib, direct mechanistic links between this compound and individual aging-related genes have not yet been reported. Thus, future studies using raw L1000 expression profiles will be necessary to more precisely characterize the quantitative reversal effects of bosutinib. 

Interestingly, previous studies have shown that bosutinib, a Src/Abl kinase inhibitor, recently shown to restore mitochondrial function, reduce pro-inflammatory gene expression, and improve physical performance [6]. These effects are thought to be mediated, at least in part, through the inhibition of Src and Abl kinase signaling pathways, which are known to regulate cellular proliferation, DNA damage responses, and senescence [38, 39]. By attenuating these kinase-driven stress and checkpoint pathways, bosutinib may help restore gene expression programs disrupted during aging. Nevertheless, it should be noted that Src inhibition has often been associated with the induction of cell cycle arrest, and therefore may not directly explain the reversal of cell cycle–related decline observed in our aging dataset. Alternatively, our findings may reflect an indirect effect, whereby bosutinib-responsive pathways influence the expression of cell cycle regulators such as KPNA2, UBE2S, and CKS2, leading to their partial restoration despite the canonical arrest-inducing role of Src inhibition. Among the five functional pathways identified in our analysis, the cholesterol homeostasis category showed the most diverse set of genes whose expression was increased by bosutinib (adjusted p-value=4.49×10-12). To our knowledge, however, there are no direct reports demonstrating that Src inhibitors regulate or enhance cholesterol metabolism genes. Nevertheless, accumulating evidence indicates that cholesterol homeostasis is tightly linked to aging processes and cellular senescence, with dysregulation of cholesterol biosynthesis and transport contributing to metabolic imbalance and age-related decline [40]. Our findings suggest that bosutinib may partially restore this pathway, the precise molecular mechanisms remain to be elucidated. This represents a limitation of the present study and underscores the need for future investigations to determine how Src/Abl inhibition interfaces with cholesterol metabolism in the context of aging. In addition, Src family kinases (SFKs) function as adaptor and auxiliary kinases downstream of TLRs, immunoreceptors (ITAMs), integrins, and the IL-6 receptor, where they cooperate with Syk and JAK to amplify IKK–NF-κB and JAK–STAT3 signaling [41]. This amplification promotes the expression of pro-inflammatory cytokines such as TNF-α and IL-6, thereby reinforcing the SASP that characterizes aging cells. This mechanistic link provides a plausible explanation for the Bosutinib-associated suppression of inflammatory pathways observed in our dataset.

While bosutinib itself has not yet been directly tested in aging-focused clinical studies, its close mechanistic relative dasatinib—a Src family kinase inhibitor—has been evaluated as a senolytic agent in both preclinical models and human trials. For example, in the ongoing clinical trial NCT04946383, intermittent dasatinib plus quercetin administration is being tested for its ability to reduce epigenetic aging and senescence-associated biomarkers in older adults. This provides precedent that targeting Src kinases may have geroprotective effects in humans. In this context, bosutinib’s ability to inhibit Src/Abl kinases and reverse aging-related transcriptomic signatures in our analysis suggests that it may share similar potential as a candidate geroprotective intervention. To improve the precision of drug screening approaches, it will be essential to incorporate a broader range of drug perturbation databases and to develop strategies for distinguishing genes that are directly affected by candidate compounds from those that are indirectly modulated. Such refinements will help clarify the mechanisms through which compounds exert their effects and enhance the translational relevance of the findings.

Although the use of passage number as a surrogate for aging in cultured cells has enabled controlled investigation of aging-related transcriptomic changes, this model presents inherent limitations. As it reflects replicative senescence under in vitro conditions, it may not fully capture the multifaceted biological, environmental, and systemic factors that contribute to organismal aging in vivo. Thus, to build upon the insights gained from this study, future research should aim to integrate more physiologically relevant models and experimental contexts that better recapitulate the complexity of aging biology.

