Bioreactor Design for Tissue Engineering

JOURNAL OF BIOSCIENCE AND BIOENGINEERING Vol. 100, No. 3, 235–245. 2005 DOI: 10. 1263/jbb. 100. 235 © 2005, The Society for Biotechnology, Japan REVIEW Bioreactor Design for Tissue Engineering Ralf Portner,1* Stephanie Nagel-Heyer,1 Christiane Goepfert,1 Peter Adamietz,2 and Norbert M. Meenen3 Technische Universitat Hamburg-Harburg, Bioprozess- und Bioverfahrenstechnik, Denickestr. 15, 21071 Hamburg, Germany,1 Universitatsklinikum Eppendorf, Institut fur Biochemie und Molekularbiologie II, Martinistr. 52, 20246 Hamburg, Germany,2 and Universitatsklinikum Eppendorf, Unfall-, Hand- und Wiederherstellungschirurgie, Martinistr. 2, 20246 Hamburg, Germany3 Received 7 March 2005/Accepted 31 May 2005 Bioreactor systems play an important role in tissue engineering, as they enable reproducible and controlled changes in specific environmental factors. They can provide technical means to perform controlled studies aimed at understanding specific biological, chemical or physical effects. Furthermore, bioreactors allow for a safe and reproducible production of tissue constructs. For later clinical applications, the bioreactor system should be an advantageous method in terms of low contamination risk, ease of handling and scalability.

To date the goals and expectations of bioreactor development have been fulfilled only to some extent, as bioreactor design in tissue engineering is very complex and still at an early stage of development. In this review we summarize important aspects for bioreactor design and provide an overview on existing concepts. The generation of three dimensional cartilage-carrier constructs is described to demonstrate how the properties of engineered tissues can be improved significantly by combining biological and engineering knowledge.

In the future, a very intimate collaboration between engineers and biologists will lead to an increased fundamental understanding of complex issues that can have an impact on tissue formation in bioreactors. [Key words: tissue engineering, bioreactor, design considerations, cartilage] The loss and damage of tissues cause serious health problems (1). In the US, almost one-half of the costs for medical treatments are spent on implant devices annually (2). Worldwide, 350 billion USD are expended for substitute of organs (3).

The substitution of tissues (such as bone or cartilage) or joints with allograft materials includes the risk of infections by viruses (such as HIV, hepatitis C) or a graft rejection. Artificial implants such as those used in knee or hip replacement, have limitations due to their limited lifespan, insufficient bonding to the bone, and allergic reactions caused by material abrasion. New therapy concepts for practical medical applications are required. To this end, tissue engineered substitutes generated in vitro could open new strategies for the restoration of damaged tissues.

The goal of tissue engineering can be defined as the development of cell-based substitutes to restore, maintain or improve tissue function. These substitutes should have organ-specific properties with respect to biochemical activity, microstructure, mechanical integrity and biostability (2). Cell-based concepts include the (i) direct transplantation of isolated cells, (ii) implantation of a bioactive scaffold for the stimulation of cell growth within the original tissue and (iii) implantation of a three di* Corresponding author. -mail: poertner@tuhh. de phone: +49-40-42878-2886 fax: +49-40-42878-2909 235 mensional (3D) biohybrid structure of a scaffold and cultured cells or tissue. Furthermore, non implantable tissue structures can be applied as external support devices (e. g. , an extracorporal liver support when a compatible donor organ is not readily available [4, 5]) or engineered tissues can be used as in vitro physiological models for studying disease pathogenesis and developing new molecular therapeutics (e. g. , drug screening [5, 6]).

The generation of 3D tissue substitutes in vitro requires not only a biological model (e. g. , an adequate source for proliferable cells with appropriate biological functions, a protocol for proliferating cells while maintