High formability Mg-Zn-Gd wire facilitates ACL reconstruction via its swift degradation to accelerate intra-tunnel endochondral ossification

Xun He ,Ye Li ,Hongwei Mio ,Jinkun Xu ,Mihel Tim-yun Ong ,Chenmin Wng ,Lizhen Zheng,Jili Wng,Le Hung,Hiyue Zu,Zhi Yo,Jie Mi,Bingyng Di,Xu Li,Ptrik Shu-hng Yung,∗,Gungyin Yun,Ling Qin,∗

aMusculoskeletal Research Laboratory of Department of Orthopaedics &Traumatology and Innovative Orthopaedic Biomaterial and Drug Translational Research Laboratory of Li Ka Shing Institute of Health Sciences, The Chinese University of Hong Kong, Hong Kong, China

b National Engineering Research Center of Light Alloy Net Forming, School of Materials Science &Engineering, Shanghai Jiao Tong University, Shanghai,China

cDepartment of Orthopaedics and Traumatology, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong, China

d Orthopaedic Department, Peking University Third Hospital, Beijing, China

e Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hong Kong, China

Abstract After reconstructing the anterior cruciate ligament(ACL),unsatisfactory bone tendon interface healing may often induce tunnel enlargement at the early healing stage.With good biological features and high formability,Magnesium-Zinc-Gadolinium (ZG21) wires are developed to bunch the tendon graft for matching the bone tunnel during transplantation.Microstructure,tensile strength,degradation,and cytotoxicity of ZG21 wire are evaluated.The rabbit model is used for assessing the biological effects of ZG21 wire by Micro-CT,histology,and mechanical test.The SEM/EDS,immunochemistry,and in vitro assessments are performed to investigate the underlying mechanism.Material tests demonstrate the high formability of ZG21 wire as surgical suture.Micro-CT shows ZG21 wire degradation accelerates tunnel bone formation,and histologically with earlier and more fibrocartilage regeneration at the healing interface.The mechanical test shows higher ultimate load in the ZG21 group.The SEM/EDS presents ZG21 wire degradation triggered calcium phosphate (Ca-P) deposition.IHC results demonstrate upregulation of Wnt3a,BMP2,and VEGF at the early phase and TGFβ3 and Type II collagen at the late phase of healing.In vitro tests also confirmed the Ca-P in the metal extract could elevate the expression of Wnt3a, β catenin,ocn and opn to stimulate osteogenesis. Ex vivo tests of clinical samples indicated suturing with ZG21 wire did not weaken the ultimate loading of human tendon tissue.In conclusion,the ZG21 wire is feasible for tendon graft bunching.Its degradation products accelerated intra-tunnel endochondral ossification at the early healing stage and therefore enhanced bone-tendon interface healing in ACL reconstruction.

Keywords: Magnesium wire;ACL reconstruction;Magnesium alloy;Biomaterials;Endochondral ossification.

1.Introduction

Anterior Cruciate Ligament (ACL) rupture usually occurs in sports with frequent twisting motions.Both conservative and surgical management are conventional for the treatment of ACL injury [1–3].Patients who have unstable knee joints with ACL rupture require reconstruction surgery to avoid rotational instability,secondary meniscus tear,articular cartilage lesion,or post-traumatic osteoarthritis(OA)[4].The American Academy of Orthopaedic Surgeons (AAOS) guideline stated that doctors should perform ACL reconstruction for young patients to protect their articular cartilage and meniscus [5].So far,ACL reconstruction achieves satisfactory outcomes in more than 75% of patients,which means that unsatisfactory healing cases are not rare at the other side [6].At the early healing stage,an unfavorable mechanical environment and biological potential induced result in poor healing outcomes characterized by tunnel widening,weaken enthesis,and graft atrophy [7].The unsatisfactory healing will delay the postoperative training and may require revision surgery for severe cases [8].Biologically,the enthesis or bone tendon junction regeneration can crucially determine the healing outcomes[9].Four-layer fibrocartilage is located at the central part of the native enthesis with Sharpy-like fibers connection distributed at the outskirts [10].Therefore,good biological bone tendon junction healing is essential for achieving good ACL reconstruction.

In the recent decade,various biodegradable and bioactive implants have been developed and tested to improve the healing outcome after ACL reconstruction.These implants stimulate bone growth or bone tendon junction healing with degradation over time [11].Clinically,polymers have been applied for internal fixation at low mechanical bearing sites as alternative materials to traditional permanent metal implants to avoid later implant removal surgery [12].As one of the major concerns,acidic product accumulation during polymer degradation may induce local inflammation and bone resorption[13,14].Besides the favorable biomechanical properties of reduced stress shielding [15],as a bio-metal,magnesium(Mg)-based implants showed excellent osteogenic potential duringin vivodegradation [16].In the ACL reconstruction model,a recent experimental study reported that Mg-based interference screw could enhance ACL reconstruction by upregulating the expression of osteogenic and angiogenic factors [17–20].

Clinically,surgeons bunch the graft to fit the tunnel size and shape before dragging the tendon graft through the bone tunnel[21,22](Video 1).However,the polymer suture threads might hinder the bone tendon junction healing due to the slow hydrolysis speed.The Mg-base wire was first developed as a suture for the ligation of ovariotomy and anastomosis [23].However,Mg-based wire exhibited poor mechanical properties for suturing even before degradation [24].To improve the mechanical strength and biodegradability of Mg-based implants,several advanced technologies such as rare earth-containing alloys or coating treatments were investigated and applied [25–30].Nowadays,Cold-drawn Mg-based wire demonstrated good mechanical properties with finer microstructure [31].Besides fabrication technologies,rare earth elements used for Mg alloys presented promising translation potential for clinical use [32].After testing different Mg-rare earth (RE) alloys,ZG21 (Magnesium-Zinc-Gadolinium,Mg-2.33Zn-0.84Gd) alloy presents excellent mechanical properties with a low tensile-compression asymmetry,which is a crucial parameter of suturing material reported in our early study [33].Furthermore,we investigated the biocompatibility of ZG21 bothin vitroandin vivo[34].Based on these studies,our team developed and tested ZG21 wire for tendon graft bunching.

