effect of surface structure on protein adsorption to biphasic calcium-phosphate ceramics in vitro...

Available online at www.sciencedirect.com

Acta Biomaterialia 5 (2009) 1311–1318

www.elsevier.com/locate/actabiomat

Effect of surface structure on protein adsorption to biphasiccalcium-phosphate ceramics in vitro and in vivo

X.D. Zhu a,*, H.S. Fan a, Y.M. Xiao a, D.X. Li b, H.J. Zhang a, T. Luxbacher c, X.D. Zhang a

a National Engineering Research Center for Biomaterials, Sichuan University, Chengdu 610064, Chinab Institute of Pharmacology & Toxicology, Sichuan Academy of Chinese Medicine Science, Chengdu, China

c Colloid Science Department, Anton Paar� GmbH, Graz, Austria

Received 25 April 2008; received in revised form 9 October 2008; accepted 25 November 2008Available online 10 December 2008

Abstract

Protein adsorption affects the function of cells and determines the bioactivity of biomaterial implants. Surface structure and propertiesof materials determine the behavior of protein adsorption. In the present study, two biphasic calcium-phosphate ceramics (BCPs) withdifferent surface structures were fabricated by pressing and H2O2 foaming methods. Their surface characteristics were analyzed and thein vitro and in vivo protein adsorption on them was investigated. Porous BCP showed higher ability to adsorb proteins, and transform-ing growth factor-b1 (TGF-b1) adsorption notably increased with increasing in vivo implantation time. The strong affinity of BCP toTGF-b1 might provide important information for exploring the mechanism of the osteoinduction of calcium phosphates.� 2008 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: BCPs; Surface structure; Protein adsorption; TGF-b1

1. Introduction

The bioactivity of biomaterials is one of the factors thatdetermine the successes of the implants. It is known thatthe initial event is the adsorption of various proteins fromblood or other body fluids to the biomaterial after implan-tation [1–3], and the adsorbed proteins would affect thebehavior of cells and then determine the bioactivity ofthe implant [4–6]. However, protein adsorption is a com-plex process and is dependent on the surface propertiesof a biomaterial, such as the chemical composition, struc-ture, surface charge and so on [7]. Therefore, the fullknowledge for the interaction between proteins and thebiomaterial surface is helpful for us to understand andreveal the biological nature of the implant.

Calcium phosphates have a similar chemical composi-tion to the inorganic phase of human bone tissue and havebeen extensively studied for a long time in the biomaterials

1742-7061/$ - see front matter � 2008 Acta Materialia Inc. Published by Else

doi:10.1016/j.actbio.2008.11.024

* Corresponding author. Tel./fax: +86 28 85410703.E-mail address: [email protected] (X.D. Zhu).

field. Since the osteoinduction of calcium phosphates wasreported in 1991 [8,9], biomaterials scientists have beenexploring its mechanism, which is not yet completely clear.It had been confirmed that the appropriate porous struc-ture is necessary for the osteoinduction of calcium-phos-phate ceramic [10–12] and the dense one could not inducebone formation [13,14]. Moreover, BCP consisting ofhydroxyapatite (HA) and b-tricalcium phosphate (b-TCP)has better osteoinduction than single phasic HA or b-TCP [15,16]. It is known that various bone growth factors,especially the super family of transform growth factorsincluding bone morphogenetic protein-2 (BMP-2) andTGF-b1, would participate in the process of bone forma-tion and remodeling [17]. Yuan et al. reported that cal-cium-phosphate ceramic loaded with BMP-2 showedenhanced osteoinduction [18,19]. Zhang et al. believed thatthe ability to concentrate various bone growth factorsunder physiological conditions should be necessary forthe osteoinduction of calcium phosphates [16].

