extruded collagen fibres for tissue-engineering applications: influence of collagen concentration...

This article was downloaded by: [McGill University Library]On: 14 November 2014, At: 10:28Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office:Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Biomaterials Science, PolymerEditionPublication details, including instructions for authors and subscriptioninformation:http://www.tandfonline.com/loi/tbsp20

Extruded Collagen Fibres for Tissue-Engineering Applications: Influence ofCollagen Concentration and NaCl AmountD. I. Zeugolis a , R. G. Paul b & G. Attenburrow ca Tissue Modulation Laboratory, National University of Singapore TissueEngineering Programme, National University of Singapore, 117510Singapore; Division of Bioengineering, Faculty of Engineering, NationalUniversity of Singapore, 117576 Singapore; Immunology Programme,Department of Microbiology, Yong Loo Lin School of Medicine, NationalUniversity of Singapore, 117456 Singaporeb Devro Plc, Glasgow G69 0JE, UKc School of Applied Sciences, The University of Northampton,Northampton NN2 7AL, UKPublished online: 02 Apr 2012.

To cite this article: D. I. Zeugolis , R. G. Paul & G. Attenburrow (2009) Extruded Collagen Fibres forTissue-Engineering Applications: Influence of Collagen Concentration and NaCl Amount, Journal ofBiomaterials Science, Polymer Edition, 20:2, 219-234, DOI: 10.1163/156856209X404505

To link to this article: http://dx.doi.org/10.1163/156856209X404505

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis, ouragents, and our licensors make no representations or warranties whatsoever as to theaccuracy, completeness, or suitability for any purpose of the Content. Any opinions and viewsexpressed in this publication are the opinions and views of the authors, and are not the viewsof or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shallnot be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages,and other liabilities whatsoever or howsoever caused arising directly or indirectly in connectionwith, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantialor systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or

http://www.tandfonline.com/loi/tbsp20

http://www.tandfonline.com/action/showCitFormats?doi=10.1163/156856209X404505

http://dx.doi.org/10.1163/156856209X404505

distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and usecan be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014

http://www.tandfonline.com/page/terms-and-conditions

Journal of Biomaterials Science 20 (2009) 219–234www.brill.nl/jbs

Extruded Collagen Fibres for Tissue-EngineeringApplications: Influence of Collagen Concentration

and NaCl Amount

D. I. Zeugolis a,b,c,∗ , R. G. Paul d and G. Attenburrow e

a Tissue Modulation Laboratory, National University of Singapore Tissue Engineering Programme,National University of Singapore, 117510 Singapore

b Division of Bioengineering, Faculty of Engineering, National University of Singapore,117576 Singapore

c Immunology Programme, Department of Microbiology, Yong Loo Lin School of Medicine,National University of Singapore, 117456 Singapore

d Devro Plc, Glasgow G69 0JE, UKe School of Applied Sciences, The University of Northampton, Northampton NN2 7AL, UK

Received 4 October 2007; accepted 1 February 2008

AbstractExtruded collagen fibres have been shown to be a competitive biomaterial for tissue-engineering applica-tions. Since different tissues are coming in different textures, as far as it is concerned their fibre diameterand consequently their mechanical properties, herein we aim to investigate the influence of the collagenconcentration and the amount of NaCl on the properties of these fibres. Scanning electron microscopy studyrevealed that the substructure of the collagen fibres was the same, regardless of the treatment. The ther-mal properties were found to be independent of the collagen concentration or the amount of NaCl utilized(P > 0.05). An inversely proportional relationship between dry fibre diameter and stress at break was ob-served. Increasing the collagen concentration yielded fibres with significant higher diameter (P < 0.002),strain (P < 0.009) and force (P < 0.001) values, whilst the stress (P < 0.008) and modulus (P < 0.009)values were decreased. For the fabrication of fibres with reproducible properties, 20% NaCl was found tobe the optimum. Overall, reconstituted collagen fibres were produced with properties similar to native orsynthetic fibres to suit a wide range of tissue-engineering applications.© Koninklijke Brill NV, Leiden, 2009

KeywordsCollagen scaffold, structural characteristics, thermal properties, mechanical properties

* To whom correspondence should be addressed. Tel.: (65) 6516-5395; Fax: (65) 6776-5322; e-mail:[email protected]

© Koninklijke Brill NV, Leiden, 2009 DOI:10.1163/156856209X404505

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014

220 D. I. Zeugolis et al. / Journal of Biomaterials Science 20 (2009) 219–234

1. Introduction

The repair of tissue defects such as anterior cruciate ligament, skin and cartilageremains a challenging clinical problem. Thus, tissue engineering aims to designand develop materials that would interact with the body to encourage tissue re-pair. A number of synthetic and natural biomaterials are currently in use as tissuescaffolds, with the ideal ones to be those that most closely mimic the naturallyoccurring environment in the host tissue matrix [1–4]. Naturally-derived collagenhas demonstrated a number of desirable features, among which are its high tensilestrength, biodegradability and low immunogenicity, making it an excellent choiceas a biomaterial [5–9].

Previous studies on extruded collagen fibres have demonstrated that such three-dimensional biomaterials comprise an excellent scaffold for soft- and hard-tissuereplacement with structural, mechanical and thermal properties similar to the nativetissue [10–13]. Furthermore, the resultant fibres not only exhibit the characteris-tic for collagen ultrastructural axially periodicity that confirms quarter-staggeredsupramolecular assemblies, but also contain aligned collagen fibrils with diameterdistributions similar to native collagen fibres [14, 15]. In addition, their morpholog-ical characteristics have been shown to facilitate fibroblast migration and neotissueformation [16, 17].

