Drp1-Mediated Mitochondrial Fission Coordinates Mineral Precursor Transport and Mitochondrial Homeostasis During Osteogenesis

Abstract

Background

During the osteogenic differentiation of human mesenchymal stem cells (hMSCs), amorphous calcium phosphate (ACP) precursors within mitochondria are transported to the extracellular matrix via PINK1/Parkin-mediated mitochondrial autophagy, thereby initiating mineralization. However, the mechanisms by which mitochondria maintain functional homeostasis in the face of such extensive consumption remain to be elucidated. The present study investigated the role of Drp1-mediated mitochondrial fission in coordinating ACP transport and maintaining mitochondrial function.

Material/Methods

hMSCs were treated with shRNA to knock down Drp1 or with pharmacological inhibitors (Mdivi-1, CsA). Mitochondrial morphology, ACP transport, energy metabolism, and osteogenic differentiation were assessed using transmission electron microscopy, ATP assays, qPCR, western blot, and eosin staining. In vivo skeletal mineralization phenotypes were validated using a zebrafish model.

Results

Following osteogenic induction, Drp1 Ser616 phosphorylation increased while Ser637 phosphorylation decreased; mitochondrial fission increased, producing fragments approximately 450 nm in diameter, matching the size of autolysosomes, which facilitates ACP encapsulation and clearance. Inhibition of Drp1 resulted in abnormal mitochondrial elongation, ACP accumulation, and reduced ATP production (decreased to 59.2% in the sh-Drp1 group), and significantly suppressed RUNX2 and ALPL expression as well as mineralization (mineralization area decreased to 36.7%). In zebrafish, treatment with Mdivi-1 or CsA similarly led to reduced craniofacial mineralization and downregulation of osteogenic gene expression.

Conclusions

Drp1-mediated mitochondrial fission strikes a balance between the efficient transport of ACP precursors and mitochondrial functional homeostasis via the fission-clearance-regeneration cycle. This process serves as a key regulatory node in osteogenic differentiation. Further investigation of Drp1-associated mitochondrial dynamics in bone metabolic disease models is warranted.

Introduction

Osteogenic differentiation in bone-derived mesenchymal stem cells is a physiological process subject to rigorous regulation. Abnormalities in this process underlie metabolic disorders such as skeletal dysplasia and osteoporosis. A significant marker of mature osteogenic differentiation is the formation of hydroxyapatite mineralization nodules within the extracellular matrix. Despite earlier studies providing evidence, the precise mechanisms by which intracellular amorphous calcium phosphate (ACP) precursors are released via exocytosis into the extracellular space to initiate collagen fiber mineralization remain to be fully elucidated. The intracellular transport mechanism by which these ACP precursors are transported from their site of synthesis to secretory vesicles is not yet fully understood. Pei et al established that PINK1/Parkin-mediated mitochondrial autophagy plays a pivotal role in this process. In osteogenically induced mesenchymal stem cells, ACP initially accumulates within the mitochondria [1]. Subsequently, these ACP-laden mitochondria are recognized and engulfed by autophagosomes via the PINK1/Parkin pathway, eventually fusing with lysosomes to form autolysosomes. Within the autolysosomes, the mitochondria are degraded, while the ACP is released and aggregates into larger mineral granules, which are ultimately exported to the extracellular space via exocytosis, providing raw materials for collagen mineralization. This finding extends the function of mitochondrial autophagy from its traditional role in organelle quality control to the intracellular transport of mineral precursors.

However, this newly discovered transport mechanism gives rise to a critical biological question: given that mitochondria are central organelles responsible for intracellular ATP production and calcium homeostasis, how does the cell maintain functional homeostasis while simultaneously consuming large quantities of mitochondria for mineral transport, thereby ensuring the continuous progression of the osteogenic process? In other words, how do cells coordinate the balance between “mitochondrial consumption” and “mitochondrial functional maintenance”? Dynamically related protein 1 (Drp1), a pivotal regulator of mitochondrial fission, may provide the solution to this enigma. Mitochondrial fission, through which damaged fragments are separated and new mitochondria are generated, is a critical component of the mitochondrial quality control cycle (mitochondrial dynamics) [2]. It has been suggested by previous studies that Drp1 can be recruited to damaged mitochondria by PINK1, thus initiating the process of fission. This, in turn, facilitates the process of autophagy, the clearance of damaged segments, while allowing healthy segments to proliferate [3]. Based on this, we propose the following scientific hypothesis: Drp1-mediated mitochondrial fission maintains the dynamic equilibrium of the mitochondrial network, ensuring a continuous supply of ACP while preventing excessive mitochondrial depletion, thereby supporting the smooth progression of osteogenic differentiation. The aim of this study was to investigate the role of Drp1 in the osteogenic differentiation of bone-derived mesenchymal stem cells and to verify the mechanistic refinement of ACP transport logic and the boundary-condition testing of Drp1 dependence. The results will enhance our understanding of the mechanisms underlying cell-mediated biomineralization. Drp1-associated mitochondrial dynamics may be a candidate pathway for further investigation in bone metabolic disease models.

Material and Methods

Experimental Materials

The animal and cells experiment protocol was approved by the Ethics Committee for Laboratory Animal Welfare of the School of Medicine, Kunming University (Approval No.: KMU2023-025).

Cell Lines

Human bone marrow mesenchymal stem cells (hMSCs) were purchased from the American Type Culture Collection (ATCC, USA; Cat. No: PCS-500-012).

Experimental Animals

Adult wild-type AB strain zebrafish (Danio rerio) were purchased from Nanjing Yishu Li-hua Biotechnology Co, Ltd.

Experimental Methods

Cell Culture

hMSCs were cultured in DMEM complete medium (Yisheng Bio, Shanghai, China; Cat. No: 40101ES), and cultured in a 37 °C, 5% CO₂ incubator. When cells reached 80% to 90% confluence, they were passaged using 0.25% trypsin-EDTA solution (Biyuntian, Shanghai, China; Cat. No.: C0201).

