1 Introduction

Regenerative medicine is recognized as one of the most promising osteoarthritis treatments and has been extensively studied over the past few decades [1, 2], and platelet-derived therapy is one of its most popular applications [3,4,5,6,7]. Growth factors released by α-granules contained in platelets are a group of bioactive polypeptides with anabolic and anti-catabolic effects that can promote cell production, regulate new cartilage formation and chondrocyte metabolism and differentiation, and promote tissue and cartilage repair [8, 9]

At present, autologous platelet-rich plasma (PRP) is widely used in clinical practice. The advantage of autologous PRP is that there is no rejection problem. After centrifuging the autologous blood, the platelet dense layer is extracted and then injected into the affected area. However, inconveniences remain in clinical use as it is time-consuming and the final PRP growth factor concentration can vary. Many studies have been carried out to ameliorate this disadvantage. The production and extraction of recombinant human growth factors is one of the sources [10], but it provides a single growth factor and carries concerns regarding carcinogenicity [11]. The development of allogeneic platelet-derived preparations has also been discussed. But the allogeneic PRP still has the problem of certain unremoved antigens, which reduces the tissue regeneration capability of the growth factors [12]. Therefore, platelet lysate (PL) is a potential application method that further removes antigens. PL is using physical methods to destroy the membrane of the platelets in PRP. Filter large proteins and platelet fragments and then heat to remove complements, so that possible antigens can be removed and obtained growth factors from platelets. Platelet lysate, as well as other autologous platelet-derived preparations as a source of growth factors and cytokines, have been successfully used in the regenerative engineering of bone and cartilage [13,14,15]. Platelet lysate has been widely used in cell culturing as a replacement for fetal bovine serum to obtain a better cell culturing effect. Autologous PL has been a replacement for PRP in clinical practice. Clinical evidence and products of autologous PL have been launched but allogeneic and xenogeneic PL still needs more research to confirm its effectiveness. Aldén et al. used porcine PL to culture monkey and hamster cells and obtained good differentiation and value-added effects [16]. There is still no paper published on the safety and efficacy of xenogeneic PL in vivo. We designed an experiment to conduct an in vivo study of xenogeneic PL to confirm its safety and efficacy. In the future, it can be developed for the treatment of cartilage damage and arthritis in animals, and further applied to the regenerative treatment of humans and other tissues with xenogeneic PL. Herein, we developed porcine PL containing undetectable antigens such as blood cells and complement. The porcine PL was injected into the rabbits’ joints to observe whether it caused an inflammatory reaction. Then, osteochondral defects were created in rabbits’ knee joints and they were injected with porcine PL to observe the repair of osteochondral defects and the development of knee joint arthritis.

2 Materials and Methods

2.1 PL Preparation

Porcine platelet lysate was produced following the protocol of Jonsdottir-Buch SM et al. [17]. Porcine whole blood was centrifuged and the supernatant was taken, leukocytes were filtered, platelets were lysed by the freeze–thaw method, heated to remove the complement, and filtered with a 0.22 µm pore size filter. The blood cell count was performed and the growth factor concentration of the lysate was detected by TGF-ß ELISA Kit. Growth factors in porcine blood PL include PDGF, EGF, TGF-α, β, VEGF, and FGF, which exist in fixed proportions. In this study, TGF-β was used as the representative for detection. The measurement of TGF-β concentration in samples was based on the recommended protocols of TGF-β DuoSet ELISA Kit (catalog number DY240-05, R&D Systems, MN, USA) and R&D DuoSet Ancillary Reagent Kit 1 (catalog number DY007, R&D Systems, MN, USA).

Figure 1 shows the TGF-β standard curve obtained by following the aforementioned protocol. Once the concentration of TGF-β was determined for each batch of extraction process of PL, quantitative freeze-dried PL with known TGF-β was weighted and diluted with normal saline for the subsequent animal experiments.

