Mitochondrial oxidative stress or decreased autophagy in osteoblast lineage cells is not sufficient to mimic the deleterious effects of aging on bone mechanoresponsiveness

Aging reduces the periosteal osteogenic response to mechanical stimulation

We first compared the response to mechanical stimulation of young (6 months) and old (21 months) female C57BL/6J mice. The axial tibial compressive loading was used with a peak force (or strain) of +1200με at the tibial midshaft (Figure 1A), previously shown to mimic the strains experienced by bones during normal daily activities [44]. The left tibia of each mouse was loaded at 4 Hertz (Hz) for 1200 cycles/day, 5 days/week for 2 weeks. Young and old mice were loaded with -7.8 and -6.6 Newtons (N), respectively, based on a priori strain analysis to achieve strain-matched loading. Mice were loaded on days 1–5 and 8–12, injected with calcein on day 5 and alizarin on day 12 to allow for the quantification of bone formation (Figure 1B). After sacrifice on day 15, both tibiae were collected. The right tibia was used as the non-loaded control. Micro-computed tomography (micro-CT) analysis at midshaft or at 5 mm proximal from the distal tibiofibular junction revealed that loading had no significant impact on cortical thickness in both young and old mice (Figure 1C). Bone formation rate measured by calcein-alizarin double labeling, was increased at the outer (periosteal) loaded tibia surface in 6-month-old mice, due to increases in both mineralizing surface and mineral apposition rate (Figure 1D, 1E). These osteogenic effects were also detected in the loaded tibia of 21-month-old mice. Yet, the magnitude of the loading-induced bone formation rate was significantly lower (p_int <0.01) in old mice, as described previously by others [17–19].

We also investigated how higher magnitudes of strain, exceeding what bones experience during normal daily activities, affect the mechanoresponsiveness of aged mice. The left tibia of 6- and 21-month-old female C57BL/6J were loaded with -10N and -8.5N, respectively, to achieve a strain magnitude of +1500με at midshaft. Two weeks of loading increased the cortical thickness at midshaft and proximal sites in the loaded tibia of young mice due to increased periosteal perimeters, while no effects were found at the endosteal surface (Figure 2A, 2B). Loading of old mice caused a similar effect. However, during the micro-CT analysis, we observed that old mice formed woven bone in their loaded tibias. This phenomenon typically occurs in response to bone damage, such as micro-fractures. Histomorphometry analysis confirmed the formation of irregular woven bone, particularly at the periosteal surface (Figure 2C). Collectively, these results indicate that the ability of the aged skeleton to form new bone in response to increased mechanical demands is considerably attenuated compared to the young skeleton.

Excessive mitochondrial ROS in osteoblast-lineage cells does not alter load-induced bone formation

We next examined whether excessive mtROS caused by Sod2 deletion could impair the osteogenic response to mechanical loading. Sod2 metabolizes the highly reactive superoxide anion free radical, a primary byproduct of mitochondrial respiration, into hydrogen peroxide [29]. Here, we generated mice lacking Sod2 in osteoblast lineage cells, designated Sod2^ΔOsx1, by crossing Sod2-flox (Sod2^f/f) mice with transgenic mice expressing the Cre recombinase under the control of Osx1-Cre regulatory elements. The Osx1-Cre transgene causes recombination in stromal cells, osteoblast progenitors, osteoblasts, and osteocytes [45]. Female Osx1-Cre littermates were used as controls. Quantitative RT-PCR analysis confirmed a reduction in Sod2 mRNA expression levels in bone marrow-derived osteoblastic cell cultures from Sod2^ΔOsx1 mice (Figure 3A). As expected, the reduction in Sod2 increased mtROS levels as measured by mitochondrial superoxide (MitoSOX) red staining (Figure 3B).

