Introduction
The sestrin family comprises three highly conserved proteins: sestrin 1 (SESN1), sestrin 2 (SESN2), and sestrin 3 (SESN3). These proteins are stress-inducible, and play crucial roles in cellular responses to various stressors, such as oxidative stress, genotoxic stress, and nutrient deprivation [1]. Sestrins primarily function by regulating antioxidant defense mechanisms, maintaining metabolic homeostasis, and modulating cell survival pathways. Notably, they inhibit the mammalian target of rapamycin complex 1 (mTORC1) signaling pathway, a key regulator of cell growth and metabolism, thereby linking nutrient sensing to cellular stress responses [2,3]. Sestrin isoforms exhibit overlapping expression patterns and functions, but they can also display distinct expression profiles, regulatory mechanisms, and functions [4]. SESN1 is ubiquitously expressed in humans, including in skeletal muscle, heart, liver, and brain. SESN2 is highly expressed in various tissues, particularly in the kidney, lungs, liver, and brain. Similarly, SESN3 is prominently expressed in skeletal muscle, intestine, liver, and brain. Notably, in skeletal muscle, SESN1 expression levels are higher than those of SESN3 [5]. The p53 tumor suppressor protein induces the expression of both SESN1 and SESN2, whereas SESN3 expression is primarily regulated by FoxO transcription factors [4]. SESN1 and SESN3 are involved in the inhibition of oxidative stress, nutrient sensing, suppression of mTORC1, and induction of autophagy. Additionally, SESN3 plays a role in the regulation of mTORC2 and the maintenance of glucose homeostasis. Beyond the functions shared with SESN1, SESN2 is also involved in the inhibition of DNA damage and endoplasmic reticulum stress, as well as the maintenance of glucose and lipid homeostasis [4].
Recent studies have highlighted the role of SESN2 as a regulator of aging and a potential therapeutic target for age-related diseases [6]. In response to oxidative stress, SESN2 is upregulated, activating antioxidant defenses and protecting cells from oxidative damage, a major contributor to aging and age-related diseases. Additionally, SESN2 may influence aging by inhibiting mTORC1, as hyperactivation of mTORC1 is linked to reduced autophagy, increased cellular senescence, and aging [7,8].
Mesenchymal stem/stromal cells (MSCs) are multipotent non-hematopoietic cells found in various tissues, including bone marrow, adipose tissue, skin, placenta, and thymus [9]. Bone marrow-derived MSCs (BMSCs) can differentiate into osteoblasts, chondrocytes, and adipocytes. However, aging impairs BMSCs differentiation into osteoblasts, the cells responsible for bone formation, while enhancing adipogenic differentiation, leading to fat cell accumulation in the bone marrow. This imbalance contributes to the reduced bone density and increased fracture risk observed in older individuals [10,11]. Various studies have identified the molecular mechanisms underlying these changes. For example, an increase in miR- 188 and a decrease in miR-130a expression in aging BMSCs are associated with bone loss and fat accumulation in bone marrow [12,13]. In addition, compared to cells from young mice, BMSCs obtained from aged mice exhibit significantly reduced lipolysis and fatty acid oxidative metabolic activity [14]. The predominant adipogenic differentiation of skeletal stem/ progenitor cells has been suggested to be associated with inflamm-aging, which in turn further triggers stem cell dysfunction [15]. Adipokines and other hormones released from mature adipocytes in aged bone marrow may create a vicious cycle that drives the differentiation fate of neighboring BMSCs toward adipocytes rather than osteoblasts [16]. Despite these findings, the mechanisms driving age-related changes in BMSCs differentiation require further investigation.
Enhancement of AMP-activated protein kinase (AMPK) activity is known to promote osteogenesis and inhibit adipogenesis [17]. Given that SESN2 activates AMPK through mTORC1 inhibition [18], this study aims to investigate whether aging regulates SESN2 expression in BMSCs and determine SESN2’ s regulatory role in osteogenic and adipogenic differentiation of BMSCs.
