Redox Biology 41 (2021) 101915

Contents lists available at ScienceDirect

Redox Biology
journal homepage: www.elsevier.com/locate/redox

Sirt3-mediated mitophagy regulates AGEs-induced BMSCs senescence and

senile osteoporosis
Yuanyuan Guo a, b, 1, Xiong Jia b, 1, Yongzhi Cui c, Yu Song c, Siyuan Wang d, Yongtao Geng c,
Rui Li c, Weihang Gao d, Dehao Fu c, *
a
Department of Pharmacy, Liyuan Hospital, Tongji Medical College, Huazhong University of Science and Technology, China
b
Department of Cardiovascular Diseases, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, China
c
Department of Orthopaedics, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, China
d
Department of Orthopaedics, Liyuan Hospital, Tongji Medical College, Huazhong University of Science and Technology, China

A R T I C L E I N F O A B S T R A C T

Keywords: Senile osteoporosis (SOP) is widely regarded as one of the typical aging-related diseases due to a decrease in bone
Mitophagy mass and the destruction in microarchitecture. The inhibition of mitophagy can promote bone marrow mesen­
Sirtuin3 chymal stem cells (BMSCs) senescence, and increasing studies have shown that interventions targeting BMSCs
Advanced glycation end products
senescence can ameliorate osteoporosis, exhibiting their potential for use as therapeutic strategies. Sirtuin-3
Senile osteoporosis
Cell senescence
(Sirt3) is an essential mitochondria metabolic regulatory enzyme that plays an important role in mitochon­
drial homeostasis, but its role in bone homeostasis remains largely unknown. This study seeks to investigate
whether advanced glycation end products (AGEs) accumulation aggravated BMSCs senescence and SOP, and
explored the mechanisms underlying these effects. We observed that AGEs significantly aggravated BMSCs
senescence, as well as promoted mitochondrial dysfunction and inhibited mitophagy in a concentration-
dependent manner. In addition, this effect could be further strengthened by Sirt3 silencing. Importantly, we
identified that the reduction of Sirt3 expression and the mitophagy were vital mechanisms in AGEs-induced
BMSCs senescence. Furthermore, overexpression of Sirt3 by intravenously injection with recombinant adeno-
associated virus 9 carrying Sirt3 plasmids (rAAV-Sirt3) significantly alleviated BMSCs senescence and the for­
mation of SOP in SAMP6. In conclusion, our data demonstrated that Sirt3 protects against AGEs-induced BMSCs
senescence and SOP. Targeting Sirt3 to improve mitophagy may represent a potential therapeutic strategy for
attenuating AGEs-associated SOP.

1. Introduction skeletal system, generally refers to the osteoporosis after the age of 70.
By 2050, the proportion of people aged 60 years or over in the total
Osteoporosis is characterized by a decrease in bone mass and the population will reach 22% worldwide [4]. With the increase of the aging
destruction in microarchitecture resulting in an increased propensity of population, SOP and related fractures not only distinctly augment the
fracture [1,2]. Aging is a multi-factorial process featured by a contin­ morbidity and mortality of elderly individuals, but also dramatically
uous loss of physiological function and biological mechanism, giving aggravate the financial burden of public health. However, the patho­
rise to enhanced vulnerability toward infections [3]. Senile osteoporosis genesis of SOP has not been fully clarified, and there is still a lack of
(SOP) is age-related bone loss and a specific biological ageing in the effective prevention and treatment measures.

Abbreviations: SOP, Senile osteoporosis; BMSCs, bone marrow mesenchymal stem cells; Sirts, Sirtuins; AGEs, advanced glycation end products; MMP, mito­
chondrial membrane potential; CCK-8, Cell Counting Kit-8; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; ROS, reactive oxygen species; rAAV, recombinant
adeno-associated viral; SA-β-gal, senescence-associated beta-galactosidase; CCCP, carbonyl cyanide m-chlorophenyl hydrazone; CsA, cyclosporin A; AZR, alizarin red
staining; ORO, oil red O; OC, osteocalcin; Col1, collagen type I; PPARγ, Peroxisome proliferation-activated receptor gamma; ALP, alkaline phosphatase; TRAP5b,
tartrate-resistant acid phosphatase 5b.
* Corresponding author. Department of Orthopaedics, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, No. 1277 Jiefang
Avenue, Wuhan, 430022, Hubei, China.
E-mail address: fudehao@hust.edu.cn (D. Fu).
1
These authors contributed equally to the study.

https://doi.org/10.1016/j.redox.2021.101915
Received 28 January 2021; Received in revised form 19 February 2021; Accepted 19 February 2021
Available online 24 February 2021
2213-2317/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Y. Guo et al. Redox Biology 41 (2021) 101915

Advanced glycation end products (AGEs), as the end products of 2. Materials and methods
glycation reactions, involved in the pathogenesis of aging-related dis­
eases. The formation of AGEs, typical resulting from Maillard reaction or 2.1. Experimental design and experimental protocol
non-enzymatic glycation process, has been known to be irreversible and
its accumulation could lead to the dysfunction of macromolecules such In vitro experiments. BMSCs were exposed to the same volume of
as nucleic acids, lipids and proteins [5]. Increasing evidence demon­ phosphate-buffered saline (PBS), bovine serum albumin (BSA, Abcam,
strated that AGEs, accumulated with advancing age, dramatically Cambridge, UK) (100 μg/mL), AGEs (50, 100 or 200 μg/mL buffered
decreased bone density and mineralization [6]. Further research using PBS; Abcam, Cambridge, UK).
confirmed that AGEs play a significant role in the impaired bone for­ Experiment related to mitophagy. The experiments were divided into
mation through triggering inflammation and bone loss in the patho­ six groups: BSA + vehicle, BSA + carbonyl cyanide m-chlorophenyl
genesis of osteoporosis [7]. hydrazone (CCCP), BSA + cyclosporin A (CsA), AGEs + vehicle, AGEs +
As one of highly conserved nicotinamide adenine dinucleotide CCCP (10μM), AGEs + CsA (5μM).
(NAD+)-dependent histone deacetylases, sirtuins (Sirts) regulate Osteogenic differentiation. To evaluate the osteogenic differentiation,
numerous physiological and pathological of processes, such as gene the experiments were divided into six groups: BSA + vehicle + Osteo­
expression, apoptosis, inflammation and healthy aging. There are seven genic induce supplements, BSA + CCCP + Osteogenic induce supple­
members in Sirts protein family, which located in different subcellular ments, BSA + CsA + Osteogenic induce supplements, AGEs + vehicle +
regions [8]. More recent studies have shown that targeting proteins Osteogenic induce supplements, AGEs + CCCP (10μM) + Osteogenic
belonging to the sirtuin family can significantly delay the occurrence induce supplements, AGEs + CsA (5μM) + Osteogenic induce
and development of ageing process and many age-related chronic dis­ supplements.
eases [9]. Sirt3, a human Sir2 homologue, translocate to the mito­
chondria under cellular stress. It is known to serve an important role in 2.2. Animals
mitochondria, mediating mitochondrial dynamics and other key bio­
logical processes of mitochondrial quality control [10]. Brown et al. Adult ICR mice were purchased from the Beijing Vital River Labo­
reported that the inhibition of Sirt3-mediated mitochondrial homeo­ ratory Animal Technology Co., China. The senescence-accelerated
stasis involved in enhanced oxidative stress in aged stem cells (SCs). The mouse strain P6 (SAMP6), as a clinically relevant model of SOP, is
study proved that Sirt3 could reverse the aging-associated degeneration characterized by rapid aging and shortened lifespan. We adapted the
when compared the quantity and quality of SCs in wild-type and Sirt3 SAMP1 mice as control mice. SAMP1 were also purchased from Beijing
knockout mice [11]. From our previous studies, the accumulation of Vital River Laboratory Animal Technology Co., China. All mice were
AGEs in nucleus pulposus tissues contributed largely to intervertebral raised in specific pathogen-free (SPF) condition and maintained at 25 ±
disc degeneration, induced an oxidative microenvironment and mito­ 1 ◦ C and 60% ± 10% humidity under a 12-h light/dark cycle during the
chondrial dysfunction [12]. Nevertheless, very little is known about the experiments. All animals were treated according to the regulations of
potential role of Sirt3 in the regulation of mitochondrial function in Chinese law and the local Ethical Committee.
osteoporosis.
Mitophagy, the selective degradation of mitochondria by autophagy, 2.3. Isolation and culture of bone marrow mesenchymal stem (BMSCs)
maintains cell homeostasis and ensure mitochondrial homeostasis via
specifically degrading damaged mitochondria [13]. There is growing BMSCs were harvested from 2-month-old ICR mice and the 6-month-
evidence that mitophagy is reportedly closely related with aging and old SAMP6 mice according to the detailed procedures in previously
age-related diseases, which impairs mitochondrial quality and function study [21]. Generally, bone marrow was isolated aseptically by flushing
[14]. Mitophagy is a normal physiological activity that occurs under the femurs and tibias of mice and then suspended in DMEM low glucose
healthy conditions, whereas it can also be altered under pathological (Hyclone) supplemented with 10% FBS (Gibico) in a humidified atmo­
conditions, or specific physiological conditions to promote the occur­ sphere incubator (Thermo) containing 95% air and 5% CO2 at 37∘C.
rence of aging-related diseases [15]. Recent evidence indicated that Non-adherent cells were removed and medium was changed after 3
abnormal mitophagy play a key role in bone metabolism disorders, the days. Medium was changed every 2 days and cells were passaged to
maintenance and differentiation of stem cell [16,17]. employ in the further experiments.
The senescence of bone marrow mesenchymal stem cells (BMSCs)
plays an important role in the occurrence and development of osteo­
2.4. Cell viability assay
porosis. The “osteogenesis-adipogenesis” differentiation balance of bone
marrow mesenchymal stem cells is essential for the maintenance of
The BMSCs were plated in 96-well plates (1 × 104cells/mL, 100μL
healthy bone homeostasis [18]. Some studies have reported that the
per well) and grown overnight. The cells were treated according to the
intrinsic properties of BMSCs such as senescence, osteogenic and adi­
above-mentioned experimental groups. 20 μL of CCK-8 solution (Beyo­
pogenic differentiation potential were markedly changed during aging
Time Biotechnology, C0038, Shanghai, China) was added to each well at
process [19,20]. Therefore, it is an important research focus to enhance
the 24 h, 48 h and 72h and then incubated the plates for 4 h in the
the osteogenic differentiation of BMSCs and improve their osteogenic
incubator. The OD value were determined at 450 nm by a microplate
efficiency through anti-aging of BMSCs.
reader (Thermo, MK3, USA) and the cell viability was assessed by the
Based on the preliminary work, Sirt3 are involved in regulating
following equation.
mitochondrial homeostasis and mitophagy, we hypothesized that AGEs
might affect mitochondrial homeostasis by interfering with Sirt3 and experimental OD value
cell viability = × 100%
tried to assess the association between mitophagy and AGEs. Thus, we control OD value
investigated the potential role of AGEs and Sirt3 on SOP in vitro and in
vivo, the involved downstream signaling pathways were also investi­ 2.5. Cell cycle analysis
gated. We used a well-established animal model of SAMP6 and over­
expressed Sirt3 by a recombinant adeno-associated viral (rAAV) vector The BMSCs were plated in 6-well plates (1 × 105 cells/mL, 2 mL per
transfection, exploring the role of Sirt3 on AGEs-induced cell senescence well). Treated according to the above-mentioned experimental groups
and SOP. Our investigations shed a new light on mechanism of the and cultured for 24 h, 48 h, 72 h or 96 h (at 37 ◦ C, 5% CO2). After 24 h,
pathogenesis and targeting Sirt3 offer a novel therapeutic intervention 48 h and 72 h, the cells were collected for cell cycle assay using the Cell
strategy against SOP. Cycle and Apoptosis Analysis Kit (BeyoTime Biotechnology, C1052,