Fig. 1
figure 1

The TGF-β standard curve obtained by following the protocol of the ELISA kit. The concentration of TGF-β in the PL sample was obtained by interpolating to this standard curve

Full size image

2.2 Inflammation Assessment After PL Injection in Rabbit Knee Joint

A biocompatibility test was designed for the New Zealand rabbit’s hind knee joint after PL injection (IACUC Approval No. 21T10-10, Master Laboratory Co. Ltd., Taiwan). The animals used in this experiment were six male New Zealand rabbits weighing 3.0–3.5 kg (N = 6). This animal experiment followed the guidelines of ISO 10993–6:2016 edition for local effects after implantation issued by International Organization for Standardization (ISO). In this study, one hind knee joint of each rabbit was randomly selected as the experimental group, and the other hind knee joint was selected as the control group (normal saline). On day 1, 2, and 4, the joints in the experimental group were injected with 0.3 mL of PL with a total content of 7.64 ng of TGF-β (three joints) and 0.3 mL of PL with a total content of 3.82 ng of TGF-β (three joints) in the experimental group. The PL was injected into the joint cavity with injection site lateral to the patella tendon. The control group were injected with 0.3 mL of normal saline. During the test, observations were made of whether the injection site was red and swollen, and whether the physiological functions (such as changes in diet and weight) were normal. On day 14 of the experiment, the experimental animals were sacrificed and the tested joints were taken for pathological section and staining to observe the inflammation condition.

Table 1 shows the inflammation scores in ISO10993-6:2006 edition. The total score for the experimental group was subtracted from the total score for the control group to obtain the inflammatory reaction rating.

Table 1 ISO 10993–6:2006 the histological evaluation system—Cell type and Response. The total score for the experimental group was subtracted from the total score for the control group to obtain the inflammatory reaction rating
Full size table

2.3 Osteochondral Defect Experiment

A New Zealand rabbit hind knee joints osteochondral defect regeneration experiment was designed (IACUC Approval No. 22T10-10, Master Laboratory Co. Ltd., Taiwan). The animals used in this experiment were three male New Zealand rabbits weighing 3.0–3.5 kg (N = 3). This animal experiment followed the guidelines of ISO 10993–6:2016 edition. In this experiment, the experimental group and the control group were randomly selected. The femur trochlea groove of the left and right hind knee joints were used to create osteochondral defects of 4 mm diameter and 4 mm depth. On the day after the operation and the second and fourth weeks after the operation, the experimental group was injected with PL 0.5 mL (the total content of TGF-β was 25 ng). In our previous studies [18, 19], 10 ng of TGF-β can promote cartilage regeneration. The control group was without the injection of TGF-β. The PL was injected into the joint cavity with injection site lateral to the patella tendon. The animals were sacrificed at the 12th week and the appearance of osteochondral regeneration and pathological sections were observed. The repaired cartilage was scored using the ICRS scoring system for macroscopic evaluation [20] and histological assessment [21] (Table 2).

Table 2 (a) The macroscopic evaluation of cartilage repair was evaluated by International Cartilage Repair Society (ICRS). (b) The regenerated cartilage quality was evaluated by ICRS Visual Histological Assessment Scale
Full size table

3 Results

3.1 PL Preparation

Porcine whole blood was centrifuged, the supernatant was taken to filter out white blood cells, the platelets were lysed by the freeze–thaw method for five times, and the complement and protein were removed by heating and then filtered through a 0.22 µm filter to complete the PL preparation process. The PL sample was performed the blood count and biochemical analysis. The blood count showed that the white blood cell, red blood cell, and platelet counts were all zero (Table 3). No residual blood cells were noted. Biochemical analysis showed that the contents of total protein and albumin were trace amounts. This method can remove most antigenic components, including blood cells and proteins.

Table 3 Blood count and biochemical analysis of the final PL sample
Full size table

In our study, it was found that the concentration of growth factor varied with the batches of porcine whole blood and the production process. For the growth factor TGF-β, which can induce the chondrogenesis of stem cells, its concentrations in three different productions (three different porcine) results were 9.6 ng/mL, 12.8 ng/mL, and 6.0 ng/mL, respectively, and their average was 9.5 ng/mL.

3.2 Inflammation Assessment After PL Injection in Rabbit Knee Joints

There was no redness, swelling, or infection in the six experimental rabbits’ joints compared to the control group, which indicated that PL obtained by our method did not cause severe joint inflammation.

In the pathological section, no obvious aggregation of inflammatory cells such as neutrophils, eosinophils, basophils, mast cells, lymphocytes or macrophages were found in cartilage, joint synovium, or meniscus (Figs. 2, 3). Only one experimental rabbit joint (rabbit 1002) injected with TGF-β 7.64 ng showed a notable moderate inflammatory reaction with polymorphonuclear and macrophage infiltration in the synovium region compared to the control group (Fig. 3). For the rabbit (1002), the inflammatory reaction ratings of the experimental joint and the control joint were 15 and 2, respectively and indicated a moderate inflammatory reaction.

