- Email: [email protected]
Modified keratin sponge: Binding of bone morphogenetic protein-2 and osteoblast differentiation
JOURNAL OF BIOSCIENCE AND BIOENGINEERING Vol. 102, No. 5, 425–429. 2006 DOI: 10.1263/jbb.102.425
© 2006, The Society for Biotechnology, Japan
Modified Keratin Sponge: Binding of Bone Morphogenetic Protein-2 and Osteoblast Differentiation Akira Tachibana,1* Yuji Nishikawa,1 Masaaki Nishino,1 Sumika Kaneko,1 Toshizumi Tanabe,1 and Kiyoshi Yamauchi1 Department of Bioengineering, Graduate School of Engineering, Osaka City University, 3-3-138 Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan 1 Received 16 May 2006/Accepted 2 August 2006
A keratin sponge was chemically modified to obtain carboxyl and amino sponges by the alkylation of a large amount of active SH group on keratin proteins with iodoacetic acid and 2-bromoethylamine, respectively. The carboxyl sponge having a large amount of carboxyl group was a scaffold that could bind significant amounts of basic bioactive proteins, such as lysozyme and bone morphogenetic protein (BMP)-2, and drugs. Lysozyme (maximum 3.7 mg), a model of basic cytokines such as BMP-2, was absorbed by the carboxyl sponge (4.8 mg), but not by the amino sponge. The lysozyme was rapidly released from the carboxyl sponge in a buffer containing at a concentration higher than 0.5 M, but at 0.15 M, near the physiological ionic strength, after initial burst (only 11%), no significant release was observed (15%, 48 h). BMP-2 also bound the carboxyl sponge. The preosteoblast cells grown inside the BMP-2-loaded sponge differentiated, whereas those grown outside the sponge did not, suggesting that no significant amount of BMP-2 leaked into the surrounding media. The effects of BMP-2 were confined inside modified keratin sponge. Therefore, using in vivo, we expect that only internal osteogenesis will be induced and that no external heterotopis ossification will be induced. [Key words: keratin sponge, scaffold-bone morphogenetic protein-2 (BMP-2) interaction, retention of bone morphogenetic protein-2 (BMP-2) in scaffolds, osteoblast scaffolds, osteoblast differentiation]
metalloproteinase; thus, the initial cell growth (within 7 d) on/in the synthetic gel was relatively poorer than that in the sponge. We consider that porous sponge biomaterials enhance the initial growth of cells, which is important for bone repair and regeneration. Wool keratins are alternative natural proteinous biomaterials for collagen (9). We demonstrated that the keratin sponges are useful scaffolds for fibroblasts (10) and osteoblasts (11), owing to their cell adhesion sequences, arginineglycine-aspartic acid (RGD) and leucine-aspartic acid-valine (LDV), biocompatibility, and high cysteine content (about 20 mol%) for modification targets. The thiol groups of the cysteine residue on the keratin sponges were good targets for bioactive protein conjugation (12, 13) and chemical modification (10, 11). Chemically modified keratin sponges, particularly a carboxylmethylated keratin sponge (called the carboxyl sponge), were used as substrates for rapid hydroxyapatite formation in a solution containing Ca and phosphate ions because of their high contents of carboxyl groups. Hydroxyapatite formation was also observed both in the protein wall and on the wall surface of the carboxyl sponge (11). This suggests the introduction of carboxyl groups both in the protein wall and on the wall surface. We expect that these carboxyl groups with negative charges will contribute to the binding of basic proteins, such as lysozyme and
Bone morphogenetic protein-2 (BMP-2) is a member of the transforming growth factor-β superfamily (1) and regulates the differentiation of various cells involved in cartilage and bone formation during fracture repair (2). Due to the very short half-life of BMP-2 when administered in solution, the use of an appropriate delivery system is required (2–5). In previous studies, collagen sponges were used as BMP-2 carrier for bone repair including osteoblast proliferation and differentiation (6, 7). Collagen and gelatin hydrogels were also used, demonstrating that the half-lives of BMP-2 ranged from 1 to more than 30 days in the hydrogels (3); it was longer than about a day in the collagen sponges. However, it is known that collagen from bovine and other sources may transmit viral and prion infections (8). Lutolf et al. (5) proposed a gel as an excellent BMP-2 carrier, which was a polyethylene-glycol-based gel harboring two oligopeptide sites, integrin-binding (RGDSP) and a substrate for matrix metalloproteinase. The amount of BMP-2 released from the gel was smaller than that from a type-I collagen sponge; however, in the presence of matrix metalloproteinase-2, it was laeger (5). BMP-2 release was considered a response to the invasion of cells producing matrix * Corresponding author. e-mail: [email protected] phone: +81-(0)6-6605-2702 fax: +81-(0)6-6605-2786 425
426
TACHIBANA ET AL.
BMP-2. We also expect that the role of the abundant negative charges of the carboxyl sponge will be the same as that of natural extracellular matrix, such as collagen, heparin, and heparan sulfate, in BMP-2 binding. In this paper, we report the tight binding of basic proteins, lysozyme and BMP-2, to a carboxyl sponge and the effects of BMP-2 loaded to the sponge on the osteoblast differentiation of cells grown inside and outside the sponge.