For the Mg degradation’s potential mechanism associated with the osteogenic effect in ACL reconstruction,Cheng et al.reported that the high-purity Mg implant degradation products could improve osteogenesis via the upregulation of BMP2 and VEGF expression[20].Huang et al.reported that Mg ions derived from Mg alloy plate and screw could stimulate Wnt3a in the fracture model [35].On the other hand,some other studies reported that Mg ions suppressed bone formation as they competed with Ca ions during callus mineralization [36,37].Therefore,unlike Mg’s role in subperiosteal ossification,it is controversial to identify the role of Mg degradation products in endochondral ossification.Also,since Mg degradation mainly promotes osteogenesis,the role of Mg implants in fibrocartilage formation remains unclear.This paper focused on assessing the biological effect of Mg-based wire applied for tendon graft bunching in ACL reconstruction.Furthermore,this study also aimed at filling up the knowledge gap between the degradation of Mg-based implants and intra-tunnel endochondral ossification after ACL reconstruction surgery.

2.Methods and materials

2.1. Materials preparation

High purity Mg (99.99%),high purity Zn (99.995%),and Mg-30 wt% Gd master alloy were melted by a resistance furnace protected by a mixed atmosphere of SF6(1 vol%)and CO2(99 vol%).Firstly,pure Mg was melted,then the Mg-Gd master alloy was added to melted Mg at 760 °C.After that,Zn was put into the melt at 700 °C.The melt was kept at 730 °C for 15 min and then poured into a preheated steel mold at 200 °C.The ingots were machined into billets with 60mm diameter and 60mm height.All the billets were preheated at 250 °C for 1 h and then were extruded with a ratio of 9:1.The ram speed was 2 mm/s.After extrusion,the 20 mm diameter rods were obtained and a blank of 20 mm diameter and 30 mm-height was prepared for a second hot extrusion.Through the second hot extrusion,1.5 mm wires were processed for further drawing and annealing.The deformation of single pass drawing was 10%,and the online annealing temperature is about 400 °C.Finally,0.15 mm diameter wires were prepared.

2.2. Microscopic assessment

According to the published protocol [34],the wire was polished with MgO suspension (the MgO suspension was filtered 3 times for polishing).Then,the polished surface was further etched by nitric acid alcohol solution (4%).The transverse plane and longitudinal plane were observed by optical microscopy (OM,Leica MEF4M,Germany).The grain size was measured by the linear intercept method (Fig.S1).

2.3. Scanning electron microscope of straight wire and knots

According to the published protocol[38],the wire was polished with MgO suspension (particle size: 1 μm).Then,the polished surface was further etched by acetic acid and nitric acid solution.The straight and knotted wires were scanned by electron microscope (SEM,S3400N,HITACHI,Japan).

2.4. Transmission electron microscope (TEM)

According to the published protocol [39],the sample was prepared by the ion thinning method.The samples were scanned by a double spherical aberration correction electron microscope (FEI-Titan G3 Cubed Themis,Thermo Fisher,USA).The working voltage of TEM is 200 kV.

2.5. Tensile test of ZG21 wire

The ZG21 wire was fixed with two wooden rollers and mounted on a mechanical machine (H25K-S,Hounsfield Test Equipment LTD,Surrey,UK).The wire (length: 150 mm,n=3) was mounted on tensile jigs to 180.The wire was stretched with a 0.1 N pre-load and 40 mm/min speed until mechanical failure.The yield loading,ultimate loading,and elongation were recorded,and the stiffness,stress,absorbed energy,Young’s modular,and energy density were obtained according to the Tensile-Elongation curve and Stress-Strain curve (Fig.1H).The mentioned curves and data were drawn and collected by Vernier Graphical Analysis 4.10.

Fig.1.Material tests.A.ZG21 wire preparation flowchart;B.microstructure of ZG21 wire,transverse plane (left) and longitudinal plane (right);C.scanning microscope image of straight wire,scale bar from left to right: 1000,500,200,100 μm;D.scanning microscope image of knotted wire,scale bar from left to right: 1000 μm,500 μm,200 μm,100 μm;E.morphology (left) and diffraction pattern (right) of the nano-scale second phase TEM image of ZG21,scale bar from left to right: 1000 μm,500 μm,200 μm,100 μm;F.atomic-scale morphology of nano-scale second phase (left) and its interface combination with matrix (right);G.tensile test sample mounting (left),middle substance breakage (middle) and tensile-elongation curve and stress-strain curve (right) of ZG21 wire,white arrow: wire breakage point.

2.6.In vitro degradations

According to the published protocol [19],the ZG21 alloy sample was immersed in Phosphate Buffer Solution (PBS).The volume of PBS (1 L solution containing 120 mmol NaCl,2.7 mmol KCl,and 10 mmol phosphate buffer salts)was calculated according to ISO10993-5 [40],which was 1.25 cm2/ml.The test was performed at 37 °C and 220 rpm.The immersion period lasted for 3 days (Fig.1A),14 days without immersion buffer refreshing (Fig.S2A),as well as immersion with buffer refreshing every 2 days (Fig.S2B).The pH value was measured daily for continuous 14 days after immersion.The initial pH value was set to 7.4.In addition,the weight changing on days 3,7,and 14 were recorded.

2.7. Histological observation of rabbit viscera

The heart,liver,spleen,lung,and kidney were harvested at week one and week two after surgery.After fixation,the sample was embedded with paraffin.The 5 μm sample was stained with hematoxylin-eosin(H&E)solution.The histological changes were observed and compared.

2.8.In vivo degradation

To standardize the assessment,a straight 10 mm ZG21 wire was fixed along with the long digital extensor graft in the femur tunnel of rabbits.On days 3,7,and 14,three rabbits were sacrificed at each time.The harvested femurs were fixed by Formalin buffer (10%) and Ethanol (70%).According to the published protocol [19],Micro-CT scanned the samples to measure the wire’s volume.The resolution was 35 μm.

2.9. Scanning electron microscope/energy-dispersive X-ray spectroscopy (SEM/EDS) analysis of degradation process

After CT scanning,the samples were embedded with Methyl methacrylate(MMA).According to the published protocol [19],samples were fixed on the carriers by conducting resin and scanned by scanning electron microscopy (SEM,S3400N,HITACHI,Japan).The scope was amplified 100 and 500 times.The EDS map analysis and EDS line analysis were performed to detect the distribution of the elements.The target elements were Mg,Zn,Gd,Ca,and P.

2.10. ZG21 extract preparation

According to ISO10993-5 [40],the ZG21 plate was immersed into Modified Eagle’s Mediumα(MEMα,chemical composition listed in Table S7) with Fetal Bovine Serum(FBS) (10%).The volume of the extraction medium was calculated as 1.25 cm2/ml.Then,the extraction system was incubated at 37°C and CO2(5 vol%).After 72 h of immersion,the extraction was collected and diluted as 100,50,and 10%specimens.Finally,the pH value and concentration of Mg,Zn,and Gd ions were recorded.