Fig. 2
figure 2

Hematoxylin & eosin staining of the control group knee joints (The magnification: 40x). The pathological sections of six rabbits 1001–1006 in the control group (injected with normal saline) showed no abnormal aggregated neutrophils, eosinophils, basophils, mast cells, lymphocytes, and macrophages in articular cartilage (a), meniscus (b) and synovium (c)

Full size image
Fig. 3
figure 3

Hematoxylin & eosin staining of the experimental group knee joint (The magnification: 40x). The pathological sections of six rabbits (1001 ~ 1006) in the experimental group (rabbits 1001–1003 injected with 7.64 ng TGF-β of PL and rabbits 1004–1006 injected with 3.82 ng TGF-β of PL). The experimental knee joint of 1002 showed moderate inflammation reaction compared to control group, aggregated polymorphonuclear, lymphocytes, and macrophages were noted in the synovium (c) (magnified circle) with cell nuclei stained blue-purple. The rest of the pathological sections showed no abnormal aggregated neutrophils, eosinophils, basophils, mast cells, lymphocytes, and macrophages in articular cartilage (a), meniscus (b) and synovium (c)

Full size image

3.3 Osteochondral Defect Experiment

An osteochondral defect of 4 mm diameter and 4 mm depth was created in the distal femur patellofemoral surface region of the rabbit hind knee joint. PL was injected into the experimental group on the day (1st) after the operation and the second (2nd) and fourth (4th) weeks after the operation. The rabbits were sacrificed after 12 weeks for the evaluation of cartilage regeneration and arthritis. In macroscopic evaluation, the ICRS cartilage repair score of the control group was 11.7 and that for the experimental group was 19.4. No cartilage regeneration was noted at the defect region in rabbit numbered 2 in the control group (Fig. 4). Both the others in control group and all those in the experimental group showed > 75% cartilage regeneration in the defect area. The most obvious difference in the overall joint appearance evaluation was the occurrence of traumatic arthritis at the distal femur patellofemoral surface. The three joints in the control group had obvious traumatic arthritis (Fig. 4), and none of the experimental group subjects showed traumatic arthritis. The results showed that the injection of porcine-derived PL could effectively slow down the occurrence of arthritis.

Fig. 4
figure 4

Macroscopic observation: the yellow arrows are where the osteochondral defects were created and the red arrows are the distal femur patellofemoral surface. a The control group: the osteochondral defects of rabbits 1 and 3 are covered with new cartilage > 75% area (yellow arrows), while rabbit 2 shows no new cartilage formation. All three control groups had obvious traumatic arthritis in the distal femur patellofemoral surface (red arrows), irregular cartilage surface and subchondral bone exposure were noted. b In the experimental group, > 75% of the osteochondral defect had newly-formed cartilage (yellow arrow). No arthritis was observed in the distal femur patellofemoral surface and the articular surface was smooth indicating and no sign of degeneration (red arrow). (Color figure online)

Full size image

Histological evaluation of osteochondral defects of those in the experimental group scored 9 points, while those in the control group scored 7.8 points. The regeneration of cartilage defects in the experimental group was slightly better than that in the control group. However, the histological results of cartilage cross section showed that most of the regenerated cartilage was fibrocartilage or incompletely differentiated hyaline cartilage (Figs. 5, 6).

Fig. 5
figure 5

Histological evaluation of the osteochondral defects of knee joints in the control group. Hematoxylin & eosin staining (The magnifications: 40 × and 100x). Red arrows are the osteochondral defect regions. The articular cartilage of rabbit 1 was hypertrophic undifferentiated hyaline cartilage, the defect in rabbit 2 had no cartilage regeneration and the defect in rabbit 3 had a thin layer of fibrocartilage

Full size image
Fig. 6
figure 6

Histological evaluation of the osteochondral defects of knee joints in the experimental group. Hematoxylin & eosin staining (the magnification: 40 × and 100x). Red arrows are osteochondral defect regions. The articular cartilages for rabbit 1–3 had undifferentiated hyaline cartilage, a thin layer of undifferentiated hyaline cartilage, and fibrocartilage, respectively