Establishment of a Stable Drp1 Knockdown Cell Line

A stable DRP1-knockdown hMSC cell line was established using lentivirus-mediated short hairpin RNA (shRNA) technology. The shRNA sequence targeting the human DNM1L (DRP1) gene was: 5′-GCTACTTTACTCCAACTTATT-3′. A negative control shRNA (sh-NC, ie, the EV group) carrying a non-targeting scrambled sequence was also constructed. Vector construction, viral packaging, and purification of lentiviral particles were all performed by Bioyuntian (Shanghai, China).

hMSCs were seeded into 12-well plates at a density of 5 × 10⁴ cells per well and cultured until 50% confluence. Lentiviral particles were added at a multiplicity of infection (MOI) of 20 to medium containing 5 μg/mL Polybrene (Biyuntian, Shanghai, China; Cat. No.: C0351). Fresh complete medium was added 12 hours after infection. Seventy-two hours after infection, medium containing 2 μg/mL puromycin (Biyuntian, Shanghai, China; Cat. No.: ST551) was added to screen for stably transfected cells. This process continued for 7 days, with fresh puromycin-containing medium replaced every 2 to 3days. The efficiency of Drp1 knockdown was verified using western blot analysis. Successfully established sh-Drp1 and sh-NC-hMSCs were used for subsequent experiments.

Experimental Groups and Osteogenic Induction Treatment

The experiment was designed with the following 5 groups: the EV-OI group: sh-NC hMSCs, conventional culture, no osteogenic induction; the EV group: sh-NC hMSCs, subjected to osteogenic induction; the sh-Drp1 group: sh-Drp1 hMSCs, subjected to osteogenic induction; the Mdivi-1 group: sh-NC hMSCs, with Mdivi-1 (Biyuntian, Shanghai, China; Catalog No.: SC0330), final concentration 10 μmol/L; and the CsA Group: sh-NC hMSCs, osteogenic induction medium supplemented with cyclosporine A (CsA, Bioyuntian, Shanghai, China; Catalog No.: Y265567), final concentration 100 nmol/L.

Preparation of osteogenic differentiation medium (ODM): Add 50 μmol/L ascorbic acid (Biyuntian, Shanghai, China; Cat. No.: Y262819), 10 mmol/L sodium β-glycerophosphate (Aladdin, Shanghai, China; Catalog No.: G108067), and 100 nmol/L dexamethasone (Biyuntian, Shanghai, China; Catalog No.: ST1254). All reagents were freshly added prior to use, and the medium was replaced every 2 to 3 days. For the drug-treated groups, fresh osteogenic differentiation medium containing the corresponding drug was replaced every 2 days.

ATP Content Assay

On day 7 of osteogenesis induction, intracellular ATP content was measured using a chemiluminescent cell viability assay kit (Yisheng Bio, Shanghai, China; Cat. No.: 40210ES). Cell culture plates were removed and allowed to equilibrate to room temperature for 10 minutes. We added 100 μL of assay reagent to each well, and the plates were shaken at room temperature for 2 minutes to lyse the cells, followed by incubation in the dark for 10 minutes. Chemiluminescence intensity was measured using a microplate reader (SpectraMax iD3, Molecular Devices, USA), with a reading time of 0.5 seconds per well.

To calculate absolute ATP content, an ATP standard curve was simultaneously established: ATP standards were diluted in PBS to create a concentration gradient of 0.1, 0.5, 1, 5, and 10 μmol/L and tested in the same batch as the samples. Protein concentrations were determined in parallel wells using a BCA protein assay kit, and ATP levels were normalized to nmol/μg protein. To facilitate intergroup comparisons, ATP levels in each group were converted to relative percentages (%), normalized with the EV group set as 100%.

Real-Time Quantitative Fluorescent PCR

Total RNA Extraction

Cell samples: Total RNA was extracted from cells using the FastPure Cell/Tissue Total RNA Isolation Kit V2 (Vazyme, Nanjing, China; Cat. No.: RC112). The procedure was performed according to the manufacturer’s instructions. A brief description follows: Cells in a 6-well plate were directly lysed with 500 μL of Buffer RL. The lysate was transferred to a FastPure gDNA-Filter Column III and centrifuged at 12 000 rpm for 30 seconds to remove genomic DNA. Anhydrous ethanol (0.5 × the volume of the filtrate) was added to the filtrate; after mixing, the mixture was transferred to a FastPure RNA Column III and centrifuged at 12 000 rpm for 30 seconds. Wash sequentially with 700 μL Buffer RW1, 700μL Buffer RW2 (with ethanol added), and 500 μL Buffer RW2; centrifuge for 2 minutes after the final wash. Centrifuge the empty column at 12 000 rpm for 1 minute to remove residual ethanol. Elute the RNA with 30 μL RNase-free ddH₂O.

Zebrafish samples: Take zebrafish larvae or tissue, add Buffer RL, and homogenize thoroughly using a tissue homogenizer; subsequent steps are the same as for cell samples. Determine RNA concentration and purity using the Nano 600 ultra-micro nucleic acid and protein analyzer (Jiapeng, Shanghai).

cDNA Synthesis

Perform reverse transcription according to the instructions for the HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China; Cat. No.: R312). Take 500 ng of total RNA, add 2 μL of 5 × gDNA Wiper Mix, and dilute to 10 μL with RNase-free ddH₂O. Incubate at 42 °C for 2 minutes to remove genomic DNA. Subsequently, add 10 μL of 2 × RT Mix, 2 μL of HiScript III Enzyme Mix, 1μL of Oligo(dT)₂₀VN (50 μmol/L), and 1μL of Random hexamers (50 ng/μL), for a total volume of 20 μL. Reaction conditions: 25 °C for 5 minutes, 50 °C for 15 minutes, 85 °C for 5 minutes. The resulting cDNA was diluted 5-fold and stored at −20 °C.

qPCR Detection

Amplification was performed using Taq Pro Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China; Catalog No.: Q712) on a CFX96 Touch real-time quantitative PCR system (Bio-Rad, USA). The 20 μL reaction mixture contained: 5 μL diluted cDNA, 10 μL 2 × Master Mix, and 0.4 μmol/L each of forward and reverse primers, with the volume made up with nuclease-free ddH₂O. Three replicates were set up for each sample. Reaction program: 30 seconds pre-denaturation at 95 °C; 10 seconds at 95 °C and 10 seconds at 60 °C, for a total of 40 cycles; primer specificity was verified by melting curve analysis.