We previously reported that various calcium phosphateshad similar protein adsorption behavior, and a basic

vier Ltd. All rights reserved.

mailto:[email protected]

1312 X.D. Zhu et al. / Acta Biomaterialia 5 (2009) 1311–1318

lysozyme with lower molecular weight showed higheradsorption affinity for calcium phosphates than the acidicbovine serum albumin [20]. In fact, biomaterials are usuallyexposed to human plasma or other body fluids that containhundreds and thousands of proteins, and a dynamic pro-tein exchange will be observed when these proteins simulta-neously adhere to the surface after implantation of abiomaterial, so the preferential adsorption of some bone-related proteins, especially bone growth factors, on calciumphosphates might be related to its osteoinduction. So far,there are many documents dealing with behaviors ofin vitro protein adsorption on biomaterials [21–27]. How-ever, in vivo studies have rarely been done. As a valuablein vivo research tool for the study of immunology, tissuegrowth and organ transplants, diffusion chambers havebeen widely used for biomaterials research [28–30]. Diffu-sion chambers can inhibit cells and permit various ions,organic small molecules and proteins to enter into it, sothe interaction of biomaterials with proteins in body fluids,not the influences of cells, can be investigated. In the pres-ent study, we make two BCP ceramics with different sur-face structures and investigate the adsorption behavior ofserum proteins, especially TGF-b1 on them in vitro andin vivo. The effect of surface structure on protein adsorp-tion is discussed in detail.

2. Materials and methods

2.1. BCP ceramics

BCP precursor powders (HA/TCP: 70/30) prepared by awet precipitation method were supplied by National Engi-neering Research Center for Biomaterials of China. TwoBCP ceramics with different structures were prepared,respectively, by pressing and H2O2 foaming methods. Therelatively dense ones (DBCP) were fabricated by placingthe precursor powders into a special mold mounted in a sin-gle-axial press and pressing it under 300 MPa for 5 min. Theporous ones (PBCP) were made by foaming BCP precursorpowders with 5% solution of H2O2. After being sintered at1100 �C, the starting DBCP disks were polished with 1200-grit sandpaper and then 1 lm of diamond powders, andthe PBCP cylinders were cut into disks and then polishedwith 1200-grit sandpaper. The obtained DBCP and PBCPdisks were cleaned by a sonic cleaner with ethanol and dd-H2O, and then dried at 80 �C in an oven. Prior to use, all sam-ples were autoclaved at 120 �C for 30 min.

2.2. Rat serum

The blood samples, which were collected from healthyadult SD rats (200 ± 20 g, supplied by Experimental Ani-mal Center, Sichuan Institute of Chinese Materia Medica)clotted for 2 h at room temperature before centrifuging for20 min at approximately 2000 rpm. The serum wasremoved and stored at 6�20 �C before being used foradsorption experiments.

2.3. Materials characterization

The phasic compositions of the two BCP ceramics weremeasured with X-ray diffraction (XRD, Philips X’Pert ProMPD). The morphologies of them were observed by scan-ning electron microscopy (SEM, JSM-5900 OL). The spe-cific surface areas of them were measured by nitrogenadsorption according to the BET method.

Zeta potentials of the two BCP ceramics were performedusing SurPASS apparatus (Anton Paar, Austria). The stream-ing currents were measured with the SurPASS using theAdjustable Gap Cell on which the samples with20 mm � 10 mm � 1 mm size were mounted in the presenceof a 10�3 M solution of KCl, and the zeta potentials were eval-uated from streaming current measurements according to

f ¼ dIdp� g�� 0

� LW � H

ð1Þ

where f is the zeta potential; dIdp is the slope of streaming cur-

rent vs. different pressure; g is the electrolyte viscosity; e isthe permittivity; e0 is the dielectric coefficient of electrolyte;and L, W, H are the length, width and height of the stream-ing channel.