2.11. MTT tests

According to ISO10993-5 [40]and our published protocol [41],the pre-osteoblast cell line MC3T3E1 was cultured with the Modified Eagle’s Mediumα(MEMα) with Fetal Bovine Serum (FBS) (10%) and seeded in a 96-well plate with 500 cells with 100 μl medium in each well.For extract cytotoxicity text,groups were treated with incubated medium,10% extraction,50% extraction,or 100% extraction.At the test time point,50 μl MTT solution was mixed with each sample and incubated for 4 h.Then,use 200 μl Dimethyl sulfoxide (DMSO) to refresh the well.After the reaction,the absorbance value was read at the 570 nm spectrum.On days 1,3,and 5,the MTT test was performed.The cell viability higher than 75% indicated a light cytotoxicity level.

2.12. Animal study design

Sixty-four New Zealand male rabbits were used forin vivostudy,with 32 in the control group and 32 in the ZG21 group.At weeks 3,6 and 9,eight rabbits in each group (n=8) were sacrificed for the radiological and histological test.At week 9,extra eight rabbits in each group were sacrificed for the mechanical test.In addition,twelve rabbits were used in the degradation test.All the animal study was approved by the ethics committee of the Chinese University of Hong Kong(Reference number: 17/179/MIS-5-C).

2.13. Surgical protocol

The rabbits were injected intramuscularly with 1 ml xylazine and 1 mL ketamine for anesthesia induction and intravenously with pentobarbital (1%) for anesthesia maintenance.An incision was made along the patella tendon’s lateral edge to explore and harvest the long digital extensor tendon.The proximal tendon graft was sutured using ZG21 wire(diameter 0.15 mm)guided by an 18-gage needle.The graft was sutured with an absorbable suture (4–0 VICRYL PLUS (ANTIBACTERIAL) VIOLET 27′′RB-1 TAPER,36 EA/BX,Johnson &Johnson) in the control group.The suture range was 10 mm in each tunnel,and the density was 2 mm between stitches.Then,the anterior cruciate ligament was exposed and completely dissected.Bone tunnels were drilled via the transtibial technique.The diameter of the tunnels was 2.5 mm.The tendon graft was transplanted from distal to proximal through the tunnel.The tendon graft’s distal end was sutured on the medial collateral ligament by 5–0 non-absorbable suture (Johnson &Johnson),and the proximal end was secured on the anatomical femur insertion of the long digital extensor tendon.The joint capsule and wound were closed by a 5–0 absorbable suture (Johnson &Johnson).After surgery,the rabbit was injected with penicillin intramuscularly to prevent intraarticular infection.The rabbit was free to move in the cage until sacrificed.The operated rabbit was examined twice a week for avoiding infections or injuries.

2.14. Sample harvesting and fixation

Following the protocol of an early published work [19],the rabbits were sacrificed by intravenous injection of pentobarbital (10%).The transplanted graft was exposed,and both femur and tibia enthesis was detected.The knee joint was amputated from the distal femur and distal tibia by an oscillating saw.The harvested knee joint was fixed with Formalin buffer solution (10%) for 48 h.Then the sample was immersed in ethanol (70%) for storage.

2.15. Gross morphology and semi-quantitative assessment

When harvesting samples,the condition of graft,enthesis of aperture,range of motion(ROM)and stability,and articular surface condition were detected [42]and assessed by macroscopic assessment tool Oswestry Arthroscopy Score (OAS)[43](Table S1).

2.16. Scanning protocol of Micro-CT

According to the published protocol [19],the Viva CT-40(Scanco Medical AG,Swiss) was used for radiological evaluation.The scan parameter was 55 kVp and 90 μA.The tube diameter was 35.8 mm and the slide resolution was 35 mm.The calibration was 55 kVp and 1200 mg HA/ccm.The evaluation task was Bone Trab.Morphometry.The gauss sigma was 1.2 and the gauss support was 2.The evaluation threshold was 150–220.The femur-tibia complex was scanned.The average transverse area was measured by image J.The tunnel area,tunnel bone volume to tissue volume ratio (BV/TV),and tunnel bone density were measured.

2.17. Decalcified histology

According to a published protocol [44],the samples were immersed in 9% Formic Acid,which was refreshed twice a week.The decalcified process was controlled according to the needle hardening test and X-ray observation.After trimming and dehydration,the samples were embedded with paraffin.After paraffin embedding,5 μm-slides were cut for Hematoxylin &Eosin (H&E) staining,Fast green &Safranin Orange staining,and polarized light observation.

2.18. Immunochemistry staining (IHC)

According to the published protocol [20],the 5 μm paraffin slides were dehydrated.The antigens were repaired by citrate acid (pH=6.0) and blocked by bovine serum albumin(5%) with Triton (1%).After washing,samples were mixed with primary antibody (BMP2,guineapig,1:40),(Wnt3a,rabbit-HRP,1:100),(VEGF,mouse,1:100),(Type II collagen,mouse,1:100),(TGFβ3,rabbit-HRP,1:100) overnight at 4 °C.Then,the samples were mixed with secondary antibodies (goat to guineapig/goat to mouse,HRP,1:400).After H2O2blocking,the samples were reacted with 3,3′-Diaminobenzidine tetrahydrochloride (DAB).The samples were stained with hematoxylin and scot tap water.Then,the slides were dehydrated and sealed by resin mounting.The histological images were collected by microscope (Leica CTR6500,Leica Microsystem,Germany).The area of positive staining area was measured.

2.19. Non-decalcified histology

According to the published protocol [19],the original MMA solution was washed by washing solution and distilled water.Then,the cleaned MMA solution was dehydrated by CaCl2(Pellets).Mixed with Butyl methacrylate,Methyl benzoate,and Polyethylene glycol-400,MMA solution I was synthesized.Added benzoyl peroxide powders,MMA solutions II and III were gained.After dehydration,the samples were immersed in MMA solutions I,II,and III.Added N,N-dimethyl-p-toluidine to accelerate the reaction.After 48 h,the MMA polymerization process was done.After embedding,200 and 10 μm slides were cut for Stevenel’s blue &Van Gieson &Alizarin Red staining,and Goldner trichrome staining.

2.20. Double fluorescence deposition for assessment of mineralization rate

According to the published protocol [17],4 rabbits for each group and time points were injected with fluorescence reagents.The Calcein green was injected 10 days ahead of sacrificing day.While Xylene orange was injected 3 days ahead of sacrificing day.The MMA-embedded samples were cut into 200 μm and observed by fluorescence light.The mineralization rate (MAR) was calculated as the average distance between the green line and red line divided by 7 days.