2
Y. Guo et al. Redox Biology 41 (2021) 101915

Table 1 BMSCs using TRIzol reagent (Invitrogen) and used for the synthesis of
The primers used in this study. first-strand cDNAs by Revert Aid first-strand cDNA synthesis kit
Description of prime Sequence of prime (Thermo Scientific). cDNA was subjected to Quantitative PCR analysis.
Quantitative PCR: qPCR was performed using an ABI PRISM 7700
Nanog (mouse)-RT-F CCCTGATTCTTCTACCAGTCCCA
Nanog (mouse)-RT-R CACAGTCCGCATCTTCTGCTTCC sequence detection system and SYBR Green qPCR kit (TOYOBO, Osaka,
Oct-4 (mouse)-RT-F CAGAAGGAGCTAGAACAGTTTGCC Japan) (See Table 1). Expression levels were normalized to GAPDH.
Oct-4 (mouse)-RT-R CGCCTACATTAAGAGCCGTGAGAT
GAPDH (mouse)-RT-F GGTGAAGGTCGGTGTGAACG
GAPDH (mouse)-RT-R CTCGCTCCTGGAAGATGGTG 2.10. Measurement of adenosine triphosphate (ATP) content
Sirt3 (mouse)-RT-F GCCCAATGTCACTCACTACTTCCTG
Sirt3 (mouse)-RT-R TCCCAGATGCTCTCTCAAGCCCGTC Briefly, the BMSCs (1 × 105/well) were cultured in 6-well culture
plates and incubated with different treatments. The cells were lysed to
Shanghai, China) according to the manufacturer’s protocol. Cytometric determine total cellular ATP using an ATP detection kit (Solarbio,
analysis was carried out by means of a flow cytometer (BD-FACSVerse, BC0300, Beijing, China). On the basis of manufacturer’s instructions,
BD Bioscience, USA). ATP was extracted from the BMSCs and measured using a UV spectro­
photometer (Beckman Coulter DU720, California, USA).
2.6. The quantification of β-galactosidase activity
2.11. Analysis of the contents of ROS
After treated with 24 h, 48 h and 72 h, the BMSCs were collected for
the quantification of β-galactosidase activity and then fixed with 4% The BMSCs were divide into various groups as mentioned above and
formaldehyde (Rich joint, Shanghai). The senescence-associated incubated with different treatments for 72h to observe the antioxidant
β-galactosidase staining assay was carried out using a Senescence properties. We selected 2′ , 7′ -Dichlorodihydrofluorescein diacetates
β-Galactosidase Staining Kit (BeyoTime Biotechnology, C0602, (DCFH-DA) (Beyotime Biotechnology, S0033 M, Shanghai, China) as a
Shanghai, China) based on a protocol published by manufacturer. useful mitochondrial superoxide indicator of ROS. DCFH-DA was dis­
solved in serum-free medium (1:1000 dilution) and diluted to a final
2.7. Immunofluorescence staining assay concentration of 10 μmol/L. The cells resuspended in diluted DCFH-DA
and kept at a concentration of 100 × 105 to 2000 × 105 cells/mL in the
The BMSCs were treated with different group as mentioned above, cell incubator for 72 h. Mixed by inversion every 5 min to ensure full
then the cells were washed with ice-cold PBS and fixed with 4% para­ contact of the probe with the cells. We also adapted Mito-SOX Red
formaldehyde for 15–30 min. Subsequently, the cells were treated with (Invitrogen, M36008, USA) to determine mitochondrial superoxide and
0.1% Triton (Rich joint, Shanghai) for 15 min to permeate cell mem­ Mito-tracker Green (Beyotime Biotechnology, C1048, Shanghai, China)
branes. After washing twice with PBS, the cells were blocked for 15 min as a mitochondrial-selective fluorescent label. 5 μmol/L Mito-SOX and
with PBS containing 5% FBS overnight at 4 ◦ C. The cells were incubated 100 nmol/L Mito-tracker Green was added to each group and incubated
overnight at 4◦ C with mouse monoclonal anti-γ-H2AX Ser139 (Abcam, for 30 min in the dark (37◦ C and 5% CO2). Washed the cells with serum-
ab176916, diluted 1:100) or anti-H3K9me3 (Abcam, ab176916), and free medium three times. DAPI was used for nuclear localization and the
then incubated with Cy3-coupled secondary antibody (Proteintech, results were observed by a microscope(OLYMPUS, IX71, Japan).
SA00009-2, Beijing, China). The final step was to stain nuclei with
Hoechst 33258 (Beyotime Biotechnology, C1052, Shanghai, China) 15 2.12. Detection of mitochondrial membrane potential
min at room temperature without exposure to light. The images ware
observed through fluorescence microscope (IX71, Olympus Corporation, The lipophilic cationic probe 5, 5′ , 6, 6′ -tetrachloro-1, 1′ , 3, 3′ -tet­
Tokyo, Japan). raethyl-imida-carbocyanine iodide (JC-1) was applied to detect the
mitochondrial membrane potential. After experimentation, 2 μM of JC-1
2.8. Western blotting assay (Beyotime Biotechnology, S0033 M, Shanghai, China) was added to the
cells and the cells were incubated for 30 min at 37◦ C in the dark. After
The BMSCs were lysed in pre-cooled RIPA buffer containing protease washing the cells three times with PBS, the nuclei were stained with
inhibitor cocktail. Proteins were separated through 10% sodium dodecyl Hochest 33258. The results were observed by fluorescence microscope.
sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (SINOPHARM,
30166428, Shanghai), and then transferred to PVDF membranes (Mil­
lipore, HATF00010, Bradford, MA, USA). After blocking with 4% non-fat 2.13. Analysis of mitophagy
milk, the membranes were incubated with the primary antibodies P16
(1:1000; Catalog No.AF5484; Affinity Biosciences), P21 (1:1000; Cata­ Mitophagy in BMSCs was analyzed using the Mitophagy detection kit
log No.AF6290; Affinity Biosciences), P53 (1:1000; Catalog No.AF0879; (Dojindo, MD0, Kumamoto, Japan). 100 nmol/L Mtphagy Dye (con­
Affinity Biosciences), LC3B (1:1000; Catalog No. AF4650; Affinity Bio­ taining 100 nmol/L Mito-tracker Green Probe) was added to each group
sciences), P62 (1:1000; Catalog No.AF5384; Affinity Biosciences), Par­ and the cells were incubated for 30 min. After experimentation, the cells
kin (1:1000; Catalog No.AF0235; Affinity Biosciences), Sirt3 (1:1000; were washed twice with Hank’s. The results were observed by fluores­
Catalog No.AF5135; Affinity Biosciences), GAPDH (1:1000; Catalog No. cence microscope.
AF7021; Affinity Biosciences) overnight at 4 ◦ C and then incubated for 1
h at 37 ◦ C with secondary antibodies (1:10,000; Beyotime Biotech­ 2.14. Osteogenic differentiation
nology; cat. no. A0208). The membranes were treated with ECL (Ding­
guo Changsheng Biotechnology, Beijing, China) and developed with BMSCs were plated in 24-well plates (1 × 105 cells/ml, 500 μL per
detection system, then exposed onto films (ChemiScope 5300, CLINX, well). When the cells grew to 90%, the medium replaced with the
Shanghai, China). osteogenic differentiation medium, which contains 5 μg/ml insulin, 100
nM dexamethasone, 10 mM sodium β-glycerophosphate and 0.2 mM
2.9. RNA isolation and qRT-PCR assays vitamin C in DMEM with 10% FBS. Treated in group as described above.
Medicated medium was changed every 3 days and differentiation was
RNA extraction and cDNA synthesis: Total RNA was extracted from induced up to 21 days.