2.4. Protein adsorption experiments in vitro and in vivo

In vitro experiments were done by incubating the DBCPor PBCP disks in 5 ml of capacity polypropylene centrifugetubes with 2 ml of rat serum at 37 �C and 5% CO2 for 3 days.As for in vivo experiments, the samples were put inside thediffusion chambers (Millipore Co., USA), which are Plexi-glas� rings 14 mm in diameter with MF-Millipore Mem-brane filters having 0.45 lm of pore sizes glued to the topand bottom rims of the rings, and then implanted into thedorsal muscle of the adult SD rats under general anesthesiaand sterile conditions. After 3 and 7 days of implantation,the diffusion chambers were retrieved and the samples wereremoved. All samples were transferred into new tubes andwashed three times with dd-H2O to remove any loose-bind-ing proteins, then the proteins adhering to the sample surfacewere desorbed by 200 ll of 0.5 M solution of Na2HPO4 withsharp stirring for 15 min. The supernatants containing thedesorbed proteins from the materials’ surface were collectedand diluted until the Na2HPO4 concentration lowered to0.1 M for the next analysis. Changes to the surface morphol-ogy and compositions of the materials were examined usingXRD and SEM, respectively.

2.5. Polyacrylamide gel electrophoresis (PAGE) analysis

The eluted protein samples (50 ll out of 300 ll in total)were subjected to separated by SDS–PAGE, which wasperformed through 5–15% gradient gels in a PROTEAN�

II xi electrophoresis system (Bio-Rad, USA) according tothe method of Laemmli [31]. The gels were stained withCoomassie brilliant blue R-250 and scanned with Chemi-DocTM XRS Systems (Bio-Rad, USA). The protein bands

Fig. 1. XRD patterns of BCP under various conditions: (a) PBCP, (b)PBCP in vitro incubation in rat serum, (c) DBCP, (d) DBCP in vitroincubation in rat serum, (e) DBCP in vivo implantation.

X.D. Zhu et al. / Acta Biomaterialia 5 (2009) 1311–1318 1313

present in the gels were analyzed by Quantity One� 1DAnalysis Software (Bio-Rad, USA).

2.6. BCA assay

The amount of the total proteins adsorbed onto DBCPor PBCP was measured using BCATM protein assay kit(Pierce, USA). After incubation at 37 �C for 2 h, the absor-bances of the mixtures of 25 ll protein sample and 200 llworking reagent were measured using a Microplate Reader(Model 550, Bio-Rad) at the wavelength of 570 nm againsta blank of 0.1 M solution of Na2HPO4. The amounts of theadsorbed total proteins were calculated according to thecalibrated curve with BSA as standards, and all experi-ments were carried out in triplicate.

2.7. ELISA

The content of TGF-b1 in blood or body fluids that can-not be detected by SDS–PAGE analysis is very low, so theamount of TGF-b1 adsorbed onto DBCP or PBCP wastested by direct ELISA. The protein samples (50 ll out of300 ll in total) were analyzed by ELISA reagent kits(Quantikine� Mouse/Rat/Porcine TGF-b1 Immunoassay,R&D, USA) according to the specification. For the stan-dard curve, serial dilutions of the reconstituted TGF-b1standard were added to final concentrations of 0–2000 pg ml�1. The plate was read in a Microplate Reader(Model 550, Bio-Rad) at the wavelength of 450 nm. Sam-ples and standards were assayed in triplicate.

3. Results

3.1. Surface characteristics of BCP

Fig. 1 shows XRD patterns of DBCP and PBCP ceramicdisks. Obviously, DBCP and PBCP had the same XRDpatterns, meaning that different fabricating methods didnot affect the phasic composition of BCP. At the sametime, after in vitro incubation or in vivo implantation, noother new specific peaks were detected. That is to say, onlydepending on the XRD patterns, any alterations occurringon their surface could not be confirmed.

The typical SEM photographs of DBCP and PBCPdisks are, respectively, shown in Figs. 2 and 3. Accordingto Fig. 2a, a large number of pores with the size distribu-tion from 100 to 500 lm diameter occurred on PBCP. Fur-thermore, many micropores distributed on the wall of themacropores could be clearly observed from the 10,000�photo, which is shown in Fig. 2b, and most of these mac-ropores were interconnected. In contrast, as shown inFig. 3a and b, DBCP has a relatively dense structure andsmooth surface, except for some tiny pores. Figs. 2c and3c show the surface morphologies after PBCP and DBCPwere in vitro incubated in rat serum for 3 days, respec-tively. Comparing with the photos before incubation, anyobvious changes could not be observed. Similarly, based

on Fig. 3d and e, DBCP had the same surface morphologyas before, after its implantation for 3 days, no depositscould be found, and the surface only became rougher thanbefore when the implantation time reached 7 days.