2.21. Histological semi-quantitative scoring

A semi-quantitative score system was used to evaluate the histological observation according to the published paper[20](Table S2).

2.22. Biomechanical test

According to a published protocol [18],the femur-grafttibia complex was harvested.Except for the tendon graft,all the knee joint’s connective tissue was removed.The complex was mounted on a mechanical machine to 90 (H25K-S,Hounsfield Test Equipment LTD,Surrey,UK).The complex was stretched with a 0.1 N pre-load and 40 mm/min speed until failure.The ultimate loading,stiffness,absorbed energy,and strength were recorded for statistical analysis.According to the demonstration of Fig.1G,the ultimate loading was recorded when the mechanical failure happened.The stiffness was the gradient of the curve’s increasing section.The absorbed energy was the area under the curve.The failure modes were documented.

2.23. Osteogenic differentiation assay

According to the published protocol [37],for MC3T3E1,around 1000 cells for each well were seeded in a 24-well plate cultured with osteogenic medium (DMEM,10–8 M Dexamethasone,5 mMβ-glycerolphosphate,50 μM,Ascorbic acid).The osteogenic medium was refreshed every 2 days.Four groups were set: Control group,Mg ions group(2.5 mM),Mg-Zn-Gd ions group (2.5 mM Mg,0.008 mM Zn,0.00007 mM Gd),and 10% metal extract groups.The induction process terminated when any group presented calcium nodes in all wells as well as 21 days after induction for Alizarin Red staining (ARS).Also,on day 4 after induction,alkaline phosphatase (ALP) staining was performed.

2.24. RT-PCR for osteogenic maker

MC3T3-E1 were plated at a density of 1000 cells/cm2 in a 24-well plate and cultured in DMEM with FBS (10%)until 90% confluence.High-glucose DMEM supplemented with 10% FBS,10 nM dexamethasone,50 μM ascorbic acid,and 20 mMβ-glycerolphosphate was used as an osteogenic induction medium.On day 6 after osteogenic induction,MC3T3-E1 was harvested and homogenized for RNA extraction.Total mRNA was reverse transcribed to cDNA.Then,1 μl of total cDNA of each sample was amplified in a 10 μl reaction system (for 384-well plate)containing the SYBR Green qPCR SuperMix-UDG and specific primers forβ-actin (F-GGCTGTATTCCCCTCCATCG;R-CCAGTTGGTAACAATGCCATGT),tgfβ1 (F-AGACCA CATCAGCATTGAGTG;R-GGTGGCAACGAATGTAGCT GT),tgfβ3 (F-CCTGGCCCTGCTGAACTTG;R-TTGATGT GGCCGAAGTCCAAC),bmp2 (F-GCTTCCGTCCCTTTCA TTTCT;R-AGCCTCCATTTTTGGTAAGGTTT),wnt3a (FAATTTGGAGGA ATGGTCTCTCGG;R-CAGCAGGTCTTCACTTCACAG),βcatenin (F-ATGGACGTGGGCGAACTTTTA;R-CGCCATC CCTGTCAATAATCTG),ocn (F-CTGACCTCACAGATCCC AAGC;R-TGGTCTGATAGCTCGTCACAAG),opn (F-AG CAAGAAACTCTTCCAAGCAA;R-GTGAGATTCGTCA GATTCATCCG),runx2 (F-TTCAACGATCTGAGATTTGTG GG;R-GGATGAGGAATGCGCCCTA),and sp7 (F-ATGGCGTCCTCTCTGCTTG;R-TGAAAGGTCAGCGTATGGCTT) using the ABI 7900HT real-time PCR system(Applied Biosystems,USA).The expression of the target gene was normalized to that of Actb.Relative gene expression was calculated using the 2-ΔCTformula.Five groups were set: control (basal osteogenic induction medium),2.5 mM Mg2+,10% extract,filtered 10% extract,and 2.5 mM Mg2+with 0.2 μg/L basal calcium phosphate (Ca-P) crystal(according to ICP,data not shown).

2.25. SEM/EDS analysis for metal extract

The prepared metal extract (according to mentioned protocol) was centrifugated.The deposition was collected and scanned by SEM/EDS (according to mentioned protocol) to analyze the chemical element of deposition.Also,the immersed metal surface was scanned by SEM/EDS.As a control group,the extract prepared with ion-free water was analyzed.

2.26.Ex vivo assessments of clinical samples

Eight cases were involved.All the hamstring tendons were collected from the operation room of the Prince of Wales Hospital,Hong Kong.The Clinical research ethics was approved by the Joint Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee (Ref no.2018.109).Five cases were used for the mechanical test.Each tendon was trimmed into two equal-sized sections with control and one sutured with ZG21 wire.Following the mentioned protocol,the tensile test was done.The ultimate load and failure mode were recorded.The difference between the pair of sections was calculated.In addition,three cases were used for the histological test.The tendon was sutured with ZG21 wire after harvest and cultured with MEMα(10%FBS) for 24 h.Then the tendon was fixed and embedded with paraffin.With H&E staining,the area of the implanttissue interface was observed.

2.27. Statistics

Data were analyzed using SPSS and GraphPad software.For non-parametric data,Kruskal-Wallis test was used.For parametric data,t-test was used for two-group comparison.One-way ANOVA was used for multi-group comparison if only one time point was used.Two-way ANOVA was used for multi-group comparison if more than one time point was used.The Tukey test was used aspost hocfor both One-way and Two-way ANOVA as the multi-comparison method.Pvalue less than 0.05 was set as statistical significance for all tests.

3.Results

3.1. Material properties guaranteed the feasibility of ZG21 wire used as surgical suture thread

The chemical compositions of the alloy were determined as Mg-2.33Zn-0.84Gd (wt%) by inductively coupled plasma(ICP-AES,iCAP6300,USA) analyzer.The ZG21 wire was manufactured as illustrated (Fig.1A).The mean grain size of the transverse plane was 4.4 ± 1.2 μm while the longitudinal plane was 4.6 ± 0.7 μm (Fig.1B).Under the view of SEM,the knotted ZG21 wire presented intact structure both at the stretch side and compression side (Fig.1D) compared with straight ZG21 wire (Fig.1C).Even when amplified 500 times,no crack was found (Fig.1D).TEM showed a fully coherent interface between the W phase particle and Mg matrix could be observed when the incident beam is parallel to[02]Wzone axes,which could support the high formability and strength of wire.The stoichiometric composition of W phase was Mg3Zn3Gd2.(Fig.1E).The ZG21 wire presented yielding load of 5.32 ± 0.03 N,ultimate load of 5.32 ± 0.05 N,elongation of 23.81 ± 0.72 mm,stiffness of 2.47 ± 0.85 N/mm,absorbed energy of 114.91 ± 1.19 MJ,and yielding stress of 300.44 ± 2.14 MPa,ultimate stress of 305.33 ± 1.27 MPa,strain of 15.86 ± 0.42%,Young’s modulus of 231.22 ± 4.94 MPa,and energy density of 4460.34 ± 178.59 MJ/m3(Fig.1G).In summary,as surgical suture threads,the ZG21 wires presented qualified mechanical properties.

