Microgravity changes aging-related gene activity controlled by the p53 system

In a DI-SMG study conducted on ten healthy volunteers, we aimed to investigate whether simulated microgravity affects the expression of aging-related genes in the p53 gene network in human peripheral blood T cells in vivo.

The DI-SMG study was conducted and global transcriptome profiling via RNA-seq was performed on T cells as described below (see. Materials and Methods). RNA-seq analysis yielded expression data for 58,676 transcripts, including 17,760 known genes and 38,430 transcripts associated with uncharacterized gene products. This dataset has been deposited in the NCBI Gene Expression Omnibus (GEO) repository under accession number GSE301964. Subsequent analysis focused specifically on interrogating the p53 gene network. Section 3

The transcript level of the TP53 tumor suppressor gene varied among all 10 volunteers in the study. At different timepoints of the experiment, TP53 was up-regulated in 9 of 10 volunteers compared to background levels (timepoint "−7 day", Figure 1), and the dynamics of its expression varied between individuals. Seven volunteers (## 1–5, 7, and 8) demonstrated an increase in TP53 transcript levels at 7 days in SMG and then showed variable expression with an overall upward trend. Volunteer #6 showed a slight decrease in TP53 transcript levels during the SMG course and then an up-regulation above the background (BG) level at the 28-day timepoint after the SMG course. Volunteer #9 showed an overall down-regulation of the TP53 transcript level with a decline in its expression at 7 days and 14 days in SMG, and then its level was partially recovered at the 21-day timepoint. Volunteer #10 also showed a decrease in TP53 expression at 7 days and 14 days in SMG, but it was increased on day 21 of SMG, and the expression level increased further after the SMG course on day 28 (Figure 2).

The statistical analysis reveals a statistically significant upward trend in TP53 gene expression over the course of the study (Figure S1). This observation is supported by a mean slope coefficient of 1.22 ± 0.88 with a 95% confidence interval that does not include zero

The expression profile of the p53 gene network was significantly altered in response to the 3-week DI-SMG intervention. The expression profile of the complete gene set (n = 75) from the KEGG p53 signaling pathway (hsa04115) was evaluated. Notably, a subset of 30 genes (Table 1) showed differential expression in all, or nearly all, of the ten volunteers. This gene set included central apoptosis and senescence regulators such as GADD45, p21, PUMA, NOXA, and p19ARF, highlighting the pathway's central role in the physiological response to simulated microgravity.

In relation to the regulation of the cell cycle, the gene expression of three B cyclins (CCNB1, CCNB2, and CCNB3) and two cyclin-dependent kinase inhibitors (CDKN1A and CDKN1B) in p53 networks exhibited dynamic changes under DI-SMG conditions in all ten subjects.

Growth arrest and DNA damage-inducible alpha, beta and gamma (GADD45A, GADD45B, and GADD45G) playing roles in cell cycle arrest, DNA repair, apoptosis, innate immunity, genomic stability, and senescence showed coordinated differential expression during the DI SMG study in all ten volunteers (Figure 3).

KEGG pathway enrichment analysis indicated that DI-SMG exposure significantly affected the p53 signaling pathway hsa04115 (). Supplementary File S1

The differential expression of p53 network genes linked to aging and cellular senescence was recorded in the SMG study. The expression of dominant active p53 leads to the constitutive expression of downstream target genes and results in premature aging [15]. There were varying changes in the TP53 transcript level observed during the experiment as described above (Section 2.2, Figure 2).

The cell cycle cyclin-dependent kinase inhibitor 1A (CDKN1A) interacts with cyclin-dependent kinases CDK2 and CDK4 and blocks their progression through the cell cycle. BH domain Bcl-2-associated protein X (Bax) and p53 up-regulated modulator of apoptosis (PUMA) are potent inducers of apoptosis. Cellular senescence and apoptosis prevent tumorigenesis. Also, both pathways have the potential to deplete stem and progenitor cell pools that leads to impaired tissue renewal, ultimately impairing organ homeostasis—a hallmark of aging [17]. Genes encoding cyclin-dependent kinase inhibitors CDKN1A and CDKN1B, Bax, PUMA, and cyclin-dependent kinases CDK4 and CDK6 demonstrated differential expression during the DI SMG experiment in all ten volunteers. Sestrins are conserved stress-inducible anti-aging genes [25]. The expression of genes encoding sestrins was changed during DI SMG in all or almost all ten volunteers (Figure 3, Table 1).

In addition, p53 affects the insulin-like growth factor 1 (IGF1) signaling pathway, which is a key determinant of aging and longevity [17]. Reduced IGF1 signaling is associated with extended lifespan in the highly conserved pathway among different organisms from nematodes to mammals [26]. IGFBP-3 plays an important role in senescence as an aging marker. It binds to IGF1, which regulates growth, survival, and aging. IGFBP-3 up-regulates the PI3K/Akt/mTOR signaling pathway during cell aging [27]. Also, p53 was elevated in IGFBP3 gene KO cells when compared to normal cells [28]. The expression of both genes: IGF1 and IGFBP3 was changed during the DI SMG study in 5 out of 10 and in all 10 volunteers, respectively (Figure 3, Table 1).

A number of key genes and encoded proteins, previously shown to be linked to aging, longevity and cellular senescence, are associated with p53 network genes and corresponding proteins as indicated by STRING (Figure 4). Among them are important regulators of cell signaling pathways: proto-oncogenes Myc [61] and Jun [62]; members of the NF-kB network (NFKB1, NFKB2, and NFKBIA) [63,64]; hypoxia-inducible factor 1-alpha (HIF1A) [65]; matrix metalloproteinase-9 (MMP9) [66]; and forkhead box P1 (FOXP1) [67].

