Animals

The 2-week-old male colored guinea pigs used in this study were obtained from Danyang Changyi Experimental Animal Breeding Co., Ltd. (Danyang, China; license no: SYXK (Shanghai) 2018-0040). The animals were kept in the Animal Experiment Center of Yueyang Hospital of Integrated Traditional Chinese and Western Medicine, Shanghai University of Traditional Chinese Medicine. This study was approved by the Laboratory Animal Ethics Committee of Yueyang Hospital of Integrated Traditional Chinese and Western Medicine, Shanghai University of Traditional Chinese Medicine (No. YYLAC-2022-158-1). The experiment followed the statement of the Association for Research in Vision and Ophthalmology (ARVO) regarding the use of animals in research. The experimental animals were housed with a diurnal cycle of 12 h, temperature of 22–24 °C, relative humidity of 40–60%, and brightness of approximately 200 lx. A free diet was provided.

Induction of the FDM guinea pig model

The guinea pigs were divided into normal control group (NCG), FDM group (FDMG) and electroacupuncture group (EAG), with 21 guinea pigs in each group. All eyes in the NCG were untreated. According to previous studies, the right eyes of both the FDMG and the EAG were covered with a latex balloon as a translucent blindfold for 8 weeks to induce FDM guinea pig model, and the opposite eyes were not covered as a control [31]. Throughout the experiment, we ensured that the right eye was completely covered but could blink freely. After 8 weeks of form deprivation, the refraction and axial Length (AL) of guinea pigs were measured to determine whether the model of FDM was successfully established.

EA treatment

The guinea pigs of EAG were treated by EA stimulation at bilateral acupoints Fengchi (GB20) and Taiyang (EX-HN5) for 30 min once daily for 4 consecutive weeks in the fifth week of modeling. The movement of the guinea pigs was restricted with a guinea pig immobilization device and their head and eyes were fully exposed for EA intervention based on previous experimental studies in animals [28]. Acupuncture needles and acupoints were wiped with 75% alcohol before EA. Stainless steel acupuncture needles (13 mm × 0.25 mm; Suzhou Medical Appliance Factory, Suzhou, Jiangsu Province, China) were inserted at Fengchi (GB20) and Taiyang (EX-HN5) bilaterally to a depth of 2–3 mm. The handle of the needle was further connected to an electrical stimulator, whose parameters were set as 30 min of EA treatment, continuous electrical pulses, 2 Hz in frequency and 2 mA in intensity. All guinea pigs were awake during the EA treatment. Acupoint localization was referenced to previous studies and published ‘‘Experimental Acupuncturology’’ [28, 32].

Assessment of refraction and axial length (AL)

Guinea pigs were treated with 0.5% compound tropicamide solution (Santen Pharmaceutical Co., Ltd., Shiga Plant., China) for mydriasis, and refraction was detected in the dark with a streak retinoscope (YZ24; 66 Vision Tech Co., Ltd., China) [33]. We take the average refractive index of the horizontal and vertical meridiens as the final spherical equivalent (SE) refractive index. Measurements were made three times for each eye and averaged. We used A-scan ultrasonography (KN-1800; Kangning, Jiangsu Province, China) to measure the axial length (AL) of guinea pigs [34]. One drop of 0.4% oxytetracaine hydrochloride (Benoxil; Santen, Osaka, Japan) was used to each guinea pig eye before testing, after which ultrasound measurements were performed. The AL of each guinea pig eye was measured ten times in duplicate to calculate the mean value.

Hematoxylin and eosin (H&E) staining

The pathological morphology of the eyeballs of guinea pigs and humans was observed via H&E staining. The eye tissue was fixed in 4% paraformaldehyde, subjected to gradient dehydration, paraffin embedding, slicing, and drying at 60 °C on a slide, after which HE staining was performed. Finally, the eye sections were sealed and then viewed with an optical microscope (Olympus Corporation, Japan).

Isolation and culture of RGCs and fluorescence identification

RGCs were extracted from guinea pig eyeballs as described previously [35, 36]. The retina was isolated from guinea pig eyeballs and crushed with ophthalmic scissors in 4 °C ice-cold Hank’s balanced salt solution (HBSS, Invitrogen). The isolated retinal fragments were digested in a solution containing crude collagenase (Sigma), papain (Sigma), 0.02% bovine serum albumin (BSA) (Sigma), and l-cysteine for 30 min at 37 °C. The digested retinal fragments were transferred to OV-1 solution and ovomucoid (Sigma) and ground, and a single-cell suspension was obtained through a 20 µm aperture screen. Culture dishes were pretreated with goat anti-rabbit IgG (HL) antibodies, and the obtained single-cell suspensions were transferred to pretreated culture dishes overnight. The cells were again transferred to culture dishes containing goat anti-mouse IgG + IgM (HL) antibodies. The liquid in the culture dish was then discarded, and the iCell primary neuronal cell culture system (iCell Bioscience, Inc.) was used. Finally, RGCs were seeded on cell culture dishes coated with poly-L-lysine and mouse laminin. According to our previous study [13], the TUJ1 antibody (Beyotime, Shanghai, China) was used for RGC immunofluorescence identification (Supplementary Fig. 1A).