In cell experiments, human GAPDH was used as an internal control to detect the expression of DRP1 (DNM1L), RUNX2, and ALPL; in zebrafish experiments, β-actin was used as an internal control to detect the expression of RUNX2 and ALPL. Relative expression levels were calculated using the 2^−ΔΔCt method. Primers were synthesized by Sangon Biotech (Shanghai); see Table 1 for sequences.

Table 1.

Primer sequences used for qRT-PCR.

Species	Gene	Forward primer (5′→3′)	Reverse primer (5′→3′)
Human	GAPDH	AATGGGCAGCCGTTAGGAAA	GCGCCCAATACGACCAAATC
	DRP1	GATGCCATAGTTGAAGTGGTGAC	CCACAAGCATCAGCAAAGTCTGG
	RUNX2	CCACCGAGACCAACAGAGTC	TCACTGTGCTGAAGAGGCTG
	ALPL	CTATCCTGGCTCCGTGCTCC	GCACCCCAAGACCTGCTTTA
Zebrafish	β-actin	TTGACAACGGCTCCGGTATG	TCCCATGCCAACCATCACTC
	runx2	CGCTCACAAACAACCACAGA	CAGGTGACGGTCTGGTTGTT
	alpl	TCAGTTGTCTGGCACCCTTG	TCCAGTAATCAGGCCGCTTC

Western Blot Analysis

Cells and zebrafish samples were homogenized using RIPA lysis buffer (Biyuntian, Shanghai, China; Catalog No.: P0013C), centrifuged at 4 °C and 12 000 rpm for 40 minutes, and the supernatant was collected. After determining protein concentration using the BCA method, equal amounts of protein (30 μg) were separated by 7.5% to 12.5% SDS-PAGE and transferred to a PVDF membrane (0.2 μm). The membrane was blocked with 5% non-fat dry milk (Cell Signaling Technology, USA) at room temperature for 1 hour, then incubated overnight at 4 °C with the following primary antibodies: anti-GAPDH, anti-Drp1, anti-ALPL, anti-Runx2, anti-Ser637-Drp1, anti-Ser616-Drp1, anti-Pink1, anti-Parkin, and anti-P62. All antibodies used in this study were purchased from Abcam (Cambridge, UK).

The next day, the membranes were washed with Tris-buffered saline with Tween, and the corresponding secondary antibodies (1: 3500, Santa Cruz Biotechnology, USA) were added for incubation at room temperature for 1 hour. Development was performed using an ECL chemiluminescence kit (Pierce Biotechnology, USA; catalog number: WP20005), and grayscale analysis was conducted using ImageJ software (ImageJ 1.5, NIH, USA) for grayscale analysis. Results were normalized using GAPDH as an internal control and expressed as a fold change relative to the EV control group.

Transmission Electron Microscopy and Transmission Electron Microscopy – Energy Dispersive Xray Spectroscopy (TEM-EDS)

Cell Sample Preparation

After cells were treated according to experimental groups until the predetermined time point, the culture medium was discarded, and the cells were washed twice with PBS (0.1 mol/L, pH 7.4). Pre-chilled fixative (0.1 mol/L sodium dimethylarginate buffer containing 2.5% glutaraldehyde and 2% paraformaldehyde, pH 7.4) was added, and samples were fixed at 4 °C for 2 hours. Samples were washed 3 times with PBS for 10 minutes each. Post-fix at 4 °C with 1% osmic acid for 1 hours, then rinse 3 times with distilled water, 10 minutes each.

Transmission Electron Microscopy Observation

After fixation, samples were dehydrated using a gradient of ethanol (30%, 50%, 70%, 80%, 90%, 95%, 100% × 2), 10 minutes per step. After a transition to acetone, the samples were infiltrated and embedded in Epon 812 epoxy resin, followed by gradient polymerization at 37 °C for 12 hours, 45 °C for 12 hours, and 60 °C for 48 hours. Ultrathin sections (70 nm) were prepared using an ultramicrotome (Leica UC7, Germany) and mounted on copper grids. The sections were stained with 2% uranyl acetate for 20 minutes and lead citrate for 5 minutes, then rinsed with distilled water and air dried. The ultrastructure of the cells was observed using a transmission electron microscope (HITACHI HT7800, Japan) at an acceleration voltage of 80 to 100 kV, and representative images were captured. In 3 independent experiments, the organelles of 3 cells were selected. According to the Guidelines for the Use and Interpretation of Assays for Monitoring Autophagy (4th edition) [3], organelles are defined as follows: Autolysosomes were defined as single-membrane or partially disrupted membrane-bound vacuoles containing heterogeneous, partially degraded cytoplasmic materials; and normal mitochondria were defined as an intact double-membrane structure. The outer membrane was smooth and continuous, while the inner membrane formed folds (cristae) that projected into the matrix. Mitochondrial fission events were identified by the presence of mitochondrial profiles showing localized constriction of both the outer and inner membranes, giving a characteristic “hourglass” or “necked” appearance. The apparent lengths of mitochondria, fragmented mitochondria, and autolysosomes were measured using FIJI software.

Transmission Electron Microscopy – Energy Dispersive Xray Spectroscopy (TEM-EDS)

Cell smears for transmission electron microscopy were dehydrated using an ethanol gradient and then dried in a cryo-dryer (Leica EM CPD300, Germany). Samples were mounted on the sample stage using conductive carbon tape, and a carbon coating (5–10 nm thick) was applied using an ion sputtering system. A field-emission scanning electron microscope (HITACHI SU8010, Japan) was used to observe cell surface morphology and matrix mineralization at an acceleration voltage of 5 to 10 kV. Energy dispersive spectroscopy (EDS) was performed using a windowless silicon drift detector (Oxford Instruments X-MaxN 80, UK) to conduct area-selective compositional analysis in regions of high electron density within the cells. Working distance: 15 mm; dead time: < 30%. Spectra were quantified using the ZAF matrix correction method, and the elements C, O, P, and Ca were identified based on characteristic X-ray peak positions. Atomic percentages (At.%) and mass fractions (wt%) were calculated.