We then loaded the left tibiae of 12-month-old female Sod2^ΔOsx1 mice and their Osx1-Cre littermates with -8.5N at +1200με for two weeks, as described above (Figure 1B). Loading increased cortical thickness at proximal tibia and midshaft similarly in Sod2^ΔOsx1 and Osx1-Cre mice (Figure 3C, 3D). Consistently, no genotype-associated differences were observed in medullary area, total area, cortical area, polar moment of inertia, or material density at the midshaft of the loaded tibia (Figure 3D). In line with these results, the increased bone formation rate, mineralizing surface, and mineral apposition rate at the periosteal and endosteal surfaces of the loaded tibia were indistinguishable between Sod2^ΔOsx1 and Osx1-Cre mice (Figure 4A–4C). Taken together, these results indicate that an increase in mtROS in osteoblast-lineage cells was not sufficient to impact the bone anabolic response to mechanical stimulation.

Loading induces bone formation independently of Atg7-dependent autophagy

To examine whether autophagy deficiency in osteoblasts could potentially account for the decreased load-induced bone formation with aging, we used mice harboring a conditional allele for Atg7 in cells of the osteoblast lineage, designated Atg7^ΔOsx1 [36]. To confirm the suppression of autophagy, we measured the levels of LC3 and p62 in osteoblastic cell cultures using immunoblot analysis (Figure 5A). During autophagosome formation, the cytoplasmic form of LC3 (LC3-I) is lipidated and incorporated into the forming autophagosome membrane as LC3-II [40]. Based on this, LC3-I to LC3-II conversion is commonly used as an indicator of autophagy levels [46]. p62, a classical autophagy substrate, brings cargo to the autophagosome for degradation and accumulates when autophagy is suppressed [46]. Consistent with previous reports [36], deletion of Atg7 from Osx1-Cre-expressing cells inhibited autophagosome formation and thereby autophagy, as indicated by decreased LC3 conversion and p62 accumulation (Figure 5A).

We have previously shown that deletion of Atg7 drastically reduces dendrite connectivity in cortical bone [36]. It has been postulated that degeneration of the osteocyte network impairs the bone anabolic responses to loading [47, 48]. Here, we confirmed the effects of Atg7 deletion on the osteocyte dendritic network. We first assessed the density of osteocyte cell projections, which form as osteocytes become embedded into the bone matrix (Figure 5B). The average fluorescence intensity normalized to the bone area, indicative of the number and size of osteocyte cellular projections, was significantly lower in the Atg7^ΔOsx1 mice compared to Osx1-Cre controls (Figure 5C). These findings confirm that deletion of Atg7 in osteoblast lineage disrupts the formation or maintenance of the osteocyte network, as described previously [36]. In addition, osteocyte cell bodies were larger in Atg7^ΔOsx1 mice than in the controls (Figure 5C). This is likely due to the accumulation of cytoplasmic components that are normally degraded during the transition from osteoblasts to osteocytes [32].

We loaded the left tibiae of 10-month-old female Atg7^ΔOsx1 mice and Osx1-Cre littermates with −7.3 and −7.6N for 1200 cycles, respectively. Two weeks of loading increased cortical thickness at midshaft and proximal tibia similarly in Atg7^ΔOsx1 mice and Osx1-Cre littermates (Figure 6A). Accordingly, no differences between genotypes were detected in medullary area, total area, polar moment of inertia, or material density at the midshaft of the loaded tibia (Figure 6B). In addition, loading increased bone formation rate, mineralizing surface, and mineral apposition rate at the periosteal and endosteal (Figure 7A–7C) surfaces to a similar extent in Atg7^ΔOsx1 and Osx1-Cre mice. These results indicate that loss of autophagy in osteoblast-lineage cells was not sufficient to alter the osteogenic response to mechanical stimulation.

