The guinea pigs used in this study were obtained from Keda (Tianjin, China) and housed in a standard laboratory setting under a 12-hour light/dark cycle and provided with a standard diet. All experimental procedures were conducted with strict adherence to international ethical guidelines for the use of laboratory animals, and the study also received approval from the Medical Laboratory Animal Ethics Committee of Tianjin Medical University Eye Hospital (TJYY2024120244).

For the purposes of this study, we used healthy 3-week-old male guinea pigs. Pre-experimentally, those with ocular diseases or anisometropia $>$1.50D were excluded after screening and the remaining animals were randomly assigned to either a control (CON) or form-deprived myopia (FDM) group. Whereas animals in the CON group received no ocular treatment, the right eyes of guinea pigs in the FDM group were completely covered with a white latex balloon helmet (the left eye, ears, nose, and mouth remained exposed), with daily inspections being conducted to ensure effective coverage. All animals were housed under standard conditions (12 h light/dark cycle) and subjected to biweekly refraction and axial length measurements. After 6 weeks, the scleral tissues were collected for histopathological and related biological analyses.

Axial length, defined as the distance from the anterior corneal apex to the anterior retinal surface, was measured using an A-mode ultrasonic ophthalmoscope (A-scan, ODM-1000A; Tianjin Medatech Co., Ltd., Tianjin, China). Prior to measurement, a drop of 0.5% Proparacaine Hydrochloride Eye Drops (Alcon, Mechelen, Belgium) was instilled into the conjunctival sac of the guinea pigs for topical anesthesia. Subsequently, the probe of the ophthalmoscope was placed vertically on the apex of the guinea pig cornea and aligned with the center of the pupil for measurement, thereby obtaining stable waveform data. For each eye, we performed five measurements, with average values being recorded.

Refractive error was measured using an infrared auto-refractometer (Photorefractor; Global Biotech Inc., Suzhou, Jiangsu, China). In a dark room, 0.5% Compound Tropicamide Eye Drops (Santen Pharmaceutical (China) Co., Ltd., Suzhou, Jiangsu, China) were administered to both eyes of the guinea pigs for mydriasis four times at 5-min intervals to paralyze the ciliary muscle. After 30 min of dark adaptation, by which time the pupils had become fully dilated, the refractive status of guinea pig eyeballs was determined using an infrared eccentric refractometer, with three measurements being performed for each eye and average values recorded.

Euthanasia was performed via an anesthetic overdose. Guinea pigs were placed in an induction chamber and anesthetized with 5% isoflurane in oxygen until losing consciousness. A deep plane of anesthesia was confirmed by the absence of a response to a firm toe pinch. Subsequently, euthanasia was achieved by the intraperitoneal administration of pentobarbital sodium (150 mg/kg) at a concentration of 30 mg/mL (3%) or continuous exposure to 5% isoflurane for at least 10 min after a cessation of breathing. Death was confirmed by the absence of spontaneous respiration, loss of the corneal reflex, or asystole upon cardiac monitoring.

To assess pathological scleral changes in response to form deprivation, we performed staining using hematoxylin and eosin (H&E). On the 42nd day of the experiment, the guinea pigs had been rendered unconscious, and having immediately excised the eyeballs, these were fixed in FAS fixative (Servicebio, Proteintech, Wuhan, Hubei, China) for 48 h. The fixed tissues were paraffin-embedded, serially sectioned (4 µm), deparaffinized with xylene, and stained with H&E for pathological evaluation.

To assess the morphology of mitochondria in scleral tissue, we performed transmission electron microscopy (TEM). Scleral tissue were fixed in 2.5% glutaraldehyde for 18–20 h, followed by fixation with 1% osmium tetroxide fixation for 1 hour dehydration, and embedding in epoxy resin. Ultra-thin sections (60 nm thick) were cut, stained with 2% uranyl acetate and lead citrate, and examined using a transmission electron microscope (Hitachi, Minato-ku, Tokyo, Japan).

