in vitro methodologies to evaluate biocompatibility: status quo and perspective

Article original

In vitro methodologies to evaluate biocompatibility:status quo and perspective

Méthodes in vitro d’évaluation de la biocompatibilité :état de l’art et perspective

C. James Kirkpatrick a,*, Kirsten Peters a, M. Iris Hermanns a, F. Bittinger a,Vera Krump-Konvalinkova b, Sabine Fuchs a, Ronald E. Unger a

a Institute of Pathology, Johannes Gutenberg University, Langenbeckstrasse 1, D-55101 Mainz, Germanyb Institute for Prophylaxis and Epidemiology of Circulatory Diseases, LMU University of Munich, Pettenkoferstrasse 9, D-80336, Germany

Received 1 April 2005; accepted 15 April 2005

Available online 08 June 2005

Abstract

The increasing use of biomaterials in clinical medicine to augment or replace failing organ function has heightened the need to applyrelevant test systems to study the safety and efficacy of new medical devices. This becomes all the more important as the field of ″tissueengineering″ develops, in which the aim is to reconstruct tissue and organ function, for example, by using the patient’s own cells seeded on toa three-dimensional (3-D) scaffold structure. In the biomaterial research field, there has been a necessary expansion of the concept of biocom-patibility to address not only the biosafety issue, that is, the exclusion of cytotoxic and other deleterious effects of biomaterials, but also thebiofunctionality component, which concerns the fulfilment of the intended function of the applied biomaterial. Careful scrutiny of this conceptleads to the conclusion that relevant test systems for biofunctionality must centre on human cells, studied under conditions relevant to thesituation in the living organism for which the medical device has been constructed. Thus, progress in biocompatibility and tissue engineeringwould today be inconceivable without the aid of in vitro techniques. In designing such biofunctionality assays, there are certain fundamentalprinciples which must be adhered to. A constant difficulty is the availability of sterile human tissue for such test systems. Also of paramountimportance is proving the maintenance of the cell phenotype in vitro. Loss of essential characteristic functions of the cultivated cells makesextrapolatory interpretations meaningless for the clinical situation. This paper gives an overview of the basic design principles for suitableassays, and various examples covering a spectrum of applications. Relevant functional parameters will be emphasised, as well as the use ofmodern methods of cell and molecular biology, with measurement of these parameters at both the gene product and transcription levels. Theseparameters include the expression of cytokines, growth factors and cell adhesion molecules. In addition, assays can be constructed to studyinflammation and the wound healing response, which includes the angiogenic reaction. Tissue remodelling around biomaterials can be studiedin vitro by using cells such as fibroblasts, endothelial cells and various inflammatory cells, important parameters reflecting control of thisremodelling being the matrix metalloproteinases and their inhibitors. The need for more co-culture and 3-D models is stressed and data fromthe authors’ own laboratory are presented to illustrate these principles. Finally, the importance of signal transduction within those cells incontact with, or in the vicinity of, biomaterials is emphasised, as this knowledge offers the scientific basis for rational therapeutic interventionto suppress negative effects and enhance positive biological responses (use of drug delivery systems). In understanding these processesmodern technologies using nucleic acid micro-arrays coupled with methods of bioinformatics will hopefully identify key genes which can betargeted. Well-designed in vitro assays have a central role to play in this endeavour.© 2005 Elsevier SAS. All rights reserved.

Résumé

L’utilisation croissante de biomatériaux en pratique clinique pour améliorer ou remplacer la fonction d’organes défaillants a fait apparaîtrela nécessité d’appliquer des tests pertinents pour étudier la sûreté et l’efficacité de nouveaux dispositifs médicaux. Cela est d’autant plus

* Corresponding author.E-mail address: [email protected] (C.J. Kirkpatrick).

ITBM-RBM 26 (2005) 192–199

http://france.elsevier.com/direct/RBMRET/

1297-9562/$ - see front matter © 2005 Elsevier SAS. All rights reserved.doi:10.1016/j.rbmret.2005.04.008

important que le domaine de l’ingénierie tissulaire est en pleine expansion, dans le but de reconstruire la fonction d’un organe ou d’un tissu,en utilisant par exemple, des cellules autologues ensemencées sur des structures matricielles 3-D. Dans le champ des biomatériaux, le conceptde biocompatibilité recouvre non seulement l’aspect biosécurité (c’est-à-dire l’exclusion d’effets cytotoxiques ou délétères des biomatériaux)mais aussi la biofonctionnalité qui évalue la fonction attendue dudit biomatériau. Pour ce faire, les systèmes les plus pertinents sont centrés surl’utilisation de cellules humaines, étudiées dans des conditions proches, autant que faire se peut, de celles qui prévalent in vivo et aujourd’huiles progrès réalisés sont tels qu’il serait inconcevable de s’appuyer sur les techniques in vitro. Cependant, il existe quelques principes de baseà prendre en compte : disponibilité de tissu stérile humain, assurance du maintien du phénotype cellulaire in vitro sans laquelle l’interprétationextrapolée à une situation clinique n’aurait pas de sens. Cet article fournit une revue générale de ces principes et quelques exemples dansdiverses applications fondés sur l’utilisation d’outils modernes de biologie cellulaire et moléculaire, incluant l’expression de cytokines, fac-teurs de croissance et molécules d’adhésion, ainsi que la mise au point de tests d’évaluation de l’inflammation, de la cicatrisation. Le remod-elage tissulaire périimplantaire peut être étudié in vitro à l’aide de divers types cellulaires : fibroblastes, cellules endothéliales, cellulesinflammatoires ; et autres paramètres reflétant le contrôle du remodelage sous l’influence des métalloprotéinases et de leurs inhibiteurs.L’importance de la signalisation dans les cellules au contact ou à proximité de biomatériaux est abordée, puisque sa modulation offre une basescientifique pour des applications thérapeutiques rationnelles afin de supprimer des effets délétères et induire des réponses biologiques posi-tives. Pour la compréhension de ces procédés, les technologies modernes utilisant les puces à ADN couplées à la bio-informatique serviront àidentifier les gènes clés.© 2005 Elsevier SAS. All rights reserved.