MATERIALS AND METHODS Fabrication of chemically modified keratin sponge Keratin was prepared from wool according to the method reported previously (11). The keratin solution (100 µl) containing 8 mg of protein was added into a 96-well plate, frozen at −20°C for 3 d, and lyophilized to form sponges. The sponges were treated with 10 ml of 0.1 M iodoacetic acid in 0.5 M Tris–HCl buffer (pH 8.5) at room temperature to produce a carboxyl sponge (10). Another modified sponge, an amino sponge, was also obtained by the treatment of 2-bromoethylamine instead of iodoacetic acid. These chemically modified sponges were washed with phosphate-buffered saline (PBS(−) containing 0.2 g of KCl, 8 g of NaCl, 1.1 g of Na2HPO4, and 0.2 g of KH2PO4 per liter, pH 7.4) at 60°C for 1 h to remove the remaining alkylating reagents and SDS. The washing procedure was repeated. The modification of keratin sponges did not lead to significant changes in properties, such as hardness. The carboxyl sponge contained carboxymethyl cystein residues at about 20 mol% (10) and native acidic amino acid residues, such as aspartate and glutamate, at about 20 mol% (9); thus, the amounts of carboxyl residues on the carboxyl sponge were 2-fold those on a nontreated sponge. The protein content in the keratin solution was measured by a modified Lowry method (14) using bovine serum albumin as the standard. Absorption and release of lysozyme The chemically modified sponges were added to the lysozyme solution (0.5 ml) in 10 mM Tris (pH 8.0). The supernatant was sampled (10–20 µl), and the lysozyme activities of the samples were assayed using the bacterium Micrococcus as previously reported (12). Release experiment was also performed. Lysozyme-loaded sponges were transferred into 1 ml of 10 mM Tris buffer (pH 8.0) containing NaCl at various concentrations. Sampling and assay were performed as described above. Lysozyme amount was calculated from the activities based on the experimental data (50,000 units per mg of lysozyme). BMP-2 binding and osteoblast differentiation assay BMP-2 (R&D Systems, Minneapolis, MN, USA; recombinant human BMP-2, CHO-expressed) was loaded to the carboxyl and amino sponges, and the effects of these sponges on preosteoblast differentiation were tested. The sponges were sterilized with 70% ethanol for 1 h, washed with PBS(−) to remove ethanol twice, and finally washed with the cultivation medium MEM-alpha (Gibco; not containing phenolred) supplemented with 2 mM glutamine, 50 IU/ml penicillin, 50 µg/ml streptomycin, and 10% fetal bovine serum (FBS). The washed sponges were transferred into 96-well plate. BMP-2 (10 ng) in the cultivation medium (100 µl) was added into the well, and the plate was maintained at 37°C for 1 d, BMP-2 in the medium was expected to be absorbed by the carboxyl sponge. Beforecell cultivation, the resulting medium was removed. Mouse pre-osteoblast MC3T3-E1 cells obtained from the RIKEN CELL BANK (Tsukuba) were grown in the above-mentioned cultivation medium. The cells were harvested by typical trypsin-EDTA treatment, washed with PBS(−) and medium, and counted. The cells
J. BIOSCI. BIOENG.,
(50,000 cells suspended in 20 µl of medium) were seeded onto the sponges, before the cultivation medium (80 µl) was added. Over 95% of the cells seeded were trapped on the sponges (10). The cultures were maintained in a humidified incubator with a mixture of 95% air and 5% CO2 at 37°C, and the medium was exchange every three days. An early osteoblast differentiation marker, alkaline phosphatase (ALP) activity, was measured using the cell-growing sponges as enzymes for cleaving p-nitrophenyl phosphate using a SigmaFast tablet by the manufacture’s method (Sigma, St. Louis, MO, USA). The resultant solutions were transferred into a 96-well plate, and the absorbance at 405 nm was measured and expressed as ALP activity. Leakage test of BMP-2 from BMP-2 loaded sponges BMP-2 is a member of the TGF-β superfamily and multifunctional (1). Therefore, the significant leakage of BMP-2 from the transplanted sponges could induce unexpected and disordered tissue processes, such as heterotopis ossification around the sponges (5). The prevention of the leakage of BMP-2 from the sponges is very important for the above-mentioned disordered tissue induction (4). We expect that a large amount of the carboxyl group of the carboxyl sponge will prevent the leakage of the basic proteins BMP-2 and lysozyme. The differentiation of the preosteoblast MC3T3-E1 inside and outside the BMP-2-loaded sponges was analyzed. When BMP-2 was maintained in the sponges, the cultures inside the sponges differentiated and those outside the sponges did not. On the other hand, when a significant amount of BMP-2 leaked from the sponges, the cultures outside the sponges differentiated. The carboxyl and amino sponges were loaded with BMP-2 (7 ng in 10 µl of the culture medium) and were placed on the membrane (pore size, 8 µm) of the cup (Transwell, Corning, New York, USA) set on a 24-well plate (Fig. 4A). After 10 min, MC3T3-E1 cells were seeded onto the BMP-2 loaded sponges (50,000 cells) and onto the well (12,500 cells), a culture outside the sponge (Fig. 4A). At 1 h, the culture medium (700 µl) was added, the cultures were maintained in a humidified incubator with a mixture of 95% air and 5% CO2 at 37°C. The positive (7 ng of BMP-2 in 700 µl of the culture medium, 12,500 cells on the 24-well dish) and negative (without BMP-2) control cultures were also incubated. At day 4, the ALP activities the cultures inside and outside the sponge were measured.