Zebrafish Rearing, Embryo Collection, and Drug Treatment

Adult wild-type AB strain zebrafish (Danio rerio) were purchased from Nanjing Yishuli Hua Biotechnology Co., Ltd. and reared in a recirculating aquaculture system (EzeRinka, model ZF-105). The system water temperature was maintained at 28 ± 0.5 °C, with a pH of 7.0 to 7.4, a conductivity of 500 to 550 μS/cm, and a photoperiod of 14 hours light/10 hours dark. Fish were fed twice daily with freshly hatched Artemia nauplii.

Embryos were collected using the natural paired mating method. The night before the experiment, parent fish were placed in a spawning tank in a 2: 2 male-to-female ratio, separated by a partition. The partition was removed at the start of the next light cycle to allow natural mating and spawning. Embryos were collected 30 minutes later and washed with an embryo culture medium containing methylene blue to remove dead eggs. The embryos were incubated at a constant temperature of 28 °C in an intelligent light-controlled incubator, with the culture medium replaced once daily.

Healthy embryos developed to 24 hours post-fertilization (24 hpf) were selected and randomly divided into the following 3 groups for drug treatment: Control group: embryo culture medium + an equal volume of DMSO (solvent control, final concentration < 0.1%); Mdivi-1 group: 1.82 mg/L Mdivi-1 (Biyuntian, Shanghai, China; Cat. No.: SC0330) in embryo culture medium; CsA group: 1.2 mg/L Cyclosporin A (CsA, Biyuntian, Shanghai, China; Cat. No.: Y265567) in embryo culture medium. Embryos in each group were continuously cultured in a 28 °C incubator until 9 days post-fertilization (9 dpf), with fresh culture medium containing the corresponding drug replaced daily.

Alizarin Red Staining

Alizarin Red Staining of Cells

After culturing hMSCs in osteogenic induction medium for 21 days according to experimental groups, the supernatant was discarded, and the cells were rinsed twice with phosphate-buffered saline. The cells were fixed with 4% paraformaldehyde at room temperature for 15 to 30 min and rinsed 3 times with distilled water. Alizarin Red S staining solution (Wuhan Saiver Biotechnology Co., Ltd., Cat. No.: G1038) was added, and the cells were incubated at room temperature for 15 to 30 min. Discard the staining solution and rinse with distilled water until the rinse solution runs clear. After the plates dry, photograph the entire plate using a digital camera. Perform semi-quantitative analysis using FIJI software (Version 2.9.0): convert the image to an 8-bit grayscale image, set a uniform threshold to define the positive staining area, measure the positive area and integrated optical density, and calculate the relative mineralization level using the EV group as a reference. The experiment was independently repeated 3 times.

Zebrafish Alizarin Red Staining

Nine-day-post-fertilization (dpf) zebrafish larvae were fixed overnight at 4 °C in 4% paraformaldehyde. They were washed 3 times with phosphate-buffered saline for 5 minutes each. The samples were sequentially dehydrated using a gradient of 75%, 90%, and 100% (twice) ethanol, with each step lasting 5 minutes, followed by a 1-minute rinse with tap water. Add alizarin red S staining solution (Wuhan Saver Biotechnology Co., Ltd., Cat. No.: G1038) to completely cover the samples and stain at room temperature for 20 to 60 minutes, observing under a stereomicroscope until clear orange-red staining appeared at calcium salt deposition sites in the craniofacial region and spine. Discard the staining solution and rinse with tap water until the background was colorless. After adjusting the posture using a postural fixative, perform bright-field imaging using a stereomicroscope (Leica M205 FA, Germany). Randomly select at least 5 juvenile fish per group and use FIJI software (Version 2.9.0) to measure the area of positive regions and the integrated optical density; calculate relative mineralization levels using the control group as a reference.

Statistical Analysis

GraphPad Prism 9.0 was used for all statistical analyses. Normality was assessed using the Shapiro-Wilk test and homogeneity of variances using Levene’s test. For data meeting parametric assumptions, one-way ANOVA followed by Tukey’s post hoc test or Dunnett’s post hoc test was applied for multiple comparisons. Otherwise, data were either log-transformed or analyzed using the Kruskal-Wallis test with Dunn’s post hoc test. Differences were considered statistically significant at * P < 0.05 (* P < 0.05, ** P < 0.01, **** P < 0.001). Western blot and qPCR data are expressed as fold changes relative to the EV control group.

Definition of independent observations (n): For qPCR, western blotting, ARS staining and ATP assay, n represents independent biological replicates (1 separately cultured cell passage or 1 individual zebrafish). For organelle quantification, n represents the number of organelles, each obtained from a distinct biological replicate and a distinct cell (3 biological replicates, 1 cell per replicate, 1 organelle per cell meeting quality criteria). Technical replicates were averaged per biological replicate prior to analysis.

Results

Changes in the Ultrastructural Features of Mitochondria and ACP Transport in hMSCs During Osteogenic Induction

To investigate the relationship between mitochondrial morphodynamics and the transport of amorphous calcium phosphate (ACP) precursors during osteogenic differentiation. Transmission electron microscopy was used to examine the ultrastructure of hMSCs at 7 and 14 days after osteogenic induction. After 7 days of osteogenic induction, ultrastructural analysis of cells in the EV group revealed an increase in the number of mitochondria (M), with some mitochondria exhibiting fission (Mf); high-electron-density ACP granules were visible within the mitochondrial fragments formed by fission (Figure 1A, red arrows). During the same period, single-membrane vesicles containing ACP particles were observed in the cytoplasm; their morphological features were consistent with those of autolysosomes (AL), suggesting that ACP derived from mitochondria was transported to autolysosomes (Figure 1B, red arrow). Fourteen days after osteogenic induction, scattered electron-dense ACP deposits were observed in the extracellular matrix of the EV group, indicating that intracellularly synthesized mineralization precursors had been released into the extracellular environment via exocytosis (Figure 1C, red arrow). N denotes the cell nucleus. Compared with the EV group, the mitochondrial morphology of cells in the sh-Drp1 group underwent significant changes. Following Drp1 knockdown, mitochondrial fission was inhibited, and mitochondria became significantly elongated, forming a highly tubular network; a large accumulation of ACP granules was observed within their matrix (Figure 1D, red arrow), suggesting that impaired mitochondrial fission led to the retention of ACP precursors within the mitochondria, preventing their effective transport.

