The Mechanism of Cartilage Regeneration by Buffy Coat and the Pre-clinical Study

Background Autologous bone marrow buffy coat transplantation possesses obvious advantages in the therapy of large cartilage defects and osteoarthritis. However, there is no denite research on the specic effective components of bone marrow buffy coat and its mechanism of cartilage regeneration. Moreover, as the crucial cartilage regenerative cell in bone marrow buffy coat, mesenchymal stem cells(MSCs) are dicult to x onto the damaged cartilage area without damaging the subchondral bone. We assessed the effect of using hyaluronic acid(HA) as a liquid scaffold mixed with autologous bone marrow buffy coat to x cartilage defect. Methods and Materials We extracted the bone marrow from the anterior superior iliac crest of the white New Zealand rabbit, centrifuged to obtain a buffy coat, and analyzed the components of buffy coat by ELISA. Buffy coat+HA group, MSC+HA group, MSC+TGF-β+HA group were cultured in vitro and observed by staining. In addition, we made damage to the femoral condyle of rabbits, and divided them into groups: HA group, buffy coat group, buffy coat with HA group. Each group was assessed for cartilage regeneration by visual observation, histological and immunohistochemical analysis at 4 weeks and 8 weeks, and biochemical analysis at 8 weeks postoperatively. One-way ANOVA and LSD were used for statistic analysis.


Introduction
Injuries to the articular cartilage are challenging to repair because cartilage possesses a limited capacity for self-repair [1]. Articular cartilage is avascular, aneural, and alymphatic, leading to di cult to access to stem cells; additionally, low cell mobility due to surrounding matrix and a limited number of progenitor cells could be another contributing factor. Remarkable efforts have been made to repair cartilage defects.
Several techniques concerning implantation of different kinds of cells or tissues aiming to instance autologous chondrocyte implantation and mosaicplasty were developed to restore the normal structure and function of cartilage [2]. These techniques have been demonstrated to be valid in enhancing cartilage repair. However, the regeneration of hyaline cartilage has not been adequately accomplished.
Numerous studies have demonstrated the therapeutic potential of mononuclear cells (MNCs) extracted from bone marrow for tissue regeneration after ischemic femoral head necrosis, myocardial infarction, and cartilage injury [2][3][4][5]. In addition, BMSCs constitute 0.001-0.01% of mononuclear cells in bone marrow aspirate after undergoing density gradient centrifugation to remove platelets, granulocytes, red blood cells, and immature myeloid precursors [6]. Chondrogenesis of MSC requires stimulation of a variety of growth factors, such as transforming growth factor-β1 (TGF-β1) and platelet-derived growth factor (PDGF) [7,8]. Jin et al. (2011) rst observed that the buffy coat containing monocytes, lymphocytes, macrophages, platelets, and various cytokines displayed satisfactory cartilage regeneration [9],as the achieved number of MSCs in the buffy coat are higher than that of the bone marrow stimulation (100 times). The buffy coat of 4 ml bone marrow contained 2400 MSCs and 1 * 10 7 MNC. Buffy coat possesses an acceptable cartilage differentiation capability in vitro [10]. However,the mechanism of cartilage regeneration of the buffy coat is unclear and is di cult to x onto the damaged cartilage area.
Hyaluronic acid (HA) is a biodegradable, biocompatible, non-toxic, non-immunogenic, and nonin ammatory linear polysaccharide [11]. In cartilage, HA chains are used as aggregated anchors of aggrecan monomers, and play a pivotal role in assuring cartilage hydration, joint surface lubrication, and synovial uid viscosity [12,13]. In OA, the molecular weight and concentration of HA are reduced, thereby lowering uid viscosity and elasticity. Protection against articular injury is compromised and OA damage ensues [14,15].
Hence, in the present study, the buffy coat components and chondrogenic ability of the buffy coat in vitro and cartilage regeneration of the mixture of hyaluronic acid (HA) and buffy coat of in vivo was analyzed.
This study may provide an important theoretical and experimental basis for the use of buffy coat in regenerating cartilage.

Materials
Healthy 16-week-old (range of weight, 3.0-3.5 kg) New Zealand white rabbits were purchased from the Experimental Animal Center of Yanbian University (Yanji, China). All of the rabbits were housed during the period of experiment under the conditions of a 12:12 h light/dark cycle at 27 ºC.