3.2.In vitro degradation of ZG21 wire presented a stable alkaline environment and the degradation products showed low cytotoxicity

For wire immersed in phosphate buffer saline (PBS),the initial pH value was 7.4.During the first three hours,the pH value elevated to 10.Then,the pH value increased slowly to 10.5 after 48 h post-degradation (Fig.2A) and remained stable at that level until the last testing time point (Fig.S1A).Based on this finding,the PBS was refreshed every 48 h,and similar pH changing trend was observed in each 48 h-cycle(Fig.S1B).On day 0,the wire was intact (100 mm length,3 mg).The left weight to the original weight afterin vitrodegradation was 91.8 ± 3.7% on day three,65.1 ± 5.7% on day seven,and 28.4 ± 3.2% on day fourteen.The averagein vitrodegradation speed was 0.15 mg/day.

Fig.2.Degradation and cytotoxicity test results.A.pH value change of in vitro degradation in 48 h, n=3;B.cytotoxicity of ZG21 alloy extraction, n=8;C.weight changing of in vitro degradation, n=4;D.volume changing of in vivo degradation, n=3;E.wire morphology of in vivo degradation scanned by Micro-CT,scale bar: 50 μm;F.H&E staining of rabbit viscera,scale bar 200 μm;For D and E,One-way ANOVA,Tukey test compared with each group;For all figures,∗p <0.05,∗∗p <0.01.

The average concentration of Mg,Zn,and Gd was 25 mM,0.08 mM,and 0.0007 mM,respectively.The 10% alloy extract presented almost 100% cell viability on days 1,3 and 5.The 50% alloy extract showed around 90% survival rate on day 1 and then dropped under 75% on days 3 and 5.The 100% alloy extraction also showed low cytotoxicity on day 1,while the viability collapsed under 50% on days 3 and 5(Fig.2B).According to these results,the ZG21 wire degraded with a stable rate within two weeksin vitro,and its products presented low cytotoxicity.

3.3. ZG21 wire degraded in vivo with controllable speed without jeopardizing other organs

On day 0,the intact wire(10 mm length,0.177 mm3)could be observed.The left volume to the original volume after wire degradation was 82.6 ± 4.4% on day three,49.3 ± 3.3% on day seven,28.2 ± 2.3% on day fourteen (Fig.2D).The averagein vivodegradation speed was 0.09 mm3/day.The wire close to the joint gap degraded faster than the tunnel exit(Figs.2E and S2).In addition,no obvious gas accumulation was found.No histological difference in rabbit viscera was observed between the control and ZG21 at both time points(Fig.2C).Based on these observations,the safety of ZG21 wirein vivodegradation had been proven.Therefore,applying ZG21 wire as surgical suture threads in the ACL reconstruction animal model is safe.

3.4. ZG21 wire implantation was safe for articular cartilage and soft tissue around the knee joint

The standardized ACL reconstruction model was done for 64 rabbits following the mentioned protocol (Fig.3A–C).After surgery,one out of 32 rabbits in the control group showed symptoms of infection.The rabbit was sacrificed and replaced by a new one.The condition of other rabbits was healthy.All the knees showed anterior-posterior stability and normal ROM.At week 3,compared with the control group,most grafts in the ZG21 group partially connected with the surrounding tissue.At week 6,the swollen graft connected with surrounding tissue in both groups.More cartilage formation at the insertion area was found in the ZG21 group.At week 9,with the tissue swollen relief,the macroscopic view was like native ACL in both groups.ZG21 group showed a relatively better outcome with more tissue connection and cartilage matrix at the tunnel aperture.(Fig.3D).The ZG21 group showed a higher macroscopic score than the control group at week 6(Fig.3E).In summary,bunching ZG21 wire with tendon graft and implanting it into the bone tunnel was safe for articular cartilage and soft tissue around the knee joint.

Fig.3.Anterior cruciate ligament (ACL) reconstruction in rabbits and macroscope of each timepoint.A.ACL reconstruction protocol in rabbits: 1.tendon graft harvesting;2.graft preparation with ZG21 wire;3.ACL exposure and dissection;4.tunnel processing;5.graft transplantation and fixation;6.joint capsule closure;B.ZG21 wire graft suturing skill;C.graft suturing standard;D.macroscope in each time point;E.semi-quantitative score, n=8,Two-way ANOVA,Kruskal-Wallis test,∗p <0.05,∗∗p <0.01.

3.5. ZG21 wire degradation stimulated more new bone formation surround the tunnel

Morphologically,the ZG21 group presented more intact tunnel walls at weeks 3 and 6 than the control group,especially at the tibia site.In addition,no significant gas effusion was observed in all samples (Fig.4A).Quantitatively,at weeks 3 and 6,the femur tunnel area in the ZG21 group was 28% (n=8,p<0.01) and 33% (n=8,p<0.05)smaller than the controls respectively.While the tibia area in the ZG21 group was significantly 28,40,and 42% smaller than control at week 3 (n=8,p<0.01),week 6 (n=8,p<0.05),and week 9 (n=8,p<0.05),respectively.(Fig.4B).

Fig.4.Micro-CT assessments of peri-tunnel bone formation.A.tunnel frontal position and bone formation around the tunnel,red scale bar: 5mm;B.tunnel area, n=8;C.femur tunnel BV/TV, n=8;D.tunnel BMD, n=8;for BCD,black bar: significant difference between groups,blue bar: significant difference between time points of control group,red bar: significant difference between time points of ZG21 group,Two-way ANOVA,Tukey test,∗p <0.05,∗∗p<0.01.

For peri–tunnel bone formation,the ZG21 group presented 69,72,and 39% higher femur tunnel BV/TV than the control at week 3 ((n=8,p<0.01)),week 6 (n=8,p<0.01),and week 9 (n=8,p<0.05).Also,ZG21 group presented 85,81,and 51% higher BV/TV at tibia tunnel at week 3 (n=8,p<0.05),week 6 (n=8,p<0.05),and week 9 (n=8,p<0.05).(Fig.4C).For bone mineral density (BMD BV)measured at both femur and tibia intra-tunnel bone formation,no difference with statistical significance was found at all the time points.(Fig.4D).All the detailed data were shown in Table S3.Based on the radiological findings,ZG21 wire degradation could stimulate more new bone growth into the bone tunnels to avoid tunnel enlargement during the healing process.