Figure 1.

Transmission electron microscopy of hMSC mitochondrial ultrastructure, ACP transport process, and ATP content analysis. (A) hMSCs induced to osteogenic differentiation for 7 days; mitochondria (M) have fragmented (Mf) and contain ACP particles (red arrows). N: Nucleus. Scale bar: 500 nm. (B) hMSCs after 7 days of osteogenic induction; ACP particles (red arrows) are present within autophagosomes (AL). Scale bar: 500 nm. (C) hMSCs after 14 days of osteogenic induction; ACP particles (red arrows) are transported to and deposited in the extracellular matrix. Scale bar: 500 nm. (D) sh-Drp1 group after 7 days of bone induction: mitochondria (M) are significantly elongated, and ACP particles (red arrows) accumulate within the matrix. Scale bar: 500 nm. (E) EDS spectral analysis shows that electron-dense inclusions within mitochondria contain Ca, P, and O elements, consistent with the elemental composition of ACP. (F) Quantitative analysis of the major axis length of organelles in each group, n = 9. EV-M: Conventional mitochondria in the EV group (801.3 ± 79.2 nm); EV-Mf: ACP-containing mitochondrial fragments in the EV group following fragmentation (446.9 ± 41.6 nm); EV-AL: autophagosomes in the EV group (426.5 ± 46.8 nm); DRP1-kd: mitochondria in the sh-Drp1 group (1626.9 ± 403.4 nm). Data are expressed as mean ± standard deviation. * P < 0.05, ** P < 0.01, **** P < 0.0001 vs EV-M group; ns, no significant difference. (G) Relative intracellular ATP content in each group. Normalized with the EV group as 100%. Data are expressed as mean ± standard deviation, n = 6. * P < 0.05, ** P < 0.01, **** P < 0.0001 vs EV group.

Electron spectroscopy analysis revealed characteristic Ca, P, and O signal peaks in both the electron-dense granules within mitochondria and the deposits within autophagosomes (Figure 1E). The calcium-to-phosphorus atomic ratio was approximately 1.50, consistent with the elemental composition of ACP, confirming that these high-electron-density deposits are calcium phosphate precursors.

To further analyze the spatial correlation between mitochondrial fission and autophagosome formation, quantitative measurements of the major axis length of these organelles were performed. The results are shown in Figure 1F: the major axis length of conventional mitochondria (EV-M) in the EV group was 801.3 ± 79.2 nm; the length of the major axis of ACP-containing mitochondrial fragments (EV-Mf) formed after fission was significantly shorter, at 446.9 ± 41.6 nm (P = 0.003 vs EV-M); whereas the length of the major axis of ACP-containing autolysosomes (EV-AL) was 426.5 ± 46.8 nm. Notably, the major axis lengths of EV-Mf and EV-AL were similar, with no statistically significant difference between them (P = 0.782), suggesting that the size of the fragments generated after mitochondrial fission matches the phagocytic capacity of autolysosomes, facilitating the encapsulation and subsequent degradation of damaged mitochondria by autolysosomes. In the sh-Drp1 group (DRP1-kd), due to inhibited fission, the major axis length significantly increased to 1626.9 ± 403.4 nm (P < 0.0001 vs EV-M).

To assess the impact of inhibited mitochondrial fission on cellular energy metabolism, intracellular ATP levels were measured using chemiluminescence on day 7 of osteogenic induction. The results showed that in Figure 1G, with ATP levels in the EV group set as 100%, the relative ATP content in the sh-Drp1 group decreased to 59.2 ± 4.1% (P < 0.0001, vs EV), the Mdivi-1-treated group to 71.5 ± 5.0% (P < 0.0001, vs EV), and the CsA-treated group to 94.9 ± 2.3% (P = 0.033, vs EV). These results indicate that inhibition of Drp1-mediated mitochondrial fission leads to reduced cellular ATP production, with the most significant decrease observed in the sh-Drp1 group, suggesting that complete Drp1 deficiency has a more severe impact on mitochondrial energy metabolism than pharmacological inhibition.

In summary, during osteogenic induction, mitochondria in hMSCs undergo Drp1-mediated fission to generate fragments of appropriate size, which are transported via the mitochondrial autophagy pathway to autophagosomes and are ultimately released into the extracellular matrix as ACP to participate in mineralization. Inhibition of Drp1 not only leads to abnormal mitochondrial elongation and the retention of ACP within mitochondria but also impairs mitochondrial energy metabolism, as evidenced by reduced ATP production.

Drp1 Phosphorylation-Regulated Mitochondrial Fission is a Prerequisite for ACP Transport

