ARTICLE IN PRESS

Biomaterials 26 (2005) 75647571 www.elsevier.com/locate/biomaterials

A three-layered nano-carbonated hydroxyapatite/collagen/PLGA composite membrane for guided tissue regeneration

Susan Liaoa,, Wei Wanga, Motohiro Uoa, Shoji Ohkawaa, Tsukasa Akasakaa, Kazuchika Tamuraa, Fuzhai Cuib, Fumio Wataria
a

Department of Oral Health Science, Graduate School of Dental Medicine, Hokkaido University, Kita Ku Kita13 Nishi 7, Sapporo 060-8586, Japan b Department of Materials Science and Engineering, Tsinghua University, Beijing 100084, PR China Received 25 March 2005; accepted 16 May 2005 Available online 11 July 2005

Abstract Functional graded materials (FGM) provided us one new concept for guided tissue regeneration (GTR) membrane design with graded component and graded structure where one face of the membrane is porous thereby allowing cell growth thereon and the opposite face of the membrane is smooth, thereby inhibiting cell adhesion in periodontal therapy. The goal of the present study was to develop a three-layered graded membrane, with one face of 8% nano-carbonated hydroxyapatite/collagen/poly(lactic-co-glycolic acid) (nCHAC/PLGA) porous membrane, the opposite face of pure PLGA non-porous membrane, the middle layer of 4% nCHAC/ PLGA as the transition through layer-by-layer casting method. Then the three layers were combined well with each other with exibility and enough high mechanical strength as membrane because the three layers all contained PLGA polymer that can be easily used for practical medical application. This high biocompatibility and osteoconductivity of this biodegraded composite membrane was enhanced by the nCHAC addition, for the same component and nano-level crystal size with natural bone tissue. The osteoblastic MC3T3-E1 cells were cultured on the three-layered composite membrane, the primary result shows the positive response compared with pure PLGA membrane. r 2005 Elsevier Ltd. All rights reserved.
Keywords: Nano-carbonated hydroxyapatite/collagen composite; GTR; Biodegrade

1. Introduction For more than a decade, thin polytetrauoroethylene (PTFE) membranes have been used as barriers for guided tissue regeneration (GTR) procedures in repairing periodontal defects. However, such known PTFE membrane barriers as Ti plate are non-resorbable. It requires the second surgical removal step, which increases the risk of patient infection and of other undesirable side effects. Moreover, it cannot be used to reconstruct large bone defects due to its bioinert properties. Conversely, biodegradable membranes such
Corresponding author. Tel.: 81 11 706 4252; fax: 81 11 706 4251.

E-mail address: liao@den.hokudai.ac.jp (S. Liao). 0142-9612/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2005.05.050

as collagen and synthetic biodegradable polymers have been studied by many researchers [14]. More recently, membrane made from non-mineralized collagen has been used in GTR applications for periodontal defect repair. Collagen membrane had excellent cell afnity and biocompatibility to regenerate tissues. However, membrane made from non-mineralized collagen is normally weak in strength and is therefore difcult to manipulate. Furthermore, the resorption rate is difcult to match with normal tissue-healing process. Thin, reabsorbed polymer membranes have been recently developed and tested in GTR application, which comprise either polylactic acid (PLA) blended with a citric acid ester or a copolymer made of glycolide and lactide polymers (PLGA). They have found widespread