3.6. ZG21 wire degradation initially promoted bone formation at the early healing phase and subsequently benefited fibrocartilage-like tissue formation at the late healing phase

At week 3,the control group presented disorganized fibrovascular-like tissue at the bone tendon interface with less gap.ZG21 group presented fibrovascular-like tissue with loosened alignment.Chondrocytes stained in blue (Stevenel’s blue) or red (Safranin O) scattered along with the wellconnected interface.At week 6,the control group presented better-organized collagen fibers bridging the interface with some gaps.For the ZG21 group,immature fibrocartilage-like tissue bridging tissue to bone was found.At week 9,immature fibrocartilage-like tissue and Sharpy-like fibers could be observed in the control group.ZG21 group presented wellorganized collagen fibers with more mature fibrocartilages structure,mimicking normal bone-tendon insertion (not as direct as normal bone-ligament enthesis yet).(Fig.5A).The semi-quantitative score of the ZG21 group was higher than the control group at week 6 (n=8,p<0.01) (Fig.5B).ZG21 group presented 52% higher tunnel bone mineralization rate (MAR) than control group at week 6 (n=4,p<0.05).(Fig.5C and Table S4).The histological findings indicated that ZG21 degradation could promote local bone formation during the early healing phase and benefit fibrocartilage-like tissue formation during the later healing phase.The fibrocartilage-like tissue played important role in buffering the mechanical loading in case the healing interface was destructed.

Fig.5.Histological assessments.A.histological observation,from left to right: Masson’s Goldner trichrome (scale bar:200 μm),Stevenel’s blue &Von Gieson&Alizarin red S (scale bar:200 μm),Calcein green &Xylene orange (scale bar:800 μm),Hematoxylin &Eosin (scale bar:200 μm),Safranin O &Fast green(scale bar:200 μm),Polarized light (scale bar:200 μm),B: bone,T: tendon graft,arrow: fibrocartilages;B semi-quantitative score, n=8,top bar: 75% value,middle bar: median value,bottom bar: 25% value,Two-way ANOVA,Kruskal-Wallis test;C.mineralization rate of 7 days, n=8,Two-way ANOVA,Tukey test;black bar: significant difference among groups,red bar: significant difference among time points of ZG21 group,∗p <0.05,∗∗p <0.01.

3.7. ZG21 wire degradation enhanced the incorporation strength at the bone-tendon interface

The femur-graft-tibia complex’s mounting angle was 90(Fig.6A).At week 9,the ZG21 group presented a higher ultimate load than the control group,which was 62.1 ± 3.8 N–49.3 ± 3.4 N (n=8,p<0.05).Also,the ZG21 group presented higher stiffness than the controls,which was 22.5 ± 0.8 N/mm–19.6 ± 1.0 N/mm (n=8,p<0.05).Furthermore,according to the measured cross-sectional area(CSA) of the graft (tunnel area of aperture),the strength of the graft was calculated.The ZG21 group’s graft strength was 30.6 ± 2.9 MPa which was higher than 19.8 ± 2.2 MPa in the control (n=8,p<0.01) (Fig.6B).The failure mode reported that 5 samples (tibia 4,both 1) presented graft pullout and 3 samples presented graft rupture in the controls,while 3 samples (tibia 2,femur 1) failed as graft pull-out and 5 samples as graft rupture in the ZG21 group (Fig.6C and Table S5).Generally,graft rupture indicated better bonetendon junction healing while graft pull-out indicated poorer bone-tendon junction healing.As the gold standard to evaluate healing outcomes,the mechanical test’s results indicated that ZG21 wire degradation could enhance the incorporation strength at the healing interface.

Fig.6.Biomechanical test.A.tensile test jigs mounted with femur-graft-tibia complex;B.ultimate loading,stiffness,absorbed energy,and graft strength,n=8, t-test,∗p <0.05,∗∗p <0.01;C.failure modes.

3.8. ZG21 wire degradation triggered the local Ca-P accumulation

After 3-day degradation,the alloy substance remained intact.A passivation layer constituted by Ca-P surrounding the ZG21 alloy was detected.(Fig.7A).After 7-day degradation,only part of the ZG21 alloy remained with a high signal.Ca-P deposited locally after Mg degradation (Fig.7B).After 14-day degradation,the Mg wire was completely degraded and replaced by Ca-P.New bone formation was induced nearby.(Fig.7C).These findings were confirmed by ARS staining(Fig.7D and E).The accumulation of Ca-P crystals with degradation might play a crucial role in bridging Mg degradation and local bone formation.The quality of Fig.7D and E was relatively low because the Ca-P crystals were too brittle to keep intact when attached to the glass slides.Observing the sample under microscopy was the only way to record this important evidence,and it was difficult to focus on the rough sample surface.

Fig.7.Dynamic tracing in vivo degradation.A.SEM-map and liner analysis of day 3;B.map and liner analysis of each element of day 3;B.SEM-map analysis of day 7;C.SEM-map and liner analysis of day 14;D.ARS staining during in vivo degradation.E.ARS staining after degradation.For A–C,white arrows: Ca and P deposition,scale bar 100 μm.

3.9. The expression of related growth factors induced by ZG21 wire degradation presented transiting process from osteogenesis to chondrogenesis

The Wnt3a positive area in the ZG21 group was significantly larger than the control group at week 3(n=4,p<0.01).At week 3,the BMP2 positive area in the ZG21 group was significantly larger than the control group at week 3(n=4,p<0.01)and week 6(n=4,p<0.05).The VEGFpositive area in the ZG21 group was significantly larger than the control group at week 3 (n=4,p<0.01).The TGFβ3 positive area in the ZG21 group was significantly larger than that of the control group at week 6(n=4,p<0.001).At week 9,the ZG21 group still presented higher expression(n=4,p<0.05).The Col II positive area in the ZG21 group was significantly larger than the control group at week 9(n=4,p<0.05).(Fig.8 and Table S6).Among these factors,Wnt3a,BMP2,and VEGF promoted the process of endochondral ossification.TGFβ3 and BMP2 regulated the formation of fibrocartilage-like tissue.Col II expression reflected the area of the cartilage matrix.In summary,Ca-P crystal accumulation induced by ZG21 wire degradation directly promoted local endochondral ossification and indirectly regulated the formation of fibrocartilage-like tissue.