To elucidate the molecular mechanisms underlying the aforementioned ultrastructural changes, qPCR and western blot analyses were performed to assess changes in Drp1 expression, Drp1 phosphorylation status, and the expression of proteins associated with mitochondrial autophagy. qPCR results showed (Figure 2A) that Drp1 mRNA levels in the sh-Drp1 group were significantly lower than those in the EV group (P < 0.0001, vs EV). Western blot analysis (Figure 2B, 2C) further confirmed that total Drp1 protein levels in the sh-Drp1 group were significantly lower than those in the EV group (P < 0.0001, vs EV), validating the knockdown efficiency; treatment with Mdivi-1 (P = 0.646, vs EV) and CsA (P = 0.208, vs EV) had no significant effect on total Drp1 protein levels. Following osteogenic induction, the p-Drp1 (Ser616)/Drp1 ratio in the EV group was significantly higher than that in the non-induced EV-OI group (P = 0.0003), while the p-Drp1 (Ser637)/Drp1 ratio was significantly lower (P < 0.0001) (Figure 2D, 2E). Concurrently, Pink1 (P < 0.0001) and Parkin (P < 0.0001) protein expression in the EV group was significantly higher than in the EV-OI group, while p62 protein levels were significantly reduced (P < 0.0001) (Figure 2F–2H), indicating that osteogenesis induction activated the PINK1/Parkin-mediated mitochondrial autophagy pathway. Compared with the EV group, the sh-Drp1 group showed further increases in Pink1 (P < 0.0001, vs EV) and Parkin (P < 0.0001, vs EV), while p62 levels were significantly elevated (P < 0.0001, vs EV). In the Mdivi-1-treated group, the p-Drp1 (Ser616)/Drp1 ratio were significantly reduced (P = 0.0127, vs EV), while the p-Drp1 (Ser637)/Drp1 ratio (p < 0.0001, vs EV) was significantly higher than in the EV group. Pink1(p = 0.0002, vs EV) and Parkin (P = 0.0005, vs EV) expression increased, and p62 levels were elevated (P < 0.0001, vs EV). In the CsA-treated group, the p-Drp1 (Ser637)/Drp1 ratio was significantly higher than that in the EV group (P < 0.0001, vs EV), while the p-Drp1 (Ser616)/Drp1 ratio (P > 0.999, vs EV) did not exhibit a significant change in the EV group. Pink1 (P < 0.0001, vs EV) and Parkin (P < 0.0001, vs EV) expression decreased, and p62 levels increased (P = 0.0003, vs EV). These results indicate that osteogenic induction activates mitochondrial fission by promoting Drp1 Ser616 phosphorylation and inhibiting Ser637 phosphorylation, thereby initiating PINK1/Parkin-mediated mitochondrial autophagy. Mdivi-1 and CsA affect mitochondrial fission and subsequent autophagy by interfering with the active state of Drp1, leading to impaired ACP transport.

Figure 2.

Drp1 phosphorylation and expression of mitochondrial autophagy-related mRNA and protein. (A) Relative expression levels of Drp1 mRNA detected by real-time quantitative PCR. Data are presented as mean ± standard deviation, n = 3. (B) Representative western blot bands. (C) Relative expression levels of Drp1 protein. (D) p-Drp1 (Ser616)/Drp1 ratio, n = 3. (E) Relative ratio of p-Drp1 (Ser637)/Drp1, n = 3. (F) Relative expression of Pink1 protein, n = 3. (G) Relative expression of Parkin protein, n = 3. (H) Relative expression of p62 protein, n = 3. Normalized against GAPDH as an internal control; relative ratios calculated with the EV group as the reference. * P < 0.05, ** P < 0.01, **** P < 0.0001 vs EV group.

Effects of Drp1-Mediated Mitochondrial Fission on the Osteogenic Differentiation Transcriptional Program and Mineralization Function

To clarify the effects of impaired mitochondrial fission on the osteogenic differentiation transcriptional program and mineralization function, real-time quantitative PCR, western blot, and alizarin red staining were performed. qPCR results showed (Figure 3A) that, compared with the EV group, RUNX2 and ALPL mRNA expression were significantly reduced in the sh-Drp1 group (decreasing to 0.50-fold and 0.71-fold, respectively; P < 0.01 and P < 0.05); in the Mdivi-1 group, they decreased to 0.66-fold and 0.80-fold, respectively (P < 0.01 and P < 0.05); in the CsA group, only RUNX2 decreased to 0.84-fold (P < 0.05), while ALPL showed no significant change (P > 0.05). Western blot analysis (Figure 3C, 3D) showed that, compared with the EV group, protein levels of RUNX2 and ALPL were significantly reduced in the sh-Drp1 group (decreasing to 0.33-fold and 0.22-fold, respectively, P < 0.001); in the Mdivi-1 group, they decreased to 0.36-fold and 0.20-fold, respectively (P < 0.001); and in the CsA group, they decreased to 0.37-fold and 0.20-fold, respectively (P < 0.001). Alizarin red staining semi-quantitative analysis (Figure 3B) showed that, with the EV group’s mineralization area set as 100%, the relative mineralization area in the sh-Drp1 group decreased to 36.7% (P < 0.01), the Mdivi-1 group decreased to 46.7% (P < 0.01), and the CsA group decreased to 50.1% (P < 0.05). These results are consistent with the trends observed at the protein level, further confirming that Drp1-mediated mitochondrial fission is essential for osteogenic differentiation and extracellular matrix mineralization.

Figure 3.

Effects of Drp1 knockdown and pharmacological inhibition on osteogenic differentiation and mineralization of hMSCs. (A) Relative mRNA expression levels of ALPL and RUNX2. Data are presented as mean ± standard deviation, n = 3. (B) Representative microplate images of eosin-stained cells and semi-quantitative analysis of mineralized area in each group, normalized to 100% for the EV group. Data are presented as mean ± standard deviation, n = 3. (C) Quantitative analysis of relative ALPL and RUNX2 protein expression. Data are presented as mean ± standard deviation, n = 3. (D) Representative western blot bands. GAPDH was used as an internal control. * P < 0.05, ** P < 0.01, **** P < 0.0001 vs EV group.