3
Y. Guo et al. Redox Biology 41 (2021) 101915

Fig. 1. Effects of AGEs in different concentrations

on the senescence of BMSCs. The BMSCs were
treated with AGEs (50–200 μg/mL) or BSA for
24–72 h. (A) SA-β-gal assay for detection of BMSCs
senescence. Scale bar: 100 μm. (B) Detection of
H3K9me3 by immunofluorescence in BMSCs. Scale
bar: 100 μm. (C) Detection of γ-H2AX by immuno­
fluorescence in BMSCs. Scale bar: 100 μm. (D–I)
Representative Western blotting assay and quanti­
tation of the level of P16, P21, P53. **p < 0.01
versus BSA.

2.15. Analysis of alkaline phosphatase (ALP) activity manufacturer’s instructions, we selected p-nitrophenylphosphate as the
substrate to measure the ALP activity.
After 21 days of induction, differentiated samples were collected to
test ALP activity. ALP activity was assayed according to an alkaline
phosphatase assay kit (Solarbio, BC2145, Beijing, China). On basis of the

4
Y. Guo et al. Redox Biology 41 (2021) 101915

Fig. 2. Effects of different concentrations of AGEs on mitochondrial function and mitophagy of BMSCs. The BMSCs were treated with AGEs (50–200 μg/mL) or BSA
for 24–72 h. (A) Representative fluorescence images with DCF (green) staining in BMSCs stimulated with AGEs. Scale bar: 50 μm. (B) Representative fluorescence
images with Mito-SOX (red) and Mito-Tracker (green) double-staining in BMSCs stimulated with AGEs. Scale bar: 50 μm. (C) The MMP was detected through JC-1
staining in BMSCs stimulated with AGEs. Scale bar: 50 μm. (D) Representative fluorescence images with Mtphagy Dye (red) and Mito-Tracker (green) double-staining
in BMSCs stimulated with AGEs. Scale bar: 50 μm. (E) Representative fluorescence images with LC3B (red) and Mito-Tracker (green) double-staining in BMSCs
stimulated with AGEs. Scale bar: 50 μm. (F) Representative Western blotting assay and quantitation of the level of LC3B, P62, Parkin, Sirt3. **p < 0.01 versus BSA.
(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

2.16. Calcification staining siRNA (30 nM) or Sirt3 siRNA (30 nM) (Guangzhou RiboBio, China) for
48 h using Hiperfect Transfection Reagent (Qiagen; cat. no. 301705)
Calcium depositions were stained with alizarin red staining. After 21 according to the manufacturer’s instructions and immediately stimu­
days of differentiation, the cells were fixed with 70% ethanol and then lated with PBS or AGEs.
added 300 μL 1% alizarin red (pH = 4.2) to each well. The plates were
incubated at room temperature for 30 min and deionized water was
2.19. Experiments in SAMP6 mice
added to stop the reaction. The results were observed by microscope.

All studies involving animals are reported in accordance with the

2.17. Oil Red-O staining (ORO) ARRIVE guidelines for reporting experiments involving animals [22]. A
total of 40 animals described were used in the experiments. The SAMP6
After 15 days of differentiation, BMSCs were fixed with 10% neutral mice (8 weeks) were fed a standard diet for 12 weeks and then were
formaldehyde and stained with ORO for 1 h in the dark. The images were randomly assigned to 2 groups, adeno-associated virus-9 (AAV9)-EGFP
observed by microscope. (enhanced green fluorescent protein) control group (n = 20) and
AAV9-EGFP-Sirt3 group (n = 20). AAV9 vectors were recombined with
2.18. Small interfering RNA (siRNA) transfection EGFP, or Sirt3 coding sequence [23]. All virus vectors were designed and
constructed by Vigene Biosciences (Rockville, MD). Each mouse was
To silence Sirt3 expression, BMSCs were transfected with scrambled received a tail vein injection (i.v.) of rAAV, 5 x 1011 vg/mouse, once.

5
Y. Guo et al. Redox Biology 41 (2021) 101915

Four weeks after AAV9-EGFP-control and AAV9-EGFP-Sirt3 injection, results showed that the expressions of P16, P21 and P53 were
SAMP6 were randomly divided into BSA group (50 μg/mouse per 2 strengthened with rising concentrations of AGEs and longer treatment
weeks, i.v., n = 10) and AGEs group (50 μg/mouse per 2 weeks, i.v., n = durations (Fig. 1D–I). Furthermore, we further studied the effects of
10). After treatment BSA or AGEs for 12 weeks, the samples were AGEs on the proliferation and cell stemness of BMSCs. Different con­
collected after euthanasia. centrations of AGEs inhibited the proliferation of BMSCs (Fig. S1A).
Oct4 and Nanog are transcription factors all essential to maintaining the
2.20. Bone micro-CT image stem cell phenotype. Different concentrations of AGEs reduced the
mRNA levels of Oct4 and Nanog in a dose-dependent manner, indicating
The trabecular microstructure and a 3D image of left femur was that AGEs damaged the stemness of stem cells (Figs. S1B and C). Those
analyzed using the Locus SP Micro-CT System (Skyscan, N.V., Belgium) results indicated that AGEs accelerated cell senescence in BMSCs.
provided by the servicebio technology CO., LTD. Samples were set in the
sample holder of the scanner and scans were made along the longitu­ 3.2. AGEs inhibited mitochondria function and mitophagy
dinal axis of the specimen. Micro-CT images were captured consecu­
tively. The image size was set at 1024 × 1024 pixels with a pixel size of Senescent cells are characterized by dramatic changes in mitochon­
25 μm × 25 μm and a distance of 25 μm between sections. Scans were drial function, metabolism and homeostasis [28]. Accumulating evi­
made using the scanning parameters: 50 kV, 400 μA, and a 40 min scan dence has demonstrated that mitochondrial dysfunction and mitophagy
period for each specimen. For better analysis the image and quantitative contribute to senescence [29]. As mentioned above, AGEs promote cell
calculations, the images were reconstructed and processed at a spatial senescence, which is closely related to mitochondrial function and
resolution of 10 μm, using circular scanning. mitophagy. Mitochondrial function is prevailingly affected by intracel­
lular ROS and mitochondrial membrane potential [30]. In our study, we
2.21. Enzyme-linked immunosorbent assay (ELISA) firstly analyzed the effect of AGEs on mitochondrial function. To mea­
sure mitochondrial function, we examined the total intracellular ROS
The bone-specific alkaline phosphatase (BALP) and tartrate-resistant levels by DCF-DA assay (Fig. 2A). Given that ROS derive predominantly
acid phosphatase 5b (TRAP-5b) levels in the serum were measured using from mitochondria, we further used Mito-SOX assays to detect mito­
ELISA kits (Elabscience). The test was performed by using the manu­ chondrial ROS (Fig. 2B). ROS staining showed that the DCF-DA (green)
facturer’s protocol. and Mito-SOX (red) fluorescent level gradually augmented with the in­
crease in AGEs concentration in BMSCs. In other words, with an increase
2.22. Ex vivo fluorescence of SAMP6 of the AGEs concentration, the production of mitochondrial ROS was
gradually enhanced. Mitochondrial membrane potential is another
After 4 weeks of infection with the adeno-associated virus with the critical indicator of mitochondrial function, and decreased mitochon­
Sirt3 plasmid, we performed live imaging under the condition of anes­ drial membrane potential is closely associated with elevated mito­
thesia with isoflurane to observe the expression of GFP fluorescence. The chondrial ROS production [31]. In this experiment, ATP levels were
fluorescence was measured by an IVIS Imaging System 200 (Caliper Life measured by ATP determination kit. AGE treatment
Sciences) using γex = 550 nm an dγem = 570 nm. The images were concentration-dependently reduced the ATP content of BMSCs (Fig. S2).
quantified for fluorescent radiant efficiency [fluorescence emission In addition, mitochondrial membrane potential (MMP) was detected by
radiance per incident excitation intensity: (p/s/cm2/sr)/(lW/cm2)] JC-1staining, a fluorescent dye, which shows red fluorescence under
using region-of-interest (ROI) function of the Living Image 4.3.1. aggregation conditions in normal mitochondria, but shows green fluo­
software. rescence when the mitochondrial membrane potential decreases. The
green fluorescence was increased and the red fluorescence was
2.23. Statistical analysis decreased with increasing AGEs concentrations (Fig. 2C). The shift from
red to green fluorescence is thus an indicator for the decrease in the
Differences between different groups were performed using Stu­ mitochondrial membrane potential.
dent’s t-test (SPSS Statistical Software). P values of <0.05 were Mitochondrial dysfunction plays an important role in the generation
considered significant. of ROS, and dysfunctional mitochondria are removed by the process
known as mitophagy [32]. The fluorescence of Mtphagy dye-stained
3. Results mitochondria, colocalized with Mito-tracker green-labeled mitochon­
drial, was observed using fluorescence microscope. The fluorescence
3.1. AGEs accelerated senescence of BMSCs intensity of Mtphagy Dye and Mito-tracker green decreased in a
dose-dependent manner after the BMSCs were treated with different
AGEs have been shown to be critical mediators both in the patho­ concentrations of AGEs (Fig. 2D). LC3B is a well-established marker for
genesis of osteoporosis and other chronic degenerative diseases related autophagy [33]. In order to investigate the effects of AGEs on mitoph­
to ageing [24]. In this study, we first detected the effect of different agy, we further performed double-immunofluorescence staining for
concentrations of AGEs on the senescence of BMSCs. It is well known Mito-tracker-green and the autophagy-associated protein LC3B. The
that β-galactosidase (SA-β-gal) is a metabolic marker of senescence, as it LC3B fluorescence and mitochondrial content decreased with increasing
is a key lysosomal enzyme which accumulates in lysosomes and auto­ AGEs concentration (Fig. 2E). Moreover, we detected the expression of
phagosomes in senescent cells [25]. We performed a autophagy-related protein LC3B, P62 and mitophagy-related protein
senescence-associated SA-β-gal assay in order to investigate cellular Parkin and Sirt3. As shown in Fig. 2F, the autophagy markers LC3 and
senescence. As shown in Fig. 1A, when the BMSCs were treated with 50 mitophagy-related protein Parkin and Sirt3 decreased, whereas auto­
μg/mL, 100 μg/mL, 200 μg/mL AGEs for 24 h, 48 h and 72 h, AGEs phagic substrate P62 increased with the increase in AGEs concentration.
significantly accelerated senescence of BMSCs in Taken together, the above results showed that AGEs inhibited mito­
concentration-dependent manner. In addition, as shown in Fig. 1B and chondria function and mitophagy.
C, the expression of H3K9me3 and γ-H2AX, which are signs of aging
damage, also enhanced with the increase of the concentration of AGEs. 3.3. The senescence of BMSCs was negatively correlated to mitophagy
Aging-related genes, such as P16, P21 and P53, are well-established
senescence markers. The overexpression of P16, P21 and P53 have Altered mitochondrial dynamics with decreased autophagy or
previously shown to cause premature cell senescence [26,27]. Our mitophagy are hallmarks to cellular senescence [34]. The