While significant strides have been made, many challenges still exist in the en-gineering of these materials that must be addressed to bring these technologies toclinical practice. For example, it has been shown that different tissues are com-posed of individual collagen fibrils that can be greater than 500 µm in length andwith diameters ranging from 20 to 300 nm; such fibrils form supramolecular as-semblies (fibril bundles or fibres) with diameters between 1 and 300 µm [18–23].Thus, herein we ventured to investigate whether by varying the collagen concen-tration; we could control the properties of these fibres and consequently manu-facture biomimetic scaffolds with properties matching those of native tissues orsynthetic biomaterials currently in use. Furthermore, although NaCl has been uti-lized extensively for the fabrication of extruded collagen fibres [15, 24–27], it isstill unclear how it influences the properties of these fibres and whether there isan ideal amount of NaCl for producing reproducible extruded collagen fibres. Assuch, in this study we explored the influence of different amounts of NaCl on theproperties of these fibres. Moreover, the vast majority of researchers has utilisedacid-soluble rat tail tendon [15, 26] or acid-soluble bovine Achilles tendon [28,29] for the production of such fibres. However, in these in vitro preparations, thenon-helical peptides remain intact and a less biocompatible scaffold could be pro-duced [5, 9, 30]. Thus, herein we utilized pepsin-soluble bovine Achilles tendoncollagen, a raw material that has been favoured for biomimetic scaffold manufac-turing.

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014

D. I. Zeugolis et al. / Journal of Biomaterials Science 20 (2009) 219–234 221

2. Materials and Methods

2.1. Materials

All chemicals, unless otherwise stated, were purchased from Sigma-Aldrich UK.The bovine Achilles tendons (BAT) were kindly provided by the BLC ResearchCentre (Northampton, UK).

2.2. Preparation of Collagen

Typical protocols for the extraction and purification of collagen were employed[31]. Briefly, tendons were manually dissected out from the surrounding fascia,minced in ice and extensively washed with distilled water and neutral salt solu-tions. Subsequently they were suspended in 0.5 M ethanoic acid for 72 h at 4◦Cin the presence of pepsin (porcine gastric mucosa pepsin, 2500 U/mg; Roche Di-agnostics, UK); enzyme to tendon wet weight ratio 1:100. The suspension wasfiltered and purified by repeated (twice) salt precipitation (0.9 M NaCl), centrifu-gation (12 × 103 × g at 4◦C for 45 min) (Gr20.22 Jouan refrigerated centrifuge,Thermo Electron, Bath, UK) and re-suspension in 1 M ethanoic acid. The final col-lagen solutions were dialyzed (8000 MW cut off) against 0.01 M ethanoic acid andkept refrigerated until use. The atelocollagen solutions purity was determined bysodium dodecyl sulfate polyacrylamide gel electrophoresis analysis (90% type I)and their concentration was subsequent adjusted to 3, 6 and 7 mg/ml (determinedby hydroxyproline assay).

2.3. Fibre Formation

The experimental set up has been described in detail previously [31] and was basedon previous publications [24, 26, 29]. Briefly, a 5-ml syringe (Terumo Medical UK)containing either of the collagen solutions (3, 6 or 7 mg/ml) was loaded into a sy-ringe pump system (KD-Scientific 200, KD Scientific, USA) connected to siliconetubing (Samco Silicone Products, UK) of 30 cm in length and 1.5 mm in internaldiameter. The pump was set to infuse at 0.4 ml/min. One end of the tube was con-nected to the syringe pump with the other end placed at the bottom of a container.Collagen solution was extruded into a fibre formation buffer (FFB) comprising118 mM phosphate buffer and either 5, 20 or 40% NaCl at pH 7.55 and 37◦C.Fibres were allowed to remain in this buffer for a maximum period of 10 min, fol-lowed by further incubation for an additional 10 min in a fibre incubation buffer(FIB) comprising of 6.0 mM phosphate buffer and 75 mM NaCl at pH 7.10 and37◦C. Thereafter the fibres were transferred into a distilled water bath for a further10 min and consequently air-dried under the tension of their own weight at roomtemperature (22◦C). Finally the fibres were stored at 65% relative humidity at 22◦Cfor a minimum of 48 h, prior to any other experiment.

2.4. Mechanical Testing and Ultrastructural Matrix Analysis

Stress–strain curves of dry reconstituted collagen fibres were determined in uniax-ial tension using an Instron 1122 Universal testing machine (Instron, UK) operated

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


at an extension rate of 10 mm/min. The gauge length was fixed at 5 cm and softrubber was used to cover the inside area of the grips to avoid damaging the fibresat the contact points. Results obtained with fibres that broke at contact points withthe grips were rejected. The cross-sectional area of each fibre was calculated bymeasuring the diameter at five places (every 1 cm) along its longitudinal axis usinga Nikon Eclipse E600 optical microscope with a calibrated eyepiece (Nikon In-struments, UK). It was assumed that the fibres were circular for the cross-sectionalarea determinations. Fracture surfaces of collagen fibres that had been extended tofailure were examined using a Hitachi S3000N Variable Pressure Scanning Elec-tron Microscope (Hitachi, UK). The following definitions were used to calculatethe mechanical data: stress at break was defined as the load at failure divided bythe original cross-sectional area (engineering stress); strain at break was defined asthe increase in fibre length required to cause failure divided by the original lengthand modulus was defined as the stress at 0.02 strain divided by 0.02.

