Biomaterials 23 (2002) 2119–2126

Development of a ‘mechano-active’ scaffold for tissue engineering

Ying Yang*, Julia L. Magnay, Leanne Cooling, Alicia J. El Haj

Centre for Science and Technology in Medicine, School of Medicine, Keele University/North Staffordshire Hospital,

Thornburrow Drive, Hartshill, Stoke-on-Trent, North Staffordshire ST4 7QB, UK

Received 16 March 2001; accepted 2 October 2001

Abstract

In this study, we investigate the potential for manipulating bone cell mechanotransducers in tissue engineering. Membrane ionchannels such as voltage operated calcium channels (VOCC) have been shown to be a critical component of the bone celltransduction pathway with agonists and inhibitors of this pathway having profound effects on the load signal. By encapsulating a

calcium channel agonist with slow release within a poly(l-lactide) (PLLA) scaffold, we can generate a ‘mechano-active’ scaffold foruse in skeletal tissue engineering. PLLA scaffolds with and without a calcium channel agonist, BAY K8644, were seeded withprimary human bone cells or the human MG63 bone cell line and cultured for 1–3 weeks followed by mechanical stimulation with a

four-point bending model. Our results show that addition of the agonist for slow release is sufficient to enhance the load-relatedresponses in bone cells within the scaffolds. Specifically, collagen type I expression and the ratio of alkaline phosphatase to proteinare elevated in response to cyclical mechanical stimulation of approximately 1000mstr which is then further enhanced in the

‘mechano-active’ scaffolds. As the agonists only act when the calcium channels are open by attenuating the calcium flux, thestimulation is specifically targeted to scaffolds subjected to load either in vitro or ultimately in vivo. Our results suggest thatmanipulating the VOCC and attenuating the opening of the calcium channels may be an effective technique to amplify matrixproduction via mechanical stimulation which may be applied to bone tissue engineering and potentially engineering of other load-

bearing connective tissues. r 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Bone; Tissue engineering; Mechanical load; Calcium channel; Biodegradable polymer

1. Introduction

The skeleton has an enormous capacity for growth,regeneration and remodelling. The mechanical environ-ment is one of the major factors involved in theseprocesses. Mechanical loading of physiologically rele-vant magnitudes has been shown directly to initiatebone modelling in animal models [1]. In contrast, lack ofload has been shown to promote tissue atrophy andbone loss [2]. In clinical practice, it is well known thatmechanical stimuli are important local regulators ofcallus and ultimately bone tissue formation duringhealing of fracture repair. Currently, during orthopaedicfracture repair, the mechanical environment is con-trolled by the use of external fixators and/or throughdefined rehabilitation programmes. As yet there hasbeen no success in defining a pharmaceutical or chemical

alternative which could act to mimic the mechanicalinduction and control the repair of bone tissue.One of the problems with manipulating mechano-

transducers is that the process involves ill-definedpathways which are involved in converting mechanicalsignals into a biochemical event within a bone cell. It isclear that there are several mechanisms which play a roleas mechanotransducers but understanding the interac-tions between these pathways or the temporal chain ofevents is not established. In our laboratory, we havefocused on the early signalling pathways for mechano-transduction [3]. We and other groups have demon-strated that membrane channels, in particular voltageoperated calcium channel (VOCC), have an importantfunction in mechano-transduction in bone cells [3–7].Walker et al. [4] have shown, using fluo-3 AMtechniques, how application of a mechanical signal byoptical tweezers or stretch to an underlying membraneof an individual human or rat osteoblast results in adramatic flux of intracellular calcium. This response ismodulated by the VOCC blocker, nifedipine, and the

*Corresponding author. Tel.: +44-1782-554253; fax: +44-1782-

747319.

E-mail address: bea04@keele.ac.uk (Y. Yang).

0142-9612/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.

PII: S 0 1 4 2 - 9 6 1 2 ( 0 1 ) 0 0 3 4 2 - 8

VOCC enhancer, BAY K8644. In addition, we haveshown that the production of matrix proteins, such asosteopontin and osteocalcin, are elevated in response tomechanical loading and that this response is stronglyinhibited by the calcium channel blocker, nifedipine, andgreatly enhanced by the calcium channel agonist, BAYK8644 [5]. Osteopontin and osteocalcin are non-collagenous matrix proteins located in the extracellularbone matrix and have been suggested to be important inboth bone formation and resorption [8–9]. Furthermore,osteopontin has previously been shown to be amechanoresponsive gene in rat calvarial cells and chickosteoblasts and our Western analysis confirms thesefindings in rat femoral osteoblasts [5,10].In the past decade, tissue engineering has developed