Keywords: Cell adhesion; Cell spreading; Cell migration; Cell proliferation; Cell function

Mots clés : Adhésion cellulaire ; Étalement cellulaire ; Migration cellulaire ; Prolifération cellulaire ; Fonction cellulaire

1. The concept of biocompatibility

The biocompatibility concept has been much discussed inthe past decades, but there is now general consensus that thistopic has two principal components. First, there is the ele-ment of absence of a cytotoxic effect and second, there is theaspect of biofunctionality. Probably the best definition of bio-compatibility is that agreed on at the ESB Consensus Con-ference I, namely ″the ability of a material to perform with anappropriate host response in a specific application.″ (Will-iams Dictionary). The corollary of this is that the type of test-ing method will depend on the intended function of the bio-material being used. It then becomes obvious that the testingstrategy must take into account the biological situation inwhich the biomaterial will find itself.

2. Established testing systems

Experimental systems for biocompatibility exist in the formof animal experimentation and in vitro techniques. The vari-ous standards organisations have established a series of guide-lines, which by and large are aimed at the biosafety issues oftesting. The publication ″Biological evaluation of medicaldevices—Parts 1–17″ from the International Organisation forStandardisation (ISO) gives a series of guidelines on the nec-essary testing of medical devices Thus, for example, cytotox-icity testing has been well addressed by ISO 10993-5, whichpresents guidelines for the choice of suitable tests and definesimportant principles underlying these tests. This has beendetailed elsewhere [27]. In the following only in vitro meth-ods will be discussed. The inclusion of additional animalexperimental possibilities would exceed the scope of thispaper.

Tissue culture offers an excellent method to screen poten-tial biomaterials, as the methodology is generally more eco-

nomical than animal experimentation. However, it must neverbe forgotten that tissue culture represents a reduction in thecomplexity of the entire organism generally to a single celltype grown as a single sheet of cells (monolayer). Thus, thebuffering capacity of complex cellular and humoural sys-tems in the intact organism are missing, so that a biomaterialmay not perform well in the in vitro test, but be biocompat-ible in vivo. Nevertheless, this is a risk which is inherent inthe system. A further problem is the fact that established (orpermanent) cell lines, although the most convenient model ofmammalian culture, may be less sensitive to toxic effects ofbiomaterials than cells which have been directly isolated fromthe tissues (primary culture) and further subcultivated for afew passages.

In addition to the choice between permanent cell lines orprimary cells there is the choice of species. As biomaterialsare intended for clinical use, the authors’ laboratory usessolely cells of human origin.Various comparative studies dem-onstrate that human cells can react in a very different mannerfrom cells of other mammals.

3. General principles for in vitro assays

The central pathogenetic element in biomaterial applica-tions is that biomaterials can modulate the activation statusof cells. This can occur directly, through contact between thebiomaterial and cells, or indirectly, via humoural mediatorsinduced by the biomaterial which then act on the cell. Thispathobiological principle is of fundamental importance inunderstanding why biomaterials fail, the corollary of thisbeing that if we possess detailed knowledge of how biomate-rials interact with human fluids and cells, we should be ableto tailor-make new biomaterials to carry out specific func-tions.

193C.J. Kirkpatrick et al. / ITBM-RBM 26 (2005) 192–199

In designing suitable assays, a further principle is that thebiological tissue employed should be derived from humantissue. In the first instance, the isolated tissue or dissociatedcells should be of a primary nature (freshly isolated), withouttransformation, although in some instances a transformed cellline must be used as compromise, because a primary humantissue is unavailable or extremely difficult to culture. Thisapplies to certain epithelial cell types, as well as to cells ofthe central nervous system. Moreover, primary isolated humancell cultures have the disadvantage that experimental repro-ducibility can be difficult to achieve, as individual donors candiffer widely with respect to functional status. Permanent celllines tend to have a more stable phenotype. Another impor-tant principle concerns the fact that human tissues are three-dimensional (3-D), although, until now, most in vitro meth-ods are applied only in two dimensions. Tissue engineeringwill only be successful if this 3-D approach is adopted. Themaintenance of the required cell phenotype is extremelyimportant and implies that such assays must necessarilyinclude a broad spectrum of relevant functional parameters.It is not sufficient to have provided proof of a pure isolationof a cell type. It is also obligatory to demonstrate that theculture conditions do not lead with time to loss of essentialcellular functions.

4. Key parameters

Implanted biomaterials, as well as those exposed to bodyfluids in extracorporeal systems, rapidly (seconds to min-utes) become coated with proteins. This process is dynamic,with adsorption and de-adsorption of proteins [13]. The con-stellation of such adsorbed molecules of varying molecularweight and structure determines the nature of the subsequentcellular response, so these early interactions probably play acentral role in sealing the fate of the biomaterial. The level ofresearch activity into this vital parameter has so far failed tomatch its scientific significance.

