Gefitinib Facilitates Bone Fracture Healing via Inhibition of the EGFR Pathway and Counteracting SOX9-driven Bone Metabolic Reprogramming

Materials and Methods

Ethics statement. This study was approved by the Committee of Experimental Animals of the Eighth People’s Hospital, Jiangsu University (Shanghai, PR China; approval no. ECM202310165). All procedures related to animal testing were conducted in accordance with the Guidance for the Care and Use of Laboratory Animals issued by the Ministry of Science and Technology Department of the People’s Republic of China.

Animals and surgical procedures. A total of 104 4-week-old male Sprague-Dawley rats (Rattus norvegicus; Experimental Animal Research Center of Hubei Province, PR China) were adaptively fed for 1 week and were subsequently anesthetized by intraperitoneal injection of chloral hydrate (300 mg/kg). An intramedullary 1.0-mm-diameter Kirschner wire was retrogradely inserted into the greater trochanter via piercing from the femoral intercondylar, and the wire was then cut at the level of the articular surface of the knee and at the margin of the greater trochanter. An incision was made at the right lateral thigh, and the right femur was exposed by separating the vastus lateralis; subsequently, a mid-femoral transverse osteotomy was carried out with a bone scissors.

After surgery under anesthesia, X-ray examination was carried out on the right femur of all rats to observe the degree of displacement and the type of fractures, which were transverse or short oblique to the middle part of the femur. The model rats were split at random into a gefitinib-treated experimental group (n=52) and a vehicle-treated control group (n=52). Rats in the gefitinib-treated group were administered gefitinib (100 ml/kg) dissolved in 0.5% methyl cellulose through oral gavage daily for 6 weeks, while rats in the vehicle-treated control group were treated with 0.5% methyl cellulose. X-Ray examination was carried out on the right femur on days 0, 7, 14, 21, 28, and 42.

Following the completion of experimental procedures, the rats were euthanized using CO₂ inhalation. The CO₂ flow rate was set to displace 20% of the chamber volume per minute, ensuring a gradual and humane euthanasia process. This was followed by cervical dislocation to confirm death before cardiac puncture was performed to collect blood samples to measure serum osteogenic markers and osteoclastic markers. The volume of blood collected from each rat was approximately 8-10 ml. The weight of the rats at the time of sacrifice was recorded, with the average weight being 220-330 g. Femurs were harvested for histological analysis and micro-computed tomography (MCT); callus within 3 mm of the proximal and distal fracture line was collected for assay by quantitative real-time polymerase chain reaction (qRT-PCR).

The criteria for humane endpoints included: Loss of 15% or more body weight; severe lameness or inability to reach food or water; signs of severe pain or distress unrelieved by analgesics; non-healing or infected wounds; moribund state. All euthanasia procedures were performed using carbon dioxide inhalation followed by cervical dislocation, ensuring a swift and humane death. Specifically, the following criteria were used to confirm death: Cessation of respiratory movements: observing the complete absence of respiratory movements for at least 2 min; absence of heartbeat: palpating the chest to ensure the absence of a heartbeat; lack of corneal reflex: testing the corneal reflex by gently touching the cornea with a sterile object to ensure no blink response; confirmation of rigor mortis: observing the onset of rigor mortis as an additional confirmation. These steps ensured that the animals were humanely euthanized, and death was confirmed before proceeding with any further procedures.

X-Ray analyses. Fractures of anesthetized rats were radiographed for 6 s using an X-ray system (Kodak DirectView DR 3500; Kodak, Rochester, NY, USA) at 65 kV on days 0, 7, 14, 21, 28, and 42 to monitor the initial pin fixation and ongoing callus formation and resorption. Lane and Sandhu radiographic scoring (15) was used to compare the fracture healing between groups by two independent orthopedic surgeons. The development of the bone callus at the fracture end was examined using Image J (National Institutes of Health, Bethesda, MA, USA) and the average value of the maximal callus diameter on the frontal and lateral X-ray images was noted.

Micro-computed tomography. Bones were imaged at a resolution of 15 μm using a Scanco UCT 50 system (Scanco Medical, Brüttisellen, Zurich, Switzerland) operating at a peak voltage of 70 kV and 114 mA. A volume of interest was delineated, which encompassed the callus over a range of 3 mm between the distal and proximal end, and crossed all 200 sections, with contours of the volume of interest drawn on sequential slices. With 3D software (3D-UCT Ray v. 3.8; Scanco Medical), the images were visualized as three-dimensional reconstructions. Using computed tomography analyzer software (UCT Tomography v. 6.3-4; Scanco Medical), the following values were obtained for the entire callus: bone volume (BV), total volume (TV), BV/TV, bone surface (BS), BS/BV, bone mineral density, trabecular thickness, trabecular separation, and trabecular number.