The densities, specific surface areas and zeta potentials ofPBCP and DBCP are summarized in Table 1. PBCP hadlower relative density than DBCP, which was consistent withSEM observation. Moreover, due to the introduction of por-ous structure, PBCP had higher specific surface area thanDBCP. However, although both PBCP and DBCP had neg-ative surface zeta potentials, it should be noted that the abso-lute value of PBCP was higher than that of DBCP.

3.2. SDS–PAGE analysis

The results of SDS–PAGE separation for the proteinsadsorbed to PBCP and DBCP are shown in Fig. 4. Obvi-ously, either in vitro incubation or in vivo implantation,there are more protein bands detected on PBCP than DBCP,meaning that PBCP adsorbed more proteins than DBCP.Comparing with the profile of rat serum (Fig. 4a), most ofthe proteins present in rat serum could adhere to PBCP,but only albumin adhering to DBCP was detected (theamounts of other proteins adhering to DBCP might be toosmall to be detected by Coomassie brilliant blue R-250 stain-ing). On the other hand, according to intensities of thedetected protein bands shown in Fig. 4b, it could be foundthat PBCP might be more favorable for binding those pro-teins with lower molecular weights.

3.3. Total protein binding capacity of BCP in vitro and

in vivo

After in vitro incubation in rat serum for 3 days, thetotal protein binding capacities of BCP calculated from

Fig. 2. Typical SEM photographs of PBCP under various conditions: (a) PBCP 50�, (b) PBCP 5000�, (c) PBCP in vitro incubation in rat serum 5000�.


the experimental data of BCA quantitative analysis aresummarized in Table 2. It was clearly found that BCP withdifferent structures showed different abilities to bind serumproteins; PBCP adsorbed far more serum proteins thanDBCP.

After in vivo implantation for 3 and 7 days, results ofthe total protein binding capacity of DBCP are shown inTable 3. Comparing with the data in Table 2, it can be seenthat DBCP adsorbed more proteins from serum in vitrothan that from body fluids in vivo. However, with thein vivo implantation time increasing from 3 to 7 days, theamount of the adsorbed total proteins on DBCP had noobvious alteration.

3.4. TGF-b1 adsorption in vitro and in vivo

Results of ELISA analysis for the adsorption of TGF-b1to BCP in vitro and in vivo were also summarized in Tables2 and 3, respectively. Corresponding to total proteins anal-ysis above, PBCP adsorbed more TGF-b1 than DBCPafter incubation in rat serum. At the same time, DBCPadsorbed greater amounts of TGF-b1 from serumin vitro than that from body fluids in vivo. More interest-ingly, the amount of the adsorbed TGF-b1 on DBCP nota-bly increased with the in vivo implantation time increasingfrom 3 to 7 days.

4. Discussion

The results of protein adsorption experiments in vitro orin vivo on two kinds of BCP reported in the present workshow that porous structure plays an important role in pro-tein adsorption. BCP ceramics with different structureshave a different ability to bind proteins, and PBCP canbind more proteins from serum or body fluids than DBCP.First, the introduction of porous structure can increase thespecific surface area according to the above BET analysis.It is known that calcium phosphates have two differentbinding sites, called positive C-sites and negative P-sites,on the crystal surface [32,33], which supply a multiple sitebinding for proteins. Thus, PBCP has more binding sitesfor protein adsorption due to its higher specific surfacearea, leading to more proteins adhering to its surface. Onthe other hand, by comparison with the dense structure,the three-dimensional porous structure of BCP is morefavorable for protein adsorption. Suh et al. had reportedthat the micropores on the surface of silicon particlesenhanced the adsorption of BSA, and its contribution tothe amount of BSA adsorption was greater than that ofthe specific surface area [34]. It had been calculated thatthe theoretical maximum adsorption amounts of BSA areabout 4.0 [35] or 2.52 [36] mg m�2, assuming a close-packed adsorption monolayer of BSA molecules covered

Fig. 3. The typical SEM photographs of DBCP under various conditions: (a) DBCP 1000�, (b) DBCP 10,000�, (c) DBCP in vitro incubation in rat serum10,000�, (d) DBCP in vivo implantation for 3 days 10,000�, (e) DBCP in vivo implantation for 7 days 10,000�.