ARTICLE IN PRESS
S. Liao et al. / Biomaterials 26 (2005) 75647571 7565

use in clinical medicine for bone xation, surgical suture, as scaffold for tissue engineering, and for controlled drug release [58]. These polymers degrade directly to acid during the resorption process. However, medical devices made from the above types of resorbable polymer membranes have been associated with an inammatory response, which may be caused by the accumulation of the acid of the degradation product. Although good adhesion of cells on aliphatic polyesters three-dimensional scaffold was demonstrated by Ishaug and Ochi et al. [9,10], the cell afnity of these polymers as membrane is not as good as the membrane with collagen coating [11]. Then researchers began to nd other methods to meet the demands of degraded GTR membrane both in mechanical property and biocompatibility. There are two ways: one is the mineralized collagen composite, the other is polymer/collagen-based composite [1113]. It has been observed that mineralized collagen exhibits different physical and mechanical properties with those of pure collagen. Moreover, mineralized collagen is considered to provide better biocompatibility toward animal and human tissue than pure collagen. As a hard tissue implant material, mineralized collagen also has a conductivity effect that enhances hard tissue (e.g., bone) growth. Further, changing the content of calcium phosphate minerals could control the bioresorption rate of mineralized collagen membrane. In recent years, the biomimetic mineralized collagen material, called nanoapatite/collagen composite, prepared by different methods attracted researchers attention [1418]. In our previous work we have prepared nano-HA/Collagen (nHAC)-based composite, similar to that of natural bone, through the self-assembled co-precipitation method [1719]. But it could not support the implantation space for its low compressive strength. In order to improve the mechanical strength and the forming ability of the material, we developed a new bone tissue engineering scaffold material: nano-HA/collagen/PLA (nHAC/PLA) by the precipitation and casting procedures [20]. The osteoblasts can adhere and grow into the inner part of the three-dimensional scaffold, nHAC and nHAC/PLA, to obtain osteoblast/composite construct for bone tissue engineering. When implanted in rabbits, these materials underwent resorption and promoted new bone formation [17,21,22]. In order to increase the resorption in vivo, nano-carbonated hydroxyapatite/ collagen (nCHAC) composite was prepared at room temperature via biomimetic self-assembly method [23]. This composite shows the same inorganic phase of natural bone with nano-sized level and low crystallinity degree, and contains 2.814.7 wt% of carbonated content [23]. Many experiments show the high solubility of carbonated apatite in vitro and in vivo [24,25]. The HA phase present in natural bone, dentin and enamel, respectively, contains approximately 7.4, 5.6, 3.5 wt% of

carbonate [26]. CHA materials have excellent biocompatibility and properties, which can be favorably compared with those of hard tissue. As we know, the higher the carbonated minerals,the higher the osteoconductive and the earlier the bioresorption. Here, we selected the nCHAC as the bioactive component, PLGA as the barrier to prepare one novel three-layered membrane comprising a pliable not brittle, substantially cell-impermeable polymeric layer, a rst cell-permeable outer layer superposed on a rst surface of the outer layer, and a second cell-permeable outer layer superposed on a second surface of the inner layer opposite the rst face. The nCHAC material facilitated reliable bone regeneration by inducing undifferentiated cells in the graft recipient site to become osteoblasts and form new bone (i.e., stimulating cellular transformation). The composite also supplied a ready source of calcium for rapid mineralization. The barrier material is quick and easy to use during surgery and is completely biodegradable, eliminating the need for a second surgery to remove the barrier material. The functional graded materials (FGM) idea may help for improvement of medical device design not only in GTR membrane, but also in the other tissue engineering [27]. Our new mineralized collagen/PLGA membrane is considered to have many medical applications because of its exibility, strong mechanical strength, easy manipulation character, excellent biocompatibility and controllable bioresorption. It can be used in medical applications such as for repairing periodontal defects, membranes for covering bone defect surgery and for bone substitutes, skin wound repair and healing, skin sealing, and as a carrier for antibiotic, bone growth factors, skin growth factors, and so forth. Then we used osteoblastic MC3T3-E1 cells co-culture to investigate the cell afnity of this composite membrane.

2. Materials and methods 2.1. Materials nCHAC composite was prepared by biomimetic method, which has been reported previously [23]. Briey, Type I atelocollagen gel (2 wt%, Koken Company, Japan) was added to 0.5 M acetic acid. Then solutions of CaCl2 and H3PO4 (Ca/ P 1.66) were gradually added separately through their respective tube pumps, and after stirring for 1 h, solution of Na2CO3 (mol ratio of CO2/PO3 3) was gradually added. 3 4 Then it was stirred, titrated with sodium hydroxide to pH 9 at room temperature. After aging the solution for more than 2 h, the nCHAC material was harvested by centrifugation and freeze-drying method. Three 5 g 75/25 PLGA solutions (0.72 dL/g inherent viscosity, Absorbable Polymers International, USA) were dissolved into 20 ml chloroform by stirring, respectively. The rst solution is for Layer 1. To the second one, 0.2 g nCHAC