2.5. Thermal Properties

The shrinkage temperature was determined by Differential Scanning Calorimetry(DSC) using the 822e Mettler-Toledo differential scanning calorimeter (Mettler-Toledo International, UK). The reconstituted collagen fibres were hydrated over-night at room temperature in 0.01 M PBS at pH 7.4. The wet fibres were removedand quickly blotted with filter paper to remove excess surface water and hermet-ically sealed in aluminium pans. Heating was carried out at a constant tempera-ture ramp of 5◦C/min in the temperature range from 15 to 95◦C, with an emptyaluminium pan as the reference probe. Thermal denaturation, an endothermic tran-sition, was recorded as a typical peak, and two characteristic temperatures weremeasured corresponding to the peak (the temperature of maximum power absorp-tion during denaturation) and onset (the temperature at which the tangent to theinitial power versus temperature line crosses the baseline) temperatures.

2.6. Statistical Analysis

Numerical data are expressed as mean ± SD. Analysis was performed using statis-tical software (MINITABTM version 13.1; Minitab). One-way analysis of variance(ANOVA) for multiple comparisons and two-sample t-test for pair-wise compar-isons were employed after confirming the following assumptions: (i) the distribu-tion from which each of the samples was derived was normal (Anderson–Darlingnormality test); and (ii) the variances of the population of the samples were equalto one another (Bartlett’s and Levene’s tests for homogenicity of variance). Non-parametric statistics were utilized when either or both of the above assumptionswere violated and consequently a Kruskal–Wallis test for multiple comparisons orMann–Whitney test for two-samples were carried out. Statistical significance wasaccepted at P < 0.05.

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


3. Results

3.1. Matrix Morphology

Figure 1 provides the surface and fracture morphology of the extruded collagenfibres produced in this study using scanning electron microscopy. Ridges andcrevices were identified to run roughly parallel to the axis of the fibres, whilst a“cavity” lying along their longitudinal axis was observed; a characteristic that wasmore obvious in fibres with larger diameter. Microscopic investigation of the failedends revealed a homogenous internal structure with closely packed inter-fibre space.

3.2. Thermal Analysis

Calorimetric analysis using Differential Scanning Calorimetry allowed measure-ments of the thermal stability of the fibres produced in this study (Table 1). Theresults obtained indicate that the hydrothermal stability of the fibres produced,was independent of the collagen concentration and the amount of NaCl utilized(P > 0.05).

Figure 1. (a) Surface morphology and (b) internal structure of extruded collagen fibres.

Table 1.Denaturation temperatures of rehydrated extruded collagen fibres de-rived from different collagen concentration or amount of NaCl (n = 3)

Treatment Onset ± SD (◦C) Peak ± SD (◦C)

20% NaCl 3 mg/ml 46.66 ± 0.85 49.06 ± 1.116 mg/ml 46.41 ± 0.86 50.10 ± 0.948 mg/ml 47.58 ± 0.10 50.72 ± 0.15

6 mg/ml 5% NaCl 48.74 ± 0.58 51.79 ± 0.6220% NaCl 46.41 ± 0.86 50.10 ± 0.9440% NaCl 46.18 ± 0.04 50.53 ± 0.18

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


3.3. Biomechanical Analysis

Overall, dry collagen fibres exhibited similar shaped stress–strain curves (Fig. 2),independent of the treatment, which consisted of a small toe region, a region ofsteeply rising stress up to a knee region, where the gradient of the curve reduced,followed by a long region of constant gradient up to the point of fracture. Thestress–strain curves showed a diameter-dependent variation; fibres of low diametershowed high stress/low strain graphs, while fibres of high diameter demonstratedlow stress/high strain graphs. Fibres of higher diameter exhibited a longer toe re-gion.

3.3.1. Effect of % NaClIn this set of experiments the effect of NaCl amount for the reconstitution of colla-gen into fibres was investigated. The mechanical properties of the fibres producedare summarized in Table 2. The fibres produced at 5% NaCl were the least ex-tendable (P < 0.001). At 20% salt content, the smallest dry diameter (P < 0.001)

Figure 2. Characteristic stress–strain curve of dry extruded collagen fibres for dry collagenous struc-tures.

Table 2.Physical and mechanical properties of reconstituted collagen fibres as a factor of NaCl content in thefiber formation buffer

Treatment Dry Stress at Strain at Force at Modulus at 2%diameter ± SD break ± SD break ± SD break ± SD strain ± SD(µm) (MPa) (N) (GPa)

5% NaCl 142 ± 17 135 ± 59 0.17 ± 0.06 2.02 ± 0.60 0.9 ± 0.5(n = 17)20% NaCl 119 ± 15 208 ± 57 0.27 ± 0.05 2.23 ± 0.32 1.6 ± 0.5(n = 25)40% NaCl 144 ± 20 142 ± 41 0.24 ± 0.07 2.23 ± 0.47 1.2 ± 0.7(n = 16)

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


and the highest stress at break values (P < 0.001) were obtained. A significant in-crease in fibre diameter (P < 0.001) and a significant decrease in stress at break(P < 0.001) values were observed from the 20 to 40% NaCl content. Between theNaCl treatments, the stiffest fibres were produced at 20% co-agent (P < 0.002).In all cases, fibres of no significant difference in force at break were obtained(P > 0.05).