Table 1Some physical and chemical properties of PBCP and DBCP.

Materials Size (mm � mm) Relativedensity(%)

Specificsurface area(m2 g�1)

Zetapotentialsa

(mV)

PBCP /7�1.5 30.0 2.761 �42.9 ± 4.11DBCP /8.8�1.5 84.9 0.964 �8.16 ± 0.67

a Zeta potential ± SD of three samples in 10�3 M KCl solution.


on the substrate surface. As shown in Table 2, the amountof the adsorbed total proteins on PBCP was far beyondthat on DBCP, meaning that the behaviors of proteinadsorption on PBCP and DBCP are different: the former

is a multilayer adsorption and the latter is a monolayerone. It could be ascribed to the hold-back effect of porousstructure on the rearrangement of protein molecules. Mostproteins would be subjected to the structural rearrange-ment after adsorption on the substrate surface [7]. How-ever, this rearrangement would be limited by those pores,especially micropores. As a result, more proteins couldbind to the surface of PBCP.

The incubation environment is also an important factorthat affects protein adsorption. Comparing the data inTable 2 with that in Table 3, it should be found that DBCPcan bind more proteins (total proteins or TGF-b1) afterin vitro incubation in rat serum. This could be explainedby the situation of protein adsorption. It is known that

Fig. 4. (a) SDS–PAGE patterns of the adsorbed proteins on both BCP that were in vitro incubated in rat serum or in vivo implanted into the bodythrough diffusion chambers for 3 days. (b) Densitometry of different protein bands present in the polyacrylamide gel.

Table 2Analysis for the proteins adsorbed on DBCP and PBCP in vitro for 3 days(n = 3).

Materials

DBCP PBCP

Amount of adsorbed TGF-b1 (pg m�2) 29.4 ± 1.4 27,219.4 ± 113.2Amount of adsorbed total proteins

(lg m�2)132.9 ± 8.3 7144.9 ± 24.7

TGF-b1 out of 100 lg of total proteins(pg (100 lg)�1)

22.2 380.8

Table 3Analysis for the proteins adsorbed on DBCP in vivo for 3 and 7 days(n = 3).

Implantation time

3 days 7 days

Amount of adsorbed TGF-b1 (pg m�2) 11.6 ± 0.9 21.8 ± 1.1Amount of the adsorbed total proteins (lg m�2) 82.8±3.5 84.2 ± 4.2TGF-b1 out of 100 lg of total proteins

(pg (100 lg)�1)13.9 25.8


body fluids are continuously flowing in the body, andDBCP would be subjected to a shearing strength fromthe flowing body fluids when implanted in vivo. However,the rat serum is static and there are no forces on it afterDBCP in vitro incubation. The shearing strength wouldundoubtedly affect the adsorption of proteins on DBCP,resulting in the fact that the amount of adsorbed proteinsin vitro was greater than that in vivo.

