Porcine fibrin sealant combined with autologous chondrocytes successfully promotes full‐thickness cartilage regeneration in a rabbit model

Abstract Xenogeneic porcine fibrin sealant (PFS), derived from porcine blood, was used as a scaffold for cartilage tissue engineering. PFS has a porous microstructure, biocompatibility and degradation, and it provides a perfect extracellular matrix environment for the adhesion and proliferation of chondrocytes. Recently, PFS in combination with autologous chondrocytes (ACs) were used to study the microstructure of PFS scaffolds and promotion effect on the proliferation and migration of ACs. In this study, we investigated the effects of PFS in combination with ACs on the healing of cartilage defects in rabbits. A full‐thickness cartilage defect was made in the femoral trochlear in rabbits, subsequently, three surgical procedures were used to repair the defect, namely: the defect was treated with microfracture (MF group); the defect was filled with PFS alone (PFS group) or in combination with ACs (PFS + ACs group); the unrepaired cartilage defects served as the control group (CD group). Three and 6 months after the operation, the reparative effect was evaluated using medical imaging, gross scoring, pathological staining, biomechanical testing and biochemical examination. The PFS group showed a limited effect on defect repair, this result was significantly worse than the MF group. The best reparative effect was observed in the PFS + ACs group. These results hinted that PFS in combination with autologous chondrocytes has broad prospects for clinical applications in cartilage tissue engineering.


| INTRODUCTION
Osteoarthritis (OA) is a prevalent musculoskeletal disease worldwide.
Approximately 303 million people suffer greatly from OA (Kloppenburg & Berenbaum, 2020). OA affects any joint in the body, especially load-bearing joints, such as the hip and knee. OA is the main cause of physical disability, and it places a significant burden on patients and the health care system, with great socioeconomic cost (Hunter & Bierma-Zeinstra, 2019). It is possible to restore joint function and prevent the further development of OA in knee joints with cartilage damage by repairing the cartilage (Taheem et al., 2020). Articular cartilage is a highly hydrated viscoelastic connective tissue without innervation, vascularization, or lymphatic circulation. It contains a low density (accounting for 1%-5% of the total tissue volume) of chondrocytes with poor proliferative activity and a dense anti-adhesion extracellular matrix (ECM) around the cells. Three type of cartilage regeneration techniques are used clinically, including microfracture (MF), autologous chondrocyte implantation (ACI), and osteochondral autogenous transplantation (OAT; Camarero-Espinosa et al., 2016). However, these methods have numerous limitations and deficiencies. For example, MF leads to the formation of fibrocartilage with poor mechanical properties (Gao et al., 2018). The fixation of ACI is not firm, and there is a defect of a large number of cells lost (McCarthy et al., 2016), and the source of donor in OAT is limited, OAT has poor integration with surrounding cartilage (Nie et al., 2020).
Recently, tissue-engineered cartilage shows incomparable advantages. There are generally two types of biomaterials used for cartilage tissue engineering, synthetic polymers produced by chemical reactions and natural polymers produced by living organisms.
Compared to synthetic polymers (e.g., polylactide, polyglycolide, and polycaprolactone), natural polymers (e.g., fibrin, collagen, and chitosan) have many unique advantages, such as biocompatibility, biodegradability, porosity and adjustable mechanical properties (Bao et al., 2020). Fibrin is a natural polymer that provides a porous matrix to easily promotes cell adhesion, maintain cell vitality and promote cell proliferation (Whelan et al., 2014). The United States Food and Drug Administration (FDA) approved fibrin sealant (FS) for clinical use in 1998 (Roberts et al., 2020), and surgeons evaluated it as excellent. Most of the fibrinogen in commercial FS is extracted from human serum. Despite the availability of advanced virus detection technologies, the potential risk of viral disease transmission remains (Roberts et al., 2020).
Xenogeneic porcine fibrin sealant (PFS) was derived from porcine blood. Pigs have high homology with humans, and they exhibit similar anatomy, physiology and genetics. The safety performance of porcine plasma is greatly improved compared to cattle blood, snake venom and other animal sources. The risk of virus transmission is also greatly reduced. Now, PFS is widely used in clinical surgery for haemostasis. Therefore, in this study, we used xenogeneic PFS as a scaffold for cartilage tissue engineering to fill rabbit knee cartilage defects.
Chondrocytes as seed cells of tissue engineering cartilage have a higher chondrogenic capacity and exhibit greater secretion of growth factors and cytokines. The FDA approved autologous cell transplantation therapy (Colombini et al., 2019). And it has been widely used in tissue engineering. Then, the aim of the present study was to systematically evaluate the reparative effects of xenogeneic PFS in combination with or without autologous chondrocytes (ACs) on the healing of cartilage defects in rabbits using medical imaging, gross scoring, pathological staining, biomechanical testing and biochemical examination.