Materials and Methods
1. Cell culture
Human MSCs were purchased from Sigma-Aldrich, Promo- Cell, and STEMCELL Technologies. The age distribution of the human MSCs sources included both young (Y; 18 and 22 years old) and aged (O; 70 to 82 years old) donors. Mouse BMSCs were extracted from 8-week-old C57BL/6 mice obtained from Korea Research Institute of Bioscience and Biotechnology. The bone marrow was extracted from the mouse femur and tibia and cultured in MesenCultTM Basal Medium (mouse) (STEMCELL Technologies) until passage 3. Subsequently, CD31/ CD45-negative cells were sorted using FACSAria Fusion (BD Biosciences). Sorted mouse BMSCs and human MSCs were cultured in alpha-minimal essential medium high glucose (Hy- Clone) supplemented with 10% fetal bovine serum (HyClone), 100 U/mL penicillin, and 100 μg/mL streptomycin.
Osteogenic differentiation was induced by supplementing the growth medium with 50 μg/mL ascorbic acid, 10 mM β-glycerophosphate, and 100 nM dexamethasone. The medium was changed every other day. At the end of the specified culture periods, cells were subjected to RNA isolation, alkaline phosphatase (ALP) activity assay, or staining for ALP or mineralized matrix (Alizarin Red S) as described previously [19].
When cell confluency reached 100%, adipogenic differentiation was induced by adding supplements (1 μM dexamethasone, 0.5 mM IBMX, 50 μM indomethacin, and 10 μg/mL insulin) to the growth medium. The medium was changed every other day, and at the end of the culture periods, cells were subjected to RNA isolation or Oil Red O staining as described previously [20].
2. Sestrin 2 knockdown using siRNA
Non-targeting siRNAs and siRNAs for human and mouse sestrin 2 genes were purchased from Santa Cruz Biotechnology (human: sc-106544; mouse: sc-153380). For mouse BMSCs, control siRNA and Sesn2 siRNA were transfected using Lipofectamine 2000 (Invitrogen). For human MSCs, siRNA transfection was performed using an electroporator with a gold tip at 990 mV for 40 ms.
3. Quantitative reverse transcription-polymerase chain reaction
Total RNA was extracted from cells using QIAzol Lysis Reagent (QIAGEN), and cDNA was synthesized using AccuPower ® RT PreMix (Bioneer).
The primer sequences for mouse genes were as follows (5′→3′): Gabph , TCAATGACAACTTTGTCAAGC and CCAGGGT T TCT TACTCCT TGG; Sesn2 , GGCGGTGGTGATGGGTCTAC and GACGACCCGGAAGTGGCCC; Alpl , CCAACTCTTTTGTGCCAG and GGCTACATTGGTGTTGAGCTTTT; Ibsp, CAGGGAGGCAGTGACTCTTC and AGTGTGGAAAGTGTGGCGTT; Bglap , CTGACAAAGCCTTCATGTCCAA and GCGCCGGAGTCTGTTCACTA; Cebpα , TGGTCCCCGTGTCCTCCTA and TCAGACCAGAAAGCTGAGTTGTG; Pparγ , CCGAAGAACCATCCGATTGAA and GCCCAAACCTGATGGCATT; Ap2 , AAAGAAGTGGGAGTGGGCTT and CTCTTGTGGAAGTCACGCCT; Plin1, TACCCTCCAGAAAAGATCGC and CTACCACCTTCTCGATGCTT; and Adipoq , TCTTGGTCCTAAGGGTGAGA and GTTGCAGTAGAACTTGCCAG.