In studying cell behaviour on biomaterials, it is possible torecognise distinct chronological steps, which can be investi-gated individually in suitable in vitro assays. A key param-eter is cell adhesion, a process which involves molecular rec-ognition of the proteins adsorbed on the biomaterial surfaceby specific receptors in the plasma membrane (cell adhesionmolecules [CAMs] of the integrin family). Depending on thespecific application, the phenomenon of cell adhesion mayor may not be desirable. Thus, for example in blood-contactingmedical devices, biomaterial surface inertness is essential tominimise adhesion and activation of blood cells, especiallyplatelets. In contrast, in other applications, complete integra-tion of the biomaterial into the tissue is a conditio sine quanon for success, as for example, in bone-contacting biomate-rials. Adherent cells then generally undergo further cytoskel-etal changes, leading to cell spreading and cell migrationon the biomaterial surface. These are functions basic toadequate colonisation of a porous 3-D matrix structure, as in

ceramics for bone contact or collagenous matrices for dermalsubstitution. If cell colonisation of a biomaterial is a centralaim, cell proliferation ranks among the key parameters whichcan be assayed. Nevertheless, a well-colonised biomaterialmatrix is no guarantee of success, as it must be demonstratedthat these cells have the desired synthetic functions (i.e. physi-ological phenotype). Thus, evaluation of cell function istherefore a vital part of any relevant testing scheme.

The adopted spectrum of functional parameters will varyfrom cell type to cell type and, in the case of certain cells,involves a multitude of functions, as for example, in mono-cytes and endothelial cells (EC). In the latter, the balancebetween procoagulant and anti-thrombotic activity must beshown to be in favour of the physiological anti-thromboticstate, if the application is intended for blood contact. In dem-onstrating this, some of the important parameters are expres-sion of tissue factor, plasminogen activator inhibitor-1 (PAI-1), and coagulation factors V and VIII on the procoagulantside, as well as prostacyclin, nitric oxide (NO) and thrombo-modulin on the anti-thrombotic side [30]. The level of expres-sion of CAMs in EC growing on biomaterials is also an impor-tant parameter of endothelial activation [8,48]. The sameprinciple applies to all cell types used in in vitro assays,namely, the need to focus on major gene products character-istic of these cells. As well as the examples given, there arecertain synthesised products produced by numerous cell typesthat have a major role to play in the host response. Theseinclude cytokines and growth factors, which control cell–cellinteractions and the expansion of particular cell populationsin a tissue [69]. In tissue remodelling around implants, theproduction of specific matrix metalloproteinases (MMPs) andtheir tissue inhibitors (TIMPs) are among the most importantparameters determining the outcome [19]. There are numer-ous metabolic products of cells, whose significance is stillunclear with respect to biomaterial testing. Among these arethe heat shock proteins, a family of intracellular proteinsessential to protein folding and unfolding, which are presentin a broad spectrum of living organisms, from microbes tomammalian cells [70,63].

Of considerable help in deciding which key parameters tochoose is a study of the morphological response to biomate-rials, known from cases of explanation at different times. Thislarge database shows unequivocally that inflammation is aconstant response to implantation. This reaction, coupled withthe formation of new blood vessels (angiogenesis), forms thebasis of the wound healing response, a fundamental physi-ological reaction to the implantation of biomaterials, whichcan, however, become pathological by leading, for example,to excessive fibrosis or abnormal tissue contraction [38]. Thecells to be studied in this context will be addressed in the nextsection. Nevertheless, there are many essential questions asyet unanswered, for example, how to define the boundarybetween a physiological and a pathological level of inflam-mation, and, in the case of massive inflammation, when tointervene. On the topic of angiogenesis, defining its optimallevel for a physiological healing response and the optimal

194 C.J. Kirkpatrick et al. / ITBM-RBM 26 (2005) 192–199

interaction with inflammatory cells remain two importantquestions to be addressed.

5. Relevant cells

The choice of cell types for biofunctionality assays in vitrodepends on the biomaterial application. Thus, for boneimplants, human osteoblasts and osteoclasts, responsible forbone formation and degradation, respectively, are of majorinterest. However, as discussed in the previous section, theuniversal inflammatory and healing reactions are also ofimportance. For the acute and chronic phases of inflamma-tion, key players are granulocytes and cells of themonocyte/macrophage lineage, respectively. In studyinghaemocompatibility, in addition to investigation of proteinadsorption, biomaterial interaction with EC and importantblood cells, such as platelets, granulocytes and monocytes,as well as whole human blood, must be a focal point of anytesting programme. In the soft tissue field, as well as in mostforms of implantation, macrophages and fibroblasts are twocentral figures in the host response [16,6].

In the angiogenic response, the key cells are EC, whichsprout from pre-existing blood vessels in the vicinity of theimplant [7]. In most models, EC are derived from large ves-sels, such as the human umbilical vein (HUVEC)[25]. Withthis model system, much better understanding has been gainedabout endothelial function. However, it must be stressed thatthe angiogenic response, as well as the inflammatory response,involves predominantly microvascular EC and not macrovas-cular EC, and relevant in vitro models should be developedaccordingly. In the past few years, we have successfully iso-lated and characterised human pulmonary microvascular EC(HPMEC) [30,63], and this cell type has been incorporatedinto our in vitro assays to study inflammation and angiogen-esis (see below).

As has been eluded to earlier permanent cell lines possessthe advantage that experimental data are more reproduciblethan with primary isolated cells, in which individual biologi-cal variability can be very marked. It is essential when usingsuch permanent cells, which in general have either under-gone spontaneous transformation [5] or have been trans-formed experimentally using, for example viral antigens orchemicals, that essential cellular functions are maintained.The corollary of this is that adequate comparative studies areperformed. From our own laboratory, in which EC is the prin-cipal cell type investigated, we have established a permanentcell line of the HPMEC using a double genetic modificationstrategy, involving SV40 T antigen and the catalytic compo-nent of the telomerase gene [37]. In addition, we have takenavailable permanent cell lines of the endothelium and com-pared numerous phenotypic parameters with relevant pri-mary isolated EC types [59].