To assess changes in the expression of molecular markers associated with mitochondrial division, we conducted immunofluorescence (IF) analysis. Eye tissue sections (prepared as described) were blocked with 10% goat serum (50 min, room temperature), and incubated overnight at 4 °C with rabbit anti-DRP1 primary antibody (1:250 dilution: 12957-1-AP, Proteintech, China). Following incubation, the sections were treated with peroxidase-conjugated secondary antibodies (1 h at room temperature) for complete immunolabeling.

Human scleral fibroblast (HSFs) were obtained from Guangzhou Keyun Selection Biotechnology Co., Ltd. (Guangzhou, China) and cultured in DMEM (Gibco, Grand Island, NY, USA) supplemented with 10% FBS (A5669701, Gibco) and 1% penicillin-streptomycin (Gibco) at 37 °C with 5% CO₂. Logarithmic-phase cells (passages 3–4) were used for subsequent experiments. The cell line used was validated based on STR profiling and tested negative for Mycoplasma.

Cells at 70%–80% confluence were exposed to hypoxic conditions for 24 consecutive hours in sealed culture bags (Cat. No. C-41, Chiyoda-ku, Tokyo, Japan) with Mitsubishi anaerobic gas-generating packs (cat. No. C-11; Japan) to ensure that oxygen concentrations within the bags were maintained at a level below 1%. An anaerobic indicator (Cat. No. C-22, Japan) was used to monitor the oxygen concentration, with the colorimetric criteria defined as pink for 0.1% O₂, blue for $>$0.5% O₂, and bluish-purple for an indeterminate intermediate range.

Cells were seeded at an appropriate density in the wells of 96-well plates. Upon reaching 70% to 80% confluence, the cell were pre-treated with different concentrations of Mdivi-1 (cat. No.: HY-15767; MCE, Monmouth Junction, NJ, USA). A 1 mM stock solution of Mdivi-1 in dimethyl sulfoxide (DMSO) was stored at –20 °C and serially diluted in medium to final concentrations of 1, 5, 10, and 15 µM. As a solvent control, we used DMSO at a final concentration $\le$0.1% to avoid solvent interference (Supplementary Fig. 1F). After pre-treatment, the plates were subjected to hypoxia for 12, 24, 36, or 48 h in sealed bags containing hypoxic gas generation packs (Supplementary Fig. 1A). At the selected time points, the medium was replaced with CCK-8-containing fresh medium (Cat. No.: CA1210; Solarbio, Beijing, China), and after incubating for 2 h, absorbances were measured at 450 nm using an INFINITE M200 microplate reader (Model) to assess cell viability under different Mdivi-1 concentrations and hypoxic durations.

For short-hairpin RNA (shRNA) treatment, HSFs at 60%–70% confluence were transfected using FectoGene 100 (Beyonpop, Tianjin, China) per manufacturer’s instructions. shRNAs included shHIF-1$\alpha$ (targeting human HIF-1$\alpha$), shNotch1 (targeting Notch1), and negative control shNC. At 48 h post-transfection, cells were treated as designed, samples collected, and HIF-1$\alpha$/Notch1 expression determined via quantitative Real-time polymerase chain reaction (qPCR) to validate the efficiency of knockdown (Supplementary Fig. 1B,C).

HSFs were lysed (20 min) in RIPA buffer (R0010, Solarbio) containing 100 mM phenmethylsulfonyl fluoride (P0100-01; Solarbio), a phosphatase inhibitor (100$\times$: P8991; Solarbio), and a protease inhibitor (100$\times$: P6370; Solarbio). The resulting lysates were centrifuged (12,000 $\times$g, 20 min, 4 °C), and the concentrations of proteins was quantified using a BCA kit (PC0020; Solarbio). Equal amounts of proteins (20–30 µg) were run on 10% SDS-PAGE gels, with the separated proteins subsequently being transferred to 0.45 µm PVDF membranes. Following transfer, the membranes were blocked for 30 min in 1$\times$ protein-free rapid blocking buffer (PS108P; Epizyme, Cambridge, MA, USA), incubated overnight at 4 °C with primary antibodies, washed, and then incubated for 1 h at room temperature with secondary antibodies. Protein bands were developed using ECL Prime (RPN418-1; Amersham, Little Chalfont, Buckinghamshire, United Kingdom) and quantified using ImageJ software (version 1.54r; National Institutes of Health (NIH), Bethesda, Maryland, USA). The antibodies used in this study are listed in Supplementary Table 1.