Conclusion

In this study, we systematically characterized transcriptomic alterations associated with cellular aging using a long-term passaging model of human primary skin fibroblast from breast. By comparing early (P6) and late (P36) passage samples, we identified five major functional pathways—G2-M Checkpoint, Myc Targets V1, Cholesterol Homeostasis, Interferon Alpha Response, and Inflammatory Response—that were consistently downregulated with increasing passage number. Integration of these aging-associated gene signatures with drug perturbation datasets led to the identification of bosutinib, a Src/Abl kinase inhibitor, as a top candidate compound capable of reversing gene expression declines across these pathways.

Acknowledgement

This study was conducted as part of a collaborative research initiative between HuGeX, Co., Ltd. and the Department of AI Medical Biotechnology at Korea Polytechnics, established through a memorandum of understanding (MOU). We gratefully acknowledge the institutional partnership that supported this work.

Availability of data and materials

The RNA-seq data analyzed in this study are publicly available from the NCBI Gene Expression Omnibus (GEO) under the accession number GSE179848 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE179848). All additional data generated or analyzed during this study are available from the corresponding author upon reasonable request.

Conflict of Interest

The authors declare that they have no competing interests.

Funding

This work was supported by the grant from the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health and Welfare (RS-2024-00345193).

Authors' contributions

KK and BY conceived the study and designed the overall analytical strategy. SEH performed the data analysis. All authors read and approved the final manuscript.

References

1. Palmer D.; Fabris F.; Doherty A.; Freitas A. A.; de Magalhães J. P. Ageing transcriptome meta-analysis reveals similarities and differences between key mammalian tissues. Aging. 2021, 13, 3313–41. https://doi.org/10.18632/aging.202648.

2. Zhu X.; Chen Z.; Shen W.; Huang G.; Sedivy J. M.; Wang H, et al. Inflammation, epigenetics, and metabolism converge to cell senescence and ageing: the regulation and intervention. Signal Transduct Target Ther. 2021, 6, 245. https://doi.org/10.1038/s41392-021-00646-9.

3. Hartmann C.; Herling L.; Hartmann A.; Köckritz V.; Fuellen G.; Walter M, et al. Systematic estimation of biological age of in vitro cell culture systems by an age-associated marker panel. Front. Aging. 2023, 4, 1129107. https://doi.org/10.3389/fragi.2023.1129107.

4. Sturm G.; Karan K. R.; Monzel A. S.; Santhanam B.; Taivassalo T.; Bris C, et al. OxPhos defects cause hypermetabolism and reduce lifespan in cells and in patients with mitochondrial diseases. Commun Biol. 2023, 6, 22. https://doi.org/10.1038/s42003-022-04303-x.

5. Yang J-H.; Petty C. A.; Dixon-McDougall T.; Lopez M. V.; Tyshkovskiy A.; Maybury-Lewis S, et al. Chemically induced reprogramming to reverse cellular aging. Aging, 2023, 15, 5966–89. https://doi.org/10.18632/aging.204896.

6. Motaln H.; Rogelj B. The Role of c-Abl Tyrosine Kinase in Brain and Its Pathologies. Cells, 2023, 12. https://doi.org/10.3390/cells12162041.

7. Chen S.; Zhou Y.; Chen Y.; Gu J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics, 2018, 34, i884–90. https://doi.org/10.1093/bioinformatics/bty560.

8. Martin M, Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011, 17, 10. https://doi.org/10.14806/ej.17.1.200.

9. Dobin A.; Davis C. A.; Schlesinger F.; Drenkow J.; Zaleski C.; Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics, 2013, 29, 15–21. https://doi.org/10.1093/bioinformatics/bts635.

10. Liao Y.; Smyth G. K.; Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics, 2014, 30, 923–30. https://doi.org/10.1093/bioinformatics/btt656.

11. Love M. I.; Huber W.; Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014, 15, 550. https://doi.org/10.1186/s13059-014-0550-8.

12. Li B.; Dewey C. N. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics, 2011, 12, 323. https://doi.org/10.1186/1471-2105-12-323.

