All procedures involving animals were approved by the Institutional Animal Care and Use Committee of our institution (Protocol No. A2022061) and reported in accordance with the ARRIVE guidelines (Animals in Research: Reporting In Vivo Experiments) [17] and each group consisted of six mice, based on previous experience. This clinical study was approved by the ethics committee of our hospital (registered at the Chinese Clinical Trial Registry) on September 24, 2020. Written informed consent was obtained from all patients before they underwent PET/CT scans.

C2C12 culture and H2O2 stimulation

C2C12 cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (HyClone, Logan, UT, USA) supplemented with 10% fetal bovine serum (#SH30396.03, Hyclone, Canada), 100 units/mL penicillin, and 100 µg/mL streptomycin at 37 °C, in a humidified atmosphere with 5% CO2. At near-confluence (90%), the C2C12 cells were transferred to DMEM containing 2% horse serum (Gibco) to induce cell differentiation. After reaching complete coverage of the visual field (at approximately 9 days), well-differentiated myotubes were treated with a H2O2 gradient (Junsei Chemical Co., Ltd.) for 24 h, followed by quantitative polymerase chain reaction (qPCR), western blot (WB), and immunofluorescence analyses.

Construction of animal models and examination plan

Based on previous studies and our preliminary experiment [18, 19], 75 µL of 1.2% BaCl2, 0.2 M NaOH, and 1.2 M NaOH were injected into murine triceps surae to simulate models of normal repair, mild repair failure, and severe repair failure, respectively. Each group consisted of six subjects. Imaging assessments were performed on postoperative days 1, 3, 7, and 14. Subsequently, the mice were euthanized, and tissue samples were collected for qPCR, WB, RNA sequencing (RNA-Seq), and pathological analyses.

qPCR and WB analysis

Cells and tissues were lysed using QIAzol Lysis Reagent and extracted using the RNeasy Mini Kit, following the manufacturer’s protocol (QIAGEN, Hilden, Germany). Subsequently, RNA was reverse-transcribed into cDNA using the RevertAid First-Strand cDNA Synthesis Kit (Invitrogen, Carlsbad, CA, USA). qPCR was conducted on a QuantStudio™ Real-time PCR system using the SYBR® Green Master Mix (Invitrogen). The relative gene expression levels were determined using the 2−ΔΔCT method. Primer sequences employed in qRT-PCR for alpha-tubulin, FAP, myosin heavy chain, and interleukin 6 are provided in Supplementary Table 1.

Cells and tissues were lysed in 100 µL and 500 µL of radioimmunoprecipitation assay buffer containing 1% phosphatase inhibitor and 2% protease inhibitor separately. After 20 min of ice-cold lysis, the lysates were centrifuged at 12,000 rpm for another 20 min. The protein concentration was determined using a bicinchoninic acid assay, and after normalizing the concentrations of each sample, sodium dodecyl sulfate was used for heating and denaturation.

WB was performed according to standard procedures (Genescript, Wuhan, China). Membranes were blocked with 3% bovine serum albumin and subsequently probed with antibodies against alpha-tubulin and FAP. The membranes were then washed in PBST and incubated with horseradish peroxidase-conjugated secondary antibodies. Immunoreactive bands were visualized by adding an enhanced chemiluminescence substrate.

Immunofluorescence and histopathology

C2C12 myotubes were fixed in 4% paraformaldehyde and permeabilized with 0.5% Triton X-100. Following a 60-min blocking step in 10% bovine serum albumin, plates were incubated overnight at 4 °C with primary antibodies against FAP (ab53066; Abcam), interleukin 6 (Invitrogen, PA5-47209), and myosin heavy chain (Sigma-Aldrich, M4276). The following day, sections were treated with secondary antibodies, including anti-goat IgG (ab150131; Abcam), anti-mouse IgG (ab1501018; Abcam), and anti-rabbit IgG (ab150073; Abcam), at room temperature for 1 h. After washing, the plates were mounted using a 4′,6-diamidino-2-phenylindole-containing mounting medium (Beyotime, C1005).

For histopathology, the mice were euthanized, and their excised muscles fixed in 4% paraformaldehyde overnight. Subsequently, the tissues were subjected to standard dehydration in ethanol for paraffin embedding. Hematoxylin & eosin and Sirius red staining was performed according to the manufacturer’s instructions. For immunohistochemical staining, paraffin-embedded sections were treated with 3% H2O2 for 15 min to inhibit endogenous peroxidase activity. Subsequently, 10% bovine serum albumin was applied for 1 h at 21 °C to block non-specific antigens. Sections were then incubated with specific primary antibodies against FAP and Col 1 overnight at 4 °C. Following this, the sections were exposed to horseradish peroxidase-conjugated secondary antibodies (ZSGB-BIO, PV-6001 and PV-9003) for 1 h at 21 °C.