ARTICLE IN PRESS
7566 S. Liao et al. / Biomaterials 26 (2005) 75647571

Layer 3

Layer 2

Layer 1

nCHAC
(a)

PLGA
(b)

100 m

(c)

2 m

(d)

1 m

(e)

10 m

Fig. 1. (a) Schematic representation of three-layered membrane. (b) The OM result of the section of the nCHAC/PLGA TLM. (c, d) The SEM results of Layer 3. The black arrow refers to the nCHAC particles covered by polymer. (e) The SEM result of the underside of Layer 1.

prepared above is added and ultrasonically mixed for Layer 2. To he third one, 0.4 g nCHAC prepared above is added and ultrasonically mixed for Layer 3. One milliliter of the rst solution was poured into a glass Petri dish, allowed to stand for 30 min at the room temperature which is enough to form Layer 1. Then 1 ml of second solution was poured into the Layer 1 wherein the surfaces were bounded by Layer 1 to form Layer 2. Thirty minutes later, the 1 ml of third solution was poured into Layer 2 wherein the surfaces were bounded by Layer 1 to form Layer 3, which is shown in Fig. 1(a). Then transfer the three-layered membrane (labeled TLM) to a vacuum oven for 2 days at the room temperature.

We selected pure PLGA membrane only with Layer 1 as control. 2.2. Characterization Scanning electron microscopic (SEM) examinations (Sirion 200, FEI Company, USA) were performed after coating the samples with gold. The section of the membrane was observed by an optical microscope (Nikon Eclipse TS100, Japan). X-ray diffraction (XRD) analysis was performed in Rigaku/Multiex diffractometer which used Ni ltered CuKa

ARTICLE IN PRESS
S. Liao et al. / Biomaterials 26 (2005) 75647571 7567

2.5. Total protein assay Osteoblastic MC3T3-E1 cells (8000 cells/cm2) were seeded on the blank control, pure PLGA membrane and nCHAC/ PLGA TLM in the 24-well plates, respectively, and cultured in the DMEM+10%FBS medium under the standard cell cultured conditions for 1, 4, 7 days. At the end of the prescribed time, osteoblasts were trypsinized using 0.25% trypsinase from each substrate. Total protein content in the cell lysates was determined using a spectrophotometer [28]. For this purpose, aliquots of each protein-containing, distilledwater supernatant were incubated with a 1.0 ml of modied Lowry reagent for 10 min at room temperature (Modied Lowry Protein Assay Kit No.23240, PIERCE, USA). Light absorbance of these samples was measured at 750 nm on a Hitachi U-1100 Spectrophotometer (Tokyo, Japan). Total intracellular protein (expressed as mg) synthesized by osteoblasts cultured in the medium was determined from a standard curve of absorbance versus known concentrations of albumin by protein assay kit. The total protein content was normalized by cell numbers and expressed as protein/cell.

Fig. 2. Tensile mechanical testing of nCHAC/PLGA TLM. (1) Clamp; (2) specimen-nCHAC/PLGA TLM; (3) sand paper adhered on the two surfaces of clamp to prevent the slip of specimen during the testing.

3. Results
radiation, in the 2y range of 101801 at a scan rate of 2 1/min, with a sampling interval of 0.021. 2.3. Mechanical testing The tensile mechanical property was tested by an Autograph IS-5000 Mechanical Tester (Shimazu Co, Japan) with 100 kg load cell as shown in Fig. 2. The specimens were rectangular, about 10 mm in width and about 20 mm in length. We used sand paper to adhere on the two surfaces of clamp, which prevent the slip of specimen during the testing. The crosshead speed was set as 2.5 mm/min, and the load was applied until ultimate fracture of the specimen. The elastic modulus was calculated as the slope of initial linear portion of the force-strain curve. The tensile strength was determined as the maximum point of the forcestrain curve. 2.4. Cell experiments Osteoblastic MC3T3-E1 cells were cultured in DMEM supplemented with 10% FBS under standard cell culture conditions. The osteoblasts were seeded into the nCHAC/ PLGA TLM, PLGA, and blank control in 24-well plates at a concentration of 8000 cells/cm2. The cultures were maintained in a humidied atmosphere consisting of 95%air/5%CO2 (v/v) at 37 1C, and the medium was changed every three days, and routinely examined using phase contrast microscopy. The cells were accounted on a hemacytometer after trypsinization at the 1st day, 4th day, and 7th day. Borosilicate glass cover slips (+ 15 mm, Matsumami, Japan) were used as a reference substrate in experiments with cells. Cover slips were degreased in ethanol, and sterilized in UV light for 4 h. The nCHAC/PLGA TLM, PLGA membranes were both adhered on the cover slips for cell experiments.