6
Y. Guo et al. Redox Biology 41 (2021) 101915

Fig. 3. The effects of CCCP and CsA on mitochondrial function and mitophagy of BMSCs. BMSCs was treated with AGEs (200 μg/mL) for 72 h in the presence or
absence of CCCP and CsA. (A) ROS production was detected with the fluorescent dye DCF (green). Scale bar: 50 μm. (B) Representative fluorescence images with
Mito-SOX (red) and MitoTracker (green) double-staining in BMSCs stimulated with AGEs in the presence or absence of CCCP (10 μM) or CsA (5 μM). Scale bar: 50 μm.
(C) The MMP was detected through JC-1 staining in BMSCs stimulated with AGEs in the presence or absence of CCCP (10 μM) or CsA (5 μM). Scale bar: 50 μm. (D)
Representative fluorescence images with Mtphagy Dye (red) and Mito-Tracker (green) double-staining in BMSCs stimulated with AGEs in the presence or absence of
CCCP (10 μM) or CsA (5 μM). Scale bar: 50 μm. (E) Representative fluorescence images with LC3B (red) and Mito-Tracker (green) double-staining in BMSCs
stimulated with AGEs in the presence or absence of CCCP (10 μM) or CsA (5 μM). Scale bar: 50 μm. (F) Representative Western blotting assay and quantitation of the
level of LC3B, P62, Parkin, Sirt3. *p < 0.05, **p < 0.01, versus BSA, ##p < 0.01. (For interpretation of the references to colour in this figure legend, the reader is
referred to the Web version of this article.)

above-mentioned data suggested that AGEs-induced BMSCs senescence showed that CCCP improved the reduction of MMP and ATP content
is closely related to mitophagy. Therefore, we ulteriorly explored the induced by AGEs, while CsA reduced the reduction of MMP and ATP
intrinsic connections within cellular senescence and mitophagy. content induced by AGEs (Fig. 3C and Fig. S3A). Changes in mito­
Carbonyl cyanide m-chlorophenyl hydrazone (CCCP) is a potent mito­ chondrial function are capable of launching mitophagy. After supple­
chondrial membrane uncoupler to promote mitophagy [35] and cyclo­ ment with CCCP and CsA separately, we used the same method to assess
sporin A (CsA) is an immunosuppressant agent to inhibit mitophagy mitophagy in BMSCs. We observed the fluorescence of Mitophagy
[36]. From above, we found that AGEs promoted the production of dye-stained mitochondria, colocalized with Mito-tracker green-labeled
cellular ROS, reduced the mitochondrial membrane potential, inhibited mitochondrial, and the autophagy-associated protein LC3B by fluores­
mitophagy, and promoted BMSCs senescence. Here we changed cence microscope. The results of the study found that CCCP enhanced
mitophagy by adding CCCP and CsA to explore the role of mitophagy in the fluorescence intensity of Mtphagy Dye and LC3B, while CsA further
BMSCs senescence. On the basis of AGEs, we added CCCP and CsA to reduced the fluorescence intensity of Mtphagy Dye and LC3B induced by
BMSCs and then analyzed the mitochondrial function, mitophagy and AGEs (Fig. 3D and E). Moreover, CCCP reversed the decreased of LC3B,
senescence of BMSCs. We assessed the intracellular levels of ROS using Parkin and Sirt3 expression and the increased of P62 expression induced
the DCF-DA assay and detected mitochondrial ROS levels by Mito-SOX by AGEs, while CsA had the opposite effect (Fig. 3F).
Red immunofluorescence staining in BMSCs. The results showed that We examined the effects of changes in mitophagy caused by CCCP
CCCP inhibited the production of intracellular and mitochondrial ROS and CsA on the senescence of BMSCs. Consistent with our previous
induced by AGEs, while CsA increased the production of intracellular findings, AGEs induced BMSCs senescence by SA-β-gal staining
and mitochondrial ROS induced by AGEs (Fig. 3A and B). We further (Fig. 4A). CCCP activated mitophagy and alleviated AGEs-induced
tested the effects of CCCP and CsA on MMP and ATP content. The results BMSCs senescence, while CsA inhibited mitophagy and further

7
Y. Guo et al. Redox Biology 41 (2021) 101915

Fig. 4. The effects of CCCP and CsA on the senescence of BMSCs. BMSCs was treated AGEs (200 μg/mL) for 72 h in the presence or absence of CCCP (10 μM) and CsA
(5 μM). (A) SA-β-gal assay for detection of BMSCs senescence. Scale bar: 100 μm. (B) Detection of H3K9me3 by immunofluorescence in BMSCs. Scale bar: 100 μm. (C)
Detection of γ-H2AX by immunofluorescence in BMSCs. Scale bar: 100 μm. (D) Representative Western blotting assay and quantitation of the level of P16, P21, P53.
*p < 0.05, **p < 0.01 versus BSA, #p < 0.05, ##p < 0.01.

8
Y. Guo et al. Redox Biology 41 (2021) 101915

Fig. 5. The effects of CCCP and CsA on osteogenic differentiation and adipogenic differentiation of BMSCs. BMSCs was treated AGEs (200 μg/mL) for 72 h in the
presence or absence of CCCP (10 μM) and CsA (5 μM). (A–B) Osteogenic and adipogenic differentiation were tested by alizarin red staining and ORO staining,
respectively. Scale bar: 100 μm. (C) ALP activity determined by ALP activity assay. *p < 0.05, **p < 0.01 versus BSA, ##p < 0.01. (D) Detection of osteogenesis
marker gene Osterix by RT-PCR. *p < 0.05, **p < 0.01 versus BSA, ##p < 0.01. (E) Detection of osteogenesis marker gene Runx2 by RT-PCR. *p < 0.05, **p < 0.01
versus BSA, #p < 0.05, ##p < 0.01. (F) Detection of adipogenic differentiation-related genes C/EBPα by RT-PCR. *p < 0.05, **p < 0.01 versus BSA, #p < 0.05. (G)
The protein levels of OC detected by Western blot analysis. **p < 0.01 versus BSA, ##p < 0.01. (H) The protein levels of Col1 detected by Western blot analysis. *p <
0.05, **p < 0.01 versus BSA, #p < 0.05, ##p < 0.01. (I) Detection of adipogenic differentiation-related genes PPARγ by RT-PCR. *p < 0.05, **p < 0.01 versus BSA,
#
p < 0.05. (J) Detection of adipogenic differentiation-related genes Twist1 by RT-PCR. **p < 0.01 versus BSA, ##p < 0.01. (For interpretation of the references to
colour in this figure legend, the reader is referred to the Web version of this article.)

promoted AGEs-induced BMSCs senescence (Fig. 3D–F and 4A). Simi­ expression of P16, P21 and P53 (Fig. 4D).
larly, the activation of mitophagy by CCCP reduced the fluorescence These results indicated that CCCP reduced BMSCs senescence by
intensity of H3K9me3 and γ-H2AX induced by AGEs, while the inhibi­ reversing AGEs-induced mitophagy inhibition, while CsA increased
tion of mitophagy by CsA increased the fluorescence intensity of BMSCs senescence by enhancing AGEs-induced mitophagy inhibition. In
H3K9me3 and γ-H2AX induced by AGEs (Fig. 4B and C). In addition, the summary, the senescence of BMSCs was negatively correlated to
activation of mitophagy alleviated the inhibitory effect of AGEs on cell mitophagy, the senescence of BMSCs could be suppressed by the acti­
proliferation and increased the mRNA expression of stem-related genes vation of mitophagy and aggravated by the inhibition of mitophagy.
Oct4 and Nanog, while the inhibition of mitophagy aggravated the
inhibitory effect of AGEs on cell proliferation and reduced the expres­ 3.4. Activation of mitophagy significantly inhibited adipogenic
sion of stem-related genes Oct4 and Nanog (Figs. S3B–D). Next, the differentiation but facilitated osteogenic differentiation in BMSCs
expression of aging-related proteins P16, P21, and P53 were also tested.
Compared with the AGEs alone, the addition of CCCP reduced the BMSCs show an age-related lineage switch between osteogenic and
expression of P16, P21 and P53, while the addition of CsA increased the adipogenic fates, which contributes to bone loss and senile osteoporosis