into a most promising direction for new therapies. Itapplies scientific and engineering principles to thedesign, construction, modification, and/or maintenanceof living tissues, thus enabling the production ofreplacement tissues for surgical repair. Tissue engineer-ing has all the advantages of an autograft therapy,but the procedure is not subject to the limitation ofsupply since the cells and tissue may be expandedprior to re-implantation. In addition, this techniqueavoids morbidity at autograft donor site becauseonly a small biopsy or bone marrow aspirate may beneeded to obtain the autologous cells for amplificationin vitro. The challenge for bone tissue engineering is thatthe newly created or repaired tissue must fulfil amechanical role. As such, skeletal tissue engineeringshould involve the production of scaffolds and remo-delled implants that are capable of load-bearingaccording to their environment. Thus, the optimalscaffold not only mimics the three-dimensional structureof porous bone, but also attracts the cells, facilitatesadhesion and spreading, and stimulates the cells toproduce new bone matrix in response to the mechanicalenvironment.Several growth factors, such as transforming growth

factor (TGF-1) [11], bone morphogenetic protein (BMP-2), [12,13] and osteogenic protein (OP-1) [14] have beenshown to play a role in bone cell proliferation,differentiation and production of new extracellularmatrix. Various bioactive scaffolds have been developedwhich either have growth factors adherent to the surfaceof the biodegradable polymer or which contain slowrelease growth factors [15]. However, in vivo studieswhich employ these bioactive scaffolds highlight some ofthe problems with this approach. Either the surfacemolecules are degraded quickly with the biodegradablepolymer, or in the slow release model, the growth factorsare released continuously without control and theiraction is restricted to a short duration after implant intothe tissue repair site [16].In this study, we have set out not to replace but to

augment the load signal, using bioactive materials which

we term ‘mechano-active’. The aim of this projectis to create a biodegradable material containingslow release agonists which enhance the load signaland stimulate increased matrix synthesis. The advantageof this system is two-fold: the matrix is laid down inthe region according to the mechanical signals, andthe control of the augmentation is limited to amechanical switch rather than continuous release as inthe case of bioactive growth factor-containing scaffolds.The calcium channel agonist used in this project isBAY K8644, an L-type Ca2+ channel agonist whichprolongs single channel open time without affecting theclose time [17].We describe our methodology for generating me-

chanically active three-dimensional scaffolds for tissueengineering via substrate and chemical mediation. Thescaffold is encapsulated with a calcium channel agonistin slow release mode, then the surface of the scaffold ismodified in order to enhance the adhesion to cells.Finally, we test in vitro the effect of cyclical mechanicalloading applied to the bone cell-seeded scaffolds onmatrix production, bone cell proliferation and pheno-type.

2. Materials and methods

2.1. Materials and construction of biodegradablescaffolds

The polymer used for fabricating the scaffold waspoly(l-lactide) (PLLA), a medical grade material,PURASORBTM supplied by Purac Biochem bv Gor-inchem, Holland with a molecular weight of 360,000Da.The calcium channel agonist, BAY K8644 (Bay), 1,4-dihydro-2,6-dimethyl-5-nitro-4-[20-(trifluoromethyl)phe-nyl]-3-pyridine-carboxylic acid methyl ester, was pur-chased from Calbiochem. Calf collagen type I solutionin 0.1% acetic acid (Sigma, UK) was used to coat thescaffolds without further modification.The porous three-dimensional scaffolds were

formed by a solvent-evaporating and salt-leachingtechnique with sodium chloride as the leachablecomponent [18]. Briefly, the PLLA was dissolvedin chloroform (20%) and mixed with the requiredamount of salt to form a composite, which wastransferred into a mould to generate the requisitescaffold dimension. After removing the chloroform byevaporation, the composite was placed in water for 2days at room temperature with stirring to leach-out thesalt. The scaffold was then dried at 501C to constantweight. The pore size of the scaffold was controlled torange from 250 to 350 mm and porosity was estimated as90% (from the weight fraction of PLLA and salt in thescaffold).

Y. Yang et al. / Biomaterials 23 (2002) 2119–21262120

The scaffold dimensions were defined by a four-point bending apparatus and were 1.5� 24� 15mm3.For the observation of the cell seeding efficiency,the cylinder and square shape scaffolds were con-structed with dimensions of 8� 8 and 10� 10� 1mm3,respectively.To enhance the adhesion of cells to the scaffolds,

the scaffolds were coated with collagen type I with7.5 mg cm�2 by soaking in the collagen solution for2 h, denoted as PLLA/C. For processing calciumchannel agonist encapsulated scaffold, the scaffoldwas soaked with Bay solution for 1 h first. Afterremoving the solvent by evaporation, the treatedscaffold was then coated with collagen type I asdescribed above, denoted as PLLA/C/Bay. Using amechanical aspiration technique, we ensured thatthe coating of collagen and Bay penetrated deeply intothe pores. The initial concentration of Bay in thescaffold was controlled at approximately 60 nm cm�3.The Bay concentration released from the scaffold duringthe incubation period into the media was measuredusing the 3000 scanning spectrophotometer, CecilInstruments Ltd. Bay K8644 has an absorbancepeak at 405 nm.