RESULTS AND DISCUSSION Lysozyme absorption and release Keratin sponges are abundant in SH groups as determined from the high contents of cysteine residues (9). The SH group is a good target of chemical modification with alkylating agents, such as iodoacetic acid (10). Thus, the carboxyl sponge is abundant in carboxyl groups. The carboxyl sponge scaffold was a mimic of matrix γ-carboxyglutamic acid protein and where hydroxyapatite was grown as demonstrated in our previous study (11). The carboxyl sponge is also abundant in negative charges, which is a useful characteristics for basic-protein binding based on mild ionic interaction. First, we examined the basic-protein binding properties of the carboxyl sponge using the model protein lysozyme. Lysozyme in solution was absorbed by the carboxyl and nontreated sponges. Keratins are relatively acidic proteins; therefore, the nontreated sponge also absorbed lysozyme (Fig. 1). On the other hand, the amino sponge slightly absorbed lysozyme (data not shown). The absorption speed of the carboxyl sponge was higher than that of the nontreated
VOL. 102, 2006
MODIFIED KERATIN SPONGE: BMP-2 BINDING
427
FIG. 1. Lysozyme absorption by carboxyl (open) and nontreated (closed) sponges. (A) Time-dependent lysozyme absorption from lysozyme solution (2.5 mg of lysozyme in 0.5 ml of 10 mM Tris, pH 8.0). The remaining lysozyme activities were analyzed. (B) Relationship between lysozyme amounts added and absorbed. The carboxyl and nontreated sponges were incubated in lysozyme solution (0.5 ml) with the indicated lysozyme amounts for 96 h, and the remaining lysozyme activities were assayed. The data subtracted from the amount of added lysozyme to that of remaining lysozyme are shown as lysozyme absorbed.
sponge. At 2 h, 37% of lysozyme added was absorbed by the carboxyl sponge, but only 5% of lysozyme added was absorbed by the nontreated sponge, as shown in Fig. 1A. The maximum lysozyme amounts of nontreated and carboxyl sponges were 0.95 to 1.4 mg and 3.7 mg, respectively (Fig. 1B). The weight of the sponge used was 4.8 mg; thus, the amount of absorbed protein (3.7 mg) was surprisingly high. In our next examination, the absorbed lysozyme was released from the carboxyl sponge at various ionic strengths. At 0.5 or 1.0 M NaCl, lysozyme was rapidly released, but at a relatively low concentration (0 to 0.15 M), only <15% of lysozyme was released. At 0.15 M, near the physiological ionic strength, an initial burst (to 30 min, 11% was released) was observed, and then no significant release was observed (15% at 48 h). The initial burst may depend on the release of lysozyme from the surface of the sponge, but not from the inner part of the sponge. We considered that the basic bioactive protein in the carboxyl sponge is maintained for several days under physiological conditions. Osteoblast differentiation on BMP-2-loaded sponges The interaction of BMP-2 with the carboxyl sponge was examined under physiological conditions, that is, in cultivation medium containing other serum proteins and acidic and basic small molecules, such as nucleosides and nucleotides. The differentiation induction by BMP-2 bound to the carboxyl sponge was also examined. BMP-2 in cultivation medium was absorbed by the sponges, and then the preosteoblast MC3T3-E1 was seeded onto the sponges. When the sponges absorbed and maintained BMP-2 inside in an active form, osteoblast differentiation was observed. Osteoblast differentiation was analyzed by measuring ALP activity, which is an early differentiation marker. As shown in Fig. 3, the cells expressed ALP activity on the BMP-2-loaded carboxyl sponge, but not on the amino sponge. The carboxyl sponge absorbed and maintained BMP-2 in cultivation medium including 10% FBS. After BMP-2 absorption, in the remaining cultivation medium,
FIG. 2. Lysozyme release from carboxyl sponge with 10 mM Tris (pH 8.0) containing NaCl at various concentrations. Lysozyme (3 mg) was absorbed by the carboxyl sponge, and then the sponge was transferred into the Tris buffer containing NaCl at 1.0 M (open squares), 0.5 M (closed squares), 0.15 M (open circles), 0.1 M (closed triangles), and 0 M (open triangles). At 0.15 M NaCl, near the physiological ionic strength, an initial burst release was observed (to 6 h), but not a significant change (6 to 48 h).
osteoblasts did not differentiate (data not shown). On the other hand, the amino sponge did not possess the abovementioned ability. These results suggest that the negative charges on the sponges are important and that the ionic interaction between the sponges and BMP-2 is sufficient in retaining BMP-2 in the sponges.
428
TACHIBANA ET AL.
J. BIOSCI. BIOENG.,
FIG. 3. Osteoblast differentiation on BMP-2-loaded sponges. The alkaline phosphatase (ALP) activities of osteoblast culture on the carboxyl sponge (open circles) and amino sponge (closed circles) were measured as early markers of osteoblast differentiation.
Outside culture of BMP-2 loaded sponge did not significantly differentiate BMP-2 was loaded to the carboxyl and amino sponges, and then ALP activities were measured for the inside and outside cultures of osteoblasts (Fig. 4). As shown in Fig. 4B, the carboxyl sponge maintained the loaded BMP-2 and differentiated osteoblasts, but the amino sponge did not. The culture outside the amino sponge expressed an ALP activity level higher than that of the negative control culture and slightly lower than that of the positive control, and significantly differentiated. The ALP acitivity level of the culture outside the carboxyl sponge was the same as that of the negative control culture, and the effects of BMP-2 were not observed. No significant amount of BMP-2 might be released from the carboxyl sponge. In conclusion, the basic bioactive protein BMP-2 was relatively strongly bound to the carboxyl sponge without the loss of the bioactivity. The carboxyl groups are acids weaker than the phosphate and sulfonate found on ion-exchange resins as tight protein binder. Other protein biomaterials, such as collagen, have also carboxyl groups as well as the carboxyl sponge and the relatively loose binding and (controlled) release of basic proteins were reported (2). Although the carboxyl sponge contained the same carboxyl groups, why were BMP-2 and lysozyme released in small amount from the carboxyl sponge? This may be responsible for the large amount of carboxyl groups on the carboxyl sponge with chemically modified carboxymethyl cysteine residues (about 20 mol% [10]) and native acidic amino acid residues (about 20 mol% [9]), such as aspartate and glutamate. The N-terminal amino acid sequence, QAKHKQRKRLK, of human BMP-2 are lysine and arginine rich and very basic. These basic residues of the N-terminus of human BMP-2 may produce multiple ionic interactions with crowded acidic residues (about 40 mol%) on the carboxyl sponge. The multiple ionic interactions may be stronger than the interactions between BMP-2 and other scaffolds, such as collagen. The multiple ionic interactions between BMP-2 and the carboxyl sponge may be prevent the release of BMP-2 from the sponge. The effects of BMP-2 are confined inside the modi-
FIG. 4. Leakage test of BMP-2 from BMP-2-loaded carboxyl sponge. (A) Schematic illustration of cultures inside and outside BMP-2 loaded sponge. (B) The cultures inside BMP-2-loaded amino and carboxyl sponges were analyzed in term of ALP acitivities. (C) The outside cultures of amino and carboxyl sponges were analyzed with ALP activities. The addition of equal amounts of BMP-2 loaded and nonaddition are shown as the positive and negative control cultures, respectively.