Effects of Inhibited Mitochondrial Fission on Skeletal Mineralization in Zebrafish

To validate the role of Drp1-mediated mitochondrial fission in skeletal mineralization at the in vivo level, a zebrafish larval model was used for alizarin red staining analysis and detection of osteogenesis-related gene expression. Healthy embryos at 24 hours post-fertilization (24 hpf) were selected and treated continuously with 1.82 mg/L Mdivi-1 or 1.2 mg/L CsA until 9 days post-fertilization (9 dpf). Skeletal mineralization in the craniofacial region and spine was observed, and mRNA expression levels of the key osteogenic transcription factor Runx2 and alkaline phosphatase (ALPL) were detected. In the control group, 9-day-post-fertilization (9 dpf) larvae exhibited distinct orange-red calcium salt deposits in the craniofacial and spinal bones, with well-defined mineralization boundaries, indicating a normal skeletal mineralization process at this developmental stage (Figure 4C, 4D). Semi-quantitative analysis results showed that, with the mineralization area of the control group set as 100%, the relative mineralization area of the craniofacial region in the Mdivi-1-treated group decreased to 81.7 ± 2.9% (P < 0.0001, vs Control, Figure 4A), and the staining intensity of the spine was also slightly reduced. In the CsA-treated group, the relative mineralization area of the craniofacial region decreased to 85.7 ± 3.2% (P < 0.0001, vs Control, Figure 4A). qPCR results showed (Figure 4B) that, compared with the control group: there was no statistically significant difference in ALPL mRNA expression levels in the Mdivi-1-treated group (p = 0.0516, vs Control, marginally significant), while Runx2 mRNA expression levels were reduced (p = 0.0469, vs Control); and in the CsA-treated group, there were no statistically significant differences in the expression levels of ALPL mRNA (P = 0.07, vs Control) or Runx2 mRNA (P = 0.169, vs Control).

Figure 4.

Alizarin red staining and qPCR results in zebrafish larvae. (A) Semi-quantitative analysis of mineralized area, normalized to 100% for the control group. Data are presented as mean ± standard deviation, n = 10. P < 0.05, P < 0.01 vs control. (B) Relative mRNA expression levels of ALPL and Runx2 (qPCR). Data are presented as mean ± standard deviation, n = 3. *P < 0.05, vs control; ns indicates no significant difference (P > 0.05). (C) Representative image of whole-body alizarin red staining in 9-day-post-fertilization (dpf) larvae. (D) Enlarged image of the craniofacial mineralization region.

These changes in gene expression were generally consistent with the reduced mineralization phenotype observed in alizarin red staining. These results indicate that both Mdivi-1 and CsA treatment can inhibit early skeletal mineralization and osteogenic gene expression in zebrafish to varying degrees. Among them, Mdivi-1 exhibited a more significant inhibitory effect on mineralization area compared to CsA, while CsA had a stronger inhibitory effect on Runx2 expression.

Discussion

This study found impaired osteogenesis at both the cellular and animal model levels following inhibition of mitochondrial fission. At the cellular level, total DRP1 expression in hMSCs remained relatively stable, except in the sh-DRP1 group. This is because DRP1, as a GTPase primarily localized in the cytoplasm, maintains relatively constant expression levels within cells; the regulation of mitochondrial fission does not depend on changes in total DRP1 levels, but rather is achieved through its recruitment to the target membrane and functional activation [4]. Upon mitochondrial activation, DRP1 assembles into a helical ring, tightens, and hydrolyzes GTP, leading to mitochondrial fission. After fission, the DRP1 ring disassembles, merges, and is released back into the cytoplasm, awaiting the next recruitment [5]. Therefore, alterations in DRP1 expression occur only under certain pathological or aging conditions [6]. In the present study, the use of Mdivi-1 resulted in only a slight decrease in DRP1 levels, yet this was sufficient to effectively inhibit mitochondrial fission, further confirming that DRP1’s function does not depend on its total expression level. Mdivi-1 primarily inhibits DRP1’s GTPase activity, thereby preventing its translocation from the cytoplasm to the mitochondria, which functionally achieves the effect of inhibiting mitochondrial fission. During osteogenesis induction, this study provides direct evidence of DRP1 functional activation: the phosphorylation level at Ser616 (active form) was significantly elevated, while the phosphorylation level at Ser637 (inactive form) was markedly reduced [7]. This activation pattern closely aligns with the expression timing of early osteogenic markers RUNX2 and ALP, and is consistent with the trend of early mitochondrial fragmentation during osteogenic differentiation reported by Suh et al suggesting that DRP1 activation is a key early event in the initiation of osteogenic differentiation [8]. It is worth noting that the timing of DRP1 activation coincides precisely with the active phase of mitochondrial fragmentation, a period that is a critical window for the transition of cellular metabolism from glycolysis to oxidative phosphorylation. Previous studies have shown that on day 7 of osteogenic induction, there is a surge in intracellular metabolites associated with energy metabolism and matrix synthesis [9]. Day 7 is regarded as a critical juncture in cell differentiation, as the energy metabolism state at this time provides the necessary energy and metabolic reserves for the extensive bone matrix synthesis and mineralization occurring between days 7 and 21 [10]. During MSC osteogenic differentiation, the metabolic pattern shifts from glycolysis (anaerobic metabolism) to oxidative phosphorylation (aerobic metabolism). The latter is characterized by an increased FAD/NAD(P)H redox ratio and elevated protein-bound NADH levels to meet the high demand for ATP. Inhibition of DRP1 resulted in a significant decrease in ATP production, further confirming that mitochondrial fission plays a central role in metabolic remodeling. These results indicate that Drp1-mediated mitochondrial dynamics are not merely a concomitant phenomenon of differentiation initiation but a necessary condition for it.

In the zebrafish model, there was a tendency towards reduced levels of Runx2 and ALP following treatment with CsA or Mdivi-1. However, no statistically significant difference was observed in the CsA group. It is acknowledged that several factors may account for the lack of statistical significance in certain comparisons. Firstly, gene regulation in vivo is subject to compensatory mechanisms that can buffer primary molecular changes, thereby attenuating expression differences when compared to in vitro conditions. Secondly, CsA and Mdivi-1 do not directly target runx2 or alp transcription; rather, both act by inhibiting DRP1 activation and downstream mitochondrial fission and mitophagy. Consequently, this finding serves to reinforce the hypothesis that DRP1 activation promotes calcification by mediating ACP transport, rather than by directly regulating the transcription of osteogenic genes. However, it is important to note that larger cohorts are required for validation.