2.2. Culture of human primary bone cells and cell lines

Human primary bone cells and the human osteoblast-like cell line, MG63 were used in this study. Theethical approval for the human tissue usage in thisproject was granted by North Staffordshire HospitalTrust ethical committee. Informed patient consentwas obtained. Primary human bone cells were isolatedfrom trabecular bone biopsies obtained from humantibial fractures as described in Walker et al. [5]. Inparticular, muscle, cartilage and blood were removed bysaline washes from biopsies prior to dissection. Bonefragments were placed into culture flasks containingaMEM, supplemented with 10% fetal calf serum and1% antibotic/antimycotic solution. Proliferating bonecells from the fragment were cultured to confluency in a5% CO2 incubator at 371C and employed followingpassage 2; MG63 cells were used between passagenumber 20–25. The isolated cells were pooled, pelleted,resuspended in a known volume of the media, andcounted using a haemocytometer. MG63 cell lines werecultured as described previously by Peake et al. [19] upto passage number 20 used in this study. Specifically, thecells were grown in the same media formula as describedfor the primary bone cells above. 1� 105 cells (unlessotherwise indicated in the text) in 100 ml aMEM mediawere seeded into each of the scaffolds (n ¼ 3) asdescribed above and cultured in supplemented aMEMmedia with 10% fetal serum and 1% antibody/antimycotic.

2.3. Application of mechanical load to bone cell-seededconstructs

A four-point bending model which applies cyclicalphysiological levels of tensile and compressive load tothe glass coverslips was used for this study as describedpreviously by Peake et al. [19]. The PLLA/C scaffoldswith and without Bay encapsulation were adhered toglass coverslips with silicone rubber glue to allow theapplied tension transfer from the coverslips to thescaffold and cells. Bulk strain applied across theconstruct was measured to be approximately 1000 mstrstrain by direct strain gauge measurements and formuladerived calculations [19].For the four-point bending samples, the cell-scaffold

constructs were cultured statically for 1 week in 5mlmedia per scaffold, the media was changed every 2 days,then mechanical tension was applied to the constructsfor 30min with 1Hz frequency and 1000 mstr strain.Each construct was loaded in the presence of 5mlmedia. Following loading the constructs were leftundisturbed for 24 h prior to cell harvesting.

2.4. Histological preparation and analysis

Cell-scaffold constructs were prepared either for waxembedding and light microscopy or for gross staining ofalkaline phosphatase (AP) and collagen using a SigmaAP diagnostic kit and Masson’s trichrome kit, respec-tively. In some cases, the constructs were cut into twoparts through a central line, allowing observation of thecell growth internally within the porous scaffolds.For the histological analysis, 1� 106 cells were seeded

in square and cylinder shape scaffolds and culturedstatically for 10 days and 3 weeks, respectively. The cell-scaffold constructs were fixed in 10% buffered formalinovernight. Standard dehydration from 50% to 100%ethanol was performed followed by immersion in Histo-clearTM (Sigma, UK), then molten paraffin was infil-trated into the constructs in a vacuum oven for 2 h. Theconstruct blocks were sectioned at 5 mm and stainedeither by hematoxylin and eosin for visualisation of cellsor by Masson’s Trichrome staining kit (Sigma, UK) fordemonstration of collagen deposition.

2.5. Biochemical analysis after mechanical loading

In some cases, cells were extracted from the scaffoldswith a 0.05% Triton solution and analysed for APactivity, DNA and protein content. The total proteinwas analysed by the Lowry method using a Sigmaprotein assay kit. The cell number in the constructs wasdetermined by DNA analysis via fluorescence spectro-scope using Hoechst 33258 dye, an assay adapted fromLabarca [20]. AP activity was measured using the APdiagnostics assay kit (Sigma, UK). Cell viability in

Y. Yang et al. / Biomaterials 23 (2002) 2119–2126 2121

response to loading was measured by lactate dehydro-genase (LDH) activity in the media using a Sigmadiagnostic kit (LDH/LD, Sigma, UK).

2.6. Protein expression by western blotting method

Cells were released from the scaffolds by adding0.05% Triton in PBS for 5min. The samples were thenpooled and added to sodium dodecyl sulphate (SDS)reducing buffer (50% v/v distilled water, 0.5m Tris HClpH 6.8, 10% v/v glycerol, 10% w/v SDS, 5% v/v 2-b-mercaptoethanol and 0.05% w/v bromophenol blue)and boiled for 4min at 971C. Protein samples wereloaded onto a 10% SDS-PAGE slab gel using adiscontinuous Laemmli buffer system running for 2 h.The gel was transferred onto a Hybond ECL transfermembrane (Amersham, UK), which was probed withcollagen type I (Chemicon) and the housekeepingprotein, vinculin.Enhanced ChemiLuminescence (ECL) Western blot-

ting protocols were used (Amersham, UK) to detectimmobilised specific antigens conjugated with horse-radish peroxidase-labelled antibodies. Collagen type Iand vinculin were detected with rabbit polyclonalantihuman antibodies (1:600 and 1:400 dilution, respec-tively). Blots were treated with Supersignal detectionsolutions (Pierce, USA) and exposed to Hyperfilm ECLautoradiography film (Amersham UK). Developed filmswere analysed by densitometry of the collagen bandsproduced in relation to vinculin using a softwarepackage of Phoretix.