fied keratin sponge, when the loaded sponge is used in vivo. Therefore, we expect that only internal osteogenesis will be induced and that no external heterotopis ossification outside will be induced. Our previous work demonstrated that the carboxyl sponge with a large amount of carboxyl groups is a mimic of osteoblastic extracellular matrix protein, such as matrix γ-carboxyglutamic acid protein, which is responsible for osteoblast calcification (11). In addition, BMP-2/4 accumulation in extracellular matrices was reported to be essential for the osteoblastic differentiation of cells in the osteoblast lineage (16). We consider that the carboxyl sponge is also a useful mimic of extracellular matrix proteins for the accumulation and binding of BMPs for osteoblast differentiation. For the clinical applications of the biomaterials, further work is re-
VOL. 102, 2006
MODIFIED KERATIN SPONGE: BMP-2 BINDING
quired to clarify bone metabolism and the detailed differentiation of osteoblasts in the carboxyl sponge tightly binding BMP-2. ACKNOWLEDGMENTS This study was partly supported by a Science and Culture grant from the Japanese Ministry of Education, Culture, Sports, Science and Technology.
REFERENCES 1. Wozney, J. M.: The bone morphogenetic protein family: multifuctional cellular regulators in the embryo and adult. Eur. J. Oral Sci., 106, 160–166 (1998). 2. Li, R. H. and Wozney, J. M.: Delivering on the promise of bone morphogenetic proteins. Trends Biotechnol., 19, 255– 265 (2001). 3. Yamamoto, M., Takahashi, Y., and Tabata, Y.: Controlled release by biodegradable hydrogels enhances the ectopid bone formation of bone morphogenetic protein. Biomaterials, 24, 4375–4383 (2003). 4. Wildemann, B., Kandziora, F., Krummrey, G., Palasdies, N., Haas, N. P., Raschke, M., and Schrnidmaier, G.: Local and controlled release of growth factors (combination of IGF-1 and TGF-beta1, and BMP-2 alone) from a polylactide coating of titanium implants does not lead to ectopic bone formation in sheep muscle. J. Control. Release, 95, 249–256 (2004). 5. Lutolf, M. P., Weber, F. E., Schmoekel, H. G., Schense, J. C., Kohler, T., Müller, R., and Hubbell, J. A.: Repair of bone defects using synthetic mimetics of collagenous extracellular matrices. Nat. Biotechnol., 21, 513–518 (2003). 6. Friess, W., Uludag, H., Foskett, S., Biron, R., and Sargeant, C.: Characterization of absorbable collagen sponges as rhBMP-2 carriers. Int. J. Pharm., 187, 91–99 (1999).
429
7. Geiger, M., Li, R. H., and Friess, W.: Collagen sponges for bone regeneration with rhBMP-2. Adv. Drug Deliv. Rev., 55, 1613–1629 (2003). 8. Doillon, C. J., Drouin, R., Cote, M. F., Dallaire, N., Pageau, J. F., and Laroche, G.: Chemical inactivatores as sterilization agents for bovine collagen materials. J. Biomed. Mater. Res., 37, 212–221 (1997). 9. Tanabe, T., Tachibana, A., and Yamauchi, K.: Keratins: prospective proteiunous biomaterial. Recent Res. Dev. Protein Eng., 1, 247–259 (2001). 10. Tachibana, A., Furuta, Y., Takeshima, H., Tanabe, T., and Yamauchi, K.: Fabrication of wool keratin sponge scaffolds for long-term cell cultivation. J. Biotechnol., 93, 165–170 (2002). 11. Tachibana, A., Kaneko, S., Tanabe, T., and Yamauchi, K.: Rapid fabrication of keratin-hydroxyapatite hybrid sponges toward osteoblast cultivation and differentiation. Biomaterials, 26, 297–302 (2005). 12. Kurimoto, A., Tanabe, T., Tachibana, A., and Yamauchi, K.: Thiolated dermal bovine collagen as a novel support for bioactive substances — conjugation with lysozyme. J. Biotechnol., 86, 1–8 (2001). 13. Kurimoto, A., Tanabe, T., Tachibana, A., and Yamauchi, K.: Keratin sponge: immobilization of lysozyme. J. Biosci. Bioeng., 96, 307–309 (2003). 14. Peterson, G. L.: A simplification of the protein assay method of Lowry et al. which is more generally applicable. Anal. Biochem., 83, 346–356 (1977). 15. Kirsch, T., Nickel, J., and Sebald, W.: BMP-2 antagonists emerge from alterations in the low-affinity binding epitope for receptor BMPR-2. EMBO J., 19, 3314–3324 (2000). 16. Suzawa, M., Takeuchi, Y., Fukumoto, S., Kato, S., Ueno, N., Miyazono, K., Matsumoto, T., and Fujita, T.: Extracellular matrix-associated bone morphogenetic proteins are essential for differentiation of murine osteoblastic cells in vitro. Endocrinology, 140, 2125–2133 (1999).