It is known that blood or other body fluids is a multi-component protein solution, and when BCP contacts with

it, different kinds of proteins would competitively adhere toits surface according to ‘‘the Vroman effect” [37]. Theabove PAGE patterns confirmed that different proteinscan adhere to the surface of BCP. More importantly, theadsorption and accumulation of bone growth factors onBCP is vital to explore the mechanism of the osteoinduc-tion of BCP. During bone regeneration, TGF-b1 can pro-mote the proliferation and differentiation of bone-formingcells [17]. Though the content of various bone growth fac-tors is only about 1–100 ng ml�1 in blood, TGF-b1 adsorp-tion on BCP was confirmed in the present study by ELISAanalysis. As mentioned above, the porous structure of BCPincreased the amount of the adsorbed total proteins on thesurface. However, as shown in Table 2, DBCP adsorbed22.2 pg of TGF-b1 out of 100 lg of the adsorbed total pro-teins, while PBCP adsorbed 380.8 pg of TGF-b1. It meansthat increasing the adsorbed TGF-b1 resulting from theporous structure of BCP is far beyond that of the adsorbedtotal proteins. First, there are many micropores distributedon the wall of those macropores of PBCP according to theabove SEM observation, and these abundant microporescould be favorable for the adsorption and accumulationof proteins with lower molecular weight, for instanceTGF-b1 and other bone growth factors. On the otherhand, the electrostatic interaction is one of the major driv-ing forces for protein adsorption. It is known that matureTGF-b1 is a basic protein and its isoelectric point is about8.59 [38]. Under physiological conditions, it has positivesurface charge. At the same time, BCP had negative zetapotential and the absolute value of zeta potential of PBCPwas higher than that of DBCP according to the above zetapotential measurements. Thus, the higher electrostatic


attractive forces between PBCP and TGF-b1 would enhancethe adsorption of TGF-b1 on PBCP.

The results of in vivo experiments give us very importantinformation for exploring the osteoinduction mechanism ofBCP. As shown in Table 3, the amount of the adsorbedTGF-b1 notably increased but that of the adsorbed totalproteins almost did not change with the in vivo implanta-tion time increasing from 3 to 7 days, meaning that TGF-b1 could concentrate on the surface after implantation ofBCP. This might be ascribed to the nature of competitiveadsorption of different proteins on the surface of biomate-rials. When a biomaterial is implanted into the body andcontacts with body fluids, different proteins compete forthe binding sites on the surface, and there occurs homomo-lecular as well as heteromolecular exchange during theadsorption process. Under in vivo conditions, the flowingof body fluids would accelerate the exchange of theadsorbed proteins with the free ones in body fluids, andTGF-b1 might be continuously adhered to the materialsurface due to its strong affinity for BCP. This result is alsoconsistent with our previous study that lysozyme was moreeasily adhered to BCP than BSA when BCP was incubatedin binary BSA/lysozyme solution [27]. So it could beassumed that more and more bone growth factors wouldaccumulate on the surface and induce the proliferationand differentiation of osteoblasts and finally enhance newbone formation after implantation of BCP. The adsorptionof the other bone growth factors on calcium phosphates andtheir effect on cell function will be explored in future work.

5. Conclusion

The surface structure of BCP plays an important role inprotein adsorption. With either in vitro incubation in ratserum or in vivo implantation, PBCP adsorbed more pro-teins (total proteins or TGF-b1) than DBCP, meaning thatthe introduction of porous structure enhanced significantlythe adsorption of proteins, especially TGF-b1 from bloodor other body fluids. The diffusion chamber model pro-vided an effective tool to investigate the in vivo proteinadsorption on biomaterials. After implantation of DBCP,the amount of the adsorbed TGF-b1 increased but thatof the adsorbed total proteins was almost unchanged withincreasing implantation time, proving that TGF-b1 hadstrong affinity for BCP and can concentrate on its surfaceunder physiological conditions. The enhanced adsorptionof TGF-b1 on PBCP might give important informationfor exploring the osteoinduction mechanism of calciumphosphates.

Acknowledgements

This work has been financially supported by NationalBasic Research Program of China (Contract Grant No.G2005cb623901) and National Natural Foundation ofChina (Contract Grant No. 50802060).

References

[1] Andrade JD. Principles of protein adsorption. In: Andrade JD,editor. Surface and interfacial aspects of biomedical polymers. NewYork: Plenum Press; 1985. p. 1–80.

[2] Norde W. Adsorption of proteins at solid–liquid interfaces. Cells Mat1995;5(1):97–112.