2.7. Statistics

All values were expressed as mean7standard error.Samples were compared using a t-test and significancewas indicated at the 5% probability level.

3. Results

3.1. Bone cell growth in the PLLA scaffolds with andwithout agonist and collagen coating

Under the light microscope, the PLLA scaffolddemonstrated a porous structure, ranging from 250 to350 mm. The pores were interconnected, randomlydispersed and were homogeneously distributed through-out the scaffold, an important feature enabling themigration of cells and diffusion of nutrients throughoutthe scaffold. The scaffolds had a high mechanicalstrength with the compression strength greater than0.1MPa. Mechanical compression applied to the un-seeded scaffold with approximately 1% strain wasreversible. Coating of the construct with Collagen type

I or incorporation of the Bay agonist did not affect thecompressive strength of the construct.The human bone cells adhered to and proliferated on

the PLLA scaffold; however, cells were more adherentand able to proliferate more widely on the collagencoated scaffolds. Figs. 1a and b show the surface of theconstruct seeded with human bone cells on PLLA alone(Fig. 1a), and collagen coated PLLA scaffolds (Fig. 1b).The addition of the calcium channel agonist to thePLLA scaffold did not affect cell proliferation, as shownin Fig. 2. MG63 bone cells proliferated and spread wellin the PLLA scaffolds with (Fig. 2b) and without(Fig. 2a) the calcium channel agonist Bay. Fig. 3a showsa surface image of the mid-region of the scaffold aftercross-section. Bone cells proliferated across the surfaceof the scaffold but also penetrated deep within thescaffold. Fig. 3b shows a section through the scaffoldseeded with MG63 cells cultured for 3 weeks. Thesections of cell-scaffolds constructs were stained with

Fig. 1. Surface light micrographs of primary bone cells cultured

statically on PLLA scaffolds for 10 days, showing the effect of surface

treatment on cell proliferation. The bone cell seeding density was

1.0� 106. (a) The PLLA scaffold without surface treatment. (b) The

PLLA scaffold coated with collagen type I. The dark areas depict AP

stained osteoblasts. The lighter areas denote the scaffolds, whilst the

white areas are caused by reflection of light from the water on the

sample’s surface.

Y. Yang et al. / Biomaterials 23 (2002) 2119–21262122

Masson’s Trichrome. There were some evidence ofmatrix production surrounding the cells and addition ofb glycerophosphate and ascorbic acid to the mediaenhanced the production of bone matrix (data notshown). Fig. 3b demonstrates how the cells grew withinthe pores although cell density varied throughoutthe scaffold and between scaffolds. No evidence ofnutrient deficiency was observed. Assay of the mediafor released LDH showed trace levels below 4 BBunits/ml/h, which did not vary between samples withand without load. This further indicates the viability ofcells grown with the scaffolds over the incubationperiod.

3.2. Effects of mechanical load applied to the constructs

The bone cell-seeded constructs were cultured for 1week and attached to glass coverslips. Initial seedingdensity was 1� 105 as described in materials and

methods. The constructs/coverslips were then placedinto the four-point bending system and loaded for30min at 1Hz at approximately 1000 mstr strain [19].After further culture for 24 h, the cells were harvestedand the level of AP, total protein and DNA(Figs. 4a and b) were assessed. With the present coatingmethod, the Bay concentration released from theconstruct to the media has been confirmed to be withinan effective range during the 2 weeks of incubation. Theresults are expressed in terms of the ratio of load tocontrol for scaffolds with and without Bay encapsula-tion. The levels of AP expressed as a percentage of thetotal protein were significantly elevated (po0:05) inresponse to load. This load-related elevation was furtherenhanced significantly by incorporation of the Bayagonist within the scaffold (Fig. 4a) (po0:05). Incontrast, the levels of DNA were not affected by theaddition of Bay to the scaffolds (Fig. 4b).Collagen type I levels in the loaded constructs were

analysed by Western blotting using vinculin as house-keeping protein. Fig. 5 shows the ratio of collagen levels

Fig. 2. Surface light micrographs of MG-63 bone cells cultured

statically on PLLA scaffolds for 3 weeks, showing the effect of the

calcium channel agonist, BAY K8644 on cell proliferation. The cell

seeding density was 1.0� 106. (a) Bone cells seeded on the PLLA

scaffold coated with collagen type I. (b) Bone cells seeded on the PLLA

scaffold coated with collagen type I and incorporated BAY K8644.