Our concept of the composite membrane is illustrated in Fig. 1(a). From Fig. 1(b) of nal prepared membrane, the optical observation of the section of the nCHAC/ PLGA TLM demonstrated that there is no signicant boundary between the transparent Layer 1, translucent Layer 2 and Layer 3. Then the possible excellent integrating of the three layers was achieved. The thickness of each layer is about 100 mm, and the total thickness of the TLM is about 300400 mm. Fig. 1(c) and (d) shows the porous structure after the nCHAC addition. The size of nCHAC particles covered with PLGA polymer was 200400 nm. The unit size of prepared nCHAC composite is about 60 nm 120 nm [23], so the polymer solution could form into nano-level coating on the surface of nCHAC particles, about 100280 nm thickness. Here the pore size is about 10 mm, and increased pore size will be attained along with the biodegraded process. As one of factors for threedimensional porous conformation in the composite layers, the coating formation was decided by the concentration of polymer solution. Maybe we could enlarge the initial pore size if we properly increase the solution concentration or change the ratio of nCHAC and polymer in our future research. The pure PLGA layer side attached on the surface of the glass Petri dish is non-porous as the underside of the TLM, as shown in Fig. 1(e). One face of the membrane (Layer 3 upside) is porous thereby allowing cell growth thereon and the opposite face of the membrane (Layer 1 underside) is smooth, thereby inhibiting cell adhesion. Phase development of the novel nCHAC/PLGA TLM was demonstrated by XRD patterns (Fig. 3). The peaks

ARTICLE IN PRESS
7568 S. Liao et al. / Biomaterials 26 (2005) 75647571

Fig. 3. XRD patterns of (a) nCHAC/PLGA TLM; (b) nCHAC; (c) pure PLGA membrane.

of nCHAC/PLGA referred to the TLM consisted of pure PLGA and nCHAC materials, as the curve (a) referred to the (b) plus (c) in Fig. 3. The preparation process of the nCHAC/PLGA TLM did not change the phase by casting technique at room temperature, which consisted of the nCHAC and PLGA. Due to this, the TLM keeps the degraded properties of the two components. It is known that the human bone is an extracellular matrix mainly composed of 65% HA nanocrystal and 25% collagen bers, in dentine about 70% HA and 20% collagen. HA nano-crystals are aligned their C-axes along collagen bers. We have prepared nHAC-based composite through the self-assembled coprecipitation method which is similar to natural bone both in main component and microstructure [1719]. Moreover, all biological apatites are carbonated apatite in vivo. Carbonated hydroxyapatite is bioresorbed faster than hydroxyapatite, so the nCHAC is easier to be bioabsorbable than nHAC. The nCHAC composite still keeps the bundle microstructure of mineralized collagen, which is the same as that of natural bone [23]. The ultimate tensile strength of the nCHAC/PLGA TLM is 9.771.7 MPa, the elastic modulus is equal to 1.270.5 GPa, which is above the tensile property of cancellous bone [33], and the elongation ratio is 2497102% from the result of the mechanical testing. After the tensile testing, there is no cleft of the layers with each other, which also demonstrates the excellent combination of the graded layer through the method used in this study. It is one of the basic factors to design the membrane with graded component which maybe helpful for medical therapy because enough mechanical property is the necessary condition for applications. Moreover the extremely high toughness of this membrane ensures the integrality of implanted membrane as practical deformation during the surgery and could be properly used in different unaccustomed-shape defects. Herein compared with the interosseous ligament tensile data, at the longitudinal direction the modulus is 0.4370.32 GPa which is only half of the TLM prepared in this study, and the ultimate strength is 52.7733.5 MPa which is 5 fold of the TLM [34]. But at