6. Effective methods

Of prime significance in designing effective methods isthe principle of quantitation. We have presented this in the

context of cytotoxicity testing [10], but the principle equallyapplies to biofunctionality. With cultured cells, adhesionassays using microtitre plates can be readily quantitated byspectrophotometry in an automated mode. A further possibil-ity is the use of blocking antibodies to investigate the role ofcertain CAMs in the adhesion of a particular cell type to aspecific biomaterial, cell monolayer or protein substratum[31]. The microtitre plate assay also lends itself to the studyof proliferation with methods to delineate certain phases ofthe cell cycle. This can be achieved by incorporation of bro-modeoxyuridine (BrdU) into DNA [10] or by studying theexpression of the nuclear marker Ki-67 to recognise cells inthe proliferating pool [15,61]. More-detailed information onthe phases of the cell cycle can be achieved by studying thecyclin-dependent kinases [54]. Although more time-consuming, cell spreading on biomaterials can be well quan-titated by scanning electron microscopy (SEM) [29].

Biosynthetic function of cells is a topic of such breadthand depth that only general principles with a few caseexamples can be presented within this brief review. Firstly, itis essential to maintain the cardinal elements of cell pheno-type in vitro. In general, this means using primary isolatedhuman cells in early passage. In our own work on human ECcultures, we have shown that the expression of important mol-ecules for cell–cell interactions, the CAMs, decreases rap-idly with increasing passage [33]. Secondly, various aspectsof the kinetics of cell biosynthetic function must be investi-gated. This involves the use of various time-points, as well asdose–response studies. As an example of the latter, in ourstudies on the effects of corrosion products of metallicimplants on the induction of inflammation, we investigatedthe effects of metal ions, such as Zi2+ and Ni2+ on humanpolymorphonuclear granulocyte (PMN) adhesion to HUVECmonolayers. Dose–response experiments revealed that evensub-micromolar ion concentrations promoted increased adhe-sion [31]. Of course, it must be stressed that cells in in vitroassays can demonstrate an increased sensitivity compared withthe in vivo state, a fact which requires careful considerationduring extrapolation to the entire organism. Thirdly, it isimportant to study biosynthetic activity not only at the geneproduct level, but also at the level of transcription. The lattercan be performed well by using molecular biological meth-ods such as northern blotting and the reverse transcriptase-polymerase chain reaction (RT-PCR) [60]. The combinationof both levels of gene expression is essential for understand-ing gene control and signal transduction within cells, withoutwhich tissue engineering will remain an empiric exercise.Thus, in our studies on metal ion induction of endothelialCAMs and cytokines, we showed that metal ions use similartransduction mechanisms as pro-inflammatory mediators [64].

A useful adjunct to the study of cell function by biochemi-cal methodology is the cellular and sub-cellular localisationof gene products by using the techniques of confocal laserscanning microscopy (CLSM) [62] and immunoelectronmicroscopy (IEM), respectively. These methods are particu-larly useful for studying cells adherent to biomaterials.


One of the important considerations in establishing effec-tive methods to study biocompatibility is the simulation ofthe three-dimensionality (3-D) which characterises our tis-sue structure. This is well illustrated by looking at the angio-genic response during healing, as sprouts of microvessels formfrom the pre-existing microcirculation and develop in morethan one plane. We have set-up in vitro assays which attemptto simulate the early (so-called 2-D assay) and later (3-Dassay) phases of angiogenesis. In the former EC begin to dis-sociate from their tight monolayer, as individual cells reorga-nise their cytoskeleton and break down cell–cell contacts,which involve various CAM such as PECAM-1 (CD31). Thiscan be simulated in a confluent endothelial monolayer byforming a thin layer of collagen type I on the apical surfaceof the cells. This acts as a stimulus, leading to monolayerreorganisation with sprout-like structures formed from thereassociation of cells. As this occurs in the plane of the mono-layer the assay is termed 2-D [44]. Using a vital fluoro-chrome, such as calcein AM and image analysis on digitisedfluorescent images, this process can be quantitated. The 3-Dassay is designed to simulate the later stages of angiogenesis,in which the EC have an invasive, migratory phenotype,coupled with cell proliferation to form sprouts through theextracellular matrix (ECM). This assay involves embeddinga single cell suspension of EC in a relevant ″healing-type″matrix (for example, a mixture of collagen type I and fibrin)along with potent pro-angiogenic molecular cues in the formof the growth factors, basic fibroblast growth factor (bFGF)and vascular endothelial growth factor (VEGF). Over a periodof 5–7 days a complex network of vessel-like sprouts withlumen formation takes place [45], and with the use of calceinAM staining and serial fluorescent images taken at differentplanes coupled with specially designed image analytical meth-ods, this process can be quantitated. We have been able todemonstrate that even this complex 3-D system has a highsensitivity towards, for example, anti-angiogenic stimuli [43].

7. Future developments

Returning to the fundamental principle that biofunctional-ity must be studied in in vivo-like assays, it is evident that ahigher degree of complexity must be introduced into existingin vitro methodology. This has already been illustrated in theprevious section, in which the feature of three-dimensionalitywas emphasised and suitable models presented. In addition,in the living organism all cells and tissues are under someform of mechanical stress, whether it be from flowing bloodor in the form of stretching, compression or torsion. In manyinstances, the simulation of this in in vitro systems is still inits infancy and can prove to be extremely difficult. Thus, forexample, the application of compression forces to culturedosteoblasts to simulate in vivo conditions is a difficult task.Tension forces which come to bear, not only on bone cells,but also on EC and soft tissue cells, such as fibroblasts, smoothmuscle cells and skin epithelial cells, can be more readily