Mechanical testing. Rats were killed in a carbon dioxide chamber on days 21, 28, or 42 of the experiment, and the femurs were dissected to remove the fixators and any soft tissues. Mechanical testing was carried out in compliance with an earlier defined methodology (16). The femurs were centered on the osteotomy line and positioned on the supports, leaving a 20 mm free length between the bending supports for the bone. Material testing equipment (5,967 MicroTester; Instron, Norwood, MA, USA) was used to provide the load in 3-point bending at a deflection rate of 0.6 mm/min. Bending was performed at the osteotomy level in an anterior-posterior direction. The bones were maintained wet with a 0.9% NaCl solution throughout the testing. The callus was subjected to a bending force, which was continually monitored in relation to the sample deflection. Furthermore, a load-displacement diagram was captured, which allowed for the determination of the failure load. The area under the curve up to F-max was used to compute the work to failure, and linear regression was used to quantify stiffness. For both the experimental group and the control group, records of the failure load, stiffness, and work to failure were made for each femur.

Histological evaluation. Fractured limbs were removed and preserved for 24 h in 4% formalin, then for 1 to 2 months in a 10% ethylene-diaminetetraacetic acid (Sigma-Aldrich, St. Louis, MO, USA) solution to decalcify them before being embedded in paraffin for histological examination. To demonstrate the creation of cartilage, decalcified longitudinal slices were cut to a size of around 5 μm and stained with either Safranin O/Fast Green (Solabio, Beijing, PR China) or hematoxylin and eosin (Solabio). Five sections equally spaced from the first to the last slices were selected for examining the presence and characteristics of cartilage formation.

Immunohistochemistry for Ki-67. Samples that had been embedded in paraffin were sectioned, and the sections were rehydrated using water and graded ethanol. Sections were heated in a microwave for 3 min at 95°C to recover antigen for antigen unmasking, after which they were chilled for 20 min at room temperature in 10 mM of sodium citrate buffer (pH 6.0). Sections were submerged in 3% H₂O₂ for 10 minutes to reduce endogenous peroxidase activity, followed by two 5-min washes in phosphate-buffered saline (PBS). Next, they were treated with 5% bovine serum albumin for 20 min at room temperature. After applying rabbit antibody to Ki-67 (1:400; ab15580; Abcam, Cambridge, MA, USA), the sections were left to incubate at 4°C for an entire night. Following PBS washing, the specimens were incubated for 1 h at 37°C with polyclonal biotinylated goat anti-rabbit secondary antibody (1:200; Boster Bio, Pleasanton, CA, USA), followed by 30 min at 37°C with streptavidin-biotin complex. After three PBS washes, the slices were treated for 30 min with 3,3-diaminobenzidine substrate, and hematoxylin was used as a counterstain. All sections were photographed under a microscope at 400× magnification.

Detection of serum markers of bone formation and resorption. During the euthanasia of the rats, blood samples were obtained using heart puncture and allowed to stand at room temperature for a minimum of 30 min. The sera were separated using a centrifuge at 200×g for 10 min, and then kept at −20°C. Serum bone-specific alkaline phosphatase (BALP), collagen type I alpha 1 chain (COL1A1), collagen type II alpha 1 chain (COL2A1), collagen type X (COL10), osteocalcin (bone gamma-carboxyglutamate protein, BGLAP), procollagen type I N-terminal propeptide (PINP), tartrate-resistant acid phosphatase and 5b (TRACP5b), and cross-linked C-telopeptide of type collagen (CTX) were measured using an enzyme-linked immunosorbent essay (ELISA) kit, rat PINP ELISA kit and rat CTX ELISA kit, respectively, from Elbscience (Wuhan, Hubei, PR China), as per the manufacturer’s instructions.

qRT-PCR analysis. Within 3 mm of the proximal and distal fracture lines, total RNA was extracted from the callus using the TRIzol reagent (Aidlab, Beijing, PR China). SuperScript III (Invitrogen, Waltham, MA, USA) was then used to reverse transcribe the extracted RNA. Table I lists the primer sequences for the SYBR green Fast PCR system (Cytiva, Marlborough, MA, USA), which was used to conduct qRT-PCR.