RGC damage assessment

The level of RGC damage was assessed by lactate dehydrogenase (LDH) assay kit (Beyotime, Shanghai, China) and TUNEL staining. For LDH analysis, the primary cells extracted from each group were seeded into 96-well cell culture plates. After 36 h of culture, the cell supernatant was aspirated, and the samples were determined immediately. The absorbance values at 490 nm were measured with a microplate reader (BioTek, USA).

For TUNEL analysis, the retinal sections from guinea pigs and humans were dewaxed and hydrated, the tissue slices were washed with PBS 3 times, and then, the membrane was disrupted by incubation with 0.1% Triton X-100 solution for 15 min at room temperature. Then, the TUNEL reaction solution was prepared, in which the TDT enzyme, dUTP, and buffer were mixed with a ratio of 1:5:50. The sections were incubated with TUNEL reaction solution for 2 h at 37 °C in the dark. The cell nuclei were stained with DAPI and finally sealed with an antifluorescence quenching sealer. Images were acquired using a fluorescence microscope (Olympus Corporation, Japan).

Donor ocular tissues

The eyeballs of healthy, myopic and highly myopic adults aged 34–65 years were obtained from Bright Eye Hospital (Shanghai, China). The donor eye tissues were excluded from syndromic connective tissue diseases (including congenital cataract, corneal opacity, metabolic cataract, etc.), advanced eye diseases (including age-related macular degeneration and glaucoma, etc.), or severe systemic diseases (such as cancer and diabetes). The eye cups were removed within 5 h after the death of the donor, and the separated retina, choroid and sclera were placed in a storage tube and stored at − 80 °C. All ocular tissues were obtained from Asian populations with balanced sex distributions. This study was approved by the Ethics Committee of Eye & ENT Hospital, Fudan University (No. 2023114) and followed the guidelines of the Declaration of Helsinki.

RNA isolation and construction of the RNA library

High-throughput sequencing of guinea pig RGCs (Oebiotech, Shanghai, China) was performed to determine lncRNA and mRNA expressions [37]. Three samples from the EAG were tested. Total RNA was extracted by an isolation kit (mirVana™ miRNA Isolation Kit (Thermo)). Next, we measured RNA concentration by NanoDrop-2000 (Thermo), and RNA integrity was assessed using a gel imaging system (Tanon 2500, Biotanon Co., Ltd.). After that, mRNA and lncRNA libraries were constructed from TruSeq-stranded total RNA with a Ribo-Zero Gold Prep Kit (Illumina). After rRNA removal, RNA was purified using magnetic beads. First- and second-strand cDNAs were synthesized, and the library fragments were purified using AMPure XP beads for cDNA screening.

Bioinformatic analysis of mRNAs and lncRNAs

We compared the RNA sequencing data of RGCs from guinea pigs with FDM after EA treatment with the RNA sequencing data of RGCs from our previous guinea pigs with FDM [13]. The reference genomes of guinea pigs were aligned using HISAT2 [38]. StringTie software was used to determine the fragments for each gene. The transcripts with coding potential were selected by Pfam, CPC, CNCI and PLEK to obtain predicted lncRNA sequences.

The sequencing reads of each sample were aligned with lncRNAs, mRNA transcripts, and predicted lncRNA sequences via Bowtie2. Gene quantitative analysis was conducted by eXpress, and the fragments per kilobase of exon model per million reads mapped (FPKM) values and counts were obtained. DESeq (2012) R package was used to analyze the differentially expressed (DE) lncRNAs and DE mRNAs. Statistical significance was defined as P values < 0.05 and |log2(FC)| values > 1. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed by hypergeometric distribution tests.

LncRNA-miRNA-mRNA network construction

By combining miRNAs from the miRanda database, we predicted potential regulation relationships between lncRNAs and miRNAs and between miRNAs and mRNAs.

The score between ceRNA relationship pairs was calculated using the ceRNA MuTATE method, and the probability of miRNAs sharing between ceRNA relationship pairs was calculated using the hypergeometric distribution algorithm. Finally, ceRNA relationship pairs with high reliability were obtained. Cytoscape software was used for ceRNA network construction.