The dark spots depict AP stained osteoblasts. The lighter areas denote

the scaffolds, whilst the white areas are caused by reflection of light

from the water on the sample’s surface.

Fig. 3. (a) Inner surface light micrograph of MG63 bone cell cultured

statically on a PLLA scaffold for 3 weeks, demonstrating penetration

of cells into the construct pores. The construct is bisected to expose the

inner surface. (b) A section light micrograph of the same cultured cell-

scaffold construct as above. Note the cells are localised and

proliferating within the pores. The section has been stained with

Masson’s Trichrome.

Y. Yang et al. / Biomaterials 23 (2002) 2119–2126 2123

in loaded constructs versus controls and the elevation incollagen expression in response to load. This load-related elevation in collagen expression is enhanced in

the Bay encapsulated scaffolds subjected to load, but notsignificantly on the 24 h period post load.

4. Discussion

The use of biodegradable polymers in bone tissueengineering may have advantages over some of the non-degrading materials which have been used extensively torepair skeletal defects [21]. To date, the majority ofstudies have investigated acellular systems with variablepore size and degradation rates in vivo [22]. In addition,PLLA type materials have been investigated for the useof cell-based strategies either as scaffolds for celltransplantation or for guided regeneration of nativeosteogenic cells [23]. Optimisation of the seeding densityand culture time in vitro prior to implantation hasdemonstrated the potential for these materials inapplications of bone tissue engineering [24]. In all thesecases, there is a clear requirement for the implantmaterials to be mechanically appropriate for theirenvironment or to degrade and enable the replacementof matrix tissue which is mechanically appropriate to theimplant site. This mechanical integrity acts in oppositionto rapid degradation times often proposed for tissueengineering. It is with this in mind that we set out toinvestigate whether we can harness the mechanicalenvironment and amplify the mechano-signalling inour mechano-active scaffolds.In this paper, we establish proof of concept that we

can manipulate the mechanical responses of bone cellswithin a PLLA scaffold by encapsulating an agonist forVOCCs which have been shown to be implicated as anearly critical step in the load pathway. AP is animportant component of mineralised matrix productsin bone [25]. Using an encapsulated release method, wecan significantly enhance AP production and elevate thematrix, collagen type I, production within cells seeded ina PLLA scaffold exposed to a short duration of cyclicalloading. Incorporation of inductive components toPLLA/PGA type scaffolds has been proposed for anumber of applications in tissue engineering. Incorpora-tion of vascular endothelial growth factor has beenproposed as a method for the induction of rapidvascular ingrowth during bone development [26]. Anumber of other chemical and bioactive agents havebeen proposed for delivery as therapeutics in slowrelease PLLA type membranes such as heparin, trypsininhibitor, insulin, etc. [27–29]. Incorporation of Bay intoPLLA type scaffolds follows these principles closely butthe purpose of using this type of agent is novel. Acriticism which is often levelled at continuous releasebioactive materials is that the agent is always presentwith time and there is no means of controlling the levelsto which the cells are exposed with time. In the case ofBay, the agonist is continuously released but only acts

0

0.5

1

1.5

2

2.5

PLLA/C PLLA/C/BAY

Rat

io lo

ad/c

ontr

ol

*

0

0.4

0.8

1.2

PLLA/C PLLA/C/Bay

Rat

io lo

ad/c

ontr

ol

(a)

(b)

Fig. 4. (a) Levels of AP in relation to total protein in the PLLA

scaffolds with and without BAY K8644 and mechanical loading

expressed as a ratio of load to control. (b) Concentration of DNA

present in the scaffolds with and without BAY K8644 and mechanical

loading expressed as a ratio of load to control. The primary bone cells-

scaffold constructs were statically cultured for 1 week and loaded for

30min in the four-point bending system. The bars indicate the mean

values and standard error (n ¼ 3) and * denotes a statistically

significant difference (po0:05) in the loading between samples with

and without Bay encapsulation. The scaffold type is shown on x-axis.

PLLA/C: collagen coated PLLA; PLLA/C/Bay: collagen coated

PLLA, encapsulated with Bay.

0

0.5

1

1.5

2

2.5

PLLA/C PLLA/C/Bay

Rat

io lo

ad/c

ontr

ol

Fig. 5. Levels of collagen type I expressed as a proportion of levels of

vinculin analysed by Western blotting expressed as a ratio of load to

control. The primary bone cells-scaffold constructs were statically

cultured for 1 week and loaded for 30min in the four-point bending

system. The bars indicate the mean values and standard error (n ¼ 3).

The scaffold type is shown on x-axis. PLLA/C: collagen coated PLLA;

PLLA/C/Bay: collagen coated PLLA, encapsulated with Bay.