Fitting a linear regression model between stress at break and dry fibre diameterfor every treatment (Fig. 3), the strongest correlation was obtained for the 20%NaCl amount (R2 values of 0.56, 0.72 and 0.51 for 5, 20 and 40% NaCl content,respectively).

3.3.2. Effect of Collagen ConcentrationTable 3 summarizes the mechanical properties of extruded collagen fibres derivedfrom PSBAT collagen of 3, 6 and 7 mg/ml collagen solution concentration. It wasfound that by increasing the collagen concentration, the fibre diameter was in-creased (P < 0.002). Fibres of low concentration and consequent low diameter,required lower forces for breaking (P < 0.001) and exhibited decreased strain atbreak values (P < 0.009). However, fibres of low diameter exhibited high stress atbreak (P < 0.008) and modulus values (P < 0.009).

Fitting a linear regression model between stress at break and dry fibre diameterfor every treatment (Fig. 4) strong correlations were obtained (R2 values of 0.88,0.72 and 0.77 for 3, 6 and 7 mg/ml collagen concentration, respectively).

4. Discussion

4.1. Structural Evaluation

The extruded collagen fibres exhibited a rough surface with undulations along theirlength and with ridges and crevices running roughly parallel to the longitudinal axis.Similar observations have been attributed to the handling and/or to the substructureof the fibres [31–33]. Such unique morphological characteristics have been shownto facilitate and promote cell attachment and fibroblast migration [16, 17]. A con-sistent closely packed inter-fibre structure was observed for the fibres studied inthis work and was independent of the collagen concentration or the NaCl amount.This observation is in agreement with previous publications, where other co-agentswere used [34, 35]. It has also been mentioned that very little free inter-fibre spaceoccurs in fibres formed in vitro [36]. The fracture mechanisms identified are consis-tent with previous publications, where they have been explained in detail [23, 27,37–39].

It is worth pointing out that occasionally air-bubbles were trapped within thefibrous structure. Such fibres, when tested to failure, tended to break at the sideof the bubble, before they were elongated appreciably. Thus, for a successful fibreproduction a completely degassed collagen solution and fibre formation buffer isrequired as has been reported before [29].

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


Figure 3. Fitting a linear regression model between stress at break and dry fibre diameter, the strongestcorrelation was obtained from the 20% NaCl in the FFB.

4.2. Thermal Analysis

When collagen in the hydrated state is heated, the helix–coil transition takes place,during which the triple helix melts and progressively dissociates into the three

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


Table 3.Physical and mechanical properties of dry extruded collagen fibres as a factor of atelocollagen con-centration

Treatment Dry Stress at Strain at Force at Modulus at 2%(20% NaCl) diameter ± SD break ± SD break ± SD break ± SD strain ± SD

(µm) (MPa) (N) (GPa)

3 mg/ml (n = 12) 77 ± 10 237 ± 104 0.23 ± 0.04 1.03 ± 0.18 1.8 ± 0.86 mg/ml (n = 25) 119 ± 15 208 ± 57 0.27 ± 0.05 2.23 ± 0.32 1.6 ± 0.57 mg/ml (n = 15) 161 ± 19 130 ± 31 0.28 ± 0.05 2.50 ± 0.18 0.9 ± 0.3

randomly coiled peptide α-chains (gelatin). The rigid triple helical molecules de-nature over a narrow range of temperatures, the mid-point of which is referred toas denaturation temperature [22, 40–45]. The fibres produced in this study exhib-ited denaturation temperatures ranging from 49 to 52◦C, which were higher thannon-cross-linked collagen gels (37–39◦C) [46–48], films (44◦C) [49] and sponges(43–44◦C) [50, 51]. We would contribute the thermal superiority of the fibres overthe other collagen preparations to the presence of salt. Salts have been shown toaffect the thermal stability of collagen in a typical ion-specific way; they alter thestructure of the solvent, which consequently modifies the solvent–macromoleculeinteraction involved in the stabilization of the natural conformation [52]. NaCl,being a strong electrolyte, could have removed water from the fibrous structure,achieving a close inter-fibre packing and elevated denaturation temperatures. Inaddition, it has been shown that when molecules aggregate to form fibres, the de-naturation temperature is higher than the denaturation temperature of gels due tothe increased energy of crystallisation derived from the strong interaction betweenthe closed packed molecules in the fibre [46–48, 51, 53, 54].

Neither the increase of collagen concentration nor the increase of the NaCl inFFB influenced significantly the thermal properties of the fibres, although it hasbeen shown that the thermal stability is dependent on both amount of hydroxypro-line [55, 56] and solute concentration [57, 58]. In both cases we would correlate thethermal properties of the fibres to the collagen–solute interaction. We would spec-ulate that under the conditions utilized herein, we obtained fibres that, proportionalspeaking, were comprised of the same amount of collagen-solute and, thus, theyexhibited no significant different denaturation temperatures.