[3] Puleo DA, Nanci A. Understanding and controlling the bone–implant interface. Biomaterials 1999;20(23–24):2311–21.

[4] Ducheyne P, Qiu Q. Bioactive ceramics: the effect of surface reactivityon bone formation and bone cell function. Biomaterials 1999;20(23–):2287–303.

[5] El-Ghannam A, Ducheyne P, Shapiro IM. Effect of serum proteins onosteoblast adhesion to surface-modified bioactive glass and hydroxy-apatite. J Orthop Res 1999;17(3):340–5.

[6] Villarreal DR, Sogal A, Ong JL. Protein adsorption and osteoblastresponses to different calcium phosphate surfaces. J Oral Implantol1998;24(2):67–73.

[7] Dee KC, Puleo DA, Bizios R. Protein–surface interactions. In: DeeKC, Puleo DA, Bizios R, editors. An introduction to tissue-biomaterial interactions. New York: John Wiley; 2002. p. 37–52.

[8] Zhang XD, Zou P, Wu C, Qu Y, Zhang JG. A study of porous blockHA ceramics and its osteogeneses. In: Ravaglioli A, Krajewski A,editors. Bioceramics and the human body. Amsterdam: Elsevier;1991. p. 408–15.

[9] Ripamonti U. The morphogenesis of bone in replicas of poroushydroxyapatite obtained from conversion of calcium carbonateexoskeletons of coral. J Bone Joint Surg Am 1991;73(5):692–703.

[10] Toth JM, Lynch KL, Hackbarth DA. Ceramic-induced osteogenesisfollowing subcutaneous implantation of calcium phosphate. Bioce-ramics, vol. 6, Zurich-Uetikon: Trans Tech Publications; 1993. p. 9–13.

[11] Klein C, Degroot K, Chen WQ, Li YB, Zhang XD. Osseoussubstance formation induced in porous calcium-phosphate ceramicsin soft-tissues. Biomaterials 1994;15(1):31–4.

[12] Yuan HP, Yang ZJ, Li YB, Zhang XD, De Bruijn JD, De Groot K.Osteoinduction by calcium phosphate biomaterials. J Mater SciMater Med 1998;9(12):723–6.

[13] Winter M, Griss P, de Groot K, Tagai H, Heimke G, von Dijk H,et al. Comparative histocompatibility testing of seven calciumphosphate ceramics. Biomaterials 1981;2(3):159–61.

[14] Misiek D, Kent J, Carr R. Soft tissue responses to hydroxyapatiteparticles of different shapes. J Oral Maxillofac Surg 1984;42(3):150–60.

[15] Yuan H, Van Den Doel M, Li S, Van Blitterswijk CA, De Groot K,De Bruijn JD. A comparison of the osteoinductive potential of twocalcium phosphate ceramics implanted intramuscularly in goats. JMater Sci Mater Med 2002;13(12):1271–5.

[16] Zhang X, Yuan H, De Groot K. Calcium phosphate biomaterialswith intrinsic osteoinductivity. The 6th World Biomaterials Congress,Hawaii, May 2000.

[17] Boontheekul T, Mooney DJ. Protein-based signaling systems in tissueengineering. Curr Opin Biotechnol 2003;14(5):559–65.

[18] Yuan H, De Bruij JD, Zhang X, Van Blitterswijk CA, de Groot K.Use of an osteoinductive biomaterial as a bone morphogeneticprotein carrier. J Mater Sci Mater Med 2001;12(9):761–6.

[19] Yuan H, Zou P, Yang Z, Zhang X, De Bruijn JD, De Groot K. Bonemorphogenetic protein and ceramic-induced osteogenesis. J Mater SciMater Med 1998;9(12):717–21.

[20] Zhu XD, Fan HS, Zhao CY, Lu J, Ikoma T, Tanaka J, et al.Competitive adsorption of bovine serum albumin and lysozyme oncharacterized calcium phosphates by polyacrylamide gel electropho-resis method. J Mater Sci Mater Med 2007;18(11):2243–9.

