T.Fujisato et al.Effect of bFGF on cartilage regeneration
Effect of basic fibroblast growth factor on cartilage regeneration
in chondrocyte-seeded collagen sponge scaffold
Toshia FUJISATO, Toshinobu SAJIKI, Qiang LIU, and Yoshito IKADA
Research Center for Biomedical Engineering, Kyoto University
53 Kawahara-cho, Shogoin, Sakyo, Kyoto 606-01, JAPAN
ABSTRACT
A chondrocyte-collagen composite was prepared in an attempt to regenerate cartilage by its subcutaneous implantation in nude mouse. When the composite was impregnated with basic fibroblast growth factor (bFGF) prior to implantation, regeneration of the cartilage tissue was remarkably accelerated. Histological staining of the implanted composites with safranin O-fast green revealed that the cells incorporated in the composites exhibited their phenotype and formed a new matured cartilage. A thin layer of fibrous capsule was observed surrounding the implanted composite and the inflammatory response of the host to the implant was mild. Specific proteoglycans were accumulated in the composite even 1 week after implantation. At 2 weeks after implantation, the chondrocytes regenerated the cartilage tissue, although still immatured, but at 4 weeks almost all of the chondrocytes transferred to the matured stage. On the contrary, such matured cartilage tissue was not noticed up to 4 weeks after implantation, if the collagen scaffold was not impregnated with bFGF. Moreover, the matured area was limited to only a small fraction of the implanted composite, unless bFGF was incorporated in it.
Keywords: bFGF, chondrocyte, cartilage regeneration, collagen, angiogenesis, tissue engineering
Correspondence to Professor Y.Ikada.
Fax: +81-75-751-4144
INTRODUCTION
Recently much attention has been paid on the use of biodegradable polymers to regenerate metabolic organs such as the liver1,2 and intestine3 and to reconstruct structural tissues like cartilage4-9, bone8-10, and urothelial structure11 by cell transplantation12-14. In clinics, cartilage replacement is also needed, especially in maxillofacial, orthopedic, and plastic surgery. Mostly, silicone prosthesis and autologous rib bones have been used for this purpose, but several problems are involved in these cartilage replacements, such as infection at the interface between the implanted material and the tissue, and deterioration of the donor site by the filling with fibrous cartilage tissue. Therefore, a preferable replacement would be to use the natural cartilage tissue. The first attempt to use the cultured chondrocytes for an articular cartilage repair was reported by Green et al.15 and it is supposed that a template is necessary for cartilage reconstruction by chondrocytes in vivo. There are some reports describing the seeding of collagen and other porous matrices with chondrocytes for this purpose4-9. Itay et al. reported the repair of bone tissue by chondrocytes incorporated in collagen gel as a template. Langer, Vacanti, and their coworkers have developed a tissue engineering technique with the use of chondrocytes of the target region. The cells will be isolated from patients, proliferated by the in vitro culture to a higher density, and then implanted after preparing a chondrocyte-polymer composite as the core for the cartilage reconstruction. It is reported that chondrocytes can be isolated from tissue and grown in culture in such a way as to maintain their phenotype15-18.
In this work, a chondrocyte-collagen composite is prepared in an attempt to regenerate the cartilage by its implantation in mouse. To diminish the difficulty associated with autograft transplantation, nude mice were used for the implantation of the composite. As a scaffold for the cartilage regeneration, a porous collagen sponge is employed. Collagen has been used as the scaffolds for tissue regeneration19,20. We have also reported that it is an excellent template to regenerate skin21 and esophageal replacement22. When a bilayer artificial skin composed of an outer layer of silicone and an inner sponge layer of collagen was placed on the skin defect on the backs of rats, it was observed that epidermal cells migrated from the edge of the wound between the two layers21. An artificial esophagus with a bilayered structure made of porous collagen sponge and silicone was studied to promote tissue regeneration by collagen, and the collagen sponge was replaced by autologous tissue and regeneration of the “neoesophagus” was observed 2 weeks after implantation22. In our previous work, chondrocyte-poly(lactic acid) (PLA) complex was prepared to study its potentiality for cartilage reconstruction, but no matured cartilage tissue was observed in 1 month after implantation23. This suggests that supplying nutrients to the seeded cells in the early stage of transplantation to maintain them alive and promote tissue regeneration is very important. An effective means for this purpose may be to induce the capillary formation around the implanted composite by giving an angiogenic factor like basic fibroblast growth factor (bFGF)24. In addition to the angiogenesis, bFGF is known to perform other important functions such as parenchymal cell proliferation, differentiation25,26, and promotion of cartilage repair in vivoalthough cartilage is an avascular tissue27,28. In this study, bFGF was applied to the chondrocyte-collagen composite prior to implantaion.