13. McInnes L.; Healy J.; Melville J. UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction. 2020. https://doi.org/10.48550/arXiv.1802.03426.

14. Liberzon A.; Birger C.; Thorvaldsdóttir H.; Ghandi M.; Mesirov J. P.; Tamayo P. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 2015, 1, 417–25. https://doi.org/10.1016/j.cels.2015.12.004.

15. Subramanian A.; Tamayo P.; Mootha V. K.; Mukherjee S.; Ebert B. L.; Gillette M. A, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005, 102, 15545–50. https://doi.org/10.1073/pnas.0506580102.

16. Fang Z.; Liu X.; Peltz G. GSEApy: a comprehensive package for performing gene set enrichment analysis in Python. Bioinformatics. 2023, 39. https://doi.org/10.1093/bioinformatics/btac757.

17. Kuleshov M. V.; Jones M. R.; Rouillard A. D.; Fernandez N. F.; Duan Q.; Wang Z, et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 2016, 44, W90-7. https://doi.org/10.1093/nar/gkw377.

18. Chen E. Y.; Tan C. M.; Kou Y.; Duan Q.; Wang Z.; Meirelles GV, et al. Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics. 2013, 14, 128. https://doi.org/10.1186/1471-2105-14-128.

19. Yan J.; Chen S.; Yi Z.; Zhao R.; Zhu J.; Ding S, et al. The role of p21 in cellular senescence and aging-related diseases. Mol Cells. 2024, 47, 100113. https://doi.org/10.1016/j.mocell.2024.100113.

20. Kumari R.; Jat P.; Mechanisms of Cellular Senescence: Cell Cycle Arrest and Senescence Associated Secretory Phenotype. Front Cell Dev Biol. 2021, 9, 645593. https://doi.org/10.3389/fcell.2021.645593.

21. Engeland K. Cell cycle regulation: p53-p21-RB signaling. Cell Death Differ. 2022, 29, 946–60. https://doi.org/10.1038/s41418-022-00988-z.

22. Rovira M.; Sereda R.; Pladevall-Morera D.; Ramponi V.; Marin I.; Maus M, et al. The lysosomal proteome of senescent cells contributes to the senescence secretome. Aging Cell. 2022, 21, e13707. https://doi.org/10.1111/acel.13707.

23. Lee B. Y.; Han J. A.; Im J. S.; Morrone A.; Johung K.; Goodwin E. C, et al. Senescence-associated beta-galactosidase is lysosomal beta-galactosidase. Aging Cell. 2006, 5, 187–95. https://doi.org/10.1111/j.1474-9726.2006.00199.x.

24. Navyasree K. V.; Ramesh S. T.; Umasankar P. K. Cholesterol regulates insulin-induced mTORC1 signaling. J Cell Sci. 2023, 136. https://doi.org/10.1242/jcs.261402.

25. Davis O. B.; Shin H. R.; Lim C-Y.; Wu E. Y.; Kukurugya M.; Maher C. F, et al. NPC1-mTORC1 Signaling Couples Cholesterol Sensing to Organelle Homeostasis and Is a Targetable Pathway in Niemann-Pick Type C. Dev Cell. 2021, 56, 260-276.e7. https://doi.org/10.1016/j.devcel.2020.11.016.

26. Wang H.; Zhang C.; Rorick A.; Wu D.; Chiu M.; Thomas-Ahner J, et al. CCI-779 inhibits cell-cycle G2-M progression and invasion of castration-resistant prostate cancer via attenuation of UBE2C transcription and mRNA stability. Cancer Res. 2011, 71, 4866–76. https://doi.org/10.1158/0008-5472.CAN-10-4576.

27. Nicolau-Neto P.; Palumbo A.; De Martino M.; Esposito F.; de Almeida Simão T.; Fusco A, et al. UBE2C Is a Transcriptional Target of the Cell Cycle Regulator FOXM1. Genes (Basel). 2018, 9. https://doi.org/10.3390/genes9040188.