[21] Kandori K, Masunari A, Ishikawa T. Study on adsorption mecha-nism of proteins onto synthetic calcium hydroxyapatites throughionic concentration measurements. Calcif Tissue Int 2005;76(3):194–206.


[22] Kandori K, Miyagawa K, Ishikawa T. Adsorption of immunogammaglobulin onto various synthetic calcium hydroxyapatite particles. JColloid Interf Sci 2004;273(2):406–13.

[23] Krajewski A, Piancastelli A, Malavolti R. Albumin adhesion onceramics and correlation with their Z-potential. Biomaterials1998;19(7–9):637–41.

[24] Laffargue P, Fialdes P, Frayssinet P, Rtaimate M, Hildebrand HF,Marchandise X. Adsorption and release of insulin-like growth factor-I on porous tricalcium phosphate implant. J Biomed Mater Res2000;49(3):415–21.

[25] Rosengren A, Pavlovic E, Oscarsson S, Krajewski A, Ravaglioli A,Piancastelli A. Plasma protein adsorption pattern on characterizedceramic biomaterials. Biomaterials 2002;23(4):1237–47.

[26] Sharpe JR, Sammons RL, Marquis PM. Effect of pH on proteinadsorption to hydroxyapatite and tricalcium phosphate ceramics.Biomaterials 1997;18(6):471–6.

[27] Zhu XD, Fan HS, Li DX, Xiao YM, Zhang XD. Protein adsorptionand zeta potentials of a biphasic calcium phosphate ceramic undervarious conditions. J Biomed Mater Res Part B Appl Biomater2007;82B(1):65–73.

[28] Ikeuchi M, Dohi Y, Horiuchi K, Ohgushi H, Noshi T, Yoshikawa T,et al. Recombinant human bone morphogenetic protein-2 promotesosteogenesis within atelopeptide type I collagen solution by combi-nation with rat cultured marrow cells. J Biomed Mater Res2002;60(1):61–9.

[29] Nawata M, Wakitani S, Nakaya H, Tanigami A, Seki T, NakamuraY, et al. Use of bone morphogenetic protein 2 and diffusion chambers

to engineer cartilage tissue for the repair of defects in articularcartilage. Arthritis Rheum 2005;52(1):155–63.

[30] Oreffo RO, Triffitt JT. In vitro and in vivo methods to determine theinteractions of osteogenic cells with biomaterials. J Mater Sci MaterMed 1999;10(10/11):607–11.

[31] Laemmli UK. Cleavage of structural proteins during assembly ofhead of bacteriophage-T4. Nature 1970;227(5259):680–5.

[32] Kawasaki T, Niikura M, Kobayashi Y. Fundamental study ofhydroxyapatite high-performance liquid chromatography: directexperimental confirmation of the existence of two types of adsorbingsurface on the hydroxyapatite crystal. J Chromatogr 1990(515):125–48.

[33] Ohta K, Monma H, Takahashi S. Adsorption characteristics ofproteins on calcium phosphates using liquid chromatography. JBiomed Mater Res 2001;55(3):409–14.

[34] Suh C, Kim M, Choo J, Kim J, Kim H, Lee E. Analysis of proteinadsorption characteristics to nano-pore silica particles by usingconfocal laser scanning microscopy. J Biotechnol 2004;112(3):267–77.

[35] Norde W, Giacomelli CE. BSA structural changes during homomo-lecular exchange between the adsorbed and the dissolved states. JBiotechnol 2000;79(3):259–68.

[36] Kandori K, Shimizu T, Yasukawa A, Ishikawa T. Adsorption ofbovine serum albumin onto synthetic calcium hydroxyapatite: influ-ence of particle texture. Coll Surf B 1995(5):81–7.

[37] Vroman L, Adams A. Findings with the recording ellipsometersuggesting rapid exchange of specific plasma proteins at liquid solidinterfaces. Surf Sci 1969;16:438–44.

[38] http://br.expasy.org.

effect of surface structure on protein adsorption to biphasic calcium-phosphate ceramics in vitro...

Documents