9
Y. Guo et al. Redox Biology 41 (2021) 101915

Fig. 6. The effect of Sirt3 silencing on mitochondrial function and mitophagy of BMSCs. BMSCs were transfected with scrambled siRNA (35 nmol/L, 72 h), Sirt3
siRNA (35 nmol/L, 72 h), treated with or without AGEs. (A) ROS production was detected with the fluorescent dye DCF (green). Scale bar: 50 μm. (B) Representative
fluorescence images with Mito-SOX (red) and MitoTracker (green) double-staining in BMSCs stimulated with AGEs. Scale bar: 50 μm. (C) The MMP was detected
through JC-1 staining in BMSCs stimulated with AGEs. Scale bar: 50 μm. (D) Representative fluorescence images with Mtphagy Dye (red) and Mito-Tracker (green)
double-staining in BMSCs stimulated with AGEs. Scale bar: 50 μm. (E) Representative fluorescence images with LC3B (red) and Mito-Tracker (green) double-staining
in BMSCs stimulated with AGEs. Scale bar: 50 μm. (F) Representative Western blotting assay and quantitation of the level of LC3B, P62, Parkin, Sirt3. *p < 0.05, **p
< 0.01 versus BSA, #p < 0.05, ##p < 0.01. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

[37]. Multilineage differentiation capacity of BMSCs was evaluated by CCCP reversed the AGE-induced reduction of ALP activity and CsA
subjecting BMSCs to osteogenic and adipogenic differentiation protocols promoted this effect (Fig. 5C). We further tested osteogenic differenti­
and dying for lipids and mineralization respectively [38]. It was re­ ation markers including Osterix, Runx2, Osteocalcin (OC) and collagen
ported that the inhibition of autophagy reduced osteogenic differentia­ type I (Col1), and lipogenic differentiation markers including C/EBPα,
tion and increased adipogenic differentiation in BMSCs [39,40]. Hence, PPARγ, and Twist1. As shown in Fig. 5C–J, the results of RT-PCR and WB
we studied the intrinsic interaction between the mitophagy and the showed that the mRNA expression levels of Osterix and Runx2 and the
balance of osteogenic differentiation and adipogenic differentiation in protein expression levels of OC and Col1 were significantly reduced by
BMSCs. In order to confirm the osteogenesis and adipogenesis of BMSCs, treatment with AGEs, while the mRNA expression levels of C/EBPa,
the BMSCs were induced to differentiate with osteogenic and adipogenic PPARγ, and Twist1 were significantly up-regulated. Stimulated with
media. Osteogenic differentiation was detected by alizarin red staining AGEs and CCCP at the same time, the mRNA expression levels of Osterix
(AZR) and adipogenic differentiation was determined by ORO staining. and Runx2 and the protein expression levels of OC and Col1 increased
As shown in Fig. 5A–B, the results showed that the osteogenic ability of significantly, while the mRNA expression levels of C/EBPα, PPARγ, and
BMSCs was significantly weakened, while the adipogenic ability was Twist1 decreased significantly. It turned out just the opposite when
significantly enhanced after treatment with AGEs. CCCP reversed the treated with CsA. The above results indicate that the senescence of
effects of AGE-induced decreased osteogenesis and increased adipo­ BMSCs regulated by mitophagy plays an important role in osteogenic
genesis, while CsA further aggravated the effects of AGE-induced and adipogenic differentiation.
decreased osteogenesis and increased adipogenesis. ALP activity, a
symbolic marker of early-stage osteogenesis, was detected using an ALP
activity assay kit. We found that AGEs reduced the activity of ALP, while

10
Y. Guo et al. Redox Biology 41 (2021) 101915

Fig. 7. The effect of Sirt3 silence on BMSCs senes­

cence. BMSCs were transfected with scrambled
siRNA (35 nmol/L, 72 h), Sirt3 siRNA (35 nmol/L,
72 h), treated with or without AGEs. (A) SA-β-gal
assay for detection of BMSCs senescence. Scale bar:
100 μm. (B) Detection of H3K9me3 by immunoflu­
orescence in BMSCs. Scale bar: 100 μm. (C) Detec­
tion of γ-H2AX by immunofluorescence in BMSCs.
Scale bar: 100 μm. (D) Representative Western
blotting assay and quantitation of the level of P16,
P21, P53. **p < 0.01 versus BSA, ##p < 0.01.

11
Y. Guo et al. Redox Biology 41 (2021) 101915

3.5. Knockdown of Sirt3 further enhanced AGEs-induced mitochondrial

dysfunction and the inhibition of mitophagy

Mitochondrial dynamics have critical roles in cellular senescence,

and their impairment represents a prominent risk factor for bone
metabolic diseases [16]. Mitochondrial deacetylase sirtuin3 (Sirt3)
contributes greatly to the prevention of redox stress and cellular
senescence [41]. Sirt3, localized to the mitochondria where its deace­
tylates and activates a number of enzymes involved in modulating the
mitochondrial function and ROS production, has been reported to
inhibit mitochondrial apoptosis [41,42]. As mitochondria are respon­
sible for senescence and Sirt3 is also closely related to mitochondrial
function and autophagy, thus, we postulated whether Sirt3 could affect
the changes in AGEs-induced mitochondrial dysfunction and mitophagy
in BMSCs. Therefore, we assessed the mitochondrial ROS, intracellular
ATP levels and MMP to analyze the functional status of mitochondria
after silencing sirt3. Then we picked the most-efficient Sirt3 silencing
siRNA by RT-PCR. Thus, we decided to continue experiments with this
siRNA (Fig. S4A). We found that the silencing of Sirt3 enhanced the
production of intracellular ROS (Fig. 6A) and mitochondrial ROS
(Fig. 6B) and further reduced the intracellular ATP content (Fig. S4B)
and MMP (Fig. 6C) induced by AGEs. Mitophagy is tightly associated
with mitochondrial health and function [43]. Mitophagy selectively
eliminates damaged mitochondria through autophagy pathway and
protects cells from the damage of mitochondrial dysfunction and
apoptosis induction [44]. We are interested in observing the effect of
Sirt3 silencing on mitophagy. Then mitophagy detection was also per­
formed using a Mtphagy Kit and a co-localization of LC3B immunoflu­
orescence with Mito-tracker green-stained mitochondria. As shown in
Fig. 6D and E, AGEs increased Mtphagy and LC3B fluorescence and
reduced mitochondrial content, while Sirt3 silencing further inhibited
mitophagy.
To further investigate the effects of Sirt3 on AGEs-induced mitoph­
agy, we detected the expression of LC3B, P62, Parkin and Sirt3. On the
basis of AGEs treatment, the silencing of Sirt3 further inhibited the
expression of LC3B, Parkin, Sirt3 and promoted the expression of
autophagy substrate P62 (Fig. 6F). Overall, Sirt3 silencing further
aggravated the AGEs-induced mitochondrial dysfunction and the inhi­
bition of mitophagy.

3.6. Sirt3 silencing enhanced the AGEs-induced senescence

Previous studies have confirmed that altered mitochondrial dy­

namics with decreased autophagy or mitophagy are hallmarks of
cellular senescence [45]. A decline in mitochondrial quality and activity
has been associated with normal aging and correlated with the devel­
opment of an abundant age-related diseases [46]. Mitophagy is an
important mitochondrial quality control mechanism that eliminates
damaged mitochondria. The above-mentioned results suggested that
Sirt3 has a crucial effect on mitochondrial function and mitophagy, so
we investigated the effect of Sirt3 on AGEs-induced senescence in
BMSCs. SA-β-gal assay showed that Sirt3 silencing aggravated
AGE-induced BMSCs senescence (Fig. 7A). H3K9Me3 and γ-H2AX are
considered to be signs of cell damage and senescence. Similarly, we Fig. 8. Observed the effect of BMSCs mitophagy and senescence after over­
expression of Sirt3 gene by adeno-associated virus vector in SAMP6 mice. (A)
observed the enhanced fluorescence of H3K9Me3 and γ-H2AX after Sirt3
The protocol used for the analysis of the effects of Sirt3. (B) SA-β-gal assay for
silencing (Fig. 7B and C). In addition, the silencing of Sirt3 inhibited cell
detection of BMSCs senescence. Scale bar: 50 μm. (C) The MMP was detected
proliferation and reduced the mRNA expression of Oct4 and Nanog in through JC-1 staining in BMSCs stimulated with AGEs. Scale bar: 50 μm. (D)
BMSCs (Figs. S5A–7C). Next, we also examined the expression of the Representative fluorescence images with Mtphagy Dye (red) and Mito-Tracker
aging-related proteins P16, P21, and P53. Our results showed that the (green) double-staining in BMSCs stimulated with AGEs. Scale bar: 50 μm. (For
silencing of Sirt3 further increased the expression of P16, P21 and P53 interpretation of the references to colour in this figure legend, the reader is
compared with the AGEs alone (Fig. 7D). Above results pointed out that referred to the Web version of this article.)
Sirt3 silencing enhanced the AGEs-induced BMSCs senescence.

