Comparison of recombinant and purified human bone morphogenetic protein

British Journal of Oral and Maxillofacial Surgery (1999) 37, 2–5 © 1999 The British Association of Oral and Maxillofacial Surgeons

BRITISH JOURNAL OF ORAL

& M A X I L L O FA C I A L S U R G E RY

Comparison of recombinant and purified human bone morphogenetic protein K. Bessho,* K. Kusumoto,‡ K. Fujimura,* Y. Konishi,* Y. Ogawa,‡ Y. Tani,† T. Iizuka* *Department of Oral and Maxillofacial Surgery, Faculty of Medicine; †Research Center for Biomedical Engineering, Kyoto University, Kyoto; ‡Department of Plastic and Reconstructive Surgery, Kansai Medical University, Osaka, Japan SUMMARY. Clinically, it would be more convenient to use recombinant than purified preparations of bone morphogenetic protein (BMP). Recently, recombinant human BMP (rhBMP) has attracted the attention of many investigators, but it has not been fully characterized. We examined the bone-inducing activity of rhBMP-2 and compared it with that of purified BMP derived from human bone matrix (phBMP). Two, 10, or 50 µg of rhBMP-2 or phBMP was mixed with 3 mg of atelopeptide type I collagen (carrier), and specimens were implanted in the calf muscles of Wistar rats (n=5 in each group). Four weeks later, new bone had formed in all the rhBMP-2- and phBMP-implanted muscles and was visible radiographically and histologically. The quantitative analysis indicated that the activity of rhBMP-2 was less than one tenth that of phBMP. It is necessary to find out why rhBMP-2 has fewer activities than phBMP.

of the BMP obtained was 17.0 kDa as measured by sodium dodecyl sulphate-polyacrylamide slab gel electrophoresis (SDS-PAGE) and its isoelectric point (pI) value was 4.9.8.9 This phBMP appeared as a single band. The rate of the BMP against human bone matrix (starting material) was 0.0003 % in weight.

INTRODUCTION Purification of bone morphogenetic protein (BMP) is difficult because it is present in much smaller amounts than other non-collagenous proteins in bone matrix and is relatively insoluble. During the 1980s, BMP was found to be soluble in 4 M guanidine hydrochloride1,2 and 6 M urea,3 and its isolation and purification have progressed rapidly. Several BMPs have become available by cloning and genetic engineering.4 However, there are still some problems with the bone-inducing activity of recombinant human BMP (rhBMP). In this study, we examined the bone-inducing activity of rhBMP-2 compared with that of our original purified bone morphogenetic protein that is derived from the human bone matrix (phBMP) using our conventional bioassays, which include the optimal concentration.5

Bioassays There were two experimental groups each being tested with three different amounts of BMP. Two, 10, or 50 µg of rhBMP-2 or phBMP were mixed with 3 mg of porcine skin-derived atelopeptide type I collagen solution (3 mg/ml, pH 3.0, Nitta Gelatin Inc., Osaka, Japan) as a carrier. Each mixed sample was then lyophilized into a disc (4 mm in diameter, 1.5 mm thick). These specimens were implanted into the calf muscles of 10-week-old male Wistar rats (weight 230–260 g, n = 5 in each group) that had been anaesthetized by an intraperitoneal injection of sodium pentobarbital. The fascia and skin were then closed. Four weeks later, soft radiographs (SRO-M50; Sofron Inc., Tokyo, Japan; 35 kV, 4 mA, 2 min) were taken of the implanted regions. The implants were then excised and the surrounding tissues removed, weighed, and homogenized in 0.25 M sucrose in a Polytron homogenizer (Bio-Mixer type ABM-1; Nissei Inc., Osaka, Japan). Alkaline phosphate (ALP) activity and protein in the resultant supernatant were measured by the p-nitrophenyl phosphate method and pyrogallol red method, respectively. The sediments were demineralized in 10 ml of 0.5 M hydrochloric acid, and the calcium contents of the soluble fractions were measured by the o-cresolphthalein complex method. The ALP activity (IU/mg protein) and the calcium content (µg/mg tissue) were used as indices of bone formation. Histological