© 2006, The Society for Biotechnology, Japan
Modified Keratin Sponge: Binding of Bone Morphogenetic Protein-2 and Osteoblast Differentiation Akira Tachibana,1* Yuji Nishikawa,1 Masaaki Nishino,1 Sumika Kaneko,1 Toshizumi Tanabe,1 and Kiyoshi Yamauchi1 Department of Bioengineering, Graduate School of Engineering, Osaka City University, 3-3-138 Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan 1 Received 16 May 2006/Accepted 2 August 2006
A keratin sponge was chemically modified to obtain carboxyl and amino sponges by the alkylation of a large amount of active SH group on keratin proteins with iodoacetic acid and 2-bromoethylamine, respectively. The carboxyl sponge having a large amount of carboxyl group was a scaffold that could bind significant amounts of basic bioactive proteins, such as lysozyme and bone morphogenetic protein (BMP)-2, and drugs. Lysozyme (maximum 3.7 mg), a model of basic cytokines such as BMP-2, was absorbed by the carboxyl sponge (4.8 mg), but not by the amino sponge. The lysozyme was rapidly released from the carboxyl sponge in a buffer containing at a concentration higher than 0.5 M, but at 0.15 M, near the physiological ionic strength, after initial burst (only 11%), no significant release was observed (15%, 48 h). BMP-2 also bound the carboxyl sponge. The preosteoblast cells grown inside the BMP-2-loaded sponge differentiated, whereas those grown outside the sponge did not, suggesting that no significant amount of BMP-2 leaked into the surrounding media. The effects of BMP-2 were confined inside modified keratin sponge. Therefore, using in vivo, we expect that only internal osteogenesis will be induced and that no external heterotopis ossification will be induced. [Key words: keratin sponge, scaffold-bone morphogenetic protein-2 (BMP-2) interaction, retention of bone morphogenetic protein-2 (BMP-2) in scaffolds, osteoblast scaffolds, osteoblast differentiation]
metalloproteinase; thus, the initial cell growth (within 7 d) on/in the synthetic gel was relatively poorer than that in the sponge. We consider that porous sponge biomaterials enhance the initial growth of cells, which is important for bone repair and regeneration. Wool keratins are alternative natural proteinous biomaterials for collagen (9). We demonstrated that the keratin sponges are useful scaffolds for fibroblasts (10) and osteoblasts (11), owing to their cell adhesion sequences, arginineglycine-aspartic acid (RGD) and leucine-aspartic acid-valine (LDV), biocompatibility, and high cysteine content (about 20 mol%) for modification targets. The thiol groups of the cysteine residue on the keratin sponges were good targets for bioactive protein conjugation (12, 13) and chemical modification (10, 11). Chemically modified keratin sponges, particularly a carboxylmethylated keratin sponge (called the carboxyl sponge), were used as substrates for rapid hydroxyapatite formation in a solution containing Ca and phosphate ions because of their high contents of carboxyl groups. Hydroxyapatite formation was also observed both in the protein wall and on the wall surface of the carboxyl sponge (11). This suggests the introduction of carboxyl groups both in the protein wall and on the wall surface. We expect that these carboxyl groups with negative charges will contribute to the binding of basic proteins, such as lysozyme and
Bone morphogenetic protein-2 (BMP-2) is a member of the transforming growth factor-β superfamily (1) and regulates the differentiation of various cells involved in cartilage and bone formation during fracture repair (2). Due to the very short half-life of BMP-2 when administered in solution, the use of an appropriate delivery system is required (2–5). In previous studies, collagen sponges were used as BMP-2 carrier for bone repair including osteoblast proliferation and differentiation (6, 7). Collagen and gelatin hydrogels were also used, demonstrating that the half-lives of BMP-2 ranged from 1 to more than 30 days in the hydrogels (3); it was longer than about a day in the collagen sponges. However, it is known that collagen from bovine and other sources may transmit viral and prion infections (8). Lutolf et al. (5) proposed a gel as an excellent BMP-2 carrier, which was a polyethylene-glycol-based gel harboring two oligopeptide sites, integrin-binding (RGDSP) and a substrate for matrix metalloproteinase. The amount of BMP-2 released from the gel was smaller than that from a type-I collagen sponge; however, in the presence of matrix metalloproteinase-2, it was laeger (5). BMP-2 release was considered a response to the invasion of cells producing matrix * Corresponding author. e-mail: [email protected] phone: +81-(0)6-6605-2702 fax: +81-(0)6-6605-2786 425
426
TACHIBANA ET AL.
BMP-2. We also expect that the role of the abundant negative charges of the carboxyl sponge will be the same as that of natural extracellular matrix, such as collagen, heparin, and heparan sulfate, in BMP-2 binding. In this paper, we report the tight binding of basic proteins, lysozyme and BMP-2, to a carboxyl sponge and the effects of BMP-2 loaded to the sponge on the osteoblast differentiation of cells grown inside and outside the sponge.