3.7. Overexpression of Sirt3 activated mitophagy and delayed senescence

Given that downregulation of Sirt3 impaired mitochondrial function

12
Y. Guo et al. Redox Biology 41 (2021) 101915

Fig. 9. Observed the effect on SOP after over­

expression of Sirt3 gene by adeno-associated virus
vector in SAMP6 mice. (A) Osteogenic differentia­
tion was detected by alizarin red staining in BMSCs.
Scale bar: 50 μm. (B) Adipogenic differentiation was
tested by ORO staining in BMSCs. Scale bar: 50 μm.
(C) Representative micro-CT images of the distal
femurs in SAMP6. (D) ALP activity determined by
ALP activity assay. *p < 0.05, versus BSA. (E)
TRAP5b activity determined by TRAP5b activity
assay. *p < 0.05, **p < 0.01 versus BSA. (For
interpretation of the references to colour in this
figure legend, the reader is referred to the Web
version of this article.)

and mitophagy process, both of which were closely associated with AAV-control-EGFP or AAV-Sirt3-EGFP overexpressing adenovirus
BMSCs senescence, we hypothesize that Sirt3 could delay the process of (Fig. 8A). Four weeks after injection of AAV-EGFP-Sirt3-Flag or
BMSCs senescence by regulating mitophagy, which affects the osteo­ AAV-EGFP-control, we first tested the transfection of AAV to ensure the
genic differentiation and adipogenic differentiation of BMSCs. The study overexpression of sirt3 in vivo. The fusion protein FLAG was expressed in
suggested that mice lacking Sirt3 develop several aging-related diseases, SAMP6 mice injected with AAV-Sirt3 and the fluorescence of EGFP was
which could be considered as a model of accelerated aging [41]. In order shown in live imaging (Figs. S6A and E). After intravenous injection of
to confirm the role of Sirt3-mediated mitophagy in AGEs-induced BMSCs AGEs (50 ug/mouse/2 weeks) for a period of time, BMSCs were
senescence, we overexpressed the Sirt3 gene in aging model mice extracted and tested for cell senescence, mitochondrial function and
through a viral system and observed the aging situation in vivo. We mitophagy. As shown in Fig. 8B, the results of SA-β-gal assay showed
adopted SAMP6 mice, a model of SOP, and transfected with that Sirt3 overexpression reversed AGEs-induced BMSCs senescence.

13
Y. Guo et al. Redox Biology 41 (2021) 101915

Moreover, the overexpression of Sirt3 reduced the expression of between carbohydrates and proteins, which are stable to enzymes and
aging-related proteins P16, P21 and P53 (Figs. S6B–D and F–H). not easily eliminated. Studies have shown that accumulation of AGEs in
Meanwhile, Sirt3 overexpression improved AGE-induced mitochondrial tissues during aging has been associated with various pathophysiolog­
dysfunction and mitophagy inhibition. Specifically, overexpression of ical consequences such as osteoarthritis and osteoporosis [50]. The
Sirt3 reversed the decrease of MMP and MtphagyDye fluorescence accumulation of AGEs in the bone marrow is a feature of aging, a process
induced by AGEs (Fig. 8C and D). Western blot results also showed that of progressive decline in the biological functions and the resistance of
the expression of LC3B and Parkin increased, while the expression of cells, tissues and organs to various stresses gradually decreases [51]. Cell
autophagy substrate P62 decreased after overexpression of Sirt3 senescence is cell cycle arrest resulting in declines of cellular function
(Fig. S6I-P). These results suggested that Sirt3 improved mitochondrial and resilience. The key signal components of the aging mechanism, such
function and activated mitophagy to reverse AGE-induced BMSCs as P16, P21, P53 and the trimethylation of histone H3 lysine 9
senescence. (H3K9me3), also serve as key regulators of stem cell function [52]. The
expression level of β-galactosidase and γ-H2AX in cells increases
3.8. Overexpression of Sirt3 inhibited osteoporosis in SAMP6 mice accompanied with the aging of cells, which are the significant basis for
the identification of cell senescence [53]. Many previous studies
The generation and accumulation of AGEs in a process of aging are demonstrated that the oxidative environment of osteoporosis facilitated
well-known nowadays. There is incremental evidence indicating that BMSCs senescence by means of inducing mitochondrial dysfunction and
AGEs can deteriorate physiological function, destroy structural integ­ excessive ROS production [54–56]. In recent years, more and more ev­
rity, elicit oxidative stress and inflammatory reactions, and eventually idences have shown that autophagy plays an important role in main­
lead to various aging-related diseases [47]. Sirt3-mediated mitochon­ taining the balance of bone metabolism, which change is an significant
drial homeostasis could rejuvenate senescence of stem cells, thus cause of osteoporosis [57]. AGEs could affect cell function and activate
reversing aging-related degeneration [11]. Bone homeostasis is depen­ autophagy via ROS, while in turn autophagy could protect cells against
dent on the balance between osteogenesis and adipogenesis of BMSCs, AGEs-caused dysfunction [58]. In our study, we found AGEs produced a
osteogenic differentiation and adipogenic differentiation seem to be large number of ROS, which could damage mitochondrial function and
competitive and reciprocal. The imbalance may give rise to osteoporosis, inhibit mitophagy, ultimately leading to cell senescence. AGEs
displaying the reduced bone formation accompanied with the cumula­ decreased the ratio of LC3B/LC3A and increased P62 expression, sug­
tion of bone fat. We examined the osteogenic and adipogenic potential of gesting that AGEs could simultaneously inhibit mitophagy.
BMSCs after overexpressing Sirt3 and then treated with AGEs in SAMP6 Increasing evidence suggested that a main cause of primary osteo­
mice. As shown in Fig. 9A and B, the osteogenic potential of BMSCs porosis was the senescence of BMSCs, which gradually affected the stem
treated with AGEs were weakened and adipogenic potential increased, cell-like properties and ultimately disturbed the balance of osteogenic
while overexpression of Sirt3 increased the osteogenic potential of and adipogenic differentiation of BMSCs [19,48]. The balance between
BMSCs and weakened the adipogenic potential. Bone quality has always osteogenesis and adipogenesis plays an important role in the occurrence
been an important indicator for evaluating osteoporosis and and development of osteoporosis [59,60]. Promoting the osteogenic
bone-related diseases. Micro-CT analysis of trabecular bone parameters. differentiation and/or inhibiting the adipogenic differentiation of
Histomorphometric analysis showed that bone density, and trabecular BMSCs are considered as promising strategies for the development of
bone volume were significantly reduced after treatment with AGEs. anti-osteoporosis. Accumulating evidence suggested that the intrinsic
However, after overexpression of Sirt3, the osteoporosis-promoting ef­ properties of BMSCs such as osteogenic and adipogenic differentiation
fect was alleviated (Fig. 9C and S6Q). Serum bone alkaline phosphatase potential were markedly changed during aging process [61]. An
(ALP) activity is an early marker of osteoblast differentiation and aggravating in cell senescence could be due to an increase in
tartrate-resistant acid phosphatase isoform 5b (TRAP5b) is an osteo­ age-dependent damage to mitochondria or an age-dependent decline in
clastic molecule. The detection of the activity of ALP and TRAP5b mitophagy, eliminating dysfunctional mitochondria. Evidence indicated
showed that AGEs inhibited the secretion of the osteoblastic molecule that the level of mitophagy markedly reduced during normal aging in
ALP and promoted the secretion of the osteoclast molecule TRAP5b, BMSCs [62]. Our results showed that the activation of mitophagy by
while the overexpression of Sirt3 promoted the secretion of ALP and CCCP reduced BMSCs senescence, enhanced osteogenic differentiation
reduced the secretion of TRAP5b, indicating that Sirt3 plays an impor­ and weakened adipogenic differentiation, while CsA showed the oppo­
tant role in the formation of osteoporosis (Fig. 9D and E). site results by inhibiting mitophagy in AGEs-treated BMSCs. In addition,
the activation of mitophagy by Sirt3 overexpression further verified the
4. Discussion significant role of mitophagy in the BMSCs senescence and osteoporosis
in vivo, indicating that osteoporosis is functionally linked with mitoph­
As a result of bone-related aging diseases, SOP is one of the most agy, which have a central role in the aging process and cell senescence.
common type of osteoporosis, which result in increasing risk to fragility Sirt3 is a major mitochondrial deacetylase and affects most key as­
fractures among the elderly [48,49]. However, the pathogenesis of SOP pects of mitochondrial homeostasis, including ROS generation, ATP
is very complicated and the precise mechanisms in the occurrence and synthesis, mitochondrial dynamics and mitophagy [63,64]. Mitochon­
development of SOP have not yet been completely illuminated. The drial localization of Sirt3 play an essential role in various mitochondrial
pathogenesis of osteoporosis involves many factors and links, among functions, such as maintaining basal ATP level and regulating ROS
which the senescence of BMSCs plays an important role. Our present generation [65,66]. Our study found that Sirt3 silencing increased ROS
study demonstrated that AGEs promoted BMSCs senescence and accel­ levels through several mechanisms, thereby inhibiting mitophagy and
erated the occurrence and development of osteoporosis by inhibiting the promoting BMSCs senescence. In addition, Sirt3 overexpression signifi­
mitophagy, which reduced the osteogenic differentiation and increased cantly activated mitophagy, inhibited cell senescence and enhanced
the adipogenic differentiation of BMSCs. Furthermore, Sirt3 promoted osteogenic differentiation in vivo, suggesting that Sirt3 regulated bone
osteogenesis and attenuates adipogenesis in AGEs-treated BMSCs by homeostasis through modulating mitophagy. Meanwhile, we further
regulating mitophagy and senescence. In conclusion, our data demon­ detected the changes of BALP and TRAB5b expression in the serum of
strated that Sirt3 protects against AGEs-induced BMSCs senescence and SAMP6. ALP and TRAB5b are the marker proteins for bone trans­
osteoporosis through the activation of mitophagy. Targeting Sirt3 to formation, which are closely associated with bone resorption, formation
enhance mitophagy may represent a potential therapeutic strategy for and mineralization. The activity of ALP decreased and the activity of
attenuating age-associated osteoporosis. TRAB5b increased, while Sirt3 overexpression reversed this
AGEs are polymers formed from nonenzymatic glycation reactions AGEs-induced effect in SAMP6 mice treated with AGEs. What’s more,