28. Lim H. H.; Surana U. Cdc20, a beta-transducin homologue, links RAD9-mediated G2/M checkpoint control to mitosis in Saccharomyces cerevisiae. Mol Gen Genet. 1996, 253, 138–48. https://doi.org/10.1007/s004380050306.

29. Han T.; Lv Y.; Wang S.; Hu T.; Hong H.; Fu Z. PPARγ overexpression regulates cholesterol metabolism in human L02 hepatocytes. J Pharmacol Sci. 2019, 139, 1–8. https://doi.org/10.1016/j.jphs.2018.09.013.

30. Sun Q.; Xing X.; Wang H.; Wan K.; Fan R.; Liu C, et al. SCD1 is the critical signaling hub to mediate metabolic diseases: Mechanism and the development of its inhibitors. Biomed Pharmacother. 2024, 170, 115586. https://doi.org/10.1016/j.biopha.2023.115586.

31. Huang Y.; Che X.; Wang P. W.; Qu X. p53/MDM2 signaling pathway in aging, senescence and tumorigenesis. Semin Cancer Biol. 2024, 101, 44–57. https://doi.org/10.1016/j.semcancer.2024.05.001.

32. Vousden K. H.; Lane D. P. p53 in health and disease. Nat Rev Mol Cell Biol. 2007, 8, 275–83. https://doi.org/10.1038/nrm2147.

33. Anerillas C.; Herman A. B.; Rossi M.; Munk R.; Lehrmann E.; Martindale J. L, et al. Early SRC activation skews cell fate from apoptosis to senescence. Sci Adv. 2022, 8, eabm0756. https://doi.org/10.1126/sciadv.abm0756.

34. Huang Y.; Che X.; Wang P. W.; Qu X. p53/MDM2 signaling pathway in aging, senescence and tumorigenesis. Semin Cancer Biol. 2024, 101, 44–57. https://doi.org/10.1016/j.semcancer.2024.05.001.

35. Kurz D. J.; Decary S.; Hong Y.; Erusalimsky J. D. Senescence-associated (beta)-galactosidase reflects an increase in lysosomal mass during replicative ageing of human endothelial cells. J Cell Sci. 2000, 113 (Pt 20), 3613–22. https://doi.org/10.1242/jcs.113.20.3613.

36. Shvedova M.; Samdavid Thanapaul R. J. R.; Thompson E. L.; Niedernhofer L. J.; Roh D. S. Cellular Senescence in Aging, Tissue Repair, and Regeneration. Plast Reconstr Surg. 2022, 150, 4S-11S. https://doi.org/10.1097/PRS.0000000000009667.

37. Tan J. X.; Finkel T. Lysosomes in senescence and aging. EMBO Rep. 2023, 24, e57265. https://doi.org/10.15252/embr.202357265.

38. Anerillas C.; Herman A. B.; Rossi M.; Munk R.; Lehrmann E.; Martindale J. L, et al. Early SRC activation skews cell fate from apoptosis to senescence. Sci Adv. 2022, 8, eabm0756. https://doi.org/10.1126/sciadv.abm0756.

39. Bohio A. A.; Wang R.; Zeng X.; Ba X. c Abl mediated tyrosine phosphorylation of DNA damage response proteins and implications in important cellular functions (Review). Mol Med Rep. 2020, 22, 612–9. https://doi.org/10.3892/mmr.2020.11156.

40. Ziegler D. V.; Czarnecka-Herok J.; Vernier M.; Scholtes C.; Camprubi C.; Huna A, et al. Cholesterol biosynthetic pathway induces cellular senescence through ERRα. npj aging. 2024, 10, 5. https://doi.org/10.1038/s41514-023-00128-y.

41. Abram C. L.; Lowell C. A. The diverse functions of Src family kinases in macrophages. Front Biosci. 2008, 13, 4426–50. https://doi.org/10.2741/3015.