investigated by employing cells grown on biomaterial mem-branes, which are stretched by various electromechanicalmeans [9,23]. Flowing blood generates shear stress, so thatbiofunctionality testing of blood-contacting medical devicesmust endeavour to simulate these conditions. This can beachieved by using, for example, a plate and cone rheometer[3] or a parallel plate flow chamber [42]. In the latter, real-time video microscopy and computerised morphometry wereused to study PMN adhesion to different blood-contactingbiomaterials under typical venous and arterial shear stress con-ditions. Nevertheless, this type of investigation is still rare inthe field of in vitro methods for biomaterial testing. How-ever, with the present progress in the concept of regenerativemedicine and research in biotechnology, sophisticated biore-actor systems are being developed [17], for example, for liver[46], heart valves [21] and blood vessels [41,53]. In addition,microfluidic systems are being developed and are alreadyyielding promising new data on how cells can be cultured ina more in vivo-like microenvironment [65,24]. Tan and Desai[56] described a 3-D microfluidics system, engineered layerby layer to yield microscale tissue-like structures, which couldbe highly useful in developing new strategies for tissue engi-neering and also gathering more knowledge on how cells inter-act with biomaterials.

Another field of activity for the future must be the appli-cation of more co-culture assays to study cell–cell interac-tions. This is absolutely necessary, since the in vivo state prac-tically never involves a single cell type in an isolated biologicalmicroenvironment. The development of hybrid organs andtissue engineering implies the establishment of biomaterialsin contact with the relevant cells characteristic of thetissue/organ under discussion. It is evident that this can onlybe achieved if the equivalent human cells are studied in invitro assays with the correct physiological proportions of thesecells. Recent studies emphasise the fact that in co-culture sys-tems one cell type can markedly influence the other cell type,as has been shown for human dermal fibroblasts and epider-mal keratinocytes [67]. Intimately related to this principle ofusing co-cultures is the principle of three-dimensionality.The in vitro reconstruction of a human tissue will thus involvenot only the relevant cells in co-culture, but also the use of invivo-like matrices. The gold standard for such matrix struc-tures is the ECM, which varies depending on the tissue topog-raphy, but generally contains various collagen macromol-ecules, proteoglycans and other glycoproteins [4]. Inbiomaterial research, porous scaffolds and synthetic hydro-gels are proving to fulfil some of the requirements of in vivoECM [49,11]. For studying angiogenesis, we have found thata collagen type I gel, supplemented with the pro-angiogenicfactor, VEGF, permits EC in 3-D culture to form vascularsprouts in an in vivo-like manner [28].

Furthermore, relevant 3-D co-culture assays could be usedas an in vitro model to design and test drug delivery systems(DDS), which in biomaterial applications represent an impor-tant therapeutic tool to promote positive biological responsesand/or suppress unwanted tissue reactions, such as an exag-


gerated inflammatory response or excessive fibrosis. We havedeveloped a co-culture system for the alveolo-capillary bar-rier in the lung [20] which could be used to test mechanismsof delivery of DDS or even nucleic acids for gene therapy orvaccine application. This type of system will hopefully yieldnew information on how such delivery systems in the form ofnanoparticles have to be constructed in order to overcome thenatural barrier in a physiological way. The same principleapplies to the blood–brain barrier (BBB), which could be tar-geted in order to treat central nervous system diseases [35].Thus, Kreuter et al. [36] used a rat model to study how theBBB can be successfully overcome to deliver CNS medica-tion. They employed nanoparticles of poly(butylcyanoacry-late) (PBCA) and demonstrated that coating with polysorbate-80, an adjuvant surfactant, was necessary to achieve effectivebrain delivery [36]. Various in vitro models of the BBB existand generally require not only capillary EC but also pericytesand/or astrocytes, as these cell–cell interactions are essentialto preserving barrier function [1,57,55]. With regard to bio-materials, various approaches are already available for suchcontrolled delivery of therapeutic substances as well as genes[12]. Before this can become reality, much more knowledgeis required on how human cells behave in 3-D matrix struc-tures, not only as a single cell type, but also in the complexmix of different cell types that is characteristic of the in vivosituation.

The evaluation of biocompatibility in any particular appli-cation necessitates a meaningful investigation of how cellsrespond to biomaterials, including bioresorbable biomateri-als, which present an even greater challenge to in vitroresearchers. The authors are convinced that the corollary ofthis is that the experimental systems adopted should providerelevant information on cellular response mechanisms. Thelatter brings us into the realm of signal transduction, whichbasically entails unravelling the molecular pathways involvedin cell communication with the surrounding microenviron-ment. An immense body of evidence exists, describing howcells use receptor–ligand binding systems to initiate cas-cades of enzymatic activity leading to the activation of tran-scription factors in the nucleus, with subsequent downstreamactivation of specific genes. Thus, for example, the mitogen-activated protein kinase (MAPK) pathway via p38 play a vitalrole in the cellular response to inflammatory cytokines andenvironmental stress [26], these being very relevant biologi-cal scenarios in biomaterial applications. Other relevant path-ways include protein kinase C in immune responses [52],GTP-binding protein cascades in response to chemokines [39],PI3K/akt in regulating the alternative pathways of cell sur-vival and apoptosis [14] and the receptor tyrosine kinase(RTK) superfamily in the response to various growth factorsmodulating cell proliferation and differentiation [58]. In addi-tion, studies in developmental biology have demonstrated theimportance of pathways such as wnt signalling for embryo-logical development [66], this being significant for the use ofstem cells for tissue engineering purposes. A further task insignal transduction and biocompatibility will be to investi-

gate how the individual pathways are integrated via cross-talk regulatory mechanisms [40]. Recently, Schneiderbaueret al. [50] showed such cross-talk between the Smad proteinsduring TGF-b1 signalling and integrin pathway signals deriv-ing from the ECM. It is also important to integrate this knowl-edge for each of the cells of interest in biocompatibility stud-ies. Thus, in the field of bone regeneration, the osteoclast hasthe important function of breaking down bone and is thusinvolved in the remodelling process. Reddy recently reviewedthe data available on regulatory mechanisms acting in thiscell type [47].