Total RNA was isolated from HSFs using the EZ-press RNA purification kit (EZBioscience, cat. B0004DP, Roseville, CA, USA), reverse-transcribed into cDNA with the Color Reverse Transcription kit (EZBioscience, CA, USA), and quantified using a Nanodrop2000 spectrophotometer. As a reference gene for normalization of mRNA levels, we used $\beta$-actin. The primers used for amplification were synthesized commercially by Sangong Biotechnology (Beijing, China), the sequences of which are listed in Supplementary Table 2.

Cells were fixed in 4% PFA for 10 min, permeabilized with 0.1% Triton X-100 (T8200, Solarbio) for 10 min, then blocked with 1% BSA (SW3015, Solarbio) + 0.1% Tween-20 (T8220, Solarbio) for 30 min. Cells were incubated overnight at 4 °C with primary antibody ($\alpha$-SMA, rabbit, 1 µg/mL; ab5694, Cambridge, UK). Having subsequently washed with phosphate-buffered saline, the cells were incubated in the dark for 1 h at room temperature with a fluorescent secondary antibody (gb2AF488, Proteintech Sanying) for 1 h at RT in the dark. Target protein distribution was visualized via Zeiss confocal microscope (LSM800, Oberkochen, Oberkochen, Baden-Württemberg, Germany), and fluorescence intensity was quantified with ImageJ.

For mitochondrial morphology analysis, cells were stained with MitoTracker (Thermo Scientific, Waltham, MA, USA) and imaged via confocal microscopy. After selecting cell regions and basic preprocessing, ImageJ was used to calculate mitochondrial morphology parameters (average branch length and standard deviation).

The levels of intracellular ROS were determined using a DCFH-DA fluorescent probe (S0033S; Beyotime, Shanghai, China). Following incubation with 10 µM DCFH-DA, cells were washed three times with serum-free medium to remove residual probe, then counterstained with Hoechst for 1 min. Fluorescence images were obtained using a Zeiss confocal microscope (LSM800, Oberkochen, Germany), and intensity was quantified with ImageJ.

HSF mitochondrial membrane potentials ($\mathrm{\Delta }$$\mathrm{\Psi }$m) were determined using JC-1 fluorescent dye (C2003S; Beyotime, China). Cells were incubated in the dark for 30 min at 37 °C with 25 µg/mL JC-1 working solution washed with buffer, and examined using a Zeiss confocal microscope. JC-1 forms either J-aggregates (red fluorescence, indicating high mitochondrial membrane potential) or J-monomers (green fluorescence, indicating low mitochondrial membrane potential), with the red/green fluorescence ratio quantified using ImageJ being used to assess $\mathrm{\Delta }$$\mathrm{\Psi }$m.

All statistical analyses were performed using GraphPad Prism software (version 9.0; GraphPad Software, USA). The normality of data distribution was initially assessed using the Shapiro–Wilk test. For comparisons between two groups, we used an unpaired two-tailed Student’s t-test. Comparisons among three or more groups were conducted using a one-way analysis of variance (ANOVA), followed by appropriate post hoc tests, as indicated in the respective figure legends. Unless otherwise specified, data are presented as the means $\pm$ SEM. The threshold for statistical significance was set at a two-sided p-value of 0.05. The levels of significance are denoted as ns (not significant), *p 0.05, **p 0.01, ***p 0.001, and ****p 0.0001. The exact sample size (n) for each experiment, which indicated the number of biological replicates, is shown in the corresponding figure legends. All experiments were independently repeated at least three times to ensure reproducibility.