MATERIALS AND METHODS
Collagen sponge
A collagen sponge as the scaffold was prepared from 0.3% hydrochloric solution of type I atelocollagen (Cell matrix®; Nitta Gelatin Co. Ltd., Osaka, Japan, pH=3.0). The collagen solution was stirred at 2000 rpm for 1 hr at 4 °C to generate small bubbles and then freeze-dried. The resulting sponge was vacuum-dried for 24 hrs at 105 °C, immersed in 0.2% acetic acid solution of glutaraldehyde for 24 hrs at 4 °C to introduce chemical crosslinking, and pressed to a sheet of 3 mm thickness. The average pore size, pore volume fraction, and density of the sponge were 86 µm, 87%, and 1.010-2 g/cm3, respectively. The sponge sheet was cut to have a round shape of 9 mm diameter. Two pieces of round sheets were overlapped each other and then sewn together with 7-0 polypropylene suture. The lapped sponge was immersed overnight in 70% ethanol for sterilization and then washed with phosphate buffered saline (PBS; Nissui pharmaceutical Co. Ltd., Tokyo, Japan, pH=7.4). Prior to implantation, the collagen sponge was impregnated with bFGF by immersing in the 80 µg/ml PBS solution of bFGF for 24 hrs at 4 °C, unless otherwise stated. bFGF was kindly supplied by Kaken Pharmaceutical Co. Ltd., Tokyo, Japan. The amount of bFGF was determined by HPLC using a heparin column.
Chondrocytes
Chondrocytes were isolated from the costal cartilage of rats by collagenase digestion29. Costae were removed from sacrificed rats and the cartilage was isolated carefully from them, so that fibrous tissues were not included. The isolated cartilage was minced by surgical scissors and immersed in 0.25% trypsin-0.05% collagenase (Amano Pharmaceutical Co. Ltd., Osaka, Japan) solution for 1 hr. After 3 times washing with PBS, the treated pieces of cartilage were immersed in 0.02% ethylenediaminetetraacetic acid (EDTA) solution for 1 hr. After washing, they were put on a petri dish for tissue culture (Corning® Type 25020; Corning Co. Ltd., NY), incubated for about 3 weeks adjusting the level of medium so as not to float the minced tissues in the medium, and then the migrated cells were collected by a cell harvesting solution (0.25% trypsin-0.02% EDTA in PBS). Eagle's MEM (Nissui Pharmaceutical Co. Ltd.) was used as culture medium with 10% fatal bovine serum (Bio Whittaker, Inc., Maryland). One hundred µl of chondrocyte suspension containing 1106 cells was carefully injected with a 27 G needle syringe into the center of lapped collagen sponge disk. The cell-injected sponge was stored in CO2 incubator for 2 hrs to allow the cells to adhere to the collagen sponge before implantation as much as possible.
Cell culture
For the in vitro study, the chondrocyte suspension or the chondrocyte-collagen composite was put into a 24 well-micro plate for tissue culture (Corning® Type 258201; Corning Co. Ltd.). The cell density was 1,000 cells/well in the case of chondrocyte suspension. Culture medium was exchanged everyday. After predetermined periods of time, the cells were counted by measuring the activity of lactate dehydrogenase (LDH) in the cells using a test kit for clinical use (LDH monotest®; Boehringer Manheim, Germany) after complete cell digestion by 0.1% polyoxyethylene(10) octylphenyl ether (Triton® X-100; Wako Pure Chemical Industries, Ltd., Osaka, Japan)30.
Implantation
Animals were carefully reared in the Research Center for Biomedical Engineering, Kyoto University, according to the guideline of Kyoto University for Animal Experiments. All animals were anesthetized with diethyl ether and pentobarbital sodium for the release of suffering from the pain during operation. Two samples of chondrocyte-collagen composite and the control without chondrocytes were subcutaneously implanted into the back of a male nude mouse. The mice were divided into 2 groups receiving collagen sponges with bFGF and without bFGF. Twelve mice were employed in each group. After predetermined periods of time, the sponges were explanted and subjected to gross and microscopic observation and a histological study to evaluate the inflammatory response of the host and cartilage matrix secretion. Explanted samples were fixed with 10% formaldehyde aqueous solution, replaced with ethanol, and embedded in paraffin. The fixed samples were sectioned to 10 µm thickness with a microtome at 3 different distances from the surface of samples, and stained with Mayer’s Haematoxylin-Eosin (H.E.) solution. In addition to the conventional H.E. staining, safranin O-fast green staining was applied for identifying the cartilage proteoglycans31.