Animal experimentation

Mice with conditional deletion of Sod2 in osteoblast-lineage cells were generated by a two-step breeding strategy. Sod2 floxed (f/f) mice [79] were recovered from cryopreserved sperm at the Genetic Models Core Facility at University of Arkansas for Medical Sciences (UAMS). Please note that during genotyping of these mice, we determined that the neomycin phosphotransferase (neo) selection cassette that was contained in the original conditional allele was missing, likely due to an unintended recombination event. Sequencing of PCR products produced using primers P1 5’-CGA GGG GCA TCT AGT GGA GAA G-3’ and P2 5’-TTA GGG CTC AGG TTT GTC CAT AA-3’ (product sizes = 500 bp for WT and 540 bp for the floxed allele) revealed that the 540 bp product contained the loxP site predicted to exist after recombination of the loxP sites flanking the neo cassette [80]. Nonetheless exon 3 of the Sod2 gene remained flanked by functional loxP sites allowing us to use this version as a conditional (floxed) allele. Offspring from the in vitro fertilization were backcrossed to C57BL/6J mice (CD45.2) for four generations. Homozygous Osx1-Cre transgenic mice (B6.Cg-Tg (Sp7-tTA, tetO-EGFP/Cre) 1Amc/J; Strain #: 6361 Jackson Laboratory) were crossed with Sod2^f/f mice to generate mice heterozygous for the Sod2 floxed allele with (Sod2^f/+; Osx1-Cre) or without (Sod2^f/+) Osx1-Cre allele. These offspring were intercrossed to generate Sod2^ΔOsx1 mice and Osx1-Cre controls. Offspring were genotyped by PCR using the following primer sequences: Sod2-flox P1 5’-CGA GGG GCA TCT AGT GGA GAA G-3’, P2 5’-TTA GGG CTC AGG TTT GTC CAT AA-3’, Cre-fwd 5’-GCG GTC TGG CAG TAA AAA CTA TC-3’ and Cre-rev 5’-GTG AAA CAG CAT TGC TGT CAC TT-3’. Mice lacking Atg7 in Osx1-Cre targeted cells were generated using a similar strategy, as previously described [36]. PCR-based genotyping was conducted using the following primer sequences: Atg7-flox P1 5’-TGG CTG CTA CTT CTG CAA TGA TGT-3’, P2 5’-TCT CCC AAG ACA AGA CAG GGT GAA-3’, P3 5’-CAG GAC AGA GAC CAT CAG CTC CAC-3’; Cre-fwd 5’-GAG AAT AGG AAC TTC GGA ATA GTA AC-3’ and Cre-rev 5’-CCC TGG AAG TGA CTA GCA TTG-3’; IntCon-fwd 5’-AGA GAG CTC CCC TCA ATT ATG T-3’ and IntCon-rev 5’-AGC CAC TTC TAG CAC AAA GAA CT-3’. Offspring from all genotypes were tail-clipped for DNA extraction at weaning (21 days) and then group-housed with same-sex littermates. Mouse genotyping was performed using REDExtract-N-Amp™ Tissue PCR Kit (MilliporeSigma, XNAT-100RXN) according to the manufacturer’s instructions. Briefly, genomic DNA was extracted from fresh tail clippings by incubation in an extraction buffer (10 min, room temperature), followed by heating (95° C, 3 min), and neutralization. Bands were visualized using a ChemiDoc XRS⁺ System (Bio-Rad, Hercules, CA, USA). All mice were housed in groups of 2-5 animals per cage under standard laboratory conditions with a 12-hour light/dark cycle, a constant temperature of 23°C, and humidity of 48%. Mice had ad libitum access to water and a standard rodent diet (Envigo, Teklad 22/5) containing 22% protein, 1.13% calcium, and 0.94% phosphorus. To avoid the bone phenotype associated with the Osx1-Cre transgene, Atg7^ΔOsx1 mice were maintained on a doxycycline diet until 3 weeks of age to prevent Cre activity. Sod2^ΔOsx1 and Atg7^ΔOsx1 mice were used between the ages of 10-12 months old to avoid both the confounding effects of skeletal growth and those caused by natural aging. Mice were euthanized by CO₂ inhalation at designated time points followed by cervical dislocation. All animal procedures (IPROTO202400000001 and IPROTO202100000238) were approved by the UAMS Institutional Animal Care and Use Committee.