4.3. Biomechanical Evaluation

Uniaxial tensile tests of dry reconstituted collagen fibres produced stress–straincurves similar to those reported for semi-crystalline polymers which yield and un-dergo plastic flow [59]. The yielding mechanism would imply some form of flowoccurring within the fibre, possibly inter-fibrillar slippage, which plays an impor-tant role in the tensile deformation of aligned connective tissue such as tendon [23,60]. It was also observed that larger diameter fibres exhibited a longer toe region,

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


Figure 4. Independent of the collagen concentration, when a linear regression model was fitted be-tween stress at break and dry fibre diameter, strong correlations were obtained.

in contrast to the smaller diameter fibres. We would attribute this difference to thefibrillar packing that is loose for thick fibres, whilst it is tight for thin fibres (see

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


below). In native fibres (tendon and ligament), the lower modulus of the non-lineartoe region is believed to reflect the un-crimping of the collagen fibrils, as well as theinitiation of stretching of the triple helix, the non-helical ends and the cross-links[61, 62]. Similar curves have been reported previously for dry reconstituted colla-gen fibres [27, 29, 31, 34, 35, 63, 64], collagen and gelatin films [22], nano-fibrousmeshes [65–67], ligaments [68] and tendons [64].

4.3.1. Effect of % NaClAlthough NaCl has been extensively for the production of extruded collagen fibres[15, 24–27], its influence on the properties the fibres has not been explained. It isworth pointing out that at the total absence of NaCl, although fibres were produced,they were too fragile to handle. Similarly, it has been shown that utilization of otherco-agents is essential in order to produce collagen fibres [34, 35] or gels [60] ofsufficient mechanical strength. These results indicate that the self-assembly of col-lagen is a process depending on the relative amount of the co-agent present, as hasbeen observed previously for collagen micro-fibrillar networks [69] and collagenfibres [34, 35].

A correlation between the amount of the NaCl and the dry fibre diameter wasnot found, although it has previously been demonstrated that the diameter of micro-spheres [70] and alumina fibres [71] could be controlled by the amount of polymerincorporated into the structure’s backbone. During precipitation experiments, it hasbeen shown that the mean particle size of the precipitates was generally reducedwhen the ionic strength was increased because of the high level of counter-ionsaround the protein in the solution [72]. In low NaCl concentration, due to thesalting-in effect, protein molecules are in the right configuration to build up a rigidand more viscoelastic material [73]. At higher salt concentrations, due to passingthe optimal amount of water removable for fibre stability, this right configurationmight be lost [74] and the particle size could be remarkably increased [72], leadingto a decrease in the effectiveness of the precipitation [75]. These observations arein accord with other work, where it was mentioned that flocs formed under poly-mer overdosing conditions are larger than flocs formed under other conditions [76].Taking as particle size the diameter of the fibre produced, proportional results wereobtained; the fibre diameter decreased as the electrolyte concentration increasedfrom 5 to 20% (under-dosing conditions) and at 40% electrolyte content (overdoseconditions) the fibre diameter was increased. As has been discussed above, highin diameter fibres exhibit low stress at break values, which is in agreement withprevious work, where it was shown that increasing dermatan sulfate concentrationhas an inhibitory effect on the mechanical strength of reconstituted collagen fib-rils [77]. It is also worth pointing out that at low ionic strengths collagen extractsproduce “structureless” fibrils, whereas extracts made at the same pH but at higherionic strength produce a mass of well-striated robust fibrils [78].

An increase in elongation-at-break and modulus values was observed as theamount of NaCl was increased from 5 to 20%. As mentioned above, the pres-ence of NaCl ensures partial removal of water and encourages collagen molecules

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


to build up a rigid and more viscoelastic material. Similarly, it has been shownthat poly(ethylene glycol) (PEG), by removing free water promotes molecular andfibrillar slippage in collagen gels, which plays an important role in the tensile defor-mation of aligned connective tissue such as tendon [60]. Modulus values appearedto decrease at 40% NaCl, which is in agreement with previous work, where it wasshown that by increasing the concentration of electrolytes, the tensile modulus val-ues were decreased [79, 80]. Finally, the poor correlations obtained between stressat strain and fibre diameter from the 5 and 40% NaCl treatments further enhancesthe utilization of 20% NaCl in the fibre formation buffer, as the optimum conditionfor producing reproducible collagen fibres, which appears to be in agreement withprevious work, where 20% PEG 8000 was identified as the optimum amount ofco-agent present in the FFB for the production of reproducible collagen fibres [35].

4.3.2. Effect of Collagen ConcentrationDuring the fibre manufacturing process, it was found that the higher the collagenconcentration, the easier it was to handle the derived fibres. Furthermore, in pre-vious experiments of our group, when the collagen concentration was lower than1 mg/ml, although fibres were produced, they were too fragile to handle. These re-sults indicate that there is a critical concentration, below which, reconstitution ofcollagen into fibres is difficult or even impossible to be achieved. This observa-tion is in agreement with other studies on in vitro fibrillogenesis, where a criticalconcentration for the reaction to occur was demonstrated [53, 81, 82]. Similarly, ithas been shown for collagen sponges, that there is a minimum effective collagenconcentration [83, 84].