RESULTS
bFGF impregnation
The processes of chondrocyte-collagen composite preparation are schematically represented in Figure 1. The collagen sponge disk was placed in plenty of 80 µg/ml bFGF solution in PBS to incorporate bFGF into the disk. Figure 2 shows the plot of bFGF amount adsorbed into the disk at 37 °C as a function of time. As is seen, the bFGF impregnation seems to come to saturation after 30 hr incubation, reaching a leveling-off value of 60 µg per mg of collagen sponge. An addition of serum albumin (BSA) to the bFGF solution had no significant effect on the bFGF impregnation. The release of the adsorbed bFGF into PBS upon immersion of the impregnated sponge in PBS at 37 °C is shown in Figure 3. The bFGF impregnation was carried out at 4 and 37 °C. Clearly, approximately 90% of the bFGF impregnated even at 4 °C for minimizing the bFGF deactivation still remains in the interior of the collagen sponge. The bFGF remaining in the sponge is expected to be released upon enzymatic degradation of the crosslinked collagen when the sponge disk is implanted in mice.
Chondrocyte seeding
Rat chondrocytes were seeded in the collagen sponge disk after trypsinization of the cultured cells isolated from the rat costal cartilage on a petri dish for tissue culture with cell harvesting solution. Figure 4 shows the in vitro growth of the chondrocytes after trypsinization on the 24 well-micro plate. Obviously, cell confluency is obtained upon incubation for about 5 days. As a preliminary study to determine the effective method for seeding the chondrocytes in the matrix, cells were seeded in the collagen sponge with three different methods; 1. injection of the cell suspension in culture medium to the sponge with a needle at 25 °C, 2. injection of the cell suspension in culture medium containing 0.3% collagen to the sponge with a needle at 4 °C, and 3. immersion of the sponge disk into the cell suspension in culture medium at 25 °C in 24 well-micro plate. Two hrs after cell seeding, the disks were washed with PBS to remove the non-seeded chondrocytes, and the number of cells was estimated from the activity measurement of LDH. Figure 5 shows the percentage of the chondrocytes still remaining in the interior of the sponge disks after washing. As can be seen, about 50% of the cells remain adhered to the collagen sponge when a needle is used, regardless of the presence of collagen in the cell suspension. In the following study we employed the 1st method for the cell seeding, that is, injection of the cell suspension containing 1106 cells and bFGF through a needle.
The chondrocytes seeded in the collagen sponge were further incubated for their stronger adhesion to the collagen surface. Figure 6 shows the result of cell growth in the collagen sponge determined by LDH activity. It is seen that the cell density slowly increases with the incubation time. The decreased cell density on day 9 is probably due to the cell injury caused after reaching confluency. The addition of bFGF to the culture medium had no effect on the in vitro cell growth. Figure 7 demonstrates a SEM microphotograph of the chondrocytes attached to the collagen surface after 3 days of incubation. As is apparent, the cell is spreading on the collagen substrate.
Composite implantation
The chondrocyte-seeded collagen composite with or without bFGF was subcutaneously implanted in nude mice to study the effect of bFGF addition. In our previous work, the collagen sponge carrying neither chondrocytes nor bFGF disappeared as a result of collagen biodegradation when implanted for longer than 25 weeks. On the other hand, remarkable angiogenesis was noticed around the collagen disk when bFGF had been incorporated in the sponge, independent of the presence of chondrocyte. A representative optical photograph of nude mouse with implanted collagen sponges is shown in Figure 8. The sponges were implanted for 4 weeks after impregnation with bFGF. As can be seen, the presence of implanted sponge is no more recognizable unless the sponge is seeded with chondrocytes, whereas we can clearly notice the sponge from the outside of the mouse if chondrocytes are seeded in the collagen sponge.
Optical photographs of sponges implanted for 1 week with and without bFGF are shown in Figure 9. Obviously, we cannot see any angiogenesis if the collagen sponge contained neither bFGF nor chondrocytes, whereas seeding of chondrocytes in the sponge induced angiogenesis even if chondrocyte was not seeded. Impregnation of the sponge with bFGF markedly enhanced angiogenesis, regardless of chondrocyte seeding. When the chondrocyte-collagen composite was implanted for 4 weeks together with bFGF, the formation of cartilage was clearly noticed. An optical photograph of the cartilage formed by 4 week implantation of the chondrocyte-collagen composite is shown in Figure 10. The soft sponge became smaller in size and less flexible. The size decrease became more prominent with the increasing density of seeded cells. The size dependence on the density of seeded cells after 2 weeks of implantation is shown in Figure 11.