| Animals
Our experimental procedures were performed in compliance with the standards for the care and use of laboratory animals. 50 New Zealand white rabbits (5 months old, weighing 2.5−3.0 kg) were purchased from the Center of Animal Experiments and the experimental scheme was approved by the Animal Experiment Ethics Committee (Chinese People's Liberation Army General Hospital). A full-thickness chondral defect was created in both knees of 30 rabbits. The left knee joint cartilage was collected from the other 20 rabbits to isolate and culture chondrocytes that were subsequently implanted into the defect in the right knee joint.

| Isolation and culture of autologous chondrocytes
Twenty rabbits were anaesthetized via an intramuscular injection of 10% xylazine hydrochloride (0.2 ml/kg, Huamu Animal Health Products Co., Ltd.), and the left knee joint cartilage was peeled away using a scalpel (Figure 1a). All rabbits were returned to the cage for the subsequent implantation of autologous chondrocytes (ACs) in the right knee joint.
These two solutions (A + B) were mixed at a 1:1 volume in a doublebarrelled syringe (with two plungers, a connecting Y-piece and casing application) until a hydrogel-like PFS was acquired after 2-4 min at room temperature.

Structure of porcine fibrin sealant
Three PFS hydrogels were randomly prepared according to the above method. The configured PFS was fixed using 2.5% glutaraldehyde and 1% osmic acid tetroxide.

| Cell proliferation detection
According to the instructions of Cell counting kit-8 (CCK-8, Dojindo Laboratory, Japan), we selected low concentration of chondrocytes for proliferation experiment. Briefly, we mixed 50 μl P3 AC suspension (1 � 10 5 /ml) with 10 μl solution B and added it to the 48-well plate, followed by 60 μl solution A to the well plate. After the PFS solidified into a hydrogel, 600 μl of DMEM/F12 (with 20% FBS) was added to each well. The same number of cells in 600 μl of DMEM/F12 (with 20% FBS) were incubated as the blank control group (three independent samples were tested in each group at each time). The CCK-8 assay (each sample had 3 vice-holes) was used to estimate the cell proliferation rate within the PFS at days 0, 3, 5, and 7. "Vice-holes" were defined as: pipetting 300 ul solution (10% CCK8 reagents in DMEM) of each sample from the 48-well plate and distributing evenly into 3 holes of 96-well plate (100 μl/hole) to measure the absorbance.

| Articular cartilage defect rabbit model
All rabbits were anesthetized via an intramuscular injection of 10% xylazine hydrochloride (0.2 ml/kg). The knee joints were disinfected, and the skin was incised to expose the femoral trochlea. A fullthickness chondral defect (3.5 mm in diameter, 1 mm in depth) in the center of the trochlear groove was drilled without damaging the subchondral bone plate using a special punch ( Figure 1c). The rabbits were divided into four groups: the defects in both knee joints of 10 rabbits were not treated (CD group); the defects in both knee joints of 10 rabbits were filled with 0.2 ml of PFS alone (PFS group); the defect in the both knee joints of 10 rabbits were treated with MF (the MF group); and the defects in the right knees of 20 rabbits was filled with 0.2 ml of a mixture of PFS and 1 � 10 7 /ml ACs originating from the rabbit's left knee joint (PFS + ACs group; Figure 1e). For the MF group, a 0.8-mm hole was drilled into the subchondral bone, and bone marrow was introduced into the defect ( Figure 1d). All animals were euthanized 3 and 6 months after surgery to collect knee joint samples for further evaluation.

Magnetic resonance imaging examination
A small-animal Magnetic resonance imaging (MRI) system (Bio-Spec70/20USR) was used to evaluate the reparative effects on cartilage. The scanning sequence was T2, and the scanning parameters were set as described in a previous report (Lu et al., 2018). Three blinded independent musculoskeletal radiologists analysed all MRI images using the Whole-Organ Magnetic Resonance Imaging Score (WORMS), which includes cartilage, marrow abnormality, and bone cysts (Peterfy et al., 2004).