Tibia axial compressive loading

An axial compressive load with a +1200 με peak strain was applied to the left tibia midshaft using an ElectroForce 5500 (TA Instruments, New Castle, DE, USA). Briefly, with anesthetized mice (2.5% isoflurane) in a prone position, the left tibiae were positioned vertically in the load testing system, with the superiorly positioned knee attached to the system’s actuator and the foot held in a static fixture inferiorly. The load was applied for 1200 cycles/day (4 Hz triangle waveform with a 0.1-second rest time between each cycle) for five consecutive days per week over two weeks, following a protocol previously shown to be anabolic [81]. The right leg served as a non-loaded contralateral control. After each loading session, mice were returned to their cages and allowed unrestricted activity. To label actively mineralizing surfaces, calcein (20 mg/kg body weight; Sigma-Aldrich) and alizarin red (20 mg/kg body weight; Sigma-Aldrich) were injected intraperitoneally 10 days and 3 days before sacrifice, respectively. Mice were euthanized on day 15, and tissues were collected for skeletal analyses.

Strain gauge measurement

To determine the load required to achieve a peak strain of +1200με in each experimental group of mice, axial compressive loading was applied to harvested tibiae ex vivo. A single-element strain gauge (C2A-06-015LW-120, VPG Micro-Measurements, Wendell, NC, USA) was attached to the anteromedial surface of the tibia, located 5 mm proximal to the distal tibiofibular junction, using the M-Bond 200 Adhesive Kit (VPG Micro-Measurements). Force-strain regressions were recorded using ElectroForce 5500 software. The same load was subsequently applied to mice in vivo, adjusted according to their age or genotype.

Skeletal analysis

Tibias were cleaned of soft tissue, fixed in 10% Millonig’s formalin (Leica Microsystems, Deerfield, IL, USA ) with 5% sucrose for 24 hours, and gradually dehydrated into 100% ethanol. Bones were loaded into a 12.3 mm diameter scanning tube, and medium-resolution scans were obtained (12 μm isotropic voxel size) by micro-computed tomography (micro-CT40, Scanco Medical, Wayne, PA, USA). Scans were integrated into 3D voxel images (1024 × 1024 pixel matrices for each planar stack), and a Gaussian filter (σ = 0.8, support = 1) was applied to reduce signal noise. Scan settings included X-ray tube potential (55 kVp), X-ray intensity (145 μA), and integration time (200 ms). Scanco Eval Program v.6.0 was used to measure bone volume. For tibial cortical thickness, 18 slices were analyzed either 5 mm proximal to the distal tibiofibular junction or at the tibia midshaft. A threshold of 260 mg/cm³ was applied for cortical analyses. Calibration and quality control were performed weekly using five density standards, and spatial resolution was verified monthly using a tungsten wire rod. Beam-hardening correction was based on the calibration records. Corrections for 200 mg hydroxyapatite were applied to all energy levels. Nomenclature conforms to the recommendations of the American Society for Bone and Mineral Research [82].