To assess the cartilage regeneration, we stained the cross-section of the explanted collagen composites with safranin-O fast green. The result for the chondrocyte-collagen composites implanted for 4 weeks in mice is given in Figure 12. The area with the regenerated cartilage should be stained strongly reddish with this dye as a result of metachromasia31. Figure 12 clearly indicates that only a very small fraction of the composite cross-section shows metachromasia unless bFGF is incorporated in the collagen sponge, whereas most of the cross-section of the composite impregnated with bFGF and seeded with chondrocytes exhibit strong metachromasia supporting the cartilage formation. The staining of explants with safranin O-fast green revealed that chondrocytes could produce specific proteoglycans only after 1 week of implantation. However, even at 2 weeks after implantation, chondrocytes could regenerate the cartilage tissue, although immatured, and at 4 weeks after implantation, almost all of the chondrocytes transferred to the matured stage. Without impregnating the collagen sponge with bFGF, such a matured cartilage tissue was not observed up to 4 weeks after implantation.
H.E. staining of the implanted composites showed that a thin layer of fibrous capsule was formed surrounding the implanted composites with a mild inflammatory response of the host to the implants (Figure 12).
DISCUSSION
To regenerate tissues and maintain their biological functions, individual cells will be probably collected from the organ or tissue of patients, followed by attachment of the cells to a bioabsorbable polymer scaffold by a culture technique. The resulting cell-polymer composite is generally implanted at a site where the cells can grow and effectively express their function. Chondrocytes are parenchymal cells like hepatocytes, but it is reported that they exhibit much more remarkable proliferation and cartilage formation even in vitro if the culture condition is adequate15-18. This suggests that reconstruction of the cartilage tissue is easier than that of the liver. In a previous work, a hepatocyte- or chondrocyte-PLA complex was prepared to study its potentiality for liver regeneration and cartilage reconstruction23. The measurement of the protein production from the cells as an index of cell function and expression of their phenotypes revealed that hepatocytes did not undergo any growth and gradually diminished their function of albumin biosynthesis on the PLA substrate in vitro. On the contrary, chondrocytes grew well even on culture dishes and produced type II collagen, which became maximal on the 10th day after cultivation. This is simply because the cell density increased to the highest level by 10 day cultivation, reaching the confluent state and decreased gradually after the 10th day. This indicates that chondrocytes produced smaller amounts of collagen under the confluent condition than prolifelative stage. In the case of hepatocytes, a multicellular colony was seen in the scaffold, in contrast to the seeded chondrocytes which spread dispersively throughout the scaffold.
When the chondrocyte-collagen composite was implanted into nude mice, injection of any immunosuppressive agent was not necessary even in the xenograft implantation. From the histological staining of the implanted composites with safranin O-fast green which can bind to negatively charged glycosaminoglycans in cartilage, it is obvious that the implanted cells exhibited their phenotype in vivo and formed a new matured cartilage in addition to the morphological characteristics (Figure 12). On the contrary, implantation of collagen scaffolds without cultured chondrocytes did not result in formation of new cartilage at all.
Quick formation of capillaries around the implanted composite seems necessary to maintain the seeded cells alive and promote tissue reconstruction. Therefore, bFGF, a well-known angiogenic factor, was incorporated into the collagen sponge, although cartilage is an avascular tissue. Expectedly, a lot of small blood vessels were observed around the matrix when the collagen scaffold was impregnated with bFGF (Figure 9). This strongly suggests that the use of angiogenic factor is very effective for blood vessel formation around the complex, which may promote the cartilage formation. Indeed, acceleration of cartilage reconstruction by bFGF incorporated in the composite was noticed as demonstrated above. It has been proposed that cartilage tissue is transformed to bone tissue if vascular invasion occurs and chondrocytes terminally differentiate to hypertrophic chondrocytes which produce high levels of alkaline phosphatase. During this study we did not observe any differentiation of chondrocytes. It is interesting to point out that bFGF is reported to promote cartilage repair in vivo27 and inhibit the terminal differentiation of chondrocytes and calcification28. The formation of small blood vessels around the collagen sponge became the most remarkable at 1 week after implantation with bFGF and then diminished gradually. The histological study suggests that chondrocytes would be in the proliferative stage in the first 2 weeks and then transferred in the matured stage. More detailed histological and long-term implantation are needed to follow the fate of the regenerated cartilage.