From Table 3, it can be observed that the collagen concentration can affect thedry fibre diameter and consequently the stress at break values of the extruded fibres.Similarly, it has been shown that the fibre diameter depends on the collagen con-centration [35, 85] and the tensile strength of rat skin on the total hydroxyprolinecontent [86]. Furthermore, it has been shown that collagen content can influence themechanical properties of collagen fibres [87]; soluble collagen content had a neg-ative correlation to tensile strength of rat skin [88]. Even the strength of gelatingels has been shown to be dependent on the concentration of gelatin [89]. The highcorrelations between stress at break and dry fibre diameter can be explained in twodimensions: (i) the tensile strength increases as the cross-sectional area decreasesbecause there is less chance for defects in thinner sections [90–92] or (ii) as thefibre diameter decreases, increased longitudinal alignment and improved packingdensity occurs that give rise to strong interactions within or between the collagenfibrils [15, 27, 28, 63]. From these observations, it is safe to assume that lower initialcollagen concentrations would result in higher fibril alignment and consequently inhigher packing densities. Fibres with higher fibril alignment have smaller diam-eter and as a result lower force at break and higher tensile strength, as has beenshown before [39, 93]. The strain at break appeared to increase and the modulusvalues to decrease with increasing collagen concentration and/or collagen fibre di-ameter. Modulus values reflect the stiffness or rigidity of the material; the higher its

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


value, the greater the load required to produce a given extension [94, 95]. These re-sults seem to be in agreement with those obtained with synthetic materials; thickersutures had lower values of tensile strength than their thinner counterparts, whilethicker sutures could sustain higher force before they broke [96]. From the strongcorrelations between the collagen concentration and the dry fibre diameter and thedry fibre diameter and the stress at break, it can be concluded that by controllingthe collagen concentration, tailored made collagen fibres can be produced to suit awide range of tissue engineering applications.

Finally, by controlling the collagen concentration, fibres of variable diameterwere produced similar to those that have been identified in vivo [18–21]. Moreover,the stress at break of native rat tail tendon [24, 38], extruded collagen fibres fromacid-soluble collagen [15, 28, 29] and synthetic sutures [32, 68, 96, 97] has beenreported to range from 120 to 366 MPa, 75 to 355 MPa and 32 to 840 MPa, re-spectively, whilst the strain at break has been reported to range from 13 to 31%,23.2 to 25.7% and 3 to 62%, respectively. The force at break has been reported tobe around 2.89 N for native rat tail tendon fibre [15], whilst reconstituted collagenfibres have been reported to break at forces ranging from 0.08 to 0.93 N, depend-ing on the tissue from which the collagen was extracted from [15, 28]. Given theabove, it appears that the fibres produced in this study demonstrate mechanical,thermal and structural properties not only similar to synthetic materials currently inuse, but also, and most important, to native tissues.

5. Conclusions

Extruded collagen fibres have been shown to be a competitive matrix for hard andsoft tissue replacement. In this study, we manufactured extruded collagen fibresfrom reduced immunogenicity atelocollagen that were characterized by thermal,physical and mechanical properties similar to those of native tissues. Furthermore,we identified 20% NaCl as the optimum amount for fabricating reproducible fibres.When the influence of the collagen concentration was put under the microscope, itwas revealed that by controlling it, competitive tailor-made materials can be pro-duced to suit a wide range of tissue-engineering applications.

Acknowledgements

The authors would like to thank Mrs. P. Potter, Ms. S. Lee, Mrs. T. Hayes andMr. L. Stathopoulos for excellent technical assistance and Dr. S. Jeyapalina andDr. P. Antunes for their useful discussion. D. I. Zeugolis is grateful to The Univer-sity of Northampton and EPSRC for their financial support.

References

1. L. J. Bonassar and C. A. Vacanti, J. Cell. Biochem. 30–31 (Suppl.), 297 (1998).2. D. W. Hutmacher and M. A. Woodruff, in: Handbook of Engineering Biopolymers: Homopoly-

mers, Blens and Composites, S. Fakirov and D. Bhattacharyya (Eds), p. 773. Hanser Publishers,Munich (2007).

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


3. J. K. Suh and H. W. Matthew, Biomaterials 21, 2589 (2000).4. D. W. Hutmacher, J. Biomater. Sci. Polymer Edn 12, 107 (2001).5. W. Friess, Eur. J. Pharm. Biopharm. 45, 113 (1998).6. P. Bunyaratavej and H. L. Wang, J. Periodontol. 72, 215 (2001).7. C. M. Kielty and M. E. Grant, in: Connective Tissue and Its Heritable Disorders: Molecular,

Genetic and Medical Aspects, P. M. Royce and B. Steinmann (Eds), p. 159. Wiley, New York, NY(2002).

8. R. G. Paul and A. J. Bailey, Sci. World J. 3, 138 (2003).9. A. K. Lynn, I. V. Yannas and W. Bonfield, J. Biomed. Mater. Res. Part B Appl. Biomater. 71B, 343

(2004).10. J. D. Goldstein, A. J. Tria, J. P. Zawadsky, K. Y. Kato, D. Christiansen and F. H. Silver, J. Bone Jt.

Surg. 71A, 1183 (1989).11. Y. P. Kato, M. G. Dunn, J. P. Zawadsky, A. J. Tria and F. H. Silver, J. Bone Jt. Surg. 73, 561 (1991).12. M. G. Dunn, A. J. Tria, Y. P. Kato, J. R. Bechler, R. S. Ochner, J. P. Zawadsky and F. H. Silver,