The primer sequences for the human genes were as follows (5′→3′): GAPDH , CCATCTTCCAGGAGCGAGATC and GCCTTCTCCATGGTGGTGAA; SESN2 , CCTGCTCAGGAGTCAGGTCA and CAACCTCCTCTGGAGGCACTT; P16 , GGGTCGGGTAGAGGAGGTG and CATCATGACCTGGATCGGC; P21 , ACCGAGACACCACTGGAGGG and CGAGGCACAAGGGTACAAGACA; ALPL , AACTTCCAGACCGGCTTGA and TTGCCGCGTGTCTT; OSX , ACCTACCCATCTGACTTTGCT and CCACTATTTCCCACTGCCTT; RUNX2 , CAGATGGGACTGTGGCTGT and GTGAAGACGGTTATGAAGG; BGLAP , CGCTACCTGTATCAATGGCTGG and CTCCTGAAAGCCGATGTGGTCA; LPL, GTGACCAAGGTAGACCAGCC and GAAGAGACTTCAGGCAGCGT; CEBPα , TCGGTGGACAAGAACAG and GCAGGCGGTCATT; CEBPβ, CGCTTACCTCGGCTACCA and ACGAGGAGGACGTGGAGAG; PPARγ , TTCTCCTATTGACCCAGAAAGC and CTCCACTTTGATTGCACTTTGG; AP2, CATCAGTGTGAATGGGGATG and GTGGAAGTGACGCCTTTCAT; and ADIPOQ, TTCACCGATGTCTCCCTTAGG and GGCATGACCAGGAAACCAC.
4. Western blot analysis
Protein samples for western blotting were extracted using PRO-PREPTM (iNtRON Biotechnology). Samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene difluoride membrane. Blocking was performed using Tris-buffered saline containing 0.1% Tween 20 and 5% skim milk. The membranes were incubated with primary antibodies, followed by incubation with horseradish peroxidase (HRP)-conjugated secondary antibodies. Pico ECL reagent was used for detection by a Bio-Image Analyzer (Vilber Lourmat).
Antibodies for p16, p21, Lamin B, osteocalcin (OCN), osterix (OSX), RUNX2, SESN2, and β-actin were purchased from Santa Cruz Biotechnology. Antibodies for γH2AX, PPARγ, and adiponectin (ADIPOQ) were obtained from Cell Signaling Technology, and HRP-conjugated secondary antibodies were from Thermo Fisher Scientific.
5. Senescence-associated β-galactosidase staining
Cultured cells were stained using the Senescence β-Galactosidase Staining Kit (Cell Signaling Technology) according to the manufacturer’s instructions. Briefly, cells were washed with phosphate buffered saline (PBS) and fixed in β-gal staining fix solution for 15 minutes at room temperature. Cells were then washed with PBS and incubated with β-gal staining solution for 20–24 hours at 37℃. After washing with PBS, cells were observed under a bright-field microscope.
6. Statistical analysis
Quantitative data are represented as the mean ± standard deviation. Statistical significance was analyzed using Student’s t-test or one-way ANOVA with Tukey-Kramer post hoc test. A p-value less than 0.05 was considered significant.
Results
1. MSCs obtained from the elderly express higher levels of cellular senescence markers
We first examined the differences in cellular senescence marker levels between human MSCs obtained from young and elderly donors. Compared to cells from young donors, MSCs from elderly donors exhibited strong senescence-associated β-galactosidase (SA β-gal) staining (Fig. 1A). Additionally, MSCs from elderly donors expressed higher levels of cellular senescence markers such as p16, p21, and γH2AX (Fig. 1B and 1C, Supplementary Table 1B). These results suggest that BMSCs obtained from elderly individuals have a higher population of cells exhibiting cellular senescence compared to those from younger individuals.
2. MSCs from the elderly express lower levels of osteogenic marker genes and higher levels of adipogenic marker genes compared to those from the young
We next examined the basal expression levels of osteogenic and adipogenic differentiation marker genes in MSCs from young and elderly donors. As shown in Fig. 2A, MSCs cultured in growth medium expressed some of these marker proteins. Western blot results indicated that MSCs from elderly donors expressed lower levels of osteogenic marker proteins (RUNX2, OSX, and OCN) while expressing higher levels of adipogenic markers (PPARγ and ADIPOQ) compared to MSCs from young donors. Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) also demonstrated similar results, showing lower levels of osteogenic marker mRNAs (ALPL, RUNX2, and OSX) and higher levels of adipogenic marker mRNAs (PPARγ and AP2) in MSCs from elderly donors (Fig. 2B, Supplementary Table 2B). These results suggest that MSCs from elderly individuals have already acquired more adipogenic characteristics with a corresponding decrease in osteogenic traits, even in the absence of exogenous differentiation signals.