RNA-Seq library preparation and data processing

Total RNA was extracted using the Total RNA Extractor (TRIzol) kit (B511311, Sangon, China) following the manufacturer’s protocol, with RNase-free DNase I treatment for genomic DNA contamination removal. The VAHTSTM mRNA-Seq V2 Library Prep Kit for Illumina® was used for library generation, starting with 1 µg RNA per sample. Pooled libraries were subjected to paired-end sequencing using NovaSeq sequencers (Illumina).

Quality evaluation of the sequenced data was performed using FastQC (version 0.11.2), and raw reads were filtered using Trimmomatic (version 0.36) to remove adaptor sequences and low-quality bases. HISAT2 (version 2.0) mapped clean reads to the reference genome, and RSeQC (version 2.6.1) and Qualimap (version 2.2.1) were used aligning result statistics and quality checks. BEDTools (version 2.26.0) was used to assess the gene coverage ratios. StringTie (version 1.3.3b) was used to compute gene expression values, and DESeq2 (version 1.12.4) was used to determine differentially expressed genes. Differentially expressed genes with q-value ≤ 0.001 and |FoldChange| ≥2 were considered significant.

18F-FDG and 68 Ga-FAPI-04 systhesis and small PET/CT imaging

18F-FDG synthesis followed the method described by Hamacher et al. [20] with modifications, using a computer-controlled apparatus. 68Ga-FAPI-04 was synthesized as previously outlined [21]. Small-animal PET scans were conducted using a Super Nova scanner (Pingseng Scientific, Jiangsu, China). Static imaging was performed by administering a dose of 3.7–7.4 MBq of 68Ga-FAPI-04 or 18 F-FDG, followed by a 5-minute PET scan taken 30 min post-injection. This timing was chosen to allow for optimal tracer uptake and image quality, facilitating accurate visualization and analysis of the biological processes under investigation. The metrics SUVmax and MTV were used for comparative analysis between groups. The Target/Background (T/B) ratio was defined as the SUVmax of the surgical side divided by the SUVmax of the contralateral normal muscle.

MRI and ultrasound

During MRI (3T MRI scanner, GE Healthcare, MR750, Milwaukee, WI), mice were anesthetized with 0.3% pentobarbital sodium (0.1 mL/10 g, i.p.) and inserted into a 16-channel mouse body coil. The MRI protocol consisted of two sequences. First, a T2-weighted rapid acquisition relaxation-enhanced sequence was acquired (repetition time [TR] = 1664 ms, time to echo [TE] = 42 ms, NEX = 6, slice thickness = 2 mm, slices = 9, field of view [FOV] = 6 × 4.5 cm2, matrix = 192 × 192, and acquisition time = 2:45 min). Next, a T2-weighted rapid acquisition relaxation-enhanced sequence with Dixon water and fat separation was acquired (TR = 1664 ms, TE = 24 ms, NEX = 2, slice thickness = 2 mm, slices = 9, FOV = 6 × 4.5 cm2, matrix = 192 × 192, acquisition time = 2:50 min).

Ultrasound (ZS3 EXP; Mindray, Shenzhen, China) assessments were performed on mice anesthetized with 3% (v/v) isoflurane. An L30-8 probe was used to detect the thickest part of the triceps surae. Muscle parameters, including thickness, texture, and echo, were measured, and scores were assigned based on the echo and texture manifestations of the muscles (Supplementary Table 2).

Clinical study

The clinical study assessed the performance characteristics of 18F-FDG and 68Ga-FAPI-04 within 1 month, 2–3 months, 5–6 months, and > 7 months after hip arthroplasty. Images were obtained from the pelvis to the knee using a time-of-flight PET/CT scanner (uMI510, United Imaging Healthcare, Shanghai, China). Scanning commenced one hour after administering 68Ga-FAPI or 18F-FDG (1.8–2.4 MBq per kilogram of body weight for 68Ga-FAPI, 4–6 MBq per kilogram of body weight for 18F-FDG), with an acquisition time of 4 min per bed position. A low-dose CT scan (120 kV, 30–50 mAs) was employed for anatomical localization and attenuation correction. The images were reconstructed using a standard ordered-subset expectation-maximization algorithm. All CT images were converted to a bone window. All PET images were transformed into SUV units by normalizing the activity concentration to the administered radiopharmaceuticals dose and the patient’s body weight after decay correction. For the delineation of MTV, regions with significantly elevated uptake on each slice were outlined individually and then merged across all slices to obtain the corresponding MTV. SUVmax was defined as the highest uptake value within the specified MTV region.

Statistical analysis

Data are presented as the mean ± standard error of the mean. For both animal experiments and clinical studies, one-way analysis of variance followed by Tukey’s multiple comparison test was employed. All tests were two-tailed, and statistical significance was set at P < 0.05. Statistical analysis was performed using SPSS 25.0 (SPSS Inc., Chicago, IL, USA), and GraphPad Prism 8.0.2 (GraphPad Software, San Diego, CA, USA) was utilized to present the results.