To investigate changes in scleral mitochondria during myopic progression, we successfully established an FDM guinea pig model. Compared with the control group, those in the FDM group were found to be characterized by pronounced myopia-associated symptoms, including an elongated axial length and a shift in refractive power towards myopia (Fig. 1A,B), along with marked thinning of the scleral tissue and a looser arrangement of collagen fibers (Fig. 1C). Further TEM examination revealed a pronounced fission of mitochondria in the sclera of the FDM group predominantly exhibited an active fission state (Fig. 1D, Supplementary Fig. 1G). Moreover, DRP1 expression was markedly elevated in the FDM group compared to the CON group (Fig. 1E,F). Comparatively, we detected no significant differences between the two groups with respect to levels of the mitochondrial fusion proteins MFN1 and MFN2, are which taken to be representative of the status of mitochondrial fusion. Contrastingly, mRNA expression levels in the FDM group were significantly lower levels of these proteins in FDM group guinea pigs than in the CON group (Fig. 1G), indicating corresponding changes at the transcriptomic level. Collectively, these findings indicate an increase in mitochondrial fission in the sclera of myopia.

Fig. 1.

Scleral hypoxia and Mitochondrial fission are active in Myopia. (A,B) Axial length and refractive error of 3-week-old guinea pigs measured at 0, 2, 4, 6 weeks post-form deprivation (n = 10). (C) H&E staining of scleral tissues after 6 weeks of form deprivation (n = 3), scale bar = 100 µm. (D) Mitochondrial morphology was observed under transmission electron microscopy (TEM) at magnifications of $\times$20.0 k (scale bar = 500 nm) and $\times$5.0 k (scale bar = 2 µm), respectively, with the results as follows: (a) In the CON group, intact cristae morphology is observable (red arrows indicate mitochondria); (b) In the FDM group, mitochondrial cristae were almost completely absent, and mitochondria appeared fragmented (red arrows indicate mitochondria, n = 3). (E) Scleral tissues were stained with DRP1 antibody (Alexa Fluor 488, green) to observe DRP1 expression (n = 3), scale bar = 20 µm. (F) Western blot and qPCR were used to analyze the protein and mRNA levels of DRP1 in scleral tissues between CON and FDM groups (n = 3). (G) Western blot was analysis of MFN1, MFN2 in CON/FDM groups (n = 6) and qPCR was analysis of MFN1, MFN2 in CON/FDM groups (n = 3). Bar chart data: mean $\pm$ SEM. Significance: ns, not significant, *p 0.05, **p 0.01, ****p 0.0001. Statistics: unpaired two-tailed Student’s t-test (C–G); one-way ANOVA with Tukey’s post hoc test (A,B). H&E, hematoxylin and eosin; CON, control; FDM, form-deprived myopia; DRP1, dynamin-related protein 1; MFN, mitochondrial fusion protein; SEM, Scanning Electron Microscopy; ANOVA, analysis of variance.

To establish the potential causes of mitochondrial fragmentation in myopic sclera, we examined the expression of HIF-1$\alpha$, and accordingly detected the elevated expression of this marker in the FDM (Fig. 2A), indicating the occurrence of hypoxia in scleral tissue during the progression of myopia. Further culture of hypoxic HSFs revealed an increase in the expression of HIF-1$\alpha$ following exposure to hypoxia for 24 h (Fig. 2B, Supplementary Fig. 1E), accompanied by mitochondrial fragmentation (Fig. 2C). Moreover, there was a close correlation between excessive mitochondrial fission and alterations in mitochondrial membrane potential ($\mathrm{\Delta }$$\mathrm{\Psi }$m). Consequently, we employed the JC-1 fluorescent probe to assess $\mathrm{\Delta }$$\mathrm{\Psi }$m. Under hypoxia conditions, we detected markedly elevated levels of intracellular ROS levels were markedly elevated in HSFs (Fig. 2D). Concomitantly, JC-1 predominantly exhibited green fluorescence in these cells (Fig. 2E), a pattern indicative of monomeric JC-1 formation and reflective of a significant reduction in mitochondrial membrane potential ($\mathrm{\Delta }$$\mathrm{\Psi }$m). Moreover, we observed increases in the levels of DRP1 protein, whereas at the protein level, there were no significant differences between MFN1 and MFN2 at the protein level (Fig. 2F). Conversely, at the mRNA level, HSFs exposed to hypoxia were found to be characterized by an elevated expression of DRP1 and reduction in the expression of MFN1 and MFN2 (Fig. 2G). These findings thus indicate that in HSFs, hypoxia induces an excessive DRP1-mediated fission of mitochondria, accompanied by the accumulation of ROS.