MATERIALS AND METHODS Fabrication of chemically modified keratin sponge Keratin was prepared from wool according to the method reported previously (11). The keratin solution (100 µl) containing 8 mg of protein was added into a 96-well plate, frozen at −20°C for 3 d, and lyophilized to form sponges. The sponges were treated with 10 ml of 0.1 M iodoacetic acid in 0.5 M Tris–HCl buffer (pH 8.5) at room temperature to produce a carboxyl sponge (10). Another modified sponge, an amino sponge, was also obtained by the treatment of 2-bromoethylamine instead of iodoacetic acid. These chemically modified sponges were washed with phosphate-buffered saline (PBS(−) containing 0.2 g of KCl, 8 g of NaCl, 1.1 g of Na2HPO4, and 0.2 g of KH2PO4 per liter, pH 7.4) at 60°C for 1 h to remove the remaining alkylating reagents and SDS. The washing procedure was repeated. The modification of keratin sponges did not lead to significant changes in properties, such as hardness. The carboxyl sponge contained carboxymethyl cystein residues at about 20 mol% (10) and native acidic amino acid residues, such as aspartate and glutamate, at about 20 mol% (9); thus, the amounts of carboxyl residues on the carboxyl sponge were 2-fold those on a nontreated sponge. The protein content in the keratin solution was measured by a modified Lowry method (14) using bovine serum albumin as the standard. Absorption and release of lysozyme The chemically modified sponges were added to the lysozyme solution (0.5 ml) in 10 mM Tris (pH 8.0). The supernatant was sampled (10–20 µl), and the lysozyme activities of the samples were assayed using the bacterium Micrococcus as previously reported (12). Release experiment was also performed. Lysozyme-loaded sponges were transferred into 1 ml of 10 mM Tris buffer (pH 8.0) containing NaCl at various concentrations. Sampling and assay were performed as described above. Lysozyme amount was calculated from the activities based on the experimental data (50,000 units per mg of lysozyme). BMP-2 binding and osteoblast differentiation assay BMP-2 (R&D Systems, Minneapolis, MN, USA; recombinant human BMP-2, CHO-expressed) was loaded to the carboxyl and amino sponges, and the effects of these sponges on preosteoblast differentiation were tested. The sponges were sterilized with 70% ethanol for 1 h, washed with PBS(−) to remove ethanol twice, and finally washed with the cultivation medium MEM-alpha (Gibco; not containing phenolred) supplemented with 2 mM glutamine, 50 IU/ml penicillin, 50 µg/ml streptomycin, and 10% fetal bovine serum (FBS). The washed sponges were transferred into 96-well plate. BMP-2 (10 ng) in the cultivation medium (100 µl) was added into the well, and the plate was maintained at 37°C for 1 d, BMP-2 in the medium was expected to be absorbed by the carboxyl sponge. Beforecell cultivation, the resulting medium was removed. Mouse pre-osteoblast MC3T3-E1 cells obtained from the RIKEN CELL BANK (Tsukuba) were grown in the above-mentioned cultivation medium. The cells were harvested by typical trypsin-EDTA treatment, washed with PBS(−) and medium, and counted. The cells
J. BIOSCI. BIOENG.,
(50,000 cells suspended in 20 µl of medium) were seeded onto the sponges, before the cultivation medium (80 µl) was added. Over 95% of the cells seeded were trapped on the sponges (10). The cultures were maintained in a humidified incubator with a mixture of 95% air and 5% CO2 at 37°C, and the medium was exchange every three days. An early osteoblast differentiation marker, alkaline phosphatase (ALP) activity, was measured using the cell-growing sponges as enzymes for cleaving p-nitrophenyl phosphate using a SigmaFast tablet by the manufacture’s method (Sigma, St. Louis, MO, USA). The resultant solutions were transferred into a 96-well plate, and the absorbance at 405 nm was measured and expressed as ALP activity. Leakage test of BMP-2 from BMP-2 loaded sponges BMP-2 is a member of the TGF-β superfamily and multifunctional (1). Therefore, the significant leakage of BMP-2 from the transplanted sponges could induce unexpected and disordered tissue processes, such as heterotopis ossification around the sponges (5). The prevention of the leakage of BMP-2 from the sponges is very important for the above-mentioned disordered tissue induction (4). We expect that a large amount of the carboxyl group of the carboxyl sponge will prevent the leakage of the basic proteins BMP-2 and lysozyme. The differentiation of the preosteoblast MC3T3-E1 inside and outside the BMP-2-loaded sponges was analyzed. When BMP-2 was maintained in the sponges, the cultures inside the sponges differentiated and those outside the sponges did not. On the other hand, when a significant amount of BMP-2 leaked from the sponges, the cultures outside the sponges differentiated. The carboxyl and amino sponges were loaded with BMP-2 (7 ng in 10 µl of the culture medium) and were placed on the membrane (pore size, 8 µm) of the cup (Transwell, Corning, New York, USA) set on a 24-well plate (Fig. 4A). After 10 min, MC3T3-E1 cells were seeded onto the BMP-2 loaded sponges (50,000 cells) and onto the well (12,500 cells), a culture outside the sponge (Fig. 4A). At 1 h, the culture medium (700 µl) was added, the cultures were maintained in a humidified incubator with a mixture of 95% air and 5% CO2 at 37°C. The positive (7 ng of BMP-2 in 700 µl of the culture medium, 12,500 cells on the 24-well dish) and negative (without BMP-2) control cultures were also incubated. At day 4, the ALP activities the cultures inside and outside the sponge were measured.