Am. J. Sports Med. 20, 507 (1992).13. T. J. Koob and D. J. Hernandez, Biomaterials 23, 203 (2002).14. D. L. Christiansen, E. K. Huang and F. H. Silver, Matrix Biol. 19, 409 (2000).15. G. D. Pins and F. H. Silver, Mater. Sci. Eng. C 3, 101 (1995).16. K. G. Cornwell, B. Downing and G. D. Pins, J. Biomed. Mater. Res. 71A, 55 (2004).17. K. G. Cornwell, P. Lei, S. T. Andreadis and G. D. Pins, J. Biomed. Mater. Res. 80A, 362 (2007).18. Y. Huang, K. M. Meek, M.-W. Ho and C. A. Paterson, Exp. Eye Res. 73, 521 (2001).19. F. H. Silver, J. W. Freeman and G. P. Seehra, J. Biomech. 36, 1529 (2003).20. T. Jarvinen, T. Jarvinen, P. Kannus, L. Jozsa and M. Jarvinen, J. Orthop. Res. 22, 1303 (2004).21. F. H. Silver, J. W. Freeman and D. DeVore, Skin Res. Technol. 7, 18 (2001).22. I. V. Yannas, J. Macromol. Sci. Part C: Polym. Rev. 7, 49 (1972).23. I. Yannas and C. Huang, J. Polym. Sci. 10, 577 (1972).24. Y. P. Kato, D. Christiansen, R. A. Hahn, S.-J. Shieh, J. D. Goldstein and F. H. Silver, Biomaterials

10, 38 (1989).25. K. S. Weadock, E. J. Miller, E. L. Keuffel and M. G. Dunn, J. Biomed. Mater. Res. 32, 221 (1996).26. M. C. Wang, G. D. Pins and F. H. Silver, Biomaterials 15, 507 (1994).27. G. D. Pins, D. L. Christiansen, R. Patel and F. H. Silver, Biophys. J. 73, 2164 (1997).28. M. G. Dunn, P. N. Avasarala and J. P. Zawadsky, J. Biomed. Mater. Res. 27, 1545 (1993).29. J. F. Cavallaro, P. D. Kemp and K. H. Kraus, Biotechnol. Bioeng. 43, 781 (1994).30. N. D. Light, in: Methods in Skin Research, D. Skerrow and C. J. Skerrow (Eds), p. 559. Wiley,

Chichester (1985).31. D. I. Zeugolis, R. G. Paul and G. Attenburrow, J. Biomed. Mater. Res. Part A, in press (2008).32. Y. Shao, J. Qiu and S. P. Shah, Cement Concrete Res. 31, 1153 (2001).33. D. L. Christiansen and F. H. Silver, Cells Mater. 3, 177 (1993).34. D. I. Zeugolis, R. G. Paul and G. Attenburrow, J. Appl. Polym. Sci., in press (2008).35. D. I. Zeugolis, R. G. Paul and G. Attenburrow, J. Biomed. Mater. Res. Part B, in press (2008).36. J. L. Brokaw, C. J. Doillon, R. A. Hahn, D. E. Birk, R. A. Berg and F. H. Silver, Int. J. Biol.

Macromol. 7, 135 (1985).37. V. Arumugam, M. D. Naresh, N. Somanathan and R. Sanjeevi, J. Mater. Sci. 27, 2649 (1992).38. K. H. Rajini, R. Usha, V. Arumugam and R. Sanjeevi, J. Mater. Sci. 36, 5589 (2001).39. G. Pins, E. Huang, D. Christiansen and F. Silver, J. Appl. Polym. Sci. 63, 1429 (1997).40. C. C. Danielsen, Mech. Ageing Dev. 15, 269 (1981).41. H. Hormann and H. Schlebusch, Biochemistry 10, 932 (1971).

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


42. A. J. Bailey, J. Soc. Leather Technol. Chem. 76, 111 (1992).43. J. Kopp, M. Bonnet and J. P. Renou, Matrix 9, 443 (1989).44. A. Veis, J. Anesey and J. Cohen, Arch. Biochem. Biophys. 94, 20 (1961).45. F. Verzar, in: Biochimie et Physiologie du Tissu Conjonctif, P. Compte (Ed.), p. 381. Centre Tech-

nique du Cuir, Lyon (1965).46. W. Friess and G. Lee, Biomaterials 17, 2289 (1996).47. A. Sionkowska, Int. J. Biol. Macromol. 35, 145 (2005).48. J. M. Orban, L. B. Wilson, J. A. Kofroth, M. S. El-Kurdi, T. M. Maul and D. A. Vorp, J. Biomed.

Mater. Res. 68A, 756 (2004).49. H. Petite, V. Frei, A. Huc and D. Herbage, J. Biomed. Mater. Res. 28, 159 (1994).50. S. Roche, M.-C. Ronziere, D. Herbage and A.-M. Freyria, Biomaterials 22, 9 (2001).51. B. Chevallay, N. Abdul-Malak and D. Herbage, J. Biomed. Mater. Res. 49, 448 (2000).52. M. Luescher, M. Ruegg and P. Schindler, Biopolymers 13, 2489 (1974).53. A. J. Bailey and N. D. Light, Connective Tissue in Meat and Meat Products. Elsevier Applied

Science, London (1989).54. I. Rault, V. Frei, D. Herbage, N. Abdul-Malak and A. Huc, J. Mater. Sci. Mater. Med. 7, 215

(1996).55. S. Jimenez, M. Harsch and J. Rosenbloom, Biochem. Biophys. Res. Commun. 52, 106 (1973).56. J. Rosenbloom, M. Harsch and S. Jimenez, Arch. Biochem. Biophys. 158, 478 (1973).57. R. Komsa-Penkova, R. Koynova, G. Kostov and B. G. Tenchov, Biochim. Biophys. Acta 1297, 171