Fig.8.Immunochemical staining for Wnt3a,BMP2,VEGF,TGFβ3 and Type II collagen at bone tendon interface.Yellow stained area: positive area,scale bar: 200 μm;black bar: significant difference between groups,blue bar: significant difference between time points of the control group,red bar: significant difference between time points of ZG21 group,Two-way ANOVA,Tukey test,∗p <0.05,∗∗p <0.01.

3.10. Bioactive degradation products of ZG21 wire presented osteogenic potential in vitro

On day 6,the extracts group presented a significantly larger ARS staining area than other groups (n=6,p<0.01 each).On Day 21,calcium nodes could be observed in all groups.The extracts group showed a significantly larger ARS area than other groups (n=6,p<0.01).Compared with the control group Mg ions group showed a smaller ARS staining area(n=6,p<0.01).(Fig.9A).The RT-PCR results have shown that cells cultured with extracts presented higher expression in Wnt3a,βcatenin,opn,and ocn,while cells cultured with 2.5 mM Mg2+presented lower expression.Interestingly,after filtering the deposition of extracts,the mentioned gene expression was close to the Mg2+group.Although adding Ca-P into the Mg2+medium could elevate the gene expression to the level of control,it was still lower than the group of 10% extract (Fig.9B).The SEM/EDS showed the major element in extract deposition was Mg,Ca,and P when immersing the alloy inαMEM with 10% FBS,while only Mg was detected when immersing the alloy in ion-free water.Although the rich Mg and O on the metal surface were detected in both immersion fluids,the crystal was completely different (Fig.9C).Putting together,the ZG21 degradation could promote Ca-P crystals deposition and promote osteogenesisin vitro,which was highly consistent with thein vivoosteogenic potential of ZG21 degradation.

Fig.9.Osteogenesis in vitro.A.ARS staining on day 6 and day 21 of MC3T3E1 cell line; n=6;B.RT-PCR results of osteogenic markers, n=3;One-way ANOVA with post hoc Tukey test,∗∗p <0.01;C.SEM/EDS analysis of extract deposition and metal surface.Scale Bar,left 500 μm;right 50 μm.

3.11. ZG21 wire maintained the mechanical strength of the human tendon without inducing tissue necrosis

For clinical samples,less than 20% ultimate load decrease was found in the ZG21 group compared with the paired control sample (Fig.10A).After 24 hex vivoculturing,no obvious necrosis area was observed at the wire-tendon interface(Fig.10B).Although it was theex vivoassessment,the results reflected that ZG21 wire degradation was not harmful to human tissue.These results indicated the ZG21 wire’s promising translation potential.

Fig.10. Ex vivo clinical samples.A. ex vivo tensile test of human hamstring tendon section;B.H&E staining of human hamstring tendon sutured with ZG21 wire,scale bar 1000 μm (25x) and 200 μm (100x).

4.Discussion

4.1. Bioactive potential of ZG21 wire in the process of bone-tendon junction healing

After the ACL reconstruction,the interface of the tendon graft usually presents narcosis since lacking nutrition,which induces a gap between the tendon and bone.This gap weakens the mechanical stability of the graft and retards the mechanical healing process [7].In this study,to replace the traditional surgical suture for graft preparation,Mg alloy wire was developed as a fast degradation bioactive metallic wire to stimulate bone growth into the gap swiftly.The tunnel bone formation at the early time point was directly related to the degradation process confirmed by the Micro-CT and MAR results.According to histological findings,a better bone tendon interface healing process in the ZG21 group later point could be observed.Also,the difference between groups at week 9 was small since the timepoint was far away from Mg degradation.Hence,the healing process related to wire degradation,early or late,should be discussed as the osteogenic phase and chondrogenic phase.

4.1.1.Ca-P crystals accumulation promoted endochondral ossification surrounding the bone tunnel at the early healing phase

Somein vivostudies reported the Wnt/β-catenin pathway was closely related to the Mg implantation-promoted osteogenesis [45,46].As forin vitro,Hung et al.indicated that implants derived Mg2+stimulate Wnt3a/β-catenin to promote bone formation [35],whereas some studies reported Mg2+reverse or suppress osteogenesisin vitro[36,37].Considering the evidence reported by these studies,our team tried to find the clues based on the degradation process and products.Gonzalez et al.described that Ca-P salts could be attached to the Mg(OH)2during the passivation period [47].In our study,the SEM/EDS furtherly detected Ca-P crystals deposited gradually with the Mg wire degradation.According to a published study,Ca-P crystals could elevate the Wnt3a/βcatenin signaling pathway to promote chondrocyte hypertrophy[48].Our IHC staining also found a high-level expression of Wnt3a related to Mg degradation.Besides,the degradation process induced an alkaline environment which might potentially provide an osteogenic-friendly niche[49,50].In addition,Cheng et al.reported Mg interference screw degradation led to elevated levels of BMP2 and VEGF [20],which was confirmed by this study (Fig.8).Upregulating both BMP2 and VEGF promoted osteogenesis while BMP2 with downregulating VEGF promoted chondrogenesis [51].In this study,at the osteogenic phase,the high-level expression of VEGF and BMP2 accompanied with Wnt3a promoted bone formation.

At the chondrogenic phase,without the expression of VEGF,the BMP2 turned to benefit fibrocartilage formation(Fig.8).Consistent within vivofindings,the Ca-P was detected in the metal extract deposition.Our RT-PCR results showed that pre-osteoblasts cultured with 10% extract expressed high level of wnt3a,βcatenin,ocn,and opn.Interestingly,these makers expression significantly decreased after filtering the deposition from extract,indicating the Ca-P played a crucial role in osteogenesis.Although adding Ca-P into culture medium could offset the suppressive effect of Mg2+,the gene expression was still lower than the group of extract,indicating there should be other factors,such as the alkaline environment and optimal Ca-P crystals size in the metal extracts to enhance the osteogenic effects of Ca-P(Fig.9).

4.1.2.More fibrocartilage-like tissue was detected after the endochondral ossification stage

Unlike the natural anatomy,there is no typical intra-tunnel bone tendon insertion development with growth.From an embryology perspective,enthesis development is driven by endochondral ossification [52,53].Generally,the relatively vertical fiber alignment with a strain-compression mechanical environment tends to promote fibrocartilage formation [54].For fibrocartilage formation,Ihh/PTHrP loop may play a crucial role to modulate chondrogenesis and local homeostasis[44,55,56].Carbone et al.assumed compressive mechanical stimulation could trigger Ihh/Gli pathway to suppress chondrocyte hypertrophy,thus maintaining the fibrocartilage structure [57].As the results showed,the Mg degradation promoted bone formation at the bone tendon interface.The newly formed bone matrix compressed tendon bundles,which provided a fitful mechanical environment.Also,the expression of SOX9 and TGFβ3 was essential to maintain the cartilage matrix [58–60].Confirming by histological and IHC staining,at the osteogenic phase,no fibrocartilage-like tissue was found since high expression of Wnt3a,BMP2,and VEGF promoted chondrocyte hypertrophy and suppressed SOX9 and TGFβ3 expression[48,51].Accompanied by lower expression of Wnt3a and VEGF,the increasing mechanical stimulation modulated by Ihh/PTHrP loop [44,55,56]and expression of SOX9 [58]and TGFβ3 (Fig.8) contrarily suppressed chondrocytes hypertrophy to keep the structure of fibrocartilage(Fig.11).