Fig. 2.

Activation of HIF-1$\alpha$ and promotion of mitochondrial fission in the effects of scleral hypoxia. (A) Western blot and qPCR were used to analyze the protein and mRNA levels of HIF-1$\alpha$ in scleral tissues between CON and FDM groups (n = 3). (B) Western blot and qPCR were used to analyze the protein and mRNA levels of HIF-1$\alpha$ in HSFs after 24 h of hypoxia (n = 3). (C) HSFs cultured under hypoxia for 24 h; MitoTracker red staining for mitochondria (n = 3), with average branch length analyzed, scale bar = 20 µm or 10 µm. (D) DCFH-DA probe detected ROS content in hypoxic HSFs (n = 5), scale bar = 100 µm. (E) JC-1 staining measured mitochondrial membrane potential in HSFs post-24 h hypoxia (n = 3), scale bar = 100 µm. (F) Western blot analyzed MFN1, MFN2, and DRP1 expression at the protein levels in hypoxic HSFs (n = 3). (G) qPCR analyzed MFN1, MFN2, and DRP1 expression at the mRNA levels in hypoxic HSFs (n = 3). Bar chart data: mean $\pm$ SEM. Significance: ns, not significant, *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001. All data were analyzed via an unpaired two-tailed Student’s t-test. HIF-1$\alpha$, hypoxia-inducible factor alpha; HSFs, human scleral fibroblasts; DCFH-DA, 2’,7’-dichlorodihydrofluorescein diacetate; ROS, reactive oxygen species; JC-1, 5,5’,6,6’-tetrachloro-1,1’,3,3’-tetraethylbenzimidazolcarbocyanine iodide.

To investigate the effects of scleral hypoxia and mitochondrial fission. In the FDM group, we observed significant reductions in the expression of type I collagen, a key component involved in the synthesis of scleral collagen, and corresponding increases in the expression of $\alpha$-SMA, a marker specific for FMT (Fig. 3A,B). Similarly, in HSFs after 24 hours of hypoxic culture. We also observed a decrease in type I collagen and an increase in $\alpha$-SMA (Fig. 3C–E). These findings thus provide evidence to indicate that hypoxia induces phenotypic transformation in HSFs and disrupts collagen metabolism, consistent with the elevated levels of FMT detected the myopic sclera.

Fig. 3.

Scleral hypoxia-induced FMT increase. (A) Western blot analysis of Col1$\alpha$1 and $\alpha$-SMA in scleral tissues between CON and FDM groups (n = 3). (B) Q-PCR analysis of Col1$\alpha$1 and $\alpha$-SMA in scleral tissues between CON and FDM groups (n = 3). (C) Q-PCR analysis of Col1$\alpha$1 and $\alpha$-SMA in HSFs after 24 h of hypoxia (n = 3). (D) Western blot detected Col1$\alpha$1 and $\alpha$-SMA protein levels in HSFs after 24 h of hypoxia (n = 3). (E) HSFs cultured under hypoxia for 24 h; $\alpha$-SMA expression visualized by immunofluorescence staining (Alexa Fluor 488, green) (n = 3), scale bar = 50 µm. Bar chart data: mean $\pm$ SEM. Significance: *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001. All data were analyzed via an unpaired two-tailed Student’s t-test. FMT, fibroblast-myofibroblast transition; Col1$\alpha$1, collagen type I alpha 1 chain; $\alpha$-SMA, $\alpha$-smooth muscle actin.

The Notch signaling pathway has been established to play a pivotal role in the regulation of mitochondrial dynamics, and we accordingly sought to determine whether Notch1 is implicated in the mitochondrial fission induced by scleral hypoxia. In the FDM group, we detected elevated levels of Notch1 expression (Fig. 4A), and similar increases in Notch1 expression were observed in HSFs exposed to hypoxia (Fig. 4B). Contrastingly, reductions in Notch1 levels were observed following the shRNA transfection-mediated suppression HIF-1$\alpha$ expression (Fig. 4C), thereby providing evidence of an association between HIF-1$\alpha$ and the expression of Notch1. On the basis of these observations, we subsequently sought to assess the role of Notch1 in mitochondrial fission and fusion, and accordingly found that the shRNA-mediated suppression of Notch1 expression was associated with a significant inhibition of DRP1 activity, whereas the levels of MFN1 and MFN2 were markedly increased at both the protein and mRNA levels (Fig. 4D). Concurrently, a shift toward red fluorescence indicated restoration of mitochondrial membrane potential to normal levels (Fig. 4E). These findings thus indicate that scleral hypoxia-induced mitochondrial fission is promoted via a regulation of Notch1 expression.