Recent advances in the fields of genomics and proteom-ics have opened up the possibility to obtain signals regardingexpression of thousands of genes simultaneously in a singleexperimental set-up. Thus, nucleic acid micro-arrays and pro-tein electophoretic systems provide huge data files, whichrequire special algorithms and other tools of bioinformaticsto extract meaningful conclusions regarding cell function [22].However, with cluster analysis and similar methodologies[2,68], it is possible to identify relevant genes involved incell–biomaterial interactions, which would otherwise haveremained unrecognised [32]. Moreover, it has been shownthat integrating the techniques of proteomics with transcrip-tional profiling can offer new insights into the control sys-tems in cells, for example general metabolism [34].

The technologies developed in molecular biology alsomake it possible to genetically manipulate individual cells inorder to study the role of specific genes. This can, for example,be performed by causing a particular gene of interest to beexpressed at much higher levels than normal. Alternatively,gene activity can be switched off by so-called anti-sense RNAtechnology, in which the mRNA transcript of a particular geneis inhibited by a short oligonucleotide sequence which hybri-dises to it by base pairing. The recent progress made in themolecular biology and application of so-called small inter-fering RNAs (siRNAs) is providing a realistic perspective tospecifically target and switch off gene activity [51,18]. Thesetechniques are now extensively used in cell and developmen-tal biology and in studying disease pathomechanisms, butshould prove beneficial both for deepening our understand-ing of cell–biomaterial interactions, as well as for future thera-peutic intervention within the scope of regenerative medi-cine.

8. Conclusions

Studying biofunctionality in vitro in the testing regimensfor the biocompatibility of biomaterials is essential for betterunderstanding of how biomaterials influence cell function andthus for scientific progress in tissue engineering, as well asthe industrial development of new generations of biomateri-als. Due to the absolute requirement to cultivate human cellsunder conditions which are much more advanced than thesimple two-dimensional monolayer culture technique, it willbe essential in the future to recruit more scientists from the


fields of cell and molecular biology to co-operate with mate-rials scientists and clinicians to help extend the interdiscipli-nary work of biomaterial research.

Acknowledgements

The author wishes to gratefully acknowledge the supportof the European Commission over the past 10 years from vari-ous grants within the scope of its biomaterial programmes,the German Research Foundation (DFG; Priority ProgrammeBiosystem 322 1100) and support from the state governmentof Rhineland-Palatinate, Germany.

References

[1] Abbott NJ. Astrocyte–endothelial interactions and blood–brain bar-rier permeability. J Anat 2002;200:629–38.

[2] Bassett Jr. DE, Eisen MB, Boguski MS. Gene expressioninformatics—it’s all in your mine. Nat Genet 1999;21(1 Suppl):51–5.

[3] Bittinger F, Mühlbayer S, Salih V, Grau M, Richter H, Klein CL, et al.Experimental studies on the suitability of human mesothelial cells forseeding vascular prostheses: shear stress resistance in vitro. J MaterSci Mater Med 1996;7:543–50.

[4] Bosman FT, Stamenkovic I. Functional structure and composition ofthe extracellular matrix. J Pathol 2003;200:423–8.

[5] Boukamp P, Petrussevska RT, Breitkreutz D, Hornung J, Markham A,Fusenig NE. Normal keratinization in a spontaneously immortalizedaneuploid human keratinocyte cell line. J Cell Biol 1988;106:761–71.

[6] Carlson MA, Longaker MT. The fibroblast-populated collagen matrixas a model of wound healing: a review of the evidence. Wound RepairRegen 2004;12:134–47.

[7] Carmeliet P. Angiogenesis in health and disease. Nat Med 2003;9:653–60.

[8] Cenni E, Granchi D, Ciapetti G, Verri E, Cavedagna D, Gamberini S,et al. Expression of adhesion molecules on endothelial cells aftercontact with knitted Dacron. Biomaterials 1997;18:489–94.

[9] Dartsch PC, Haemmerle H. Orientation response of arterial smoothmuscle cells to mechanical stimulation. Eur J Cell Biol 1986;41:339–46.

[10] Dekker A, Panfil C, Valdor M, Pennartz G, Richter H, Mittermayer C,et al. Quantitative methods for in vitro cytotoxicity testing of bioma-terials. Cells Mater 1994;4:101–12.

[11] Drury JL, Mooney DJ. Hydrogels for tissue engineering: scaffolddesign variables and applications. Biomaterials 2003;24:4337–51.

[12] Dubruel P, Dekie L, Christiaens B, Vanloo B, Rosseneu M, Van DeKerckhove J, et al. Poly-l-glutamic acid derivatives as multifunctionalvectors for gene delivery. Part B. Biological evaluation. Biomacro-molecules 2003;4:1177–83.

[13] Elwing H. Protein adsorption and ellipsometry in biomaterialresearch. Biomaterials 1998;19:397–406.

[14] Franke TF, Hornik CP, Segev L, Shostak GA, Sugimoto C. PI3K/Aktand apoptosis: size matters. Oncogene 2003;22:8983–98.

[15] Gerdes J, Lemke H, Baisch H, Wacker HH, Schwab U, Stein H. Cellcycle analysis of a cell proliferation-associated human nuclear antigendefined by the monoclonal antibody Ki-67. J Immunol 1984;33:1710–5.

[16] Gretzer C, Eriksson AS, Alldén B, Ericson LE, Thomsen P. Monocyteactivation on titanium-sputtered polystyrene surfaces in vitro: theeffect of culture conditions on interleukin-1 release. Biomaterials1996;17:851–8.