(1996).58. C. A. Miles and T. V. Burjanadze, Biophys. J. 80, 1480 (2001).59. G. E. Attenburrow and D. C. Bassett, J. Mater. Sci. 14, 2679 (1979).60. D. P. Knight, L. Nash, X. W. Hu, J. Haffegee and M. W. Ho, J. Biomed. Mater. Res. 41, 185 (1998).61. F. H. Silver, D. L. Christiansen, P. B. Snowhill and Y. Chen, J. Appl. Polym. Sci. 79, 134 (2001).62. F. H. Silver, D. L. Christiansen, P. B. Snowhill and Y. Chen, Connect. Tissue Res. 41, 155 (2000).63. E. Gentleman, A. N. Lay, D. A. Dickerson, E. A. Nauman, G. A. Livesay and K. C. Dee, Bioma-

terials 24, 3805 (2003).64. B. J. Rigby, N. Hirai, J. D. Spikes and H. Eyring, J. Gen. Physiol. 43, 265 (1959).65. W. He, Z. Ma, T. Yong, W. E. Teo and S. Ramakrishna, Biomaterials 26, 7606 (2005).66. W. He, T. Yong, Z. W. Ma, R. Inai, W. E. Teo and S. Ramakrishna, Tissue Eng. 12, 2457 (2006).67. I. K. Kwon and T. Matsuda, Biomacromolecules 6, 2096 (2005).68. D. G. Hepworth and J. P. Smith, Composites Part A: Appl. Sci. Manuf. 33, 797 (2002).69. F. Jiang, H. Horber, J. Howard and D. Muller, J. Struct. Biol. 148, 268 (2004).70. J. Fu, J. Fiegel, E. Krauland and J. Hanes, Biomaterials 23, 4425 (2002).71. R. Jancic and R. Aleksic, Mater. Lett. 42, 350 (2000).72. W.-S. Kim, I. Hirasawa and W.-S. Kim, Chem. Eng. Sci. 56, 6525 (2001).73. E. Tsaliki, S. Pegiadou and G. Doxastakis, Food Hydrocolloids 18, 631 (2003).74. J. K. Candlish and G. R. Tristram, Biochim. Biophys. Acta 88, 553 (1964).75. J. A. M. Ramshaw, J. F. Bateman and W. G. Cole, Anal. Biochem. 141, 361 (1984).76. W. Chen and J. C. Berg, Chem. Eng. Sci. 48, 1775 (1993).77. C. C. Danielsen, Connect. Tissue Res. 9, 219 (1982).78. J. Gross, J. H. Highberger and F. O. Schmitt, Proc. Natl. Acad. Sci. USA 41, 1 (1955).79. B. Loret and F. M. F. Simoes, Mech. Mater. 36, 515 (2004).80. C. Agrawal, J. McKinney, D. Lanctot and K. Athanasiou, Biomaterials 21, 2443 (2000).81. K. E. Kadler, Y. Hojima and D. J. Prockop, J. Biol. Chem. 262, 15696 (1987).

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014


82. D. T. Woodley, M. Yamauchi, K. C. Wynn, G. Mechanic and R. A. Briggaman, J. Invest. Dermatol.97, 580 (1991).

83. M. Hanthamrongwit, M. H. Grant and R. Wilkinson, J. Biomed. Mater. Res. 28, 213 (1994).84. G. Chen, T. Ushida and T. Tateishi, J. Biomed. Mater. Res. 51, 273 (2000).85. D. Laude, K. Odlum, S. Rudnicki and N. Bachrach, J. Biomech. Eng. 122, 231 (2000).86. H. Vogel, Connect. Tissue Res. 2, 177 (1974).87. C. C. Danielsen and T. T. Andreassen, J. Biomech. 21, 207 (1988).88. G. W. Dombi, R. C. Haut and W. G. Sullivan, J. Surg. Res. 54, 21 (1993).89. H. Babin and E. Dickinson, Food Hydrocolloids 15, 271 (2001).90. Y. Saito, K. Minami, M. Kobayashi, Y. Nakao, H. Omiya, H. Imamura, N. Sakaida and A. Oka-

mura, J. Thorac. Cardiovasc. Surg. 123, 161 (2002).91. D. Hull and T. W. Clyne, An Introduction to Composite Materials. Cambridge University Press,

Cambridge (1996).92. J. L. Thomason, Composites Sci. Technol. 59, 2315 (1999).93. N. C. Bleach, S. N. Nazhat, K. E. Tanner, M. Kellomaki and P. Tormala, Biomaterials 23, 1579

(2002).94. C. S. Osborne, J. C. Barbenel, D. Smith, M. Savakis and M. H. Grant, Med. Biol. Eng. Comput.

36, 129 (1998).95. J. R. Field and R. M. Stanley, Injury 35, 243 (2004).96. P. Makela, T. Pohjonen, P. Tormala, T. Waris and N. Ashammakhi, Biomaterials 23, 2587 (2002).97. D. Goupil, in: Biomaterials Science. An Introduction to Materials in Medicine, B. D. Ratner,

A. S. Hoffman, F. J. Schoen and J. E. Lemons (Eds), p. 356. Academic Press, London (1996).

Dow

nloa

ded

by [

McG

ill U

nive

rsity

Lib

rary

] at

10:

28 1

4 N

ovem

ber

2014

extruded collagen fibres for tissue-engineering applications: influence of collagen concentration...

Documents