Histomorphometry

After micro-CT analysis, the tibia was embedded undecalcified in methyl methacrylate, and 80-μm-thick longitudinal sections were obtained at 5 mm proximal to the distal tibiofibular junction. Sections were left unstained to assess fluorescent calcein and alizarin red labeling. Quantitative histomorphometry was performed on periosteal and endocortical surfaces using the OsteoMeasure Analysis System (OsteoMetrics, Inc., Atlanta, GA, USA) interfaced with an Axio image M2 (Carl Zeiss, White Plains, NY, USA). The following primary measurements were made: bone surface (BS), single-label surface (sL.S), double-label surface (dL.S), inter-label thickness (Ir.L.Th), and inter-label time (Ir.L.t). The following derived indices were calculated: mineralizing surface (MS, %) [(dLS + sLS/2)/BS], mineral apposition rate (MAR, μm/d) [Ir.L.Th/Ir.L.t], and bone formation rate (BFR, μm³/μm²/d) [MAR*(MS/BS)]. When MAR was equal to zero an arbitrary value of 0.3 mm was attributed. During dynamic histomorphometric analysis, we observed that two Atg7^ΔOsx1 and two Osx1-Cre mice were missing one of the injections since no double labels were observed. These mice were excluded from the analysis. All histology measurements were made in a blinded fashion. The results are reported using the terminology recommended by the Histomorphometry Nomenclature Committee of the American Society for Bone and Mineral Research [82].

Osteocyte network analysis and immunofluorescence

Phalloidin staining of actin was performed on the femurs. First, femurs were fixed in buffered 10% formalin for 24 hours, decalcified in 14% EDTA pH 7.1 for one week, stored in 30% sucrose solution, and embedded in Cryo-Gel™ (Electron Microscopy Sciences, Hatfield, PA, USA) for frozen sectioning. Frozen sections, 20 μm thick, were cut and rinsed 3 times in PBS for 10 minutes. Phalloidin staining of the osteocytes dendrites was performed on the femurs as previously described [36]. Briefly, sections were incubated in 0.2% Trition-X100 for 20 minutes with agitation, followed by washing in PBS. Sections were then incubated in 2% BSA for 30 minutes and incubated with Alexa Fluor™ 488-phalloidin (Invitrogen, Waltham, MA, USA) in 0.5% BSA for 2 hours at 4° C. At the end of the incubation, the sections were rinsed with PBS and cover-slipped with Vectashield® Mounting Medium containing DAPI (Invitrogen). Images of phalloidin-stained bone sections were acquired using a Zeiss LSM 510 Meta/AxioVert 200 confocal microscope using a 40 × or 63 × oil objective and z-stacks were obtained using ZEN 2009 software. Measurements of osteocyte dendrites and osteocyte cell body diameter were performed on flattened z-stacks (10 μm for femur) using ImageJ software. To quantify osteocyte dendrites, first, a region of interest (ROI) containing only cortical bone was drawn. Then osteocyte cell bodies were excluded from the ROI. Finally, the total dendrite fluorescence inside the ROI was calculated by subtracting the background fluorescence in a region lacking osteocytes or dendrites from the integrated intensity of the entire region. This value was then divided by the area of the region of ROI. Three fields of cortical bone from each sample were quantified and 3–4 samples per genotype were used.

cell culture

Bone marrow-derived mesenchymal stem cells (BMSCs) were harvested from the long bones of 5-6 pooled mice per group. Briefly, bone marrow stromal cells were flushed from the tibiae and femora, depleted of red blood cells with ammonium-chloride-potassium (ACK) buffer (0.01 mM EDTA, 0.011 M KHCO3, and 0.155 M NH4Cl, pH 7.3) and filtered through a 70 μM cell strainer. The isolated cells were then cultured for 7 days in 100 mm dishes containing osteogenic medium (α-MEM supplemented with 20% fetal bovine serum (FBS), 1% penicillin/streptomycin (PSG), and 50 μg/ml ascorbic acid) with a half-medium change on day 4. Adherent stromal cells were then trypsinized using Trypsin-EDTA solution 0.25% (Sigma-Aldrich) and used for subsequent analyses. To measure mitochondrial reactive oxygen species, 2×10⁴ cells were re-plated in a black-walled 96-well plate and cultured for two days. To perform qPCR assays, adherent BMSCs were plated in triplicate in 12-well plates at 2×10⁵ cells per well in α-MEM supplemented with 10% FBS, 1% PSG, and 50 μg/ml ascorbic acid. Upon reaching confluency, cells were cultured for an additional 7 days with 10 mM β-glycerophosphate, and media changes every 3 days to induce osteogenic differentiation.