Component analysis of buffy coat
In the present study, 8 rabbits were selected for the experiment. Moreover, 2 ml of bone marrow was extracted from the anterior superior iliac crest of rabbits using a 20 ml syringe containing 2 ml of 0.9% sodium chloride solution. The collected bone marrow was placed in a tube (17-1440-02; Amersham Biosciences, Sweden), and centrifuged at 1000 g for 30 min at 4°C.

In vitro study
The mixing ratio of brinogen (Shanghai RAAS Blood Products Co., Ltd., Shanghai, China) to HA (Hunan Jingfeng Pharmaceutical Co., Ltd., Shanghai, China) was 10:1. Cells with density of 5 * 10 6 cells/well were then mixed homogeneously with aprotinin, 60 U/mL thrombin, brin stabilizing Factor XIII, and 50 mM CaCl 2 . Then, 250 µL of the brin/HA mixture was dropped into an empty Petri dish to form a gel [30].
Buffy coat was re-suspended in an α-minimum essential medium (α-MEM) supplemented with 1% antibiotic-antimycotic and 10% fetal bovine serum (FBS). Buffy coat was cultured in an incubator with 5% CO 2 at 37°C. After 6 days, non-adherent cells were removed and adherent cells were further cultured. After 12 days of cultivation, the adherent MSCs were retrieved by trypsin treatment and re-cultivated at density of 3 * 10 5 cells/plate (100 mm) for expansion. The second-generation cells were harvested into brin/HA composite gel (5 * 10 6 MSCs/scaffold). After 90 min of incubation, they were transferred into a chondrogenic medium. The bone marrow aspirates (2 ml) were obtained aseptically from the iliac crest, and buffy coat was obtained by Ficoll gradient centrifugation method.
After placing the buffy coat into a 1.5 ml tube, it was then centrifuged (1000 g for 5 min); after removing 450 µl of the supernatant, 50 µl of buffy coat was seeded into brin/HA composite gel (5 * 10 6 MNCs/scaffold). After 90 min of incubation, they were transferred into the chondrogenic medium.

In vivo experiment
Herein, xylazine hydrochloride (0.1 ml/kg) was injected intramuscularly, and 2% pentobarbital sodium (2 ml/kg) was injected intraperitoneally into white New Zealand rabbits. After completing anesthetization of these rabbits, the operation was carried out. An arthrotomy was made through a midline longitudinal incision on the lateral parapatellar with the patella dislocated laterally to expose the femoral condyle. A surgical drill (5 mm in diameter) was used for the creation of an osteochondral defect of 2 mm in depth in the patella groove.
For the in vivo study, we used a polyglycolic acid (PGA) (Albany International Inc. Mans eld, MA, USA) scaffold after attening to cover the cartilage defect. Absorbable sutures were cross-secured simply to prevent the graft from falling into the joint cavity.
The rabbits were randomly divided into three groups: 1. Injecting HA (Hunan Jingfeng Pharmaceutical Co., Ltd., Shanghai, China) after appearance of defect, and the defect was left untreated as a negative control (HA group); 2. Supplemented autologous buffy coat was isolated from the iliac crest (buffy coat group); 3. Buffy coat and HA composite were injected into the cartilage defect site (buffy coat + HA group). Regeneration of cartilage was evaluated by macroscopic, histological, and immunohistochemical analyses at 4th and 8th weeks post-operation, and additionally by biochemical analysis at 8th week postoperation.
The design of experiment was conducted based on "Animal research: reporting in vivo experiments: the ARRIVE guidelines" [31].

Macroscopic and histological evaluations
At 4th and 8th weeks post-operation, the rabbits were euthanized by injection of an over-dose of pentobarbital. Femoral condyles were subsequently retrieved. Macroscopic images of the condyles were initially observed; the samples were then embedded into para n, sectioned, and processed for performing Safranin-O staining, Sirius Red staining, and immunohistochemically analysis.

Histological scoring
To minimize the effects of subjective bias, the Safranin-O stained slides were graded using the Bern Score, consisting of seven categories (range of score: 0-9) [32]. The quality of the regenerated articular cartilage of the defects and a modi ed version of the histological grading scale system were used for the International Cartilage Repair Society (ICRS) score [33]. The ICRS consists of seven categories, and a score ranging from 0 to 18 was accordingly assigned.