the transverse direction is far lower than longitudinal, the modulus is 1.1271.62 MPa and the ultimate strength is 0.1470.05 MPa, which are both far lower than that of TLM. In the middle range of mechanical values, the TLM could be well adapted to the force from the neighboring interosseous ligament. The osteoblastic cells showed afnity both with the PLGA and nCHAC/PLGA TLM membranes as the preliminary biological testing has shown. From the optical microscopic observation, the cells were normally adhered and grew, the morphology of cells changed from round shape to spindle shape, then to polygonal shape. There was no signicant difference between the blank control, PLGA and nCHAC/PLGA except for the control and PLGA at the 4-day from the cell proliferation result in Fig. 4(a). However, the difference in the average values of three groups was indicated, the control increased on the 4th day, then decreased on the 7th day, the PLGA group decreased from the 1st day to 7th day, nCHAC/PLGA TLM increased from the 1st day to 4th day, after that remained the same on the 7th day. Then the smooth surface of PLGA worked as cell barrier for the non-proliferation, while the open structure of void on the surface of nCHAC/PLGA spaces was capable of accommodating tissue ingrowth if we implanted it in vivo. As one kind of biocompatible polymers, pure PLGA has been used in many elds of medical device, such as bone plate and screw, suture, etc. The opposite face of the membrane is smooth as the pure PLGA membrane thereby inhibiting cell adhesion thereon. After the addition of nCHAC, higher bone tissue compatibility was achieved in the outside surface of this TLM membrane. The GTR membrane allows the propagation of bone and periodontal ligament cells by precluding epithelial cells and gingival connective tissue cells which propagate at greater rate. This TLM membrane would be as one of the promising selections for GTR membrane when used in dentistry and orthopedics. The corresponding protein result is shown in Fig. 4(b), where there is no signicant difference between the three groups. The average protein/cell of PLGA remained the same on the 1st day and 4th day, then increase on the 7th day about 3 fold. But blank control and nCHAC/PLGA rstly decreased on the 4th day to about 30% of that on the 1st day, then increased on the 7th day, which is opposite to that of PLGA group. The possible reason relied on the same adhered ratio of three groups on the 1st day, while no distinct proliferation was shown in PLGA on the 4th day and 7th day, but in nCHAC/PLGA group after the cells proliferation, the normal protein expression of cells came back. The average value of the PLGA TLM on the 7th day is the highest in the three groups, which also demonstrated the biocompatibility with continued growth of the adhered cells. However, the possible increased protein content of control and nCHAC/

ARTICLE IN PRESS
S. Liao et al. / Biomaterials 26 (2005) 75647571 7569

280 6 Control PLGA nCHAC/PLGA 260 240

control PLGA nCHAC/PLGA

Protein/Cell (g/ml/104)
1 4 7

220 200 180 160 140 120 100 80 60

Cells (X104)

40 20

0 1 4 7

(a)

Time (day)

(b)

Time (day)

Fig. 4. MC3T3-E1 cells cultured on the control (coverslip), pure PLGA membrane and nCHAC/PLGA TLM. (a) The cell proliferation from 1 to 7 days. (b) The total protein assay result normalized by cell numbers on 1, 4, and 7 days. n 4, *Po0.05.

PLGA would be achieved in the longer culture period. Then the cell amount with protein results demonstrated that relatively higher afnity of osteoblasts on the nCHAC/PLGA than PLGA membrane.

4. Discussions The nCHAC/PLGA material was designed to be superimposed on each other and afxed together in a laminar fashion so that the implantable bioabsorbable article had one surface that provided an open structure of void spaces capable of accommodating tissue ingrowth while the opposite surface provided a cellbarrier. The cell-barrier sheet material is typically initially non-porous, PLGA polymer is bioabsorbable in vivo, is also porous so long as the pore sizes are small enough to substantially preclude cell passage and ingrowth, when passage of nutrients or gasses across the material may be important. Epple et al. [29] reported their one geometrically structured implant for cranial reconstruction made of PLA and calcium phosphate/ calcium carbonate prepared by a combination of hot pressing and gas foaming with two layers. On the inside, the implant consisted of a macroporous and fast degradable material (PDLA+CaCO3) to allow the ingrowth of bone cells. On the outside, the implant consisted of a compact and slower biodegradable material (PLA and calcium phosphate) to ensure the mechanical stability and protection. The same aim as our design of this implant is the polyester which combined with basic lling materials (calcium salts) to