14
Y. Guo et al. Redox Biology 41 (2021) 101915

Fig. 10. Schematic illustrating the mechanism of Sirt3-mediated mitophagy regulated AGEs-induced BMSCs senescence and senile osteoporosis.

the decrease in the number of trabecular bones and the increase in Appendix A. Supplementary data
trabecular spacing caused by AGEs were reversed by the overexpression
of Sirt3. Our preliminary data showed Sirt3 induced autophagy, leading Supplementary data to this article can be found online at https://doi.
to retardation of BMSCs senescence mediated osteoporosis. However, org/10.1016/j.redox.2021.101915.
the specific mechanisms of activating mitophagy by Sirt3 and the
changes of related signaling pathways need to be further studied. With References
in-depth studies, the relationship between BMSCs senescence, mitoph­
agy, and aging-associated diseases including osteoporosis will be [1] T. Sözen, L. Özışık, Başaran NÇJEjor, An overview and management of
osteoporosis 4 (1) (2017) 46.
clarified. [2] K.N. Tu, J.D. Lie, C.K.V. Wan, M. Cameron, A.G. Austel, J.K. Nguyen, K. Van, D.J.
P. Hyun, Therapeutics: Osteoporosis: a review of treatment options 43 (2) (2018)
92.
5. Conclusion [3] R. Schmitt, Melk AJKi, Molecular mechanisms of renal aging 92 (3) (2017)
569–579.
The current study represented the first demonstration of the pro­ [4] C. Boz, S.H. Ozsarı, The causes of aging and relationship between aging and health
expenditure: an econometric causality analysis for Turkey, Int. J. Health Plann.
tective role of Sirt3 in AGEs-induced osteoporosis (Fig. 10). Previous Manag. 35 (1) (2019) 162–170.
data showed that mitophagy played an important role in cellular [5] J. Chaudhuri, Y. Bains, S. Guha, A. Kahn, D. Hall, N. Bose, A. Gugliucci, P. Kapahi,
senescence and senescence-related osteoporosis [67,68]. We further The role of advanced glycation end products in aging and metabolic diseases:
bridging association and causality, Cell Metabol. 28 (3) (2018) 337–352.
clarified that Sirt3 could delay BMSCs senescence and present favorable
[6] Y. Tabara, T. Ikezoe, M. Yamanaka, K. Setoh, H. Segawa, T. Kawaguchi, S. Kosugi,
anti-osteoporotic effect by activating mitophagy. The importance of T. Nakayama, N. Ichihashi, T. Tsuboyama, Advanced glycation end product
these findings is that it supplied a potential target for the therapeutic accumulation is associated with low skeletal muscle mass, weak muscle strength,
intervention of SOP. and reduced bone density: the Nagahama study, J. Gerontol.: Series A 74 (9)
(2019) 1446–1453.
[7] D.-H. Yang, T.-I. Chiang, I.-C. Chang, F.-H. Lin, C.-C. Wei, Y.-W. Cheng, Increased
levels of circulating advanced glycation end-products in menopausal women with
osteoporosis, Int. J. Med. Sci. 11 (5) (2014) 453.
Declaration of competing interest
[8] Y. Wang, J. He, M. Liao, M. Hu, W. Li, H. Ouyang, X. Wang, T. Ye, Y. Zhang,
L. Ouyang, An overview of Sirtuins as potential therapeutic target: structure,
All authors have approved the final version of the manuscript and function and modulators, Eur. J. Med. Chem. 161 (2019) 48–77.
declared that no competing interest exists. [9] W. Grabowska, E. Sikora, A. Bielak-Zmijewska, Sirtuins, a promising target in
slowing down the ageing process, Biogerontology 18 (4) (2017) 447–476.
[10] J.M. Marcus, S.A. Andrabi, SIRT3 regulation under cellular stress: making sense of
Acknowledgements the ups and downs, Front. Neurosci. 12 (2018) 799.
[11] K. Brown, S. Xie, X. Qiu, M. Mohrin, J. Shin, Y. Liu, D. Zhang, D.T. Scadden,
D. Chen, SIRT3 reverses aging-associated degeneration, Cell Rep. 3 (2) (2013)
This study was supported by the National Natural Science Founda­ 319–327.
tion of China (81672216 and 81874026).