Full size image

4 Discussion

Among inflammatory cytokines, interleukin 1 beta (IL-1β) and tumor necrosis factor-alpha (TNF-α) are considered the major factors in the pathogenesis of osteoarthritis (OA) and leading to chondrocyte death and joint degeneration [22]. Elevated levels of IL-1β and TNF-α have been found in joint fluid affected by OA [23]. Alpha-granules in platelets contain growth factors and proteins have benefits in terms of tissue regeneration [24]. To combat the catabolic milieu of OA-affected joints, PRP is thought to inhibit the catabolic cytokines of IL-1β and TNF-α [25, 26] and to promote the interaction between the cartilage matrix and the paracrine-related factors including fibroblast growth factor, transforming growth factor-beta, vascular endothelial growth factor, and epidermal growth factor [27]. Intra-articular injections of PRP can effectively delay cartilage erosion, but the more effective repair of cartilage defects is to use a composite of PRP and fibrin clots or scaffolds [28].

Platelet lysate contains various growth factors involved in cartilage repair and exerts chondrocyte protection and extracellular matrix modification in a dose-dependent manner. In vitro PL significantly restored TNF-α-inhibited anabolic gene expression (Col2 and aggrecan) and TNF-α increased catabolic gene expression (Col10, Mmp13, Adamts5, and Adamts9) [29] in cartilage repair and arthritis prevention and can hence achieve the same safety and effect as PRP [30].

Camargo Garbin et al. [31] carried out an in vitro cell culture of allogeneic platelet lysate and obtained a good cell proliferation effect. Aldén et al. [16] used porcine PL to culture monkey and hamster cells and obtained good differentiation and value-added effects.

Whether leukocytes should be removed has been discussed in the application of autologous PRP [32]; it is generally believed that high concentrations of leukocytes in PRP preparations can cause inflammation and joint degeneration, and negatively impact the extracellular matrix of cartilage tissue. On allogeneic or xenogeneic platelet-derived preparations, potential antigen removal is very important, especially the removal of leukocytes.

In this study, the preparation process of porcine PL removed leukocytes with a leukocyte filter after centrifugation for removal of erythrocytes. Afterward, a freeze–thaw method was employed to break platelets and to release growth factors, followed by heating the lysates at 56 °C to deactivate the complement. Lastly, a filter of 0.22 μm was used to remove platelet fragments and large proteins. With this process, most antigens including red blood cells, white blood cells, platelets, complement, and large proteins can be effectively removed. As a result, the injection of porcine-derived PL into the knee joint of rabbits did not cause rejection and inflammation reaction. In our osteochondral defect experiment, the injection of porcine-derived PL was significantly effective in preventing the occurrence of arthritis in patellofemoral joints, but for comparison with the control group, the regeneration of the osteochondral defect site had no significant difference, and the treatment of local osteochondral defects with the use of local scaffolds or gels should be able to obtain better therapeutic effects.

The production process of PRP and PL cannot accurately control the amount of platelets or the concentration of growth factors, so each clinical treatment has its own variability. To improve the clinical drawback, freeze-dried PL is one of the solutions. By measuring the concentration of growth factors after PL production, the growth factors required for the injection can be calculated. In addition, freeze-dried PL can maintain good growth factor activity and regeneration ability [29], and can maintain the necessary qualitative and quantitative factors for each injection to achieve a stable therapeutic effect.

Regarding the inflammatory reaction in the rabbit experiment, one rabbit in the experimental group had a moderate inflammatory reaction in the pathological cross section. No redness or swelling was noted in appearance. The other five rabbits in the experimental group showed no inflammatory reaction. The reason may be either that this rabbit was allergic to xenogeneic PL or it sustained injuries due to external forces during the injection process. Future animal experiments can further clarify the problem.

Our experiments provide preliminary evidence for the application of xenogeneic PL in clinical treatment. By removing blood cells, complement, and large proteins in the xenogeneic blood, xenogeneic PL can be safely and effectively used in the treatment of arthritis.

5 Conclusion

The use of xenogeneic PL for regenerative therapy is feasible. Compared with PRP, the PL production process removes possible antigens such as white blood cells, red blood cells, platelets, complement, and large proteins, leaving only the required growth factors to avoid the rejection of xenogeneic sources. Our experimental results show that xenogeneic PL is a safe and effective method in the treatment of arthritis, which can be used as a research basis for future medical applications.