MATERIAL AND METHODS rhBMP-2 and phBMP We obtained rhBMP-2 (Genetics Institute Inc., Massachusetts, USA) from the Yamanouchi Pharmaceutical Co. Ltd (Tokyo, Japan). It was dissolved in a buffer (pH 6.5) containing 0.5 M L-arginine and 10 mmol L-histidine, and stored at –80°C. phBMP was obtained using Bessho’s BMP purification method;6 we used normal human bone extracted from patients during operations for fractures of the femoral neck and similar procedures. We did not use bone extracted from patients with tumours or infectious diseases. phBMP was extracted chemically from human bone matrix with 4 M guanidine hydrochloride (pH 5.2) and purified by liquid chromatography as described previously.7 The molecular weight 2

Comparison of recombinant and purified human bone morphogenetic protein

3

characteristics of the excised tissues were evaluated after being fixed in 10% formalin neutral buffer solution (pH 7.4), demineralized in EDTA, and embedded in paraffin. Sections 4 µm thick were cut on a microtome, stained with haematoxylin and eosin, and examined under a light microscope. The control, type I collagen alone, was implanted and evaluated in a similar manner.

RESULTS Soft radiographic findings Radiographs, radio-opaque shadows were visible in the radiographs of all of the rhBMP-2- and phBMP-implanted muscles. The opaque images in each of the rhBMP-2- and phBMP-implanted muscles were morphologically identical throughout the implanted material (Fig. 1). These radio-opaque images were not visible in the controls. Histological findings Light microscopy showed new bone surrounded by immature mesenchymal-type cells in all the rhBMP-2and phBMP-implanted muscles. However, there was no evidence of similar osteoinduction anywhere in the controls, and the maturity of rhBMP-2- and phBMPinduced bones differed. Implantation of 2 µg of rhBMP-2 resulted in new bone formation in only part of the area on the outermost edge of the intramuscular implant (Fig. 2A). Implantation of 10 µg of rhBMP-2 resulted in new bone formation on the outermost edge of the implant around almost the entire circumference, and immature mesenchymal type cells surrounded the new bone (Fig. 2B). Implantation of 50 µg of rhBMP-2 resulted in new bone formation not only on the outermost edge but also deep within the implant. Immature mesenchymal-type cells were present both inside and outside the new bone, and there was also bone marrow containing angioid tissue and fatty marrow (Fig. 2C). Implantation of 2, 10, and 50 µg of phBMP resulted in new bone formation not only on the outermost edge but also deep within the implant. The bone induced by phBMP had little bone marrow or rich bone matrix, and there were quantitative differences between these induced bones depending on the amount of phBMP implanted. There were immature mesenchymal-type cells surrounding the new bone (Fig. 2D). Quantitative analysis The volume of rhBMP-2- and phBMP-induced bone tended to depend on the amount of BMP implanted (Fig. 3). The mean ALP activities in the implanted regions were under 0.1 (control), 0.7 (2 µg rhBMP-2), 4.9 (10 µg rhBMP-2), 16.6 (50 µg rhBMP-2), 9.6 (2 µg phBMP), 51.0 (10 µg phBMP), and 170.7 (50 µg phBMP) IU/mg protein. The mean calcium contents

Fig. 1 – Soft radiographs taken 4 weeks after implantation. Radioopaque shadows are the lyophilised specimens of phBMP (A) obtained by Bessho’s BMP purification method, and rhBMP-2 (B) obtained from Genetics Institute Inc. with atelopeptide type I collagen as a carrier implanted into the right calf muscles of male Wistar rats.

of the implanted regions were under 0.1 (control), 21.3 (2 µg rhBMP-2), 26.1 (10 µg rhBMP-2), 36.8 (50 µg rhBMP-2), 35.4 (2 µg phBMP), 84.2 (10 µg phBMP), and 309.4 (50 µg phBMP) µg/mg tissue. These values indicate that the activity of rhBMP-2 was below one tenth of that of phBMP.