overcome the problems like inammatory reactions caused by acidic degradation products of polyester [29,30]. Moreover, the membrane in this study is not only used in cranial defect repair, but also in other medical applications for guided tissue regeneration, such as periodontal defect repair, skin wound repair, etc. Or it can be used combined with other bone ller porous material when the defect size is too large. The extended concept of the graded material may be a useful guide to develop proper biomimetic medical device in different sites, including self-assembly system application. Apatite-based composite without collagen can be produced by hot temperature technique such as hot pressing. But with collagen, limited technology can be used for the collagen-based material fabrication because collagen cannot endure temperature higher than that of the body. Here we selected the former self-assembled method of apatite and collagen cooperated with later solution casting method ensured the initial biomimetic chemical and physical properties of nano-composite membrane. Design of biomaterials with properties similar to physiological bone (characterized by surface grain sizes in the nanometer regime) would undoubtedly aid in the formation of new bone at the tissue/ biomaterial interface and, therefore, improve orthopedic/dental implant efcacy [28]. The reasons rely on the osteoblast, osteoclast and osteocyte attraction by similar component and microstructure, following which the nano-sized minerals could be used as the mineralized template and source for new bone tissue regeneration. Murphy and Mooney [31] and Shin et al. [32] reported that the mineralized surface formation on PLGA

ARTICLE IN PRESS
7570 S. Liao et al. / Biomaterials 26 (2005) 75647571

scaffold in higher ion concentration simulated body uid (41.00 SBF), especially phase size of minerals in 2.00 SBF is lower than that in 0.751.75 SBF after 16 days of incubation. In other words, the high Ca2+ and H2PO ions play a key role for new minerals 4 formation on the PLGA surface in vitro. The ion source may be very important for bone mineral regeneration in vivo, the possible quick mineral formation in defect sites could be achieved after this composite membrane implantation. Similar cell results published by Webster et al. [28] showed increased osteoblast adhesion on alumina, titania, and hydroxyapatite with nanometer grain sizes when compared to conventional grain sizes. This change in grain size was shown to signicantly inuence both the type and conformation of absorbed proteins, which led to increased osteoblast function [35]. Here the nCHAC in TLM is nanometer grain sized, and the carbonated component increased the bioabsorbability in vitro and in vivo which might be the impetus for osteoblasts function. More specically, a bonelike mineral has been shown to be a prerequisite to bonding of the orthopedic implant materials with native bone tissue (osteoconductivity), and may drive osteogenic differentiation of adult human stem cells (osteoinductivity) [3638]. The casting technique can be applied to the dissoluble polymer, including the PLA, PGA, PLGA, etc., which can help us to adjust the bioresorption behavior for optimal osteogenesis in different sites for defect repairs. The degraded character of TLM was investigated in vitro by simulated experiments, and the 3 months period of TLM full degradation was properly used for GTR in periodontal therapy (data not shown). In vivo animal experiment that can nally prove the osteo-conductivity of this TLM material for bone regeneration is in progress.

ment of Material Science and Engineering) for helping us to do SEM experiments. Thanks to Dr. Hiroko Takita (Hokkaido University, Graduate School of Dental Medicine) for freeze-drying the samples. Thanks to Professor Matthias Epple (University of DuisburgEssen, Institute of Inorganic Chemistry) for gaving us helpful suggestions.