To further confirm that MSCs from elderly donors have higher potential for adipogenic differentiation but lower potential for osteogenic differentiation, the cells were cultured for 14 days under osteogenic (Fig. 2C, Supplementary Table 2C) or adipogenic conditions (Fig. 2D, Supplementary Table 2D). qRTPCR results demonstrated that under differentiation-inducing conditions, MSCs from elderly donors exhibited significantly enhanced expression levels of adipogenic marker genes and suppressed levels of osteogenic marker genes compared to MSCs from young donors. These results confirm that during aging, the differentiation potential of MSCs shifts from the osteogenic to the adipogenic lineage.
3. The expression levels of SESN2 are decreased in MSCs from the elderly
Bulk RNA sequencing (RNA-seq) was performed using MSCs from young and elderly donors. RNA-seq results showed that among the sestrin family members, the average FPKM levels of SESN2 were significantly lower in MSCs from elderly donors compared to those from young donors (Fig. 3A). To confirm the RNA-seq results, mRNA and protein samples were prepared from six batches of human MSCs obtained from 18-, 22-, 70-, 77-, and 82-year-old donors. Both qRT-PCR and western blot analyses (Fig. 3B and 3C, Supplementary Table 3B) demonstrated that MSCs from elderly donors expressed lower levels of SESN2 compared to MSCs from young donors. These results suggest that SESN2 expression in MSCs declines with aging.
4. Knockdown of SESN2 in young human MSCs suppresses osteogenic marker gene expression while enhancing the expression of adipogenic and senescence marker genes
To examine whether the decrease in SESN2 expression contributes to the phenotypic changes observed in MSCs from elderly donors, SESN2 knockdown was performed in young MSCs using siRNAs. Knockdown of SESN2 resulted in more than a three-fold induction of senescence marker gene (P16 and P21) expression (Fig. 4A, Supplementary Table 4A). Furthermore, SESN2 knockdown in young MSCs altered the basal expression levels of osteogenic and adipogenic marker genes in the absence of exogenous differentiation-inducing stimuli (Fig. 4B and 4C, Supplementary Table 4B and 4C). Similar to the phenotype of MSCs from elderly donors, SESN2 - knockdown young MSCs expressed lower levels of osteogenic marker gene mRNAs but higher levels of adipogenic marker gene mRNAs. These results suggest that the decline in SESN2 expression levels in MSCs during aging plays a role in shifting the differentiation fate of BMSCs from the osteogenic to the adipogenic lineage.
5. S esn2 knockdown inhibits osteogenic differentiation but enhances adipogenic differentiation of mouse BMSCs
To further confirm the regulatory effect of SESN2 loss on the osteogenic and adipogenic differentiation potential of BMSCs, similar experiments were performed using mouse BMSCs prepared from 8-week-old mice. Knockdown of Sesn2 in mouse BMSCs significantly decreased the mRNA levels of osteogenic marker genes, including Alpl, Ibsp, and Bglap (Fig. 5A, Supplementary Table 5A). Additionally, ALP activity and matrix mineral deposition were reduced in Sesn2-knockdown BMSCs (Fig. 5B and 5C, Supplementary Table 5B and 5C).
When mouse BMSCs were incubated under adipogenic conditions, Sesn2 knockdown significantly enhanced the expression of adipogenic marker mRNAs, including Cebpα, PPARγ, Adipoq, and Plin1 (Fig. 6A, Supplementary Table 6A). Furthermore, Oil Red O staining showed that intracellular lipid accumulation was increased by Sesn2 knockdown (Fig. 6B). These results suggest that SESN2 in BMSCs plays a role in regulating differentiation fate in favor of the osteogenic lineage.
Discussion
In this study, we demonstrated that SESN2 expression levels are decreased in MSCs from elderly donors compared to those from young donors. Furthermore, SESN2 knockdown in MSCs from young donors altered their cellular phenotypes to mimic those of MSCs from elderly donors, resulting in enhanced adipogenic potential and cellular senescence marker expression, but diminished osteogenic potential.