Fig.11.The potential mechanism of Mg alloy wire degradation facilitates ACL reconstruction.During the osteogenic phase,Mg wire degradation specifically triggered Ca-P crystals deposition to stimulate endochondral ossification.In the chondrogenic phase,the newly formed bone provided a beneficial mechanical and biological microenvironment for fibrocartilage regeneration.

4.2. Material advantages and safety of ZG21 wire

Although the traditional Mg alloys,such as AZ31,improve mechanical strength more than pure Mg wire [61,62],the contained Al element potentially jeopardizes neurons[63].To replace Mg-Al alloy,many research teams developed Mg-Zn alloys [64]and Mg-RE alloys [65].Some RE elements like Pr and Ce are toxic,while the Gd and Nd in safe dosages are biologically acceptable [65].The precipitated nanoscale I-phase of Mg-Zn-Gd-based alloys can facilitate both strength and elongation of Mg alloy [66].Based on the Mg-2.0Zn-RE structure,Miao et al.reported that Mg-2.0-1.0Gd (0.84 wt.%),behaved best in microstructures,corrosion resistance,and mechanical properties [34].Applying the cold-drawing technique,the transverse grain size and longitudinal grain size of ZG21 wire were similar,which significantly elevated the compression stress close to the level of tensile stress,which guarantees the formability as surgical suture threads [67].High-resolution TEM indicated the interfaces between nanoscale W phase particles and Mg matrix possessed good atomic match and low lattice mismatching strain,which were generally considered unlikely to be the source of micro-cracks during the deformation process.These microstructures presented sufficient mechanical strength of 0.15 mm diameter ZG21 wire compared with thicker Mgbased wire [33,68].In bone metabolism,Zn contributes to protecting bone mass [69].In this study,the Zn concentration of the extract was lower than the physiological level.Rear Earth elements like Gd will benefit to control the degradation rate of Mg-based alloys greatly and it will be safe enough if the releasement is within the allowable scope for the human body in unit time.Just for this reason,the German company-Syntellix AG,has been awarded the CE certificate for their Magnezix Mg alloy bone screws which contain Gd in 2013,has been widely used as biodegradable bone implants in European Countries [70].According to our previous study,the degradation products of the Gd-rich nanoparticles are determined to be Mg3Gd [39].These compounds can be decomposed into micron-scaled particles further with time and can be recognized,swallowed,and digested by Macrophages [71].In clinical,the dosage of Gd in contrast medium is 0.1 mmol/kg or even higher [72].In this study,the Gd dosage is about 0.01 mmol/kg which is significantly lower than clinical application.SEM/EDS also presented a relatively low level of Gd detaining at bone and tendon graft(Fig.7).

4.3. Limitations

There are a few limitations in this study.Firstly,the gait analysis was unfeasible to rabbits since looping was difficult to validate.Secondly,the level of SOX9 and Ihh/PTHrP related factors were not detected since no primary antibody reacted with rabbit has been developed and is available commercially.Also,technically it was difficult to measure the compression force between tendon and bone.Lacking these data,we could not demonstrate the ZG 21 wire degradation’s indirect biological effect in the later healing phase with iron evidence.Thirdly,scanning the wire directly might cause some errors since it was not at a horizontal plane.It was because the structure of the crosssection sample without embedding could be destroyed by the cutting blades,while the wire degraded during the embedding process.Therefore,an axial plane was difficult to present.

5.Conclusion

The high formability Mg alloy wire (ZG21) was developed and tested for its feasibility for tendon graft bunching in ACL reconstruction.In conclusion,this study elucidated the following points.Firstly,the ZG21 wire could be applied as surgical suture threads according to its mechanical properties.Secondly,The ZG21 wire degraded steadily and its degradation process as well as its products were safe for implantation.Thirdly,suturing ZG21 wire around the tendon graft could promote bone-tendon junction healing with more new bone growth,more fibrocartilage-like tissue formation,and stronger incorporation strength.Finally,ZG21 degradation triggered local Ca-P crystals accumulation to enhance the intra-tunnel endochondral ossification,and successively promoted bone tendon interface healing with the regeneration of fibrocartilage connection and stronger mechanical integrity of the healing complex.

Declaration of competing interest

We declare that we have no conflicts of interest in the authorship or publication of this contribution.

CRediT authorship contribution statement

Xuan He:Conceptualization,Methodology,Formal analysis,Investigation,Data curation,Writing– original draft.Ye Li:Conceptualization,Methodology,Formal analysis,Data curation,Writing– original draft.Hongwei Miao:Conceptualization,Methodology,Investigation,Writing– original draft.Jiankun Xu:Methodology,Formal analysis,Data curation.Michael Tim-yun Ong:Methodology,Formal analysis,Data curation.Chenmin Wang:Methodology,Formal analysis,Data curation.Lizhen Zheng:Methodology,Formal analysis,Data curation.Jiali Wang:Methodology,Formal analysis,Data curation.Le Huang:Methodology,Formal analysis,Data curation.Haiyue Zu:Methodology,Formal analysis,Data curation.Zhi Yao:Methodology,Formal analysis,Data curation.Jie Mi:Methodology,Formal analysis,Data curation.Bingyang Dai:Methodology,Formal analysis,Data curation.Xu Li:Methodology,Formal analysis,Data curation.Patrick Shu-hang Yung:Supervision,Writing– review &editing,Project administration,Data curation.Guangyin Yuan:Supervision,Writing– review &editing,Data curation,Project administration,Funding acquisition.Ling Qin:Supervision,Writing–review&editing,Data curation,Project administration,Funding acquisition.

Acknowledgments

This study is fully supported by Theme-based research scheme of Hong Kong Research Grant Council (RGC Ref:T13-402/17-N) and National Natural Science Foundation of China (No.U1804251).

Supplementary materials

Supplementary material associated with this article can be found,in the online version,at doi:10.1016/j.jma.2022.12.006.