Chemical assay of cartilage tissue engineering
For chemical assays of the repaired tissue, the cartilage specimens were minced by a surgical knife. The harvested tissue was dried and dissolved in papain solution (5 mM L-cysteine, 100 mM Na2HPO4, pH 6.4, 5 mM EDTA, and 125 mg/ml papain type III) at 60°C for 24 h, and then centrifuged at 12,000 g for 10 min.
The GAG contents were measured by subjecting the supernatant to 1,9-Dimethylmethylene Blue (DMB) assay.

Statistical analysis
Statistical analysis for comparing the experimental data was carried out by using(one-way ANOVA), and speci c inter-data differences between mean values were determined by the LSD test. P-values less than 0.05 were considered statistically signi cant.

In vitro evaluation
On examination with naked eyes, it was found that there was no difference in the size of the three groups after inoculation. After 21 days of culture, it was revealed that the MSC group was generally smaller with blur edges. There was no remarkable change in the size of the buffy coat group and the MSC + TGF-β1 group, both with round and clear edges and high cartilage transparency (Fig. 1A). Cells on the brin/HA composite gel were stained with Safranin-O, and a signi cant intergroup difference was noted. On microscopic observation, cells in the outer region of the MSC group displayed differentiation into chondrocytes and production of GAG; however, the central cells did not differentiate into chondrocytes, and apoptosis was noted. On the contrary, in the buffy coat group, cells in the center and surrounding areas displayed active differentiation into chondrocytes and production of GAG. In the MSC + TGF-β1 group, all aspects of cartilage differentiation were active, while the surrounding cartilage was well differentiated, and the central part couldn't enter into the culture uid, resulting in a necrotic cavity in the tissue center (Fig. 1B). Analysis of these sections using the Bern score supported this observation.
The cells constructed with high cell density showed the increased effect of Safranin-O staining in the buffy coat group and MSC + TGF-β1 group, which demonstrated a rounded cell shape and higher production of extracellular matrix (ECM) (Fig. 1B). A signi cant difference in the Bern score among the buffy coat, MSC + TGF-β1, and MSC groups was noted. The mean score of MSC group was lower than that of the latter two groups, with signi cant statistical difference. The buffy coat group, however, had a lower mean score than the MSC + TGF-β1 group, but there was no statistical difference. It might be due to the large standard deviation and small sample size. (n ≥ 4; ** P = 0.002; *** P < 0.001) (Fig. 2). From this, it can be seen that the MSC contained in buffy coat has a good differentiation ability and various growth factors that can promote the cartilage differentiation.

In vivo evaluation
At 4th week post-operation, the cartilage defect site was partially regenerated by white tissue. The HA group has signi cantly less regenerative tissue than the other groups. The buffy coat group and buffy coat + HA group regenerated tissue surfaces, covering the brous tissue. At 8th week, all the groups regenerated more tissue than at 4th week, and the regenerated tissue in the buffy coat group and buffy coat + HA group was closer to the surrounding normal cartilage (Fig. 3A). Safranin-O staining showed that the HA group had a very thin cartilage regeneration, while the GAG content was very small and was separated from the subchondral bone at 4th week. Buffy coat group and buffy coat + HA group were also rich in GAG regeneration, whereas regenerated tissues were immature. At 8th week, the buffy coat group showed to be very mature, while the surface was somewhat degraded. A large amount of GAG was produced in the buffy coat + HA group, and the regenerated cartilage was mature as well. The chondrocytes were parallel to the articular surface, and the chondrocyte of the deep zone was perpendicular to the articular surface (Fig. 3B).
The ICRS scores were signi cantly increased in all the groups over time. There was a statistically signi cant difference in the scores between the HA group and the buffy coat + HA group at 4th week (n ≥ 5; **P = 0.001). The average score in the buffy coat group was notably higher than that in the HA group, which was close to that in the buffy coat + HA group. At 8th week, there was a signi cant difference in the ICRS scores between the HA group and buffy coat group (n ≥ 5; **P = 0.004; ***P < 0.001). The mean ICRS score was lower in buffy coat compared with buffy coat + HA group, while no signi cant difference was noted between these two groups (Fig. 4). Therefore, it can be concluded that HA can remarkably improve the stability of buffy coat in the treatment of cartilage damage.
In Sirius Red staining, at 4th week, HA group showed a small, very ne collage, which was formed and arranged irregularly. In buffy coat group and buffy coat + HA group, there were numerous regenerated collagens, while the collagen bers were very thin and irregularly arranged. In 8th week, the HA group had a small amount of collagen formed, and the ber was thicker than that at 4th week, which was irregularly arranged. The buffy coat group and the buffy coat + HA group regenerated several collagens, and the ber was also thick and grew perpendicular to the surface, demonstrating that the regenerated cartilage was very mature (Fig. 5A). In immunohistochemical staining of type II collagen, at 4th week, the regenerated tissues in HA group were not stained, but buffy coat and buffy coat + HA group were stained slightly. At 8th week, regenerative tissue was partly stained in HA group, while regenerative tissue was remarkably stained in the buffy coat group and buffy coat + HA group (Fig. 5B).
At 8th week post-operation,the content of GAG in the buffy coat group and buffy coat + HA group was signi cantly higher than that in the HA group, while there was a gap compared with normal cartilage (Fig. 6). The standard deviation in the buffy coat group was very large, whereas the standard deviation in the buffy coat + HA group was small, and the regeneration of cartilage was found to be very stable (n = 6; **P = 0.001; ***P < 0.001). Therefore, it was found that buffy coat + HA greatly improved the stability of cartilage regeneration, thereby overcoming the problem of the buffy coat to x to damaged cartilage.