Micro-computed tomography (μ-CT) analysis
After MRI, the repaired areas of the knee joints were exposed. The degree of subchondral bone repair in the knee joints was evaluated using a μ-CT imaging system (EXPLORE LOCUS, GE Healthcare). The scanning parameters were set as previously reported (Lu et al., 2018).
Axial, coronal, and sagittal images of the femoral condyle were reconstructed and analysed using GE Health Care MicroView ABA 2.1.2 software. A cylindrical (diameter, 3.5 mm; height, 1.5 mm) 3D region of interest (ROI) was established to ensure the inclusion of repaired tissue in the defect area in the ROI. Histomorphometric parameters of the subchondral bone, including the bone volume fraction (BVF), trabecular thickness (Tb.Th) and trabecular spacing (Tb.Sp), were calculated and analysed.
Three blinded independent pathologists evaluated images of the staining results using the ICRS II Visual Histological Assessment Scale (Mainil-Varlet et al., 2010).

| Biomechanical testing
The repaired cartilage surface was placed perpendicular to the axis of indentation, and a cylindrical flat-end indenter with a diameter of 1 mm (BOSE3220 EM USA) was used to continuously record the force and displacement applied to the repaired cartilage at a speed of 0.005 mm/s for 300 s. The elastic modulus of the repaired cartilage was calculated from the stress-strain curve.

| Biochemical analyses
The regenerated tissue was removed from the knee defect using a corneal trephine for GAG and total collagen quantification according to the instructions of a GAG kit (GenMed Scientifics, Inc.) and hydroxyproline kit (Nanjing Jiancheng Bioengineering Institute), which was used to quantify the total collagen content.

| Statistical analyses
All experimental data were statistically analysed using SPSS 23.0 software, and the values of the results are expressed as the means ± standard deviation. Date were analyzed by t-test and Oneway analysis of variance (ANOVA) followed by LSD post hoc test.
p < 0.05 was considered a significant difference.

| Characterization of porcine fibrin sealan
PFS is transparent and soft, but it has a certain toughness

| Migration effect of porcine fibrin sealan autologous chondrocytes
The migration effect of PFS (Figure 2g) on ACs was significantly stronger than the control group (Figure 2f), and there was a statistical difference (Figure 2h) between the two groups.

| Magnetic resonance imaging evaluation of cartilage repair
Three months after the operation (Figure 3a), the defect in the CD group was obvious, with an interrupted and sunken linear structure with a low signal on the defect surface and no tissue regeneration. Six months after the operation, the low-signal linear structure in the defect area in the CD group remained unrecovered with no tissue regeneration. The defect area in the PFS group was significantly smaller than before. The defect was completely repaired in the MF and PFS + AC groups. The linear structure with a low signal intensity was restored, and a uniform layer of tissue with a medium signal intensity was observed on the surface of the defect. The WORMS in all groups decreased dramatically (p < 0.05) compared to 3 months after the operation. All scores in the CD, PFS, MF, and PFS + AC groups decreased in sequence with significant differences (p < 0.05), which means that the PFS + AC group exhibited the best reparative effect. Six months after the operation, the defect in the CD group remained obvious with little subchondral bone regeneration. The regenerated tissue in the PFS and PFS + AC groups fused with the surrounding normal subchondral bone, with a smooth defect surface and regular bone trabeculae. These two groups showed no significant differences (p > 0.05). The proliferation of subchondral bone in the MF group was more evident than at 3 months. BVF and Tb.Th were highest in the MF group, and the Tb.Sp was the lowest (p < 0.05), which corresponds to the earlier results.