15
Y. Guo et al. Redox Biology 41 (2021) 101915

[12] Y. Song, S. Li, W. Geng, R. Luo, W. Liu, J. Tu, K. Wang, L. Kang, H. Yin, X. Wu, [41] R.A.H. van de Ven, D. Santos, M.C. Haigis, Mitochondrial sirtuins and molecular
Sirtuin 3-dependent mitochondrial redox homeostasis protects against AGEs- mechanisms of aging, Trends Mol. Med. 23 (4) (2017) 320–331.
induced intervertebral disc degeneration, Redox biology 19 (2018) 339–353. [42] R.M. Parodi-Rullan, X.R. Chapa-Dubocq, S. Javadov, Acetylation of mitochondrial
[13] J.S. Kerr, B.A. Adriaanse, N.H. Greig, M.P. Mattson, M.Z. Cader, V.A. Bohr, E. proteins in the heart: the role of SIRT3, Front. Physiol. 9 (2018) 1094.
F. Fang, Mitophagy and alzheimer’s disease: cellular and molecular mechanisms, [43] J. Zhang, M.L. Culp, J.G. Craver, V. Darley-Usmar, Mitochondrial function and
Trends Neurosci. 40 (3) (2017) 151–166. autophagy: integrating proteotoxic, redox, and metabolic stress in Parkinson’s
[14] M. Broda, A.H. Millar, O. Van Aken, Mitophagy: a mechanism for plant growth and disease, J. Neurochem. 144 (6) (2018) 691–709.
survival, Trends Plant Sci. 23 (5) (2018) 434–450. [44] S.M. Yoo, Y.K. Jung, A molecular approach to mitophagy and mitochondrial
[15] J.M. Bravo-San Pedro, G. Kroemer, L. Galluzzi, Autophagy and mitophagy in dynamics, Mol. Cell. 41 (1) (2018) 18–26.
cardiovascular disease, Circ. Res. 120 (11) (2017) 1812–1824. [45] N. Sun, R.J. Youle, T. Finkel, The mitochondrial basis of aging, Mol. Cell 61 (5)
[16] S. Wang, Z. Deng, Y. Ma, J. Jin, F. Qi, S. Li, C. Liu, F.J. Lyu, Q. Zheng, The role of (2016) 654–666.
autophagy and mitophagy in bone metabolic disorders, Int. J. Biol. Sci. 16 (14) [46] S. Srivastava, The mitochondrial basis of aging and age-related disorders, Genes 8
(2020) 2675–2691. (12) (2017).
[17] P.P. Naik, A. Birbrair, S.K. Bhutia, Mitophagy-driven metabolic switch reprograms [47] R. Kold-Christensen, M. Johannsen, Methylglyoxal metabolism and aging-related
stem cell fate, Cell. Mol. Life Sci. 76 (1) (2019) 27–43. disease: moving from correlation toward causation, Trends Endocrinol. Metabol.
[18] I.L. Ivanovska, J.W. Shin, J. Swift, D.E. Discher, Stem cell mechanobiology: diverse 31 (2) (2020) 81–92.
lessons from bone marrow, Trends Cell Biol. 25 (9) (2015) 523–532. [48] A. Qadir, S. Liang, Z. Wu, Z. Chen, L. Hu, A. Qian, Senile osteoporosis: the
[19] J.L. Pierce, D.L. Begun, J.J. Westendorf, M.E. McGee-Lawrence, Defining osteoblast involvement of differentiation and senescence of bone marrow stromal cells, Int. J.
and adipocyte lineages in the bone marrow, Bone 118 (2019) 2–7. Mol. Sci. 21 (1) (2020).
[20] J. Ma, Y. Guo, J. Hu, Y. Pan, X. Qi, H. Wang, L. Jia, The positive effect of chick [49] E. Curtis, A. Litwic, C. Cooper, E. Dennison, Determinants of muscle and bone
embryo and nutrient mixture on bone marrow- derived mesenchymal stem cells aging, J. Cell. Physiol. 230 (11) (2015) 2618–2625.
from aging rats, Sci. Rep. 8 (1) (2018) 7051. [50] N.L. Reynaert, P. Gopal, E.P.A. Rutten, E.F.M. Wouters, C.G. Schalkwijk, Advanced
[21] Y. H, W. L, L. Y, Y. S, Y. L, C. C, X. J, G. L, L. X, De-osteogenic-differentiated glycation end products and their receptor in age-related, non-communicable
mesenchymal stem cells accelerate fracture healing by mir-92b, J.Orthop.Translat. chronic inflammatory diseases; Overview of clinical evidence and potential
27 (2021) 25–32. contributions to disease, Int. J. Biochem. Cell Biol. 81 (2016) 403–418.
[22] J.C. McGrath, G.B. Drummond, E.M. McLachlan, C. Kilkenny, C.L. Wainwright, [51] B.F. Wang, J.J. Yu, T. Wang, Y. Shen, D.D. Lin, X. Xu, Y.Q. Wang, Identification of
Guidelines for reporting experiments involving animals: the ARRIVE guidelines, Br. megakaryocytes as a target of advanced glycation end products in diabetic
J. Pharmacol. 160 (7) (2010) 1573–1576. complications in bone marrow, Acta Diabetol. 55 (5) (2018) 419–427.
[23] M. Li, C. Wu, J.S. Muhammad, D. Yan, K. Tsuneyama, H. Hatta, Z.G. Cui, [52] M. Milanovic, D.N.Y. Fan, D. Belenki, J.H.M. Dabritz, Z. Zhao, Y. Yu, J.R. Dorr,
H. Inadera, Melatonin sensitises shikonin-induced cancer cell death mediated by L. Dimitrova, D. Lenze, I.A.M. Barbosa, et al., Senescence-associated
oxidative stress via inhibition of the SIRT3/SOD2-AKT pathway, Redox Biol 36 reprogramming promotes cancer stemness, Nature 553 (7686) (2018) 96–+.
(2020) 101632. [53] V. Labrie, O.J. Buske, E. Oh, R. Jeremian, C. Ptak, G. Gasiunas, A. Maleckas,
[24] R. Sanguineti, A. Puddu, F. Mach, F. Montecucco, G.L. Viviani, Advanced glycation R. Petereit, A. Zvirbliene, K. Adamonis, et al., Lactase nonpersistence is directed by
end products play adverse proinflammatory activities in osteoporosis, Mediat. DNA-variation-dependent epigenetic aging, Nat. Struct. Mol. Biol. 23 (6) (2016)
Inflamm. 2014 (2014) 975872. 566–573.
[25] S.K. Sharma, R.M. Leblanc, Biosensors based on beta-galactosidase enzyme: recent [54] G. Ye, Z. Xie, H. Zeng, P. Wang, J. Li, G. Zheng, S. Wang, Q. Cao, M. Li, W. Liu, et
advances and perspectives, Anal. Biochem. 535 (2017) 1–11. al., Oxidative stress-mediated mitochondrial dysfunction facilitates mesenchymal
[26] C. Chi, D.J. Li, Y.J. Jiang, J. Tong, H. Fu, Y.H. Wu, F.M. Shen, Vascular smooth stem cell senescence in ankylosing spondylitis, Cell Death Dis. 11 (9) (2020) 775.
muscle cell senescence and age-related diseases: state of the art, Biochim. Biophys. [55] P.V.S. Vasileiou, K. Evangelou, K. Vlasis, G. Fildisis, M.I. Panayiotidis,
Acta (BBA) - Mol. Basis Dis. 1865 (7) (2019) 1810–1821. E. Chronopoulos, P.G. Passias, M. Kouloukoussa, V.G. Gorgoulis, S. Havaki,
[27] L. Chen, R. Yang, W. Qiao, W. Zhang, J. Chen, L. Mao, D. Goltzman, D. Miao, 1,25- Mitochondrial homeostasis and cellular senescence, Cells 8 (7) (2019).
Dihydroxyvitamin D exerts an antiaging role by activation of Nrf2-antioxidant [56] B. Fougere, E. Boulanger, F. Nourhashemi, S. Guyonnet, M. Cesari, Chronic
signaling and inactivation of p16/p53-senescence signaling, Aging Cell 18 (3) inflammation: accelerator of biological aging, J Gerontol A Biol Sci Med Sci 72 (9)
(2019), e12951. (2017) 1218–1225.
[28] L. Habiballa, H. Salmonowicz, J.F. Passos, Mitochondria and cellular senescence: [57] K.H. Kim, M.S. Lee, Autophagy–a key player in cellular and body metabolism, Nat.
implications for musculoskeletal ageing, Free Radic. Biol. Med. 132 (2019) 3–10. Rev. Endocrinol. 10 (6) (2014) 322–337.
[29] D.R. Green, L. Galluzzi, G. Kroemer, Mitochondria and the autophagy- [58] M.A. Lumba, L.M. Willis, S. Santra, R. Rana, L. Schito, S. Rey, B.G. Wouters,
inflammation-cell death axis in organismal aging, Science 333 (6046) (2011) M. Nitz, A beta-galactosidase probe for the detection of cellular senescence by mass
1109–1112. cytometry, Org. Biomol. Chem. 15 (30) (2017) 6388–6392.
[30] X.S. Hou, H.S. Wang, B.P. Mugaka, G.J. Yang, Y. Ding, Mitochondria: promising [59] M. Bethel, B.R. Chitteti, E.F. Srour, M.A. Kacena, The changing balance between
organelle targets for cancer diagnosis and treatment, Biomater Sci 6 (11) (2018) osteoblastogenesis and adipogenesis in aging and its impact on hematopoiesis,
2786–2797. Curr. Osteoporos. Rep. 11 (2) (2013) 99–106.
[31] T. Wai, T. Langer, Mitochondrial dynamics and metabolic regulation, Trends [60] S. Benvenuti, I. Cellai, P. Luciani, C. Deledda, S. Baglioni, C. Giuliani, R. Saccardi,
Endocrinol. Metabol. 27 (2) (2016) 105–117. B. Mazzanti, S. Dal Pozzo, E. Mannucci, et al., Rosiglitazone stimulates
[32] T. Kanki, D.J. Klionsky, K. Okamoto, Mitochondria autophagy in yeast, adipogenesis and decreases osteoblastogenesis in human mesenchymal stem cells,
Antioxidants Redox Signal. 14 (10) (2011) 1989–2001. J. Endocrinol. Invest. 30 (9) (2007) RC26–30.
[33] X. Bai, X. Yang, X. Jia, Y. Rong, L. Chen, T. Zeng, X. Deng, W. Li, G. Wu, L. Wang, et [61] A. Infante, C.I. Rodriguez, Osteogenesis and aging: lessons from mesenchymal stem
al., CAV1-CAVIN1-LC3B-mediated autophagy regulates high glucose-stimulated cells, Stem Cell Res. Ther. 9 (1) (2018) 244.
LDL transcytosis, Autophagy 16 (6) (2020) 1111–1129. [62] F. Zhang, W. Peng, J. Zhang, W. Dong, J. Wu, T. Wang, Z. Xie, P53 and Parkin co-
[34] Y. Hou, X. Dan, M. Babbar, Y. Wei, S.G. Hasselbalch, D.L. Croteau, V.A. Bohr, regulate mitophagy in bone marrow mesenchymal stem cells to promote the repair
Ageing as a risk factor for neurodegenerative disease, Nat. Rev. Neurol. 15 (10) of early steroid-induced osteonecrosis of the femoral head, Cell Death Dis. 11 (1)
(2019) 565–581. (2020) 42.
[35] M.P.M. Soutar, L. Kempthorne, E. Annuario, C. Luft, S. Wray, R. Ketteler, M.H. [63] W. Yu, B. Gao, N. Li, J. Wang, C. Qiu, G. Zhang, M. Liu, R. Zhang, C. Li, G. Ji, et al.,
R. Ludtmann, H. Plun-Favreau, FBS/BSA media concentration determines CCCP’s Sirt3 deficiency exacerbates diabetic cardiac dysfunction: role of Foxo3A-Parkin-
ability to depolarize mitochondria and activate PINK1-PRKN mitophagy, mediated mitophagy, Biochim. Biophys. Acta (BBA) - Mol. Basis Dis. 1863 (8)
Autophagy 15 (11) (2019) 2002–2011. (2017) 1973–1983.
[36] Y. Zhang, X. Xi, Y. Mei, X. Zhao, L. Zhou, M. Ma, S. Liu, X. Zha, Y. Yang, High- [64] M.N. Sack, The role of SIRT3 in mitochondrial homeostasis and cardiac adaptation
glucose induces retinal pigment epithelium mitochondrial pathways of apoptosis to hypertrophy and aging, J. Mol. Cell. Cardiol. 52 (3) (2012) 520–525.
and inhibits mitophagy by regulating ROS/PINK1/Parkin signal pathway, [65] J. Wu, Z. Zeng, W. Zhang, Z. Deng, Y. Wan, Y. Zhang, S. An, Q. Huang, Z. Chen,
Biomedicine & pharmacotherapy = Biomedecine & pharmacotherapie 111 (2019) Emerging role of SIRT3 in mitochondrial dysfunction and cardiovascular diseases,
1315–1325. Free Radic. Res. 53 (2) (2019) 139–149.
[37] C.J. Li, Y. Xiao, M. Yang, T. Su, X. Sun, Q. Guo, Y. Huang, X.H. Luo, Long [66] N. Klimova, A. Fearnow, A. Long, T. Kristian, NAD(+) precursor modulates post-
noncoding RNA Bmncr regulates mesenchymal stem cell fate during skeletal aging, ischemic mitochondrial fragmentation and reactive oxygen species generation via
J. Clin. Invest. 128 (12) (2018) 5251–5266. SIRT3 dependent mechanisms, Exp. Neurol. 325 (2020) 113144.
[38] L.Y. Li, X.L. Wang, G.S. Wang, H.Y. Zhao, MiR-373 promotes the osteogenic [67] E.R. Stead, J.I. Castillo-Quan, V.E.M. Miguel, C. Lujan, R. Ketteler, K.J. Kinghorn,
differentiation of BMSCs from the estrogen deficiency induced osteoporosis, Eur. I. Bjedov, Agephagy - adapting autophagy for health during aging, Front Cell Dev
Rev. Med. Pharmacol. Sci. 23 (17) (2019) 7247–7255. Biol 7 (2019) 308.
[39] C. Vidoni, A. Ferraresi, E. Secomandi, L. Vallino, C. Gardin, B. Zavan, [68] H.C. Chiu, C.Y. Chiu, R.S. Yang, D.C. Chan, S.H. Liu, C.K. Chiang, Preventing
C. Mortellaro, C. Isidoro, Autophagy drives osteogenic differentiation of human muscle wasting by osteoporosis drug alendronate in vitro and in myopathy models
gingival mesenchymal stem cells, Cell Commun. Signal. 17 (1) (2019) 98. via sirtuin-3 down-regulation, J Cachexia Sarcopenia Muscle 9 (3) (2018)
[40] H.K. Choi, H. Yuan, F. Fang, X. Wei, L. Liu, Q. Li, J.L. Guan, F. Liu, Tsc1 regulates 585–602.
the balance between osteoblast and adipocyte differentiation through autophagy/
notch1/β-catenin cascade, J. Bone Miner. Res. 33 (11) (2018) 2021–2034.

16