Y. Yang et al. / Biomaterials 23 (2002) 2119–21262124

when the calcium channels are open in response to themechanical load. It is also possible that the mechanicalforces on the scaffold will increase the release of Bayfrom the PLLA scaffold. Hence, the agent will act toamplify the load signal preferentially on the cells whichare exposed to load rather than on those which are not.This infers that the amplification will occur specificallytailored to the load environment to which the implant issubjected.Membrane ion channels have been proposed as one of

the early critical activation steps in response tomechanical loading in a number of cell types [30,3,5].Calcium channels both voltage-activated and mechano-active have attracted particular interest due to the roleof calcium as a principal secondary messenger. Electro-physiological investigations have highlighted the pre-sence of plasma membrane calcium channels in manybone cell types. In particular, VOCCs have beendetected in primary cultures of rat calvarial, osteo-blast-like cells [31], human osteoblasts [4], chickenosteoclasts [32], rat bone marrow stromal cells [6,33]and in virtually all osteoblast-like cell lines that havebeen studied [34,35]. The role of Ca2+ signalling in theload response has been widely noted, with increases inintracellular free calcium (IFC) found to be due to eitherrelease of intracellular Ca2+ stores or opening of Ca2+

channels in the plasma membrane [4]. Such increases inIFC have been experimentally induced by short periodsof cyclical mechanical loading of rat periosteal-derivedosteoblasts [5]. Calcium influx through a specific voltageoperated (l-type) calcium channel has been demon-strated to be involved in the up-regulation of the bone-matrix protein osteopontin [5]. Manipulation of thesechannels as suggested in our studies presents a novelmeans for influencing the load signal to the cell.It is clear that mechano-transduction is a complex

cascade of temporal responses. Our results do notexclude the involvement of other mediators of load suchas cytoskeletal and integrin components in the mem-brane [36]. Application of mechanical strain to bonetissue in vivo and bone-forming cells in vitro has beenshown to increase the concentrations of a variety ofsecond messengers in addition to calcium messengers,including cAMP, cGMP [37,38] and IP3 [39] to stimulateactivity of phospholipases and nitric oxide [38,40,41],and to enhance release and synthesis of prostaglandins[37,42–48]. Our concept sets out to enhance one step inthis pathway which may then potentially furtherinfluence the cascade of responses which are involved.The production of implant biomaterials in the past

has focused upon the production of stable new bonetissue by the incorporation of direct modulators of boneformation and turnover in a non-loaded model. How-ever, the production of load-bearing new bone may beaided by the development of biomaterials where theinteractions between tissues and implants have also been

defined under mechanical loading conditions, in parti-cular the incorporation of chemical and biologicalagents into the biomaterial which may enhance the loadresponse in tissue surrounding the implant. Ourpreliminary studies to produce such a ‘mechano-active’biomaterial as reported in this study with the incorpora-tion of modulators of the calcium signalling pathwayinto a new biomaterial have successfully achieved andenhanced effects of the ‘mechano-active’ scaffolds on cellresponses under the loading conditions observed.Further work is now in progress to observe effects overlonger incubation period and to determine the orienta-tion of the matrix within the loaded cell-seededconstructs.

Acknowledgements

This project is financially supported by EuropeanCommission Vth framework programme, BITES,QLRT-1999-00559.

References

[1] Rubin CT, Lanyon LE. Regulation of bone formation by applied

dynamic loads. J Bone Jt Surg Am 1984;66:397–402.

[2] Leblanc AD, Schneider VS, Evans HJ, Englebertson DA, Krebs

JM. Bone mineral loss and recovery after 17 weeks of bed rest.

J Bone Miner Res 1990;5:843–50.

[3] El Haj AJ, Walker LM, Preston MR, Publicover SJ. Mechano-

transduction pathways in bone: calcium fluxes and the role of

voltage-operated calcium channels. Med Biol Eng Comp 1999;

37:403–9.

[4] Walker LM, Holm A, Cooling L, Maxwell L, Oberg A, Sundqvist

T, El Haj AJ. Mechanical manipulation of bone and cartilage cells

with ‘optical tweezers’. FEBS Lett 1999;459:39–42.

[5] Walker LM, Publicover SJ, Preston MR, Said Ahmed MA, El

Haj AJ. Calcium-channel activation and matrix protein upregula-

tion in bone cells in response to mechanical strain. J Cell Biochem

2000;79:648–61.

[6] Preston MR, El Haj AJ, Publicover SJ. Expression of voltage-

operated Ca2+ channels in rat bone marrow stromal cells in vitro.

Bone 1996;19:101–6.

[7] Publicover SJ, Preston MR, El Haj AJ. Voltage-dependent

potentiation of low-voltage activated Ca2+ channel currents in

cultured rat bone marrow cells. J Physiol 1995;489:649–61.

[8] Duncan RL, Hruska KA, Misler S. Parathyroid hormone

activation of stretch-activated cation channels in osteosarcoma

cells (UMR106.01). FEBS Lett 1992;307:219–23.