Mitochondrial ROS

Mitochondrial superoxide (O₂•⁻) levels were measured using the mitochondrial-specific probe MitoSOX™ Red (Thermo Fisher Scientific, M36008). Briefly, cells were washed twice with phosphate-buffered saline (PBS) and incubated with 2.5 μM MitoSOX™ Red for 20 minutes at 37° C. Following two additional PBS washes, fluorescence (Ex/Em = 510/580 nm) was measured using a Cytation™ 5 microplate reader (BioTek) to assess MitoSOX oxidation, indicative of superoxide levels. Nuclei were counterstained with Hoechst dye (Thermo Fisher Scientific, H3570), and MitoSOX fluorescence was normalized to nuclear fluorescence intensity.

Quantitative RT-PCR

Total RNA was purified from cultured BMSCs using TRIzol™ Reagent (Thermo Fisher Scientific, Waltham, MA, USA). After extraction, RNA was quantified using a NanoDrop™ instrument (Thermo Fisher Scientific), and 1 μg of RNA was then used to synthesize cDNA with the High-Capacity cDNA Archive Kit (Applied Biosystems) according to the manufacturer’s instructions. Transcript abundance in the cDNA was quantified by quantitative PCR (qPCR) using TaqMan™ Universal PCR Master Mix (Thermo Fisher Scientific). The primers and probe for Sod2 (Mm00449726_m1) were obtained from the TaqMan™ Gene Expression Assays service (Applied Biosystems, Waltham, MA, USA). Relative mRNA expression levels were normalized to the housekeeping gene mitochondrial ribosomal protein S2 (Mm00475528_m1) using the ΔCt method [83].

Immunoblot analysis

Protein was extracted from humeri shafts, which were snap-frozen in liquid nitrogen and stored at −80° C. After pulverization, the resulting bone powder was suspended in RIPA buffer (Thermo Fisher Scientific, PI89901) containing protease and phosphatase inhibitors (Cell Signaling Technology, 5872S), following the manufacturer’s instructions. Proteins were separated on 4–20% or 4–15% Mini-PROTEAN TGX™ gels (Bio-Rad, 4561093 and 4561083, respectively) and transferred onto Trans-Blot Turbo™ midi-size nitrocellulose membranes (0.2 μm pore size, Bio-Rad, 1704271). The membranes were blocked for 30 minutes with LI-COR Blocking Buffer-PBS (LI-COR, 927-70001) and incubated overnight at 4° C with primary antibodies on a rocking platform. The primary antibodies used were anti-LC3 (Cell Signaling Technology, 12741T; 1:1000 dilution), anti-p62 (Cell Signaling Technology, 23214S; 1:1000 dilution), and β-actin (MilliporeSigma, A5316; 1:4000 dilution). After incubation, membranes were rinsed three times with PBS (5 minutes each) and incubated for 45 minutes with secondary antibodies conjugated to IRDye® 680 or IRDye® 800 dyes (LI-COR, 1:2000 dilution). Following three additional 5-minute PBS washes, the membranes were scanned and analyzed using the Odyssey IR Imaging System (LI-COR) and Image Studio™ Software (Version 5.2).

Statistical analysis

The statistical tests and number of replicates for each experiment are indicated in the figure legends. Data are presented as mean ± standard deviation (SD). The interaction terms of the two-way ANOVA factors (load, age, and genotype) are reported in the corresponding graph. Statistical analysis and graphical design were performed using GraphPad Prism 10.4.1 (GraphPad Software, Inc., La Jolla, CA, USA). Group means were compared using a two-sided Student’s t-test or two-way repeated measures ANOVA, as appropriate. Normality and homogeneity of variances were assessed using the D’Agostino and Pearson test and Shapiro-Wilk test, respectively. Pairwise multiple comparisons were performed with p-values adjusted by the Holm–Sidak method. Statistical significance was defined as p<0.05.