Fig. 4.

HIF-1$\alpha$ regulates mitochondrial dynamics through Notch1. (A) Western blot and qPCR were used to analyze the protein and mRNA levels of Notch1 in scleral tissues between CON and FDM groups (n = 3). (B) Western blot and qPCR were used to analyze the protein and mRNA levels of Notch1 in HSFs after 24 h of hypoxia (n = 3). (C) Western blot analysis of Notch1 and HIF-1$\alpha$ protein levels in HSFs transfected with sh-NC or shHIF-1$\alpha$ for 48 h (n = 3). (D) Western blot analysis of Notch1, DRP1, MFN1, and MFN2 protein levels in HSFs transfected with sh-NC or sh-Notch1 for 48 h (n = 3). (E) Mitochondrial membrane potential measured via JC-1 staining in HSFs transfected with sh-NC or sh-Notch1 for 48 h (n = 3), scale bar = 100 µm. Bar chart data: mean $\pm$ SEM. Significance: ns, not significant, *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001. All data were analyzed via an unpaired two-tailed Student’s t-test.

As an intervention for myopia, to evaluate the therapeutic potential of targeting mitochondrial fission to intervene in myopia, we examined the effects of inhibiting DRP1-dependent mitochondrial fission on extracellular matrix remodeling in HSFs. The results demonstrated that treatment with mitochondrial fission inhibitor 1 (Mdivi-1) restored mitochondrial morphology in HSFs to a typical filamentous network structure (Fig. 5A). Consistently, JC-1 fluorescence staining enhanced red fluorescence, indicating restoration of mitochondrial membrane potential (Fig. 5B). Furthermore, the intracellular levels of ROS showed no significant difference from those detected in the negative control group (Fig. 5C). Further analysis of ECM-associated markers revealed that Mdivi-1 treatment contributed to a significant upregulation of type I collagen expression and markedly suppressed expression of $\alpha$-SMA compared with the CON group (Fig. 5D,E; Supplementary Fig. 1D). These findings indicate that Mdivi-1 can effectively suppress the trans-differentiation of scleral fibroblasts to myofibroblasts and promote remodeling of the scleral extracellular matrix.

Fig. 5.

Inhibition of mitochondrial fission suppresses FMT. (A) HSFs treated with negative control or Mdivi-1 (15 µM, Supplementary Fig. 1F) under hypoxia for 24 h; mitochondria labeled with MitoTracker Red, with length quantified (n = 3), scale bar = 20 µm or 10 µm. (B) Mitochondrial membrane potential measured via JC-1 staining in HSFs treated with Mdivi-1 (15 µM) under hypoxia for 24 h (n = 4), scale bar = 50 µm. (C) ROS content detected via DCFH-DA labeling in HSFs treated with Mdivi-1 (15 µM 24 h) under hypoxia (n = 3), scale bar = 100 µm. (D) $\alpha$-SMA expression determined by IF in HSFs treated with Mdivi-1 (15 µM) under hypoxia for 24 h (n = 3), scale bar = 50 µm. (E) Western blot analysis of Col1$\alpha$1, DRP1, and $\alpha$-SMA protein levels in HSFs treated with Mdivi-1 (15 µM) under hypoxia for 24 h (Col1α1, n  = 3; DRP1 and α‑SMA, n  = 4). Bar chart data: mean $\pm$ SEM. Significance: ns, not significant, *p 0.05, ***p 0.001, ****p 0.0001. Statistics: unpaired two-tailed Student’s t-test (A–D); one-way ANOVA with Tukey’s post hoc test (E). IF, immunofluorescence.