[9] Weinreb M, Shinar D, Rodan GA. Different pattern of alkaline

phosphatase, osteopontin, and osteocalcin expression in develop-

ing rat bone visualized by in situ hybridization. J Bone Miner Res

1990;5:831–42.

[10] El Haj AJ, Minter SL, Rawlinson SC, Suswillo R, Lanyon LE.

Cellular responses to mechanical loading in vitro. J Bone Miner

Res 1990;5:923–32.

[11] Beck LS, Amento EP, Xu Y, Deguzman L, Lee WP, Nguyen T,

Gellet NA. TGF-beta 1 induces bone closure of skull defects:

temporal dynamics of bone formation in defects exposed to

rhTGF-beta 1. J Bone Miner Res 1993;8:753–61.

Y. Yang et al. / Biomaterials 23 (2002) 2119–2126 2125

[12] Wozney JM. The bone morphogenetic protein family and

osteogenesis. Mol Reprod Dev 1992;32:160–7.

[13] Takuwa Y, Ohse C, Wang EA, Wozney JM, Yamashita K. Bone

morphogenetic protein-2 stimulates alkaline phosphatase activity

and collagen synthesis in cultured osteoblastic cells, MC3T3-E1.

Biochem Biophys Res Commun 1991;174:96–101.

[14] Sampath TK, Maliakal JC, Hauschka PV, Jones WK, Sasak H,

Tucker RF, White KH, Coughlin JE, Tucker MM, Pang RHL,

Corbett C, Ozkaynak E, Oppermann H, Rueger DC. Recombi-

nant human osteogenic protein (hOP)-1 induces new bone

formation in vivo with a specific activity comparable with natural

bovine osteogenic protein and stimulates osteoblast proliferation

and differentiation in vitro. J Biol Chem 1992;267:20352–62.

[15] Whang K, Tsai DC, Nam EK, Aitken M, Sprague SM, Patel PK,

Healy KE. Ectopic bone formation via rhBMP-2 delivery from

porous bioabsorbable polymer scaffolds. J Biomed Mater Res

1998;42:491–9.

[16] Bessho K. Ectopic osteoinductive difference between purified

bovine and recombinant human bone morphogenetic protein. In:

Lindholm TS, editor. Bone morphogenetic proteins: biology,

biochemistry and reconstructive surgery. Geogetown, TX: Landes

RG Co., 1996. p. 105–11.

[17] Satoh H, Katoh H, Velez P, Fill M, Bers DM. Bay K8644

increases resting Ca2+ spark frequency in ferret ventricular

myocytes independent of Ca influx: contrast with caffeine and

ryanodine effects. Circ Res 1998;83:1192–204.

[18] Thomson RC, Shung AK, Yaszemski MJ. Polymer scaffold

processing. In: Lanza RP, Langer R, Vacanti J, editors. Principles

of tissue engineering. USA: Academic Press, 1997. p. 251–62.

[19] Peake MA, Cooling LM, Magnay JL, Thomas PBM, El Haj AJ.

Cellular responses to mechanical stress: selected contribution:

regulatory pathways involved in mechanical induction of c-fos

gene expression in bone cells. J Appl Physiol 2000;89:2498–507.

[20] Labarca C, Paigen K. A simple, rapid and sensitive DNA assay

procedure. Anal Biochem 1980;102:344–52.

[21] Burg KJL, Porter S, Kellam JF. Biomaterial development for

bone tissue engineeering. Biomaterials 2000;21:2347–59.

[22] Meinig RP, Buesing CM, Helm J, Gogolewski S. Regeneration of

diaphyseal bone defects using resorbable poly(l/dl lactide) and

poly(d-lactide) membranes in the Yucatan pig model. J Orthop

Trauma 1997;11:551–8.

[23] Putnam AJ, Mooney DJ. Tissue engineering using synthetic

extracellular matrices. Nat Med 1996;2:824–6.

[24] Holy CE, Shoichet MS, Davies JE. Engineering three-dimensional

bone tissue in vitro using biodegradable scaffolds: investigating

initial cell-seeding density and culture period. J Biomed Mater

Res 2000;51:376–82.

[25] Pavlin D, Dove SB, Zadro R, Gluhak-Heinrich J. Mechanical

loading stimulates differentiation of periodontal osteoblasts in a

mouse osteoinduction model: effect on type I collagen and

alkaline phosphatase genes. Calcif Tissue Int 2000;67:163–72.

[26] Murphy WL, Peters MC, Kohn DH, Mooney DJ. Sustained

release of vascular endothelial growth factor from mineralized

poly(lactide-co-glycolide) scaffolds for tissue engineering. Bioma-

terials 2000;21:2521–7.

[27] Kreitz MR, Webber WL, Galleti PM, Mathiowitz E. Controlled

delivery of therapeutic from microporous membranes: II In vitro

degradation and release of heparin loaded poly(dl-lactide-co-

glycolide). Biomaterials 1997;18:597–603.