Recent studies indicate that senescent cells accumulate in the bone microenvironment with aging and that targeting these cells with senolytic and/or senomorphic drugs prevents aging-related bone loss, at least in mice [21]. Various cell types in the bone microenvironment, including osteoblast progenitors, osteoblasts, osteocytes, myeloid cells, and B and T cells, increase the expression of senescence-associated secretory phenotype factors with aging. Consistent with previous reports, we found that MSCs from elderly donors exhibited an increased number of SA β-gal-positive senescent cells and higher levels of p16, p21, and γH2AX expression [22]. Additionally, MSCs from elderly individuals showed higher basal levels of adipogenic genes and increased adipogenic marker gene expression under adipogenic conditions, while exhibiting lower levels of osteogenic marker gene expression both in the absence and presence of osteogenic stimuli. These results align with previous findings and the bone phenotypes observed in the elderly individuals [10,11].
In this study, bulk RNA-seq analysis of MSCs from young and elderly donors revealed a common decline in SESN2 mRNA levels in MSCs from elderly individuals. Sestrins are induced by various stressors and negatively regulate agingassociated biological processes. While no prior reports specifically noted a decline in SESN2 expression in MSCs from aged donors, some studies have linked reduced SESN2 levels to aging-associated musculoskeletal diseases. For instance, the number of SESN2-positive chondrocytes in articular cartilage declines with aging and osteoarthritis in both humans and mice [23]. Additionally, serum SESN2 levels are significantly reduced in sarcopenic and frail older subjects compared to non-sarcopenic and non-frail ones, respectively [24,25]. These findings, along with our results, suggest that a decline in SESN2 is associated with aging-induced loss of bone and muscle.
To explore the functional role of SESN2 in MSC phenotypic maintenance, we performed knockdown experiments. SESN2 knockdown in MSCs from young donors increased the expression levels of p16 and p21 mRNAs. Previous studies reported that mouse embryonic fibroblasts from SESN2-deficient mice exhibit cellular senescence phenotypes, including enhanced SA β-gal staining and decreased cell proliferation [26]. Additionally, SESN2 overexpression reduces SA β-gal activity and levels of p16 and p21 proteins in macrophages [27]. These findings, together with our results, suggest that maintaining SESN2 expression plays a role in preventing senescence of MSCs.
SESN2 knockdown also altered the differentiation potential of young MSCs, favoring the adipogenic lineage. Similarly, SESN2 knockdown significantly enhanced intracellular lipid accumulation and adipogenic marker gene expression in mouse BMSCs in this study. These results are consistent with reports showing that a loss of Sesn in Drosophila leads to triglyceride accumulation, reversible by AMPK activation or TOR inhibition [2].
Consistent with enhanced adipogenic differentiation, SESN2 knockdown inhibited osteogenic differentiation and matrix mineralization in mouse BMSCs. These results suggest that SESN2 plays a role in maintaining the osteogenic differentiation potential of MSCs in bone marrow. Although direct regulatory effects of SESN2 on MSC osteogenic differentiation have not been reported, SESN2 in aortic valve macrophages inhibits oxidized low-density lipoprotein-induced M1 polarization, indirectly suppressing the osteogenic differentiation of co-cultured valvular interstitial cells [28]. The inhibitory effect of SESN2 on osteogenic differentiation of valvular interstitial cells is mediated through its antioxidant and anti-inflammatory effects in macrophages, suggesting a potential indirect regulatory mechanism. Therefore, direct regulatory effect of SESN2 on osteogenic differentiation of valvular interstitial cells remains elusive.
In this study, we did not investigate the molecular mechanisms underlying SESN2 deficiency-induced phenotypic changes in MSCs. However, previous studies have shown that oxidative stress and mTORC1 activation favor adipogenic lineage, while AMPK activation favors osteogenic lineage in MSC differentiation [29-31]. Thus, it is speculated that SESN2 regulates MSC differentiation potential by inhibiting oxidative stress and mTORC1 activity and activating AMPK. Further studies are needed to elucidate these regulatory mechanisms in MSCs.
In conclusion, we demonstrated for the first time that a decline in SESN2 expression during aging contributes to the shift in MSC differentiation fate from osteogenic to adipogenic lineage. These results enhance our understanding of the pathophysiological mechanisms underlying aging-associated bone loss.