Western blot analysis. In short, proteins were extracted from periosteal stem cells (PSCs) using RIPA lysis buffer, and their quantities were determined using a BCA assay kit (Thermo Fisher Scientific, Waltham, MA, USA). After that, the proteins were put on a polyvinylidene difluoride membrane and kept at 4°C for incubation overnight with primary antibodies to SOX9, BGLAP, COL2A1, osteocalcin, Sp7 transcription factor (SP7), Runt-related transcription factor-2 (RUNX2), glucose transporter type 1 (GLUT1), hexokinase 2 (HK2), pyruvate dehydrogenase kinase 1 (PDK1), pyruvate kinase M2 (PKM2) and EGFR (Cell Signaling Technology, Danvers, MA, USA) at a dilution of 1:1,000; and glyceraldehyde 3-phosphate dehydrogenase (Abcam) at a dilution of 1:3,000. The membrane was exposed to the secondary antibody for 1 h at room temperature and was then developed using an enhanced chemiluminescence system (Thermo Fisher Scientific).

Cell isolation and culture. PSCs were isolated from intact femora and tibiae of 10-week-old rats. For the isolation of PSCs, femurs and tibias were carefully dissected free of muscle and connective tissue under sterile conditions. The epiphyses were protected from digestion by immersing them in 5% low melting point agarose (Invitrogen). Periosteal cells were then isolated by enzymatic digestion using 3 mg/ml collagenase II (Gibco, Grand Island, NY, USA) and 4 mg/ml dispase (Gibco) in α-minimum essential medium (α-MEM; Gibco) supplemented with 1% penicillin/streptomycin (Gibco). Cells from the first 10-min digestion were discarded, as they primarily contained residual muscle and connective tissue. PSCs were subsequently harvested after a 1-h digestion. The cell suspension was filtered through a 70 μm nylon mesh (BD Falcon; BD Biosciences, San Jose, CA, USA), washed twice, and cultured in α-MEM with 1% penicillin/streptomycin and 10% fetal bovine serum (FBS; Gibco) in a humidified incubator at 37°C with 5% CO₂. All experiments were conducted using passage 2-3 PSCs.

Differentiation assays. The direction of PSCs differentiation plays a crucial role in fracture healing. To investigate the impact of gefitinib on PSCs differentiation, we adopted the method of literature to simulate the effect of lipid environment on PSCs (17), we utilized lipid-reduced serum (LRS; Thermo Fisher Scientific) medium to simulate the differentiation milieu of PSCs in the fracture site. To evaluate chondrogenic differentiation, 150,000 PSCs were resuspended in 10 μl of control medium and seeded as micromasses at the center of a 24-well plate. After a 1-h incubation at 37°C to allow cell attachment, 0.5 ml of FBS, LRS or LRS with 4 μM gefitinib medium containing 10 ng/ml recombinant human transforming growth factor-β1 (PHG9211; Gibco), 50 μM L-ascorbic acid 2-sulphate (Sigma-Aldrich), and 20 μM Rho kinase inhibitor Y-27632 (Sigma-Aldrich) was added to each well. The medium was refreshed every other day, and after 9 days, the micromasses were used for RNA or protein isolation.

Cell transfection. PSCs were seeded in six-well plates at a density of 2×10⁵ cells/well and cultured in α-MEM supplemented with 10% FBS and 1% penicillin-streptomycin under standard conditions (37°C, 5% CO₂) until they reached approximately 70-80% confluence. For Sox9 plasmid transfection, 2 μg of the pcDNA3.1-SOX9 plasmid (V79020; Invitrogen) was diluted in 250 μl of Opti-MEM (Thermo Fisher Scientific) and mixed with 5 μl of Lipofectamine 3000 reagent (Invitrogen) in 250 μl of Opti-MEM. After a 5-min incubation at room temperature, the DNA-reagent complex was added to the cells which were then incubated for 6 h. The medium was then replaced with fresh α-MEM containing 10% FBS and 1% penicillin-streptomycin. For SOX9 knockdown, PSCs were transfected with 20 nM of Sox9-specific siRNA using Lipofectamine RNAiMAX (Invitrogen). The siRNA (20 nM) was diluted in 250 μl of Opti-MEM, and 5 μl of Lipofectamine RNAiMAX in 250 μl of Opti-MEM was added. The siRNA-Lipofectamine complex was incubated for 20 min at room temperature before being added to the cells. After 6 h of incubation, the medium was replaced with fresh α-MEM containing 10% FBS and 1% penicillin-streptomycin. The transfection efficiency for both plasmid and siRNA was evaluated by qPCR analysis at 48 h post-transfection. PSCs were then harvested for downstream analysis at 48 h post-transfection. The Sox9 siRNA target sequence was as follows: 5′- GCGACAACTTTACCAGTTTCAGT-3′. SOX9 siRNA and non-targeted siRNA were purchased from RiboBio (Guangzhou, PR China).