Discussion
The results of the present study indicated that buffy coat is appropriate for cartilage regeneration. It contains several MSCs, growth factors, and cytokines. Therefore, it can differentiate into engineered cartilage without TGF-β1 in vitro. After addition of HA, it promotes the cartilage regeneration capability of buffy coat, and the problem that buffy coat is di cult to adhere to the cartilage defect could be solved as well.
In addition, TGF-β1 at the range of 0.1-25 ng/mL is often essential for proper regulation of MSC proliferation, differentiation, and ECM synthesis, while a higher concentration may inhibit the process [16,17]. All the values of TGF-β1 obtained are, in fact, lower than the expected ones. According to a research performed by Veselý et al. [18], FGF-2 is normally undetectable in serum or with very low values; and the experimental value obtained by the buffy coat was 1.326 ng/ml. In a recently conducted study on a rabbit, a low dose of FGF-2 (10 µg/mL) soaked in a highly porous scaffold showed improved healing of osteochondral lesions, whereas higher dose of FGF-2 (100 µg/mL) or scaffold without FGF-2 demonstrated an inverse response [19].
The chondrogenic differentiation of MSCs is induced by IGF-1, while it is enhanced when IGF-1 and TGF-β1 are used in combination. As the value of IGF-1 was found to be as low as 0.195 ng/ml in the present study, we hypothesized that such a low concentration of IGF-1 might not be able to suppress the chondrogenic differentiation of MSCs.
Moreover, PDGF released from platelets stimulates chemotaxis and mitosis of broblasts, synthesis of collagen, as well as remodeling of the extracellular matrix. In an experiment carried out by Schmidt et al. [20], the effect of PDGF increased proliferation of chondrocytes with concentrations ranging from 4.7 to 300 ng/mL with a maximum number of cells at 75 ng/ml. In the current research, the concentration of PDGF, 1.514 ng/ml, is neither very high nor very low than Schmidt et al.'s ndings, thus it is reasonable to hypothesize that the PDGF present in the isolated buffy coat has enough potential for proliferation of chondrocytes.
Generally, VEGF is not found in adult human articular cartilage under physiological conditions and has been reported as one of the most signi cant factors that can cause arthritis. VEGF in the minimum amount can trigger the formation of a new blood vessel and increase blood ow to the site of injury, thereby protecting the proliferated MSCs [21]. Another growth factor, EGF stimulates epithelization and decreases the healing process, in which a high level of EGF causes thinning, loss of integrity, and degradation of the articular cartilage [22]. Since the concentration of EGF is insigni cant in the isolated buffy coat, it is reasonable to hypothesize that the chances of losing the integrity and degradation of the articular cartilage are very low, while its' presence can stimulate angiogenesis of endothelial cells and mitosis of mesenchymal stem cells.
Additionally, IL-1β and TNF-α are present in joints where cartilage is undergoing repair or regeneration due to injury or disease. In an in vitro study carried out by Schuerwegh et al. [23], it was uncovered that very high concentrations of IL-1 (100 ng/mL) and TNF-α induced apoptosis of bovine chondrocytes. In addition, no effect of IL-6 was noted on the function of cartilage with respect to chondrocyte apoptosis, viability, and proliferation. To our knowledge, in ammatory cytokines are inevitable after tissue injury or disease, hence, we hypothesized that the presence of very low concentration (0.219 ng/ml) of in ammatory cytokines by the buffy coat is negligible [23].
Furthermore, Behrendt et al. found that the effect of IL-10 was highly tangible at 100 pg/ml compared with 1000 pg/ml, which prevented posttraumatic cell death, matrix degradation, and chondrocyte dedifferentiation in mechanically injured cartilage [24]. The value obtained in the present research was as low as 1.969 ng/ml, which is near to the concentration of IL-10 reported by Behrendt et al. [24].
A number of previous studies found that IL-4 has an inhibitory effect on degradation of proteoglycans in the articular cartilage, and also reduces the variation in the production of proteoglycans that are visible in the course of OA [25]. The concentrations of IL-4 obtained in the current experiment was 1.115 ng/ml, which might inhibit the effect of degradation of injured cartilage.