RESULTS AND DISCUSSION Lysozyme absorption and release Keratin sponges are abundant in SH groups as determined from the high contents of cysteine residues (9). The SH group is a good target of chemical modification with alkylating agents, such as iodoacetic acid (10). Thus, the carboxyl sponge is abundant in carboxyl groups. The carboxyl sponge scaffold was a mimic of matrix γ-carboxyglutamic acid protein and where hydroxyapatite was grown as demonstrated in our previous study (11). The carboxyl sponge is also abundant in negative charges, which is a useful characteristics for basic-protein binding based on mild ionic interaction. First, we examined the basic-protein binding properties of the carboxyl sponge using the model protein lysozyme. Lysozyme in solution was absorbed by the carboxyl and nontreated sponges. Keratins are relatively acidic proteins; therefore, the nontreated sponge also absorbed lysozyme (Fig. 1). On the other hand, the amino sponge slightly absorbed lysozyme (data not shown). The absorption speed of the carboxyl sponge was higher than that of the nontreated
VOL. 102, 2006
MODIFIED KERATIN SPONGE: BMP-2 BINDING
427
FIG. 1. Lysozyme absorption by carboxyl (open) and nontreated (closed) sponges. (A) Time-dependent lysozyme absorption from lysozyme solution (2.5 mg of lysozyme in 0.5 ml of 10 mM Tris, pH 8.0). The remaining lysozyme activities were analyzed. (B) Relationship between lysozyme amounts added and absorbed. The carboxyl and nontreated sponges were incubated in lysozyme solution (0.5 ml) with the indicated lysozyme amounts for 96 h, and the remaining lysozyme activities were assayed. The data subtracted from the amount of added lysozyme to that of remaining lysozyme are shown as lysozyme absorbed.
sponge. At 2 h, 37% of lysozyme added was absorbed by the carboxyl sponge, but only 5% of lysozyme added was absorbed by the nontreated sponge, as shown in Fig. 1A. The maximum lysozyme amounts of nontreated and carboxyl sponges were 0.95 to 1.4 mg and 3.7 mg, respectively (Fig. 1B). The weight of the sponge used was 4.8 mg; thus, the amount of absorbed protein (3.7 mg) was surprisingly high. In our next examination, the absorbed lysozyme was released from the carboxyl sponge at various ionic strengths. At 0.5 or 1.0 M NaCl, lysozyme was rapidly released, but at a relatively low concentration (0 to 0.15 M), only <15% of lysozyme was released. At 0.15 M, near the physiological ionic strength, an initial burst (to 30 min, 11% was released) was observed, and then no significant release was observed (15% at 48 h). The initial burst may depend on the release of lysozyme from the surface of the sponge, but not from the inner part of the sponge. We considered that the basic bioactive protein in the carboxyl sponge is maintained for several days under physiological conditions. Osteoblast differentiation on BMP-2-loaded sponges The interaction of BMP-2 with the carboxyl sponge was examined under physiological conditions, that is, in cultivation medium containing other serum proteins and acidic and basic small molecules, such as nucleosides and nucleotides. The differentiation induction by BMP-2 bound to the carboxyl sponge was also examined. BMP-2 in cultivation medium was absorbed by the sponges, and then the preosteoblast MC3T3-E1 was seeded onto the sponges. When the sponges absorbed and maintained BMP-2 inside in an active form, osteoblast differentiation was observed. Osteoblast differentiation was analyzed by measuring ALP activity, which is an early differentiation marker. As shown in Fig. 3, the cells expressed ALP activity on the BMP-2-loaded carboxyl sponge, but not on the amino sponge. The carboxyl sponge absorbed and maintained BMP-2 in cultivation medium including 10% FBS. After BMP-2 absorption, in the remaining cultivation medium,
FIG. 2. Lysozyme release from carboxyl sponge with 10 mM Tris (pH 8.0) containing NaCl at various concentrations. Lysozyme (3 mg) was absorbed by the carboxyl sponge, and then the sponge was transferred into the Tris buffer containing NaCl at 1.0 M (open squares), 0.5 M (closed squares), 0.15 M (open circles), 0.1 M (closed triangles), and 0 M (open triangles). At 0.15 M NaCl, near the physiological ionic strength, an initial burst release was observed (to 6 h), but not a significant change (6 to 48 h).
osteoblasts did not differentiate (data not shown). On the other hand, the amino sponge did not possess the abovementioned ability. These results suggest that the negative charges on the sponges are important and that the ionic interaction between the sponges and BMP-2 is sufficient in retaining BMP-2 in the sponges.
428
TACHIBANA ET AL.
J. BIOSCI. BIOENG.,
FIG. 3. Osteoblast differentiation on BMP-2-loaded sponges. The alkaline phosphatase (ALP) activities of osteoblast culture on the carboxyl sponge (open circles) and amino sponge (closed circles) were measured as early markers of osteoblast differentiation.
Outside culture of BMP-2 loaded sponge did not significantly differentiate BMP-2 was loaded to the carboxyl and amino sponges, and then ALP activities were measured for the inside and outside cultures of osteoblasts (Fig. 4). As shown in Fig. 4B, the carboxyl sponge maintained the loaded BMP-2 and differentiated osteoblasts, but the amino sponge did not. The culture outside the amino sponge expressed an ALP activity level higher than that of the negative control culture and slightly lower than that of the positive control, and significantly differentiated. The ALP acitivity level of the culture outside the carboxyl sponge was the same as that of the negative control culture, and the effects of BMP-2 were not observed. No significant amount of BMP-2 might be released from the carboxyl sponge. In conclusion, the basic bioactive protein BMP-2 was relatively strongly bound to the carboxyl sponge without the loss of the bioactivity. The carboxyl groups are acids weaker than the phosphate and sulfonate found on ion-exchange resins as tight protein binder. Other protein biomaterials, such as collagen, have also carboxyl groups as well as the carboxyl sponge and the relatively loose binding and (controlled) release of basic proteins were reported (2). Although the carboxyl sponge contained the same carboxyl groups, why were BMP-2 and lysozyme released in small amount from the carboxyl sponge? This may be responsible for the large amount of carboxyl groups on the carboxyl sponge with chemically modified carboxymethyl cysteine residues (about 20 mol% [10]) and native acidic amino acid residues (about 20 mol% [9]), such as aspartate and glutamate. The N-terminal amino acid sequence, QAKHKQRKRLK, of human BMP-2 are lysine and arginine rich and very basic. These basic residues of the N-terminus of human BMP-2 may produce multiple ionic interactions with crowded acidic residues (about 40 mol%) on the carboxyl sponge. The multiple ionic interactions may be stronger than the interactions between BMP-2 and other scaffolds, such as collagen. The multiple ionic interactions between BMP-2 and the carboxyl sponge may be prevent the release of BMP-2 from the sponge. The effects of BMP-2 are confined inside the modi-
FIG. 4. Leakage test of BMP-2 from BMP-2-loaded carboxyl sponge. (A) Schematic illustration of cultures inside and outside BMP-2 loaded sponge. (B) The cultures inside BMP-2-loaded amino and carboxyl sponges were analyzed in term of ALP acitivities. (C) The outside cultures of amino and carboxyl sponges were analyzed with ALP activities. The addition of equal amounts of BMP-2 loaded and nonaddition are shown as the positive and negative control cultures, respectively.