Metabolic reprogramming assays. Glucose uptake assay: A glucose uptake colorimetric test kit (Abcam) was used to measure glucose uptake. After being transfected with the chosen vector and plated in a 6-well plate, the cells were cultured for 24 h. After transfected cells were gathered, the number of cells was determined. A 96-well cell culture plate was then seeded with 1×10⁴ cells and was incubated at 37°C overnight. The cells were deprived of glucose for 2 h the next day. After 40 min of incubation with 100 μl Krebs-HEPES, each well received 10 μl of 10 mM 2-deoxy-D-glucose, and was then incubated for 20 min. After that, the cells were harvested utilizing extraction buffer, and the absorption of glucose was tested. The optical density was used to measure glucose absorption at a wavelength of 412 nm.

Lactate and adenosine triphosphate (ATP) production assay: Lactate production was measured using an L-lactate assay kit (colorimetric method, Abcam). Transfected cells were seeded in a 96-well cell culture plate and incubated overnight at 37°C. After treatment, the culture medium was collected for lactate measurement. For lactate determination, L-lactate assay kit (colorimetric method, Abcam) was used according to the manufacturer’s protocol. Briefly, 50 μl of culture supernatant was added to the assay well, and the lactate reagent was mixed with the sample. The absorbance at 570 nm was measured using a microplate reader. ATP levels were quantified using a commercial ATP assay kit (Abcam), following the manufacturer’s instructions. The resulting luminescence was measured using a luminometer, and ATP levels were normalized to protein concentration to assess metabolic activity.

Extracellular acidification rate (ECAR) assay: Seahorse XF glycolysis stress test kit (Agilent Technologies, Santa Clara, CA, USA) and Seahorse XF Cell Mito Stress Test Kit (Agilent Technologies) were used to determine ECAR and extracellular acidification rate [by measuring the oxygen consumption rate (OCR)], respectively. The transfected and gefitinib-treated PSCs were cultured for overnight at 37°C in a 96-well plate supplemented with 10% FBS. For the purpose of measuring ECAR, glucose (10 mM), oligomycin (1 μM), and 2-deoxy-D-glucose (50 mM) were successively introduced to each well after baseline measurements were made. For OCR determination, rotenone (2 μM), antimycin A (2 μM), oligomycin (1 μM), and carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (1 μM) were introduced in that order. Seahorse XF-96 Wave software (Agilent Technologies) was used to analyze the findings.

Mitochondrial function and fatty acid oxidation (FAO) assay: FAO was assessed by measuring the OCR in the presence of palmitate (30 μM). PSCs were seeded onto a 96-well XF96 type plate (Agilent Technologies) and incubated at 37°C overnight. Prior to measurement, the cells were incubated for 1 h at 37°C without CO₂ in XF Assay Modified DMEM culture medium (1 g/ml glucose, pH 7.4; Seahorse Biosciences), replacing the whole mixture. Etomoxir (Sigma-Aldrich) was added at a final concentration of 0.1 mM. Fatty acid-linked OCR was measured following the sequential addition of oligomycin (1 μM), 4 (trifluoromethoxy) phenylhydrazone (1 μM), rotenone (2 μM), and antimycin A (2 μM) to achieve the indicated final concentrations. Measurements were performed according to the kit protocol for FAO. Seahorse XF-96Wave software was used to analyze the results.

Dual luciferase reporter gene assay. To assess the transcriptional activity of Sox9 and its downstream target genes, dual luciferase reporter gene assays were performed. PSCs were seeded in 96-well plates and transfected with either the pGL3 control plasmid (Promega, Madison, WI, USA) or the pGL3-3-Sox9 reporter construct, which contains the Sox9 promoter region driving the firefly luciferase gene. After 24-48 hof treatment with gefitinib or vehicle, the cells were harvested for luciferase activity measurement. Renilla luciferase activity was used as an internal control for normalization. The dual luciferase assay was carried out using the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions. The activity of firefly luciferase was measured using a Luminoskan Ascent luminometer (Thermo Fisher Scientific), and relative luciferase activity was calculated by normalizing firefly luciferase activity to Renilla luciferase activity. In addition, PSCs were transfected with reporter constructs containing the pGL3 vector coupled to the promoters of Glut1, Pdk1, Hk2, and Pkm2, which are known downstream targets of Sox9. The cells were then treated with gefitinib or vehicle. Luciferase activity was measured, and the relative activity of the reporter constructs was calculated to assess the impact of Sox9 on the transcription of these target genes under gefitinib treatment.

Statistical analysis. The mean±standard deviation of measurements made independently at least three times is used to represent findings. To assess the statistical difference between the control and treatment groups, Student’s t-test was used. Statistical tests were conducted with SPSS software 19.0 (IBM Corp., Armonk, NY, USA), with a significance level of p<0.05. One-way analysis of variance test or Student’s t-test was used to compare the mean values across the groups of rats at various intervals.