DISCUSSION It would be more convenient to use recombinant rather than purified preparations of BMP clinically from the points of view of both quality and safety, provided that recombinant BMP could easily be synthesized and obtained. BMP-1 to BMP-910 have been cloned by genetic engineering techniques (the amino acid sequences of our original purified BMP are not the same as the reported amino acid sequences of BMP-1 to BMP-9),9 and rhBMP-2 has been studied in particular detail. However, there are still problems that must be solved; the most important of these is that the boneinducing activity of recombinant BMP is lower than that of purified BMP.11 Quantitative analysis of the results of bioassays showed that the bone-inducing activity of rhBMP-2 was less than one-tenth of that of phBMP. Because this is consistent with the previous

4

British Journal of Oral and Maxillofacial Surgery

Fig. 2 – Photomicrographs of bones induced by 2 µg (A), 10 µg (B), 50 µg (C), of rhBMP-2, or phBMP (D), in the calf muscle pouch of Wistar rats four weeks after implantation (haematoxylin-eosin; original magnification × 40). M= calf muscle of host, NB= newly-formed bone, BM= bone marrow, and C= collagen.

study,11 BMP heterodimers are 5–10 times more potent than BMP-2 in inducing cartilage and bone in vivo,12 and transforming growth factor (TGF)-β (BMP is a subgroup of the TGF-β gene superfamily) heterodimer exists naturally in a living body,13 the possibility exists that phBMP consists of heterodimer. To find out the reasons for this large difference between recombinant and purified preparations, it is necessary to examine the effects of other cytokines (including other rhBMPs), or carriers of the bone-inducing activity of rhBMP-2 and other rhBMPs, or both, and compare them with those of purified BMPs. It is important to resolve this problem, because a great deal of bone is required for the clinical application of small quantities of rhBMP. In addition, because rhBMPs are soluble in vivo and disperse shortly after implantation, they do not induce bone formation in large quantities without a carrier. In addition to rhBMPs, therefore, a carrier capable of acting as a slow-delivery system is required to allow sufficient bone induction in vivo. Many research workers have caused bone induction by implanting rhBMPs with the residue of the extract from demineralized bone using guanidine hydrochloride as a carrier.14–16 However, as we cannot exclude the possibility that this extract residue contained a factor related to bone formation, we need to evaluate the bone-inducing activities using a purer carrier. Recently, new bone formation was induced by

implanting rhBMP-2 with collagen as a carrier.5.17 We showed that atelopeptide type I collagen was also useful as a delivery system for rhBMP-2, but the main type of collagen in mineralized cartilage at cartilaginous ossification is type II collagen.18 As type I collagen was adopted as a carrier, it may have been difficult to induce cartilage formation. Under conditions of high oxygen density, it is easier to induce more bone than cartilage.19 We saw no cartilage formation in the present study. Histological findings showed that the bone tissue induced by 50 µg of rhBMP-2 consisted of rich bone marrow containing angioid tissue and comparatively poor bone matrix, while that induced by 50 µg of phBMP had little bone marrow and no angioid tissue and had a rich bone matrix. Atelopeptide type I collagen therefore seems to be the carrier of choice for phBMP but not for rhBMP-2. It will be necessary to find a carrier for rhBMP-2 that is capable of facilitating better bone formation than atelopeptide type I collagen. However, whole bone tissue induced by rhBMP-2 received a continuous blood supply and has the potential to become self-supporting bone, in which bone tissue is remodelled continuously and the form is maintained over the long term. BMPs have important physiological activities other than bone induction.20.21 However, being a multifunctional factor may prove to be a clinical disadvantage and therefore caution is required.

Comparison of recombinant and purified human bone morphogenetic protein A

B

Fig. 3 – (A) Alkaline phosphatase activity (IU/mg protein) and (B) calcium content (µg/mg tissue) in each bioassay (n = 5 in each group). Data are expressed as mean (SEM); the solid bars indicate rhBMP-2 and the hatched bars indicate phBMP.