Platelet-rich plasma (PRP) research has become more prevalent in recent years. It is a safe treatment for autologous blood products, containing high concentrations of platelets that can be treated by minimally invasive arthroscopy. It has also been reported that the repaired cartilage by PRP is brocartilage in nature, while the frequency and duration of treatment have not been reported yet [26,27]. Therefore, treatment by PRP needs to be further studied to eliminate the existing clinical barriers.
It is noteworthy that BMC contains a variety of biologically active molecules and cells, including platelets, lymphocytes, neutrophils, and monocytes, in addition to MSC. A number of scholars demonstrated that BMC is a product of PRP plus MSC. In fact, BMC and PRP contain different cytokines and growth factors. It was reported that BMC contains several platelets, total monocytes, and CD34 positive cells [28]. Cassano et al. [28] con rmed that BMC is rich in interleukin 1 receptor antagonist (IL-1ra), which is a natural IL-1 receptor antagonist, and it can treat autoimmune diseases and rheumatoid arthritis. Scholars also expressed that buffy coat is a BMC without red blood cell (RBC), that is, there is no RBC that hinders adhesion and proliferation of MSC, thus, buffy coat has a superior cartilage regeneration capability [10]. When buffy coat was extracted, we accidentally scraped numerous RBC samples, in which the quality of tissue-engineered cartilage was not eventually satisfactory. Horn et al. also reported that RBCs interferes with adhesion and proliferation of MSCs [27].
The concentration of MSC in bone marrow extracted from rabbit iliac crest was previously reported equal to 607.8/ml [9]. Thus, there are approximately 1216 MSCs and 4.6 × 10 6 MNCs in 2 mL of bone marrow. The MSC concentration in bone marrow is very low. Therefore, the same amount of MSCs cannot be inoculated, although the same number of MSCs and MNCs must be inoculated. After PGA plus HA, the buffy coat was easily xed on the defected cartilage. We used PGA to cover cartilage defects, similar to a membrane. And it was disclosed that cartilage regeneration is generally stable and the appearance is acceptable after coverage of PGA. Siciari et al. demonstrated that a cell-free PGA-HA scaffold has an appropriate cartilage regeneration effect [29].
The present in vivo study revealed that some individuals with poor cartilage regeneration had hyper osteogeny around the joints and had traumatic arthritis as well. These may be related to surgery or individual differences. However, the cartilage regeneration effect in buffy coat + HA group was found to be very satisfactory and stable. When the bone marrow was pumped out for a long period of time, the bone marrow was easily solidi ed in the syringe. When the buffy coat was extracted with such bone marrow, the buffy coat likely contained several RBCs or MSCs. The same situation was also observed when the ambient temperature was high.
However, the buffy coat was extracted by Ficoll gradient centrifugation, Ficoll couldn't be used in clinical practice. Thus, it is essential to develop a new method for extracting buffy coat over the Ficoll gradient centrifugation, which is the key to the application of buffy coat in the future.
In the current research, we developed a new method for extracting buffy coat without Ficoll. The buffy coat extracted by the new method and the buffy coat extracted by Ficoll had no difference in growth factor and cytokine content, and the MSC content was known as well. Therefore, our next study is to apply the new method to extract the buffy coat to regenerate cartilage for clinical application of buffy coat.

Conclusions
In summary, a number of components of buffy coat were herein studied, and it was revealed that there are several growth factors and cytokines that promote cartilage regeneration. The capability of buffy coat to produce cartilage was also demonstrated in in vitro and in vivo experiments. The problem that the buffy coat is di cult to x onto the cartilage defect due to the liquid state was solved, and the cartilage regeneration effect in the buffy coat + HA group was veri ed as well. The clinical application of buffy coat is another main challenge in the treatment of cartilage defect.