[17] Griffith LG, Naughton G. Tissue engineering—current challenges andexpanding opportunities. Science 2002;295:1009–14.

[18] Hannon GJ, Rossi JJ. Unlocking the potential of the human genomewith RNA interference. Nature 2004;431:371–8.

[19] Hembry RM, Bagga MR, Reynolds JJ, Hamblen DL. Stromelysin,gelatinase A and TIMP-1 in prosthetic interface tissue: a role formacrophages in tissue remodelling. Histopathol 1995;27:149–59.

[20] Hermanns MI, Unger RE, Kehe K, Peters K, Kirkpatrick CJ. Lungepithelial cell lines in co-culture with human pulmonary microvascu-lar endothelial cells: development of an alveolo-capillary barrier invitro. Lab Invest 2004;84:736–52.

[21] Hoerstrup SP, Sodian R, Daebritz S, Wang J, Bacha EA, Martin DP,et al. Functional living trileaflet heart valves grown in vitro. Circula-tion 2000;102:III44–III49.

[22] Hood L, Galas D. The digital code of DNA. Nature 2003;421:444–8.[23] Hu Y, Böck G, Wick G, Xu Q. Activation of PDGF receptor a in

vascular smooth muscle cells by mechanical stress. FASEB J 1998;12:1135–42.

[24] Hung PJ, Lee PJ, Sabounchi P, Aghdam N, Lin R, Lee LP. A novelhigh aspect ratio microfluidic design to provide a stable and uniformmicroenvironment for cell growth in a high throughput mammaliancell culture array. Lab Chip 2005;5:44–8.

[25] Ilan N, Mahooti S, Madri JA. Distinct signal transduction pathwaysare utilized during the tube formation and survival phases of in vitroangiogenesis. J Cell Sci 1998;111:3621–31.

[26] Johnson GL, Lapadat R. Mitogen-activated protein kinase pathwaysmediated by ERK, JNK, and p38 protein kinases. Science 2002;298:1911–2.

[27] Kirkpatrick CJ, Bittinger F, Wagner M, Köhler H, van Kooten T,Klein CL, et al. Current trends in biocompatibility testing. Proc InstMech Eng H 1998;212:H75–H84.

[28] Kirkpatrick CJ, Krump-Konvalinkova V, Unger R, Bittinger F,Otto M, Peters K. Tissue response and biomaterial integration: theefficacy of in vitro methods. Biomol Eng 2002;19:211–7.

[29] Kirkpatrick CJ, Otterbach T, Anderheiden D, Schiefer J, Richter H,Höcker H, et al. Quantitative scanning electron microscopy (SEM) tostudy the adhesion and spreading of human endothelial cells tosurface-modified poly(carbonate urethane)s. Cells Mater 1992;2:169–77.

[30] Kirkpatrick CJ, Wagner M, Hermanns I, Klein CL, Köhler H, Otto M,et al. Physiology and cell biology of the endothelium: a dynamicinterface for cell communication. Int J Microcirc 1997;17:231–40.

[31] Kirkpatrick CJ, Wagner M, Köhler H, Bittinger F, Otto M, Klein CL.The cell and molecular biological approach to biomaterial research : aperspective. J Mater Sci Mater Med 1997;8:131–41.

[32] Klapperich CM, Bertozzi CR. Global gene expression of cellsattached to a tissue engineering scaffold. Biomaterials 2004;25:5631–41.

[33] Klein CL, Bittinger F, Köhler H, Wagner M, Otto M, Hermanns I,et al. Comparative studies on vascular endothelium in vitro: 3. Effectsof cytokines on the expression of E-selectin, ICAM-1 and VCAM-1 by cultured human endothelial cells obtained from different pas-sages. Pathobiology 1995;63:83–92.

[34] Korke R, Gatti Mde L, Lau AL, Lim JW, Seow TK, Chung MC, et al.Large scale gene expression profiling of metabolic shift of mamma-lian cells in culture. J Biotechnol 2004;107:1–17.

[35] Kreuter J. Influence of the surface properties on nanoparticle-mediated transport of drugs to the brain. J Nanosci Nanotechnol2004;4:484–8.

[36] Kreuter J, Ramge P, Petrov V, Hamm S, Gelperina SE, Engelhardt B,et al. Direct evidence that polysorbate-80-coated poly(butylcyano-acrylate) nanoparticles deliver drugs to the CNS via specific mecha-nisms requiring prior binding of drug to the nanoparticles. Pharm Res2003;20:409–16.

[37] Krump-Konvalinkova V, Bittinger F, Unger RE, Peters K, Lehr H-A,Kirkpatrick CJ. Generation of human pulmonary microvascularendothelial cell lines. Lab Invest 2001;81:1717–27.

[38] Leask A, Abraham DJ. TGF-beta signaling and the fibrotic response.FASEB J 2004;18:816–27.


[39] Moser B, Wolf M, Walz A, Loetscher P. Chemokines: multiple levelsof leukocyte migration control. Trends Immunol 2004;25:75–84.

[40] Nelson WJ, Nusse R. Convergence of Wnt, beta-catenin, and cadherinpathways. Science 2004;303:1483–7.

[41] Niklason LE, Gao J, Abbott WM, Hirschi KK, Houser S, Marini R,et al. Functional arteries grown in vitro. Science 1999;284:489–93.

[42] Otto M, Klein CL, Köhler H, Wagner M, Röhrig O, Kirkpatrick CJ.Dynamic blood cell contact with biomaterials: validation of a flowchamber system according to international standards. J Mater SciMater Med 1997;8:119–29.