References 1. Takaoka K, Ono K, Amitani K, Kishimoto R, Nakata Y. Solubilization and concentration of a bone-inducing substance from a murine osteosarcoma. Clin Orthop 1980; 148: 274–280. 2. Hanamura H, Higuchi Y, Nakagawa M, Iwata H, Nogami H, Urist MR. Solubilized bone morphogenetic protein (BMP) from mouse osteosarcoma and rat demineralized bone matrix. Clin Orthop 1980; 148: 281–290. 3. Urist MR, Leitze A, Mizutani H et al. A bovine low molecular weight bone morphogenetic protein (BMP) fraction. Clin Orthop 1982; 162: 219–232. 4. Wozney JM, Rosen V, Byrne M, Celeste AJ, Moutsatsos I, Wang EA. Growth factors influencing bone development. J Cell Sci Suppl 1990; 13: 149–156. 5. Fujimura K, Bessho K, Kusumoto K, Ogawa Y, Iizuka T. Experimental studies on bone inducing activity of composites of atelopeptide type I collagen as a carrier for ectopic osteoinduction by rhBMP-2. Biochem Biophys Res Commun 1995; 208: 316–322. 6. Bessho K, Iizuka T. Changes in bone inducing activity of bone morphogenetic protein with aging. Ann Chir Gynecol 1993; 82: 49–54. 7. Bessho K, Tagawa T, Murata M. Comparison of bone matrixderived bone morphogenetic proteins from various animals. J Oral Maxillofac Surg 1992; 50: 496–501. 8. Bessho K. Purification and characterization of bone morphogenetic protein. Mie Medical Journal 1990; 40: 61–71. 9. Bessho K, Tagawa T, Murata M. Analysis of bone morphogenetic protein (BMP) derived from human and bovine bone matrix. Clin Orthop 1991; 268: 226–234.

5

10. Celeste AJ, Song JJ, Cox K, Rosen V, Wozney JM. Bone morphogenetic protein-9, a new member of the TGF-β superfamily. J Bone Miner Res 1994; 9 (suppl 1): S136. 11. Wozney JM. Bone morphogenetic proteins. Prog Growth Factor Res 1989; 1: 267–280. 12. Israel DI, Nove J, Kerns KM et al. Heterodimeric bone morphogenetic proteins show enhanced activity in vitro and in vivo. Growth Factors 1996; 13: 291–300. 13. Cheifetz S, Bassols A, Stanley K, Ohta M, Greenberger J, Massague J. Heterodimeric transforming growth factor β. J Biol Chem 1988; 263: 10783–10789. 14. Wang EA, Rosen V, D’Alessandro JS et al. Recombinant human bone morphogenetic protein induces bone formation. Proc Natl Acad Sci USA 1990; 87: 2220–2224. 15. Luyten FP, Cunningham NS, Vukicevic S, Paralkar V, Ripamonti U, Reddi AH. Advances in osteogenin and related bone morphogenetic proteins in bone induction and repair. Acta Orthop Belgica 1992; 58 (suppl 1): 263–267. 16. Gerhart TN, Kirker-Head CA, Kriz MJ et al. Healing segmental femoral defects in sheep using recombinant human bone morphogenetic protein. Clin Orthop 1993; 293: 317–326. 17. Boden SD, Schimandle JH, Hutton WC. Experimental spine fusion with recombinant human bone morphogenetic protein (rhBMP-2). J Bone Miner Res 1994; 9 (suppl 1): S225. 18. Boskey AL. Mineral-matrix interactions in bone and cartilage. Clin Orthop 1992; 281: 244–274. 19. Bassett CA, Becker RO. Generation of electric potentials by bone in response to mechanical stress. Science 1962; 137: 1063–1064. 20. Vainio S, Karavanova I, Jowett A, Thesleff I. Identification of BMP-4 as a signal mediating secondary induction between epithelial and mesenchymal tissues during early tooth development. Cell 1993; 75: 45–58. 21. Suzuki A, Nishimatsu S, Murakami K, Ueno N. Differential expression of xenopus BMPs in early embryos and tissues. Zool Sci 1993; 10: 175–178.

The Authors Kazuhisa Bessho DDS, DMSc Assistant Professor Kazuma Fujimura DDS, DMSc Assistant Professor Yasuzo Konishi DDS Research Fellow Tadahiko Iizuka DDS, DMSc Professor and Chairman Department of Oral and Maxillofacial Surgery Faculty of Medicine Kyoto University 54 Kawahara-cho, Shogoin Sakyo-ku, Kyoto 606 Japan Yoshiaki Tani DDS, DMSc Professor and Head Research Center for Biomedical Engineering Kyoto University 53 Kawahara-cho, Shogoin Sakyo-ku, Kyoto 606 Japan Kenji Kusumoto MD, DMSc Associate Professor Yutaka Ogawa MD, DMSc Professor and Director Department of Plastic and Reconstructive Surgery Kansai Medical University 10–15 Fumizono-cho Moriguchi, Osaka 570 Japan Correspondence and requests for offprints to: Kazuhisa Bessho Paper received 21 February 1997 Accepted 11 April 1998