fied keratin sponge, when the loaded sponge is used in vivo. Therefore, we expect that only internal osteogenesis will be induced and that no external heterotopis ossification outside will be induced. Our previous work demonstrated that the carboxyl sponge with a large amount of carboxyl groups is a mimic of osteoblastic extracellular matrix protein, such as matrix γ-carboxyglutamic acid protein, which is responsible for osteoblast calcification (11). In addition, BMP-2/4 accumulation in extracellular matrices was reported to be essential for the osteoblastic differentiation of cells in the osteoblast lineage (16). We consider that the carboxyl sponge is also a useful mimic of extracellular matrix proteins for the accumulation and binding of BMPs for osteoblast differentiation. For the clinical applications of the biomaterials, further work is re-
VOL. 102, 2006
MODIFIED KERATIN SPONGE: BMP-2 BINDING
quired to clarify bone metabolism and the detailed differentiation of osteoblasts in the carboxyl sponge tightly binding BMP-2. ACKNOWLEDGMENTS This study was partly supported by a Science and Culture grant from the Japanese Ministry of Education, Culture, Sports, Science and Technology.
REFERENCES 1. Wozney, J. M.: The bone morphogenetic protein family: multifuctional cellular regulators in the embryo and adult. Eur. J. Oral Sci., 106, 160–166 (1998). 2. Li, R. H. and Wozney, J. M.: Delivering on the promise of bone morphogenetic proteins. Trends Biotechnol., 19, 255– 265 (2001). 3. Yamamoto, M., Takahashi, Y., and Tabata, Y.: Controlled release by biodegradable hydrogels enhances the ectopid bone formation of bone morphogenetic protein. Biomaterials, 24, 4375–4383 (2003). 4. Wildemann, B., Kandziora, F., Krummrey, G., Palasdies, N., Haas, N. P., Raschke, M., and Schrnidmaier, G.: Local and controlled release of growth factors (combination of IGF-1 and TGF-beta1, and BMP-2 alone) from a polylactide coating of titanium implants does not lead to ectopic bone formation in sheep muscle. J. Control. Release, 95, 249–256 (2004). 5. Lutolf, M. P., Weber, F. E., Schmoekel, H. G., Schense, J. C., Kohler, T., Müller, R., and Hubbell, J. A.: Repair of bone defects using synthetic mimetics of collagenous extracellular matrices. Nat. Biotechnol., 21, 513–518 (2003). 6. Friess, W., Uludag, H., Foskett, S., Biron, R., and Sargeant, C.: Characterization of absorbable collagen sponges as rhBMP-2 carriers. Int. J. Pharm., 187, 91–99 (1999).
429
7. Geiger, M., Li, R. H., and Friess, W.: Collagen sponges for bone regeneration with rhBMP-2. Adv. Drug Deliv. Rev., 55, 1613–1629 (2003). 8. Doillon, C. J., Drouin, R., Cote, M. F., Dallaire, N., Pageau, J. F., and Laroche, G.: Chemical inactivatores as sterilization agents for bovine collagen materials. J. Biomed. Mater. Res., 37, 212–221 (1997). 9. Tanabe, T., Tachibana, A., and Yamauchi, K.: Keratins: prospective proteiunous biomaterial. Recent Res. Dev. Protein Eng., 1, 247–259 (2001). 10. Tachibana, A., Furuta, Y., Takeshima, H., Tanabe, T., and Yamauchi, K.: Fabrication of wool keratin sponge scaffolds for long-term cell cultivation. J. Biotechnol., 93, 165–170 (2002). 11. Tachibana, A., Kaneko, S., Tanabe, T., and Yamauchi, K.: Rapid fabrication of keratin-hydroxyapatite hybrid sponges toward osteoblast cultivation and differentiation. Biomaterials, 26, 297–302 (2005). 12. Kurimoto, A., Tanabe, T., Tachibana, A., and Yamauchi, K.: Thiolated dermal bovine collagen as a novel support for bioactive substances — conjugation with lysozyme. J. Biotechnol., 86, 1–8 (2001). 13. Kurimoto, A., Tanabe, T., Tachibana, A., and Yamauchi, K.: Keratin sponge: immobilization of lysozyme. J. Biosci. Bioeng., 96, 307–309 (2003). 14. Peterson, G. L.: A simplification of the protein assay method of Lowry et al. which is more generally applicable. Anal. Biochem., 83, 346–356 (1977). 15. Kirsch, T., Nickel, J., and Sebald, W.: BMP-2 antagonists emerge from alterations in the low-affinity binding epitope for receptor BMPR-2. EMBO J., 19, 3314–3324 (2000). 16. Suzawa, M., Takeuchi, Y., Fukumoto, S., Kato, S., Ueno, N., Miyazono, K., Matsumoto, T., and Fujita, T.: Extracellular matrix-associated bone morphogenetic proteins are essential for differentiation of murine osteoblastic cells in vitro. Endocrinology, 140, 2125–2133 (1999).