[43] Peters K, Schmidt H, Unger RE, Kamp G, Pröls F, Berger BJ, et al.Paradoxical effects of hypoxia-mimicking divalent cobalt ions inhuman endothelial cells in vitro. Molec Cell Biochem 2005;157–66.

[44] Peters K, Schmidt H, Unger RE, Otto M, Kamp G, Kirkpatrick CJ.Software-supported image quantification of angiogenesis in an invitro culture system: application to studies of biocompatibility. Bio-materials 2002;23:3413–9.

[45] Peters K, Troyer D, Kummer S, Kirkpatrick CJ, Rauterberg J. Apop-tosis causes lumen formation during angiogenesis in vitro. MicrovascRes 2002;64:334–8.

[46] Powers MJ, Domansky K, Kaazempur-Mofrad MR, Kalezi A, Capi-tano A, Upadhyaya A, et al. A microfabricated array bioreactor forperfused 3D liver culture. Biotechnol Bioeng 2002;78:257–69.

[47] Reddy SV. Regulatory mechanisms operative in osteoclasts. Crit RevEukaryot Gene Exp 2004;14:255–70.

[48] Rémy M, Valli N, Brethes D, Labrugère C, Porté-Durrieu MC,Dobrova NB, et al. In vitro and in situ intercellular adhesionmolecule-1 (ICAM-1) expression by endothelial cells lining a polyes-ter fabric. Biomaterials 1999;20:241–51.

[49] Salgado AJ, Coutinho OP, Reis RL. Novel starch-based scaffolds forbone tissue engineering: cytotoxicity, cell culture, and protein expres-sion. Tissue Eng 2004;10:465–74.

[50] Schneiderbauer MM, Dutton CM, Scully SP. Signaling ″cross-talk″between TGF-beta1 and ECM signals in chondrocytic cells. CellSignal 2004;16:1133–40.

[51] Schutze N. siRNA technology. Mol Cell Endocrinol 2004;213:115–9.[52] Spitaler M, Cantrell DA. Protein kinase C and beyond. Nat Immunol

2004;5:785–90.[53] Stegemann JP, Nerem RM. Phenotype modulation in vascular tissue

engineering using biochemical and mechanical stimulation. AnnBiomed Eng 2003;31:391–402.

[54] Swanton C. Cell-cycle targeted therapies. Lancet Oncol 2004;5:27–36.

[55] Tahti H, Nevala H, Toimela T. Refining in vitro neurotoxicitytesting—the development of blood–brain barrier models. Altern LabAnim 2003;31:273–6.

[56] Tan W, Desai TA. Layer-by-layer microfluidics for biomimetic three-dimensional structures. Biomaterials 2004;25:1355–64.

[57] Terasaki T, Ohtsuki S, Hori S, Takanaga H, Nakashima E, Hosoya K.New approaches to in vitro models of blood–brain barrier drug trans-port. Drug Discov Today 2003;8:944–54.

[58] Tran KT, Griffith L, Wells A. Extracellular matrix signaling throughgrowth factor receptors during wound healing. Wound Repair Regen2004;12:262–8.

[59] Unger RE, Krump-Konvalinkova V, Peters K, Kirkpatrick CJ. In vitroexpression of the endothelial phenotype: comparative study of pri-mary isolated cells and cell lines, including the novel cell line,HPMEC-ST1.6R. Microvasc Res 2002;64:384–97.

[60] Unger RE, Peters K, Huang Q, Funk A, Paul D, Kirkpatrick CJ.Vascularization and gene regulation of human endothelial cells grow-ing on porous polyethersulfone (PES) hollow fiber membranes. Bio-materials 2005;26:3461–9.

[61] van Kooten TG, Klein CL, Kirkpatrick CJ. Cell cycle control incell–biomaterial interactions. Expression of p53 and Ki-67 in humanumbilical vein endothelial cells (HUVEC) in direct contact andextract testing of biomaterials. J Biomed Mater Res 2000;52:199–209.

[62] Van Kooten TG, Klein CL, Wagner M, Kirkpatrick CJ. Focal adhe-sions and assessment of cytotoxicity. J Biomed Mater Res 1999;46:33–43.

[63] Wagner M, Hermanns I, Bittinger F, Kirkpatrick CJ. Induction ofstress proteins in macro- and microvascular human endothelial celltypes by heavy metal ions and heat shock. Am J Physiol Lung CellMol Physiol 1999;277:L1026–L1033.

[64] Wagner M, Klein CL, van Kooten TG, Kirkpatrick CJ. Mechanisms ofcell activation by heavy metal ions. J Biomed Mater Res 1998;42:443–52.

[65] Walker GM, Zeringue HC, Beebe DJ. Microenvironment design con-siderations for cellular scale studies. Lab Chip 2004;4:91–7.

[66] Wang J, Wynshaw-Boris A. The canonical Wnt pathway in earlymammalian embryogenesis and stem cell maintenance/differen-tiation. Curr Opin Genet Dev 2004;14:533–9.

[67] Witte RP, Kao WJ. Keratinocyte–fibroblast paracrine interaction: theeffects of substrate and culture condition. Biomaterials 2005;26:3673–82.

[68] Wu X, Dewey TG. Cluster analysis of dynamic parameters of geneexpression. J Bioinform Comput Biol 2003;1:447–58.

[69] Yang GP, Lim IJ, Phan TT, Lorenz HP, Longaker MT. From scarlessfetal wounds to keloids: molecular studies in wound healing. WoundRepair Regen 2003;11:411–8.

[70] Young JC, Agashe VR, Siegers K, Hartl FU. Pathways of chaperone-mediated protein folding in the cytosol. Nat Rev Mol Cell Biol2004;5:781–91.


in vitro methodologies to evaluate biocompatibility: status quo and perspective

Documents