| Macroscopic observation of cartilage repair
Three and 6 months after the operation, gross observation of cartilage repair in the knee joints of rabbits in each group was performed ( Figure 4a). Three months after surgery, the defect in the CD group was clearly visible, and there was almost no tissue regeneration. The defect in the PFS group was partially filled with regenerated tissue, which had a plane lower than the normal cartilage with a rough surface and a clear boundary between the new tissue and the normal cartilage tissue. Most of the defects in the MF and PFS + ACs groups were filled with regenerated tissue, with a smooth surface and blurred border. The ICRS scores (Figure 4b) in the CD and PFS groups were significantly lower than the other two groups (p < 0.01).
The score in the CD group was lower (p < 0.05) than the PFS group, and the score in the PFS + ACs group was higher than the MF group. The results of F-H are presented as the mean ± SD (n = 4, *p < 0.05, **p < 0.01) MRI, Magnetic resonance imaging; WORMS, whole-organ magnetic resonance imaging score was slightly lower than the normal cartilage, with a smooth surface and blurred boundary. The defect in the PFS + ACs group was completely filled with transparent and smooth regenerated tissue, which integrated with the normal cartilage tissue. The scores in the PFS + ACs and MF groups were significantly higher than the other two groups (p < 0.01). The score in the CD group was lower (p < 0.05) than the PFS group, and the score in the PFS + ACs group was higher (p < 0.05) than the MF group.

| Biomechanical analysis and biochemical components of regenerated cartilage
Analyses of the elastic modulus (Figure 4c were significantly lower (p < 0.01) than those in the MF and PFS + AC groups. There was no significant difference (p > 0.05) between the MF and PFS + AC groups. Six months after the surgery, the elastic modulus, GAG and total collagen content were increased in all groups (p < 0.05). However, the differences in the elastic modulus, GAG and total collagen content between the CD group and the other groups were the same as 3 months after the operation. Notably, the elastic modulus, GAG and total collagen content increased more rapidly in the PFS + AC group than the MF group, and these values were highest in the PFS + AC group.

| Histological evaluation of regenerated cartilage
Three and 6 months postoperatively, H&E, safranin O/fast green and type II collagen immunohistochemical staining of the repaired cartilage in each group was analysed (Figure 5a). At 3 months, the defect area in the CD group remained concave, without tissue formation and negative GAG and collagen II staining results. The defect region in the PFS group was filled with fibrous tissue with a few disordered fibroblast-like cells. The GAG and type II collagen Six months after the operation, the defect in the CD group was filled with a small amount of fibrous tissue, which was negative for GAG deposition and type II collagen. The new tissue in the PFS and MF groups contained columnar and clustered chondroid cells, nonuniform GAG staining results, and weak expression of type II collagen. However, the new tissue in the PFS group showed uneven thickness and poor continuity compared to the MF group. Notably, the defect in the PFS + ACs group was completely filled with cartilage-like tissue, in which a large number of typical chondrocytes arranged in columns similar to normal cartilage was observed. GAG content and type II collagen expression were consistent with the surrounding normal cartilage, and the integration of the new tissue was continuous and smooth.
The results of the ICRS II Visual Histological Assessment Scale showed the best reparative effect in the PFS + ACs group 3 and 6 months after the operation, with a significant difference compared to the other groups (Figure 5b).