[28] Athanasiou KA, Singhal AR, Agrawal CM, Boyan BD. In vitro

degradation and release characteristics of biodegradable

implants containing trypsin inhibitor. Clin Orthop 1995;

315:272–81.

[29] Kwong AK, Chou S, Sun AM, Sefton MV, Goosen MFA.

In vitro and in vivo release of insulin from poly(lactic acid)

microbeads and pellets. J control Release 1986;4:47–62.

[30] Sacchs F, Morris CE. Mechanosensitive ion channels in

nonspecialized cells. Rev Physiol Biochem Pharmacol 1998;

132:1–77.

[31] Chesnoy-Marchais D, Fritsch J. Concentration-dependent mod-

ulations of potassium and calcium currents of rat osteoblastic cells

by arachidonic acid. J Membr Biol 1994;138:159–70.

[32] Miyauchi A, Hruska KA, Greenfield EM, Duncan R, Alvarez J,

Barattolo R, Colucci S, Zambonin-Zallone A, Teitelbaum SL,

Teti A. Osteoclast cytosolic calcium, regulated by voltage-gated

calcium channels and extracellular calcium, controls podosome

assembly and bone resorption. J Cell Biol 1990;111:2543–52.

[33] Publicover SL, Thomas GP, El Haj AJ. Induction of a low

voltage-activated, fast-inactivating Ca2+ channel in cultured bone

marrow stromal cells by dexamethasone. Calcif Tissue Int

1994;54:125–32.

[34] Duncan R, Misler S. Voltage-activated and stretch-activated

Ba2+ conducting channels in an osteoblast-like cell line (UMR

106). FEBS Lett 1989;251:17–21.

[35] Morain P, Peglion JL, Giesen-Crouse E. Ca2+ channel inhibition

in a rat osteoblast-like cell line, UMR 106, by a new

dihydropyridine derivative, S11568. Eur J Pharmacol 1992;

220:11–7.

[36] Ingber D. Integrin as mechanochemical transducers. Curr Opin

Cell Biol 1991;3:841–8.

[37] Rodan GA, Bourret LA, Harvey A, Mansi T. cAMP and cGMP

mediators of the effects on bone remodeling. Science 1975;

189:467–9.

[38] Binderman I, Zor U, Kaye AM, Shimshoni Z, Harell A, Somjen

D. The transduction of mechanical force into biochemical events

in bone cells may involve activation of phospholipase A2. Calcif

Tissue Int 1988;42:261–6.

[39] Brighton CT, Sennett B, Farmer JC, Iannotti JP, Hansen CA,

Williams JL, Williamson J. The inositol phosphate pathway as a

mediator in the proliferative response of rat calvarial bone cells to

cyclical biaxial strain. J Orthop Res 1992;10:385–93.

[40] Pitsillides AA, Rawlinson SC, Suswillo RF, Bourrin S, Zaman G,

Lanyon LE. Mechanical strain-induced NO production by bone

cells: a possible role in adaptive bone (re)modeling? FASEB J

1995;9:1614–22.

[41] Jones DB, Nolte H, Scholubbers JG, Turner E, Velter D.

Biochemical signal transduction of mechanical strain in osteo-

blast-like cells. Biomaterial 1991;12:101–10.

[42] Somjen D, Binderman I, Berger E, Harell A. Bone remodelling

induced by physical stress is prostaglandin E2 mediated. Biochem

Biophys Acta 1980;627:91–100.

[43] Rawlinson SC, El Haj AJ, Minter SL, Tavares IA, Bennett A,

Lanyon LE. Loading-related increases in prostaglandin produc-

tion in cores of adult canine cancellous bone in vitro: a role for

prostacyclin in adaptive bone remodeling? J Bone Miner Res

1991;6:1345–51.

[44] Chow JW, Chambers TJ. Indomethacin has distinct early and late

actions on bone formation induced by mechanical stimulation.

Am J Physiol 1994;267:E287–92.

[45] Thomas GP, El Haj AJ. Bone marrow stromal cells are load

responsive in vitro. Calcif Tissue Int 1996;58:101–8.

[46] Zaman G, Suswillo RF, Cheng MZ, Tavares IA, Lanyon LE.

Early responses to dynamic strain change and prostaglandins in

bone-derived cells in culture. J Bone Miner Res 1997;12:769–77.

[47] Klein-Nulend J, Semeins CM, Burger EH. Prostaglandin

mediated modulation of transforming growth factor-beta meta-

bolism in primary mouse osteoblastic cells in vitro. J Cell Physiol

1996;168:1–7.

[48] Klein-Nulend J, Burger EH, Semeins CM, Raisz LG, Pilbeam

CC. Pulsating fluid flow stimulates prostaglandin release and

inducible prostaglandin G/H synthase mRNA expression in

primary mouse bone cells. J Bone Miner Res 1997;12:45–51.

Y. Yang et al. / Biomaterials 23 (2002) 2119–21262126