Quantitative real-time PCR (qRT‒PCR)

Total RNA was extracted from guinea pig RGCs and human retinas by TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The cDNA synthesis of lncRNAs and mRNAs was performed using the PrimeScript RT Reagent Kit with gDNA Eraser (TaKaRa, Japan), and the cDNA synthesis of miRNA was performed using an EZ-press microRNA Reverse Transcription Kit (EZBioscience, USA). The cDNA was amplified on a Roche Light Cycler 480 II using TB Green™ Premix Ex Taq™ (Tli RNaseH Plus) (TaKaRa, Japan) for lncRNAs and mRNAs and using an EZ-press microRNA qPCR Kit (EZBioscience, USA) for miRNAs. The relative amount of RNA expression was analyzed by the 2−ΔΔCt method. GAPDH and U6 were used as controls. The guinea pig and human sequences of primers used are listed in Supplementary Tables 1–2. The primers were obtained from the Shanghai Generay Biotech Co., Ltd.

Western blot (WB)

Total protein was extracted from guinea pig RGCs and the human retina. The same amount of protein was separated by SDS‒PAGE and subsequently transferred to a PVDF membrane (Millipore, Boston, MA, USA). The PVDF membrane was blocked with 5% BSA. Subsequently, primary antibodies against Mfn2 (sc-100560, Santa Cruz, MA, USA), Optn (sc-166576, Santa Cruz, MA, USA), LC3B (ab192890, Abcam, Cambridge, UK), p62 (ab109012, Abcam, Cambridge, UK), PINK1 (ab186303, Abcam, Cambridge, UK), Parkin (ab77924, Abcam, Cambridge, UK), PARL (sc-514836, Santa Cruz, MA, USA) and β-Actin (AF2811, Beyotime, Shanghai, China) were incubated with the PVDF membranes overnight. On the second day, the membranes were incubated with secondary antibody (A0216, A0208, 1:1000; Beyotime, Shanghai, China). Protein was detected by an enhanced chemiluminescence (ECL) kit (Beyotime, Shanghai, China). Protein gray values were examined using ImageJ software.

Immunofluorescence (IF) and MitoTracker staining

Sections of the eyeball were deparaffinized, and antigen retrieval was performed. The slices were blocked with 5% BSA for 30 min. Primary antibodies against PINK1 (GB114934, Servicebio, Wuhan, China) and LC3A/LC3B (sc-398822, Santa Cruz, MA, USA) were added to the sections, and the sections were incubated overnight at 4 °C in a black wet box. After washing several times in PBS, the sections were incubated with secondary antibodies in a black wet box for 60 min at room temperature. Mitochondria were stained with 200 nM MitoTracker® Red CMXRos (9082S, CST, Danvers, MA, USA) at 37 °C for 30 min. Then, the slices were incubated with DAPI staining solution for 10 min at room temperature. Finally, photos were taken under a fluorescence microscope (Olympus Corporation, Japan). The IOD intensity was analyzed using ImageJ software.

Coimmunoprecipitation (Co-IP)

Coimmunoprecipitation analysis was performed using the IP/Co-IP kit (Thermo Fisher, Waltham, USA). Primary guinea pig RGCs were lysed with IP lysate containing protease inhibitors for 30 min on ice. The protein concentration was determined by the BCA assay. Then, the antibodies were added, and the samples were incubated for 2 h at 24 °C. The immunoprecipitates were incubated with Protein A/G Plus Agarose at 4 °C. Then, the protein complexes were washed and eluted with Lane Marker Sample Buffer containing 10 mM DTT at 100 °C for 10 min. The eluate was collected after centrifugation. The samples were subjected to SDS‒PAGE and transferred to PVDF membranes for subsequent western blot analysis.

Immunohistochemistry

For immunohistochemical staining of human samples, eye sections were incubated with an anti-LC3A/LC3B antibody (sc-398822, Santa Cruz, MA, USA) overnight at 4 °C and then sections were incubated with a secondary antibody for 50 min. Subsequently, human eye sections were visualized using a DAB kit for color development. Finally, the human eye sections were stained with hematoxylin for 3 min, and images were acquired using a light microscope.

Statistical analysis

Statistical analysis was performed with SPSS 25.0 statistical software (IBM, Armonk, NY, USA) and GraphPad Prism 10 (GraphPad Software, San Diego, CA, USA). All the data are expressed as the mean ± standard error of the mean (SEM). Student’s t test was performed for differences between two groups, and one-way ANOVA was performed for differences among three or more groups. P < 0.05 was considered to indicate statistical significance.