References
[1] Piattelli A, Scarano A, Russo P, Matarasso S. Evalution of guided bone regeneration in rabbit tibia using bioresorble and nonresorbable membranes. Biomaterials 1996;17:7916. [2] Ozmeric N, Bal B, Oygur T, Balos K. The effect of a collagen membrane in regenerative therapy of two-wall intrabony defects in dogs. Periodontal Clin Invest 2000;22:2230. [3] Dupoirieux L, Pourquier D, Picot MC, Neves M. Comparative study of three membranes for guided bone regeneration of rat cranial defects. Int J Oral Maxillofac Surg 2001;30:5862. [4] Kikuchi M, Koyama Y, yamada T, Imamura Y, Okada T, Shirahama N, Akita K, Takuda K, Tanaka J. Development of guided regeneration membrane composed of b-tricalcium phosphate and poly(L-lactide-co-glycolide-co-e-caprolactone) composites. Biomaterials 2004;25:597986. [5] Ashammakhi N, Rokkanen P. Absorbable polyglycolide devices in trauma and bone surgery. Biomaterials 1997;18:39. [6] Osther PJ, Gjode P, Mortensen BB, Bartholin J, Gottrup F. Randomized comparison of polyglycolic acid and polyglyconate sutures for abdominal fascial closure after laparotomy in patients with suspected impaired wound healing. Br J Surg 1995;82: 10802. [7] Ishaug-Riley SL, Crane GM, Gurlek A, Miller MJ, Yasko AW, Yaszemski MJ, Mikos AG. Ectopic bone formation by marrow stromal osteoblast transplantation using poly(DL-lactid-co-glycolic acid) foams implanted into the rat mesentery. J Biomed Mater Res 1997;36:18. [8] Langer R. Drug delivery and targeting. Nature 1998;392(Suppl.): 510. [9] Ishaug SL, Crane GM, Miller MJ. Bone formation by threedimensional stromal osteoblast culture in biodegrable polymer scaffolds. J Biomed Mater Res 1997;37:1728. [10] Ochi K, Chen GP, Ushida T, Gojo S, Segawa K, Tai H, Ueno K, Ohkawa H, Mori T, Yamaguchi A, Toyama Y, Hata JI, Umezawa A. Use of isolated mature osteoblasts in abundance acts as desired-shaped bone generation in combination with a modied poly DL-lactic-co-glycolic acid (PLGA)-collagen sponge. J Cellular Physiology 2003;194:4553. [11] Kim SY, Kanamori T, Noumi Y, Wang PC, Shinbo T. Preparation of porous poly(D,L-lactide) and poly(D,L-lactide-coglycolide) membranes by a phase inversion process and investigation of their morphological changes as cell culture scaffolds. J Appl Polym Sci 2004;92:208292. [12] Liu ST. Thin mineralized collagen membrane and method of making same. US Patent No. 6417166, 2002. [13] Pachence JM, Frenkel S, Menche D. Multi-stage collagen-based template or implant for use in the repair of cartilage lesion. US Patent No. 6080194, 2000. [14] Kikuchi M, Itoh S, Ichinose S, Shinomiya K, Tanaka J. Selforganization mechanism in a bone-like hydroxyapatite/collagen nanocomposite synthesized in vitro and its biological reaction in vivo. Biomaterials 2001;22:170511. [15] Chang MC, Ikoma T, Kikuchi M, Tanaka J. Preparation of a porous hydroxyapatite/collagen nanocomposite using glutaraldehyde as a crosslinkage agent. J Mater Sci Lett 2001;20:1199201.

5. Conclusions In this study, the three-layered graded membrane consisted of nCHAC and PLGA polymer was prepared by biomimetic method and casting technique. With one side of porous structure and other side with smooth non-porous structure resulting in the different biodegradation rate, and the enough high strength for practical use, the composite membrane is a promising material for guided tissue regeneration under the concept of FGM. It provides an improved design for medical device applications.

Acknowledgements This research was funded by JSPS (Japan Society for the Promotion of Science). Thanks to Associate Professor Guofu Xu (Central South University, Depart-