| DISCUSSION
Fibrin is a natural nanoscaffold that is used to initiate haemostasis and providing a temporary structure to promote cellular activity and the deposition of new ECM (Shiu et al., 2014). It has extraordinary biocompatibility, biodegradability, inherent biological activity and many other unique properties, which makes it a very distinctive biomaterial for cartilage tissue engineering (Noori et al., 2017).
SEM images showed that a large number of fibres in the PFS were superimposed on and intertwined with each other to form a 3D reticular structure, which yielded a high area/volume ratio. This unique network structure facilitates cell attachment, cell-matrix interaction, and nutrient or metabolite transport. The fibre thickness, pore diameter and pore depth may be adjusted by changing the concentration of fibrinogen, thrombin and XIII factor (Kurniawan et al., 2014). Kim et al. found that 30 mg/ml fibrinogen and 100 IU/ml thrombin were the best conditions for chondrogenesis. Our study used 30 mg/ml fibrinogen to fill cartilage defects, which is consistent with the Kim et al. study. To make PFS fit better with the shape of the cartilage defect in a liquid state and control the timing of PFS polymerization into gel, we determined that 42 IU/ml thrombin was ideal and had little effect on cell morphology. Lower thrombin activity provides a more suitable environment for cartilage repair (Irwin et al., 2019). Moreover, the change in thrombin concentration had little effect on cell morphology (Kim et al., 2020).  (Yue et al., 2021). Three-dimensional culture in FS did not affect the phenotype of ACs (Bianchi et al., 2019).
Preparing a mixed cell-fibrin hydrogel prevents cell leakage, and the hydrogel does not completely harden in a short time. This approach ensures that the initially expected hydrogel volume stays in the defect and perfectly integrates with the surrounding tissue according to the shape and size of the defect. During the process of polymerization, the gel achieves a uniform cell distribution via the process of rapid immobilization. FS enables mechanical adhesion, and it is beneficial for the biological integration of implants. FS scaffolds carry cells and promote the formation of host-scaffold interfacial tissue (Drobnic et al., 2018). No inflammatory reaction was observed in any of the repaired sites 3 or 6 months after the operation in the PFS and PFS + ACs groups, which supports the perfect biocompatibility of the xenogeneic PFS in the rabbit knee joint. We did not use an additional graft fixation technique to create the defect, even in these highly mobile experimental animals, and PFS maintained its stability in situ, which confirmed its excellent adhesion performance and ability to adapt to cartilage defects. Barros et al. reported that a similar FS derived from snake venom had sufficient viscosity, stability, and no adverse reactions in animal experiments and promoted cartilage repair (de Barros et al., 2016).
The new cartilage formed by fibrin scaffolds has histological characteristics similar to natural cartilage (Cakmak et al., 2013).
The mechanical properties of cartilage tissue largely depend on the composition and structure of the ECM (primarily type II collagen and proteoglycan; Ahmed et al., 2011). Type II collagen provides the strength for cartilage to resist tension, and proteoglycan allows cartilage to resist compression (Vinatier et al., 2009). Our animal experimental results showed that the contents of total collagen and GAG were highest in the PFS + ACs group, and the safranin O/fast green and immunohistochemical type II collagen staining results were the strongest. These results indicate that the regenerated cartilage in this group contained the most type II collagen and proteoglycan, and its mechanical properties should be the YANG ET AL. best, which was confirmed in the biomechanical testing. Although the mechanical properties of PFS are poor, the mechanical properties of the regenerated cartilage in the PFS + ACs groups were significantly better than the traditional MF treatment group.
Khanmohammadi et al. wrapped menstrual blood-derived stem cells in FS to repair osteochondral defects of the rabbit knee joint, and the regenerated cartilage tissue was superior to FS alone in the content of GAG and type II collagen and integration (Khanmohammadi et al., 2019). Choi also used allogeneic chondrocytes and autologous bone marrow cells in combination with fibrin to repair full-thickness cartilage defects in rabbits, and good quality hyaline cartilage was produced (Choi et al., 2018). Almeida et al. used granular cartilage ECM combined with functionalized fibrin hydrogel to form excellent cartilage tissue in vivo and in vitro (Almeida et al., 2016).
The 3D reticular structure of FS has no specific orientation in its natural state. After the application of continuous strain to the viscoelastic FS, it showed a fibre bundle structure that was consistent with the strain direction (Matsumoto et al., 2007). Cell proliferation and matrix mineral deposition in strained FS are also limited in the same direction. The pathological results in the PFS + ACs group in our experiment showed that the vertical arrangement of chondrocytes in the regenerated tissue was consistent with the normal cartilage tissue.
However, PFS alone is limited by rapid degradation and poor mechanical properties, and it cannot provide all of the properties needed for cartilage tissue engineering scaffolds. PFS may be properly processed, chemically modified or combined with other polymers (organic/inorganic nanomaterials) and biomolecules (growth factors) via various processing or synthesis methods or a variety of crosslinking methods to obtain the desired properties.
PFS may be injected with minimally invasive implantation in clinical treatment, which shortens the healing time and reduces patient discomfort, postoperative complications, and medical costs.
Although PFS has some disadvantages, such as poor mechanical properties and fast degradation, these problems may be solved with the development of fibrinogen and thrombin as recombinant proteins and the application of biomolecules binding to fibrin nanomaterials (Noori et al., 2017). PFS will be more widely used in the fields of tissue regeneration in the near future.
In conclusion, PFS has a wide range of sources, low production cost, low risk of disease transmission, unique 3D network structure, good biocompatibility, and a controllable degradation process. It also provides a suitable ECM microenvironment for chondrocytes and promotes chondrocyte adhesion and proliferation. PFS may be perfectly matched with cartilage defects during surgery using a minimally invasive injection, and the regenerated hyaline cartilage tissue is similar to natural cartilage tissue.