This study is subject to several limitations. Firstly, the cohort consisted of a small sample size (n = 10) of young, healthy male volunteers. Consequently, the findings may lack generalizability to female, older, or clinically compromised populations. Secondly, while the three-week DI-SMG protocol effectively models short-duration spaceflight, it may not fully recapitulate the physiological adaptations associated with longer mission or cumulative exposures. Finally, although microgravity was the variable of interest, the potential confounding influences of stress, physical inactivity, and confinement cannot be entirely discounted. To isolate the specific effects of microgravity simulation, future investigations should incorporate a parallel control group, such as subjects undergoing horizontal bed rest without SMG intervention. The transcriptomic changes identified here suggest a potential association with senescence. The direct measurement of senescence hallmarks, such as the secretion of SASP factors and the assessment of telomere attrition, will be critical for confirming the functional onset of senescence in future studies.

All participated volunteers provided informed consent for the use and sharing of their fully anonymized data, according to the Helsinki Code of Medical Ethics for human samples. The Biomedical Ethics Committee of the IBMP RAS and Section of Physiology of the Bioethics Committee of the UNESCO National Bioethics Commission approved this study (Meeting No. 483 took place on 3 August 2018).

An experiment with three weeks of exposure to DI-SMG without any countermeasures was performed at the Institute of Biomedical Problems (Figure 1A) during a period of 8 months from September 2018 to April 2019 with the participation of ten healthy men aged from 24 to 32 years as described [68].

In total, five samples of 10 mL peripheral blood were collected in sodium heparin tubes from each volunteer at five timepoints during the time course of the DI SMG study: at 7 days prior to Day 0 of DI (timepoint "−7 day"; background, BG), after 7 days, 14 days and 21 days in DI SMG, and on the 7th day after DI SMG (timepoint "28 days") (Figure 1B).

Peripheral blood mononuclear cells (PBMCs) were obtained using the Ficoll density gradient centrifugation isolation protocol optimized for Ficoll-Paque™ PLUS (Cytiva, Uppsala, Sweden). Human T cells were enriched by the positive CD3 cell selection kit (EasySep™ Human CD3 Positive Selection Kit II, Stem Cell Technology, Vancouver, British Columbia, Canada) with a purity of about 95%.

Total RNA was purified from sorted cells using the RNeasy kit and treated with the RNAse-free DNAse kit (##74104; 79254 QIAGEN, Hilden, Germany). The total RNA concentration was measured (NanoDrop2000, Thermo Scientific, Wilmington, DE, USA) and RNA integrity and purity were evaluated by gel electrophoresis analysis (1 × TAE 1% UltraPureAgarose, #16500500 Invitrogen, Waltham, MA, USA). Gel images were created with the ImageQuant LAS 4000 LAS4000 Image system using ImageQuant LAS 4000 Control Software (GE Healthcare, Freiburg, Germany) as described [69]. All selected isolated RNA samples were subjected to and passed both the internal lab quality control (QC) and QC performed by a sequencing company (Novogene, Hongkong).

Sequencing libraries were constructed from ribodepleted RNA using a stranded protocol. RNA sequencing and data quality control were performed with the Illumina HiSeq-PE150 Platform using the HiSeq 2500 Sequencing System at Novogene (Hongkong). Sequences were mapped to the reference genome with Tophat, v.2.0.12. Reads were aligned to the human reference genome assembly in December 2013 (GRCh38/hg38). The quality of the resulting RNA-seq data was assessed, with key metrics summarized in Supplementary Table S1. All bioinformatic software applications, their versions and statistical parameters are listed in Supplementary Table S2. The method, including software and statistical parameters, for differential gene expression analysis is detailed in Supplementary Table S3. In gene expression analysis by Novogene (Hongkong), H-cluster, K-means, and SOM were used to cluster log2 (ratios). Genes with an adjusted p-value (q-value) of <0.05 and an absolute log2 fold change of >1 were considered significantly differentially expressed. The gene expression heatmap was composed in NASQAR v1.0 (New York University (NYU) Center for Health Informatics and Bioinformatics, New York, NY, USA) using log2 normalized count matrix RNA-seq reads. The functional classification of refined data for peak-related genes was performed using NCBI Gene resource www.ncbi.nlm.nih.gov↗ (accessed on 22 June 2025) (Bethesda, MD, USA), as described previously [70,71,72].

A statistical analysis was performed to highlight the p53 gene expression trend over time course of the DI-SMG study. A linear trendline was calculated for each subject, and the resulting slope coefficients were analyzed using Excel (Microsoft 365 version (Microsoft Corporation, Redmond, WA, USA). The statistical inference of a significant upward trend was based on the statistical property that the 95% confidence interval around the mean slope was greater than zero.

A network of protein–protein interactions (PPIs) was constructed using STRING (v12.0, Swiss Institute of Bioinformatics (SIB), Lausanne, Switzerland; string-db.org accessed on 19 September 2023) with all active evidence channels enabled, including Experiments, Databases, Co-expression, Neighborhood, Gene Fusion, Co-occurrence, and Textmining, and interactions were filtered to include only those with a high confidence score of ≥0.700; subsequently, KEGG pathway enrichment analysis was performed, with statistical significance defined by a corrected p-value of <0.05 using the Bonferroni method. The PPI network was generated with the input gene set comprising molecular components annotated in the KEGG p53 signaling pathway (hsa04115). The combined confidence scores, which integrated evidence from multiple channels, were calculated by STRING using a default phylogenetically curated collection of genomes as the statistical background.