ARTICLE IN PRESS
S. Liao et al. / Biomaterials 26 (2005) 75647571 [16] Du C, Cui FZ, Zhang W, Feng QL, Zhu XD, de Groot K. Formation of calcium phosphate/collagen composites through mineralization of collagen matrix. J Biomed Mater Res 2000; 50:51827. [17] Du C, Cui FZ, Zhu XD, de Groot K. Three-dimensional nanoHAp/ collagen matrix loading with osteogenic cells in organ culture. J Biomed Mater Res 1999;44:40715. [18] Zhang W, Liao SS, Cui FZ. Hierarchical self-assembly of nanobrils in mineralized collagen. Chem Mater 2003;15:32216. [19] Liao SS, Cui FZ, Zhang W, Feng QL. Hierarchically biomimetic bone scaffold materials: nano-HA/collagen/PLA composite. J Biomed Mater Res Part B 2004:15865. [20] Liao SS, Cui FZ, Zhu Y. Osteoblasts adherence and migration through three-dimensional porous mineralized collagen based composite: nHAC/PLA. J Bioact Compat Polym 2004;19:11730. [21] Liao SS, Guan K, Cui FZ, Shi SS, Sun TS. Lumbar spinal fusion with a mineralized collagen matrix and rhBMP-2 in a rabbit model. Spine 2003;28:195460. [22] Du C, Cui FZ, Feng QL, Zhu XD, de Groot K. Tissue response to nano-hydroxyapatite/collagen composite implants in marrow cavity. J Biomed Mater Res 1998;42:5408. [23] Liao SS, Watari F, Uo M, Ohkawa S, Tamura K, Wang W, Cui FZ. The preparation and characteristics of a carbonated hydroxyapatite/collagen composite at the room temperature. J Biomed Mater Res Part B 2005, in press. [24] Schanek WJ, Yoshimura M. Processing and properties of hydroxyapatite-based biomaterials for use as hard tissue replacement implants. J Mater Res 1998;13:94117. [25] Doi Y, Shibutani T, Moriwaki Y, Kajimoto T, Iwayama Y. Sintered carbonate apatites as bioresorbable bone substitutes. J Mater Res 1998;39:60310. [26] LeGeros RZ. Calcium phosphates in oral biology and medicine. Basel, Switzerland: Karger AG; 1991. [27] Watari F, Yokoyama A, Omori M, Hirai T, Kondo H, Uo M, Kawasaki T. Biocompatibility of materials and development to functionally graded implant for bio-medical application. Compos Sci Technol 2004;64:893908. 7571 [28] Webster TJ, Siegel RW, Bizios R. Osteoblast adhesion on nanophase ceramics. Biomaterials 1999;20:12217. [29] Schiller C, Rasche C, Wehmoller M, Bechmann F, Eunger H, Epple M, Weihe S. Geometrically structured implants for cranial reconstruction made of biodegradable polyesters and calcium phosphate/calcium carbonate. Biomaterials 2004;25: 123947. [30] Linhart W, Peters F, Lehmann W, Schwarz K, Schiling AF, Amling M, Rueger JM, Epple M. Biologically and chemically optimized composite of carbonated apatite and polyglycolide as bone substitution materials. J Biomed Mater Res 2001;54: 16271. [31] Murphy WL, Mooney DJ. Bioinspired growth of crystalline carbonate apatite on biodegradable polymer substrata. J Am Chem Soc 2002;124:19107. [32] Shin K, Jayasuriya AC, Luong L, Outslay M, Kohn DH. Formation and dissolution behavior of biomimetic mineral on 3-dimensional poly(lactide-co-glycolid) scaffold. Seventh world biomaterials congress, 2004. p. 1841. [33] Black J, Hastings GW. Handbook of biomaterials properties. London, UK: Charpman and Hall; 1998. [34] Stabile KJ, James Pfaefe H, Weiss JA, Gabriel MT, Tomaino MM, Fischer KJ. Longitudinal and transverse mechanical properties of the interosseous ligament of the forearm. Bioengineering Conference ASME 2001:3634. [35] Webster TJ, Ergun C, Doremus RH, Siegel RW, Bizios R. Specic proteins mediate enhanced osteoblast adhesion on nanophase ceramics. J Biomed Mater Res 2000;51:47583. [36] Hench LL. Bioceramicsfrom concept to clinic. J Am Ceram Soc 1991;74:1487510. [37] Ohgushi H, Caplan AI. Stem cell technology and bioceramics: from cell to gene engineering. J Biomed Mater Res Part B 1999; 48:91327. [38] Krebsbach PH, Kuznetsov SA, Satomura K, Emmons RV, Rowe DW, Robey PG. Bone formation in vivo: comparison of osteogenesis by transplanted mouse and human marrow stromal broblasts. Transplantation 1997;63:105969.