Through electron microscopy and energy-filtered spectroscopy analysis, this study found that early-stage osteogenic MSCs contain a large number of amorphous calcium phosphate (ACP) particles, which are often colocalized with mitochondria in a fission state; by day 14, these ACP particles migrate into the extracellular matrix and complete mineralization. In contrast, in DRP1-knockdown cells, mitochondria exhibited an abnormally fused, elongated state, with large amounts of electron-dense material (ACP) accumulating within them, making effective clearance difficult. This phenomenon suggests that Drp1-mediated mitochondrial fission performs a critical “breaking down the whole into parts” function: physically separating ACP-laden mitochondrial fragments from healthy ones, thereby reducing the former to a size suitable for engulfment by autophagosomes. Research by Yamashita et al confirmed that a diameter of less than 600 nm is the critical size for effective phagosome engulfment and degradation of mitochondrial fragments; this is because the limited opening of the phagosome membrane naturally restricts the size of ingestible fragments during the contraction and engulfment process, ensuring that only appropriately sized fragments are ingested and degraded [11]. This mechanism allows cells to excise only the small segment of the mitochondria bearing damage or containing ACP, without disrupting the entire mitochondrial network. In this study, the post-fission diameter of mitochondria and the average diameter of autolysosomes measured by electron microscopy were both less than 500 nm, which precisely validates this mechanism. However, it should be noted that these measurements were obtained from a limited number of organelles and thus represent preliminary observations. To establish the size distribution with greater precision and to validate the proposed size-dependent mitophagy model, larger-scale quantification is required. Furthermore, the accumulation of ACP within elongated mitochondria observed following DRP1 inhibition, accompanied by a scarcity of autophagosome envelopment, similarly confirms the indispensability of the physical mitochondrial separation process. Therefore, we propose that DRP1’s cleavage function provides conformally suitable substrates for mitochondrial autophagy and serves as a key prerequisite for the extracellular transport of ACP from mitochondria.

DRP1 and the PINK1/Parkin Pathway Synergistically Constitute a Complete Mitochondrial Quality Control Network

PINK1 is a serine/threonine kinase with a mitochondrial targeting sequence and transmembrane domains, while Parkin is a cytoplasmic ubiquitin E3 ligase; both play synergistic roles in mitochondrial quality control [12]. This study found that in DRP1-inhibited cells, PINK1 accumulates, but downstream autophagic clearance is impaired, as evidenced by elevated P62 levels. This phenomenon reveals an important principle: the accumulation of PINK1 merely serves as a molecular marker for mitochondria overloaded with ACP, whereas the physical fragmentation of mitochondria is a necessary prerequisite for effective autophagosome engulfment. Previous studies have shown that calcium ions can activate the kinase CAMK1, which phosphorylates PINK1 (Ser228) to activate the PINK1/Parkin signaling pathway [13], thereby precisely molecularly marking the overloaded regions and guiding autophagosome recognition. In this synergistic process, DRP1 is responsible for physical fragmentation, providing autophagosomes with substrates of appropriate size; both are indispensable and together constitute the complete ACP transport mechanism. This finding is consistent with the research by Han et al, who reported that PINK1 can phosphorylate the Ser616 site of DRP1, promoting its translocation to mitochondria [14]. On the other hand, Ser637 acts as a “braking” site; its phosphorylation inhibits DRP1’s GTPase activity and prevents its binding to mitochondrial receptors, thereby suppressing mitochondrial fission [15,16]. This site can be dephosphorylated by calmodulin: when intracellular calcium ion concentrations rise, Ca²⁺-dependent serine/threonine protein kinases are activated, removing the inhibitory phosphate group added by PKA at Ser637, thereby lifting the inhibition on DRP1 [17]. In this study, treatment with calmodulin-dependent protein kinase (CDPK) inhibitors resulted in increased Ser637 phosphorylation levels, which was corroborated by impaired bone formation. To interpret these static and endpoint observations, a working hypothesis of a “fission-clearance-regeneration” cycle is proposed, in which the obtained data is integrated into a coherent sequence. The hypothesized cycle is outlined as follows: (1) DRP1-mediated fission separates fragments of aged or ACP-laden mitochondria; (2) Small mitochondrial fragments are marked by PINK1 and cleared via mitochondrial autophagy; (3) ACP is released into autophagosomes and accumulates at extracellular mineralization sites; (4) Healthy mitochondrial fragments regenerate through fission, providing new mitochondria for the next cycle of ACP. This cycle ingeniously reconciles 2 seemingly conflicting demands in osteogenic differentiation: on the one hand, the need for extensive mitochondrial consumption for mineral transport; on the other hand, the necessity to maintain sufficient mitochondrial function to support continuous energy supply. DRP1 acts as the switch for this cycle; its functional deficiency disrupts the cycle, manifesting as impaired mineral transport, disrupted energy metabolism, and failed osteogenic differentiation.

Limitations of the Study

Firstly, the pharmacological inhibitors utilized are known to possess off-target effects, which necessitate a cautious interpretation of the results. Mdivi-1, utilized in this study to inhibit DRP1, is also a reversible inhibitor of mitochondrial complex I, and this may affect cellular metabolism independently of mitochondrial fission [18]. CsA, the agent employed to inhibit calcineurin in the present study, has also been shown to inhibit cyclophilin D and modulate the opening of the mitochondrial permeability transition pore (mPTP). Consequently, the phenotypes observed with CsA may partly reflect changes in mitochondrial permeability beyond calcineurin inhibition [19]. Secondly, the zebrafish model does not fully recapitulate mammalian bone remodeling. Thirdly, the proposed “fission-clearance-regeneration” cycle was inferred from static imaging and endpoint assays; real-time dynamic tracking is lacking.

Conclusions

In contrast to previous studies, which chiefly concentrated on the function of mitochondrial metabolic remodeling in stem cell differentiation, this study combined 3 processes—mitochondrial dynamics, mitochondrial autophagy, and mineral transport—to demonstrate the crucial role of DRP1 as a central hub in biomineralization. The “fission-clearance-regeneration” model proposed here should be regarded as a working hypothesis that provides a novel perspective on the cellular basis of osteogenic differentiation in mesenchymal stem cells and suggests potential new avenues for interventions in bone metabolic diseases such as osteoporosis. By targeting mitochondrial dynamics, particularly the activation state of DRP1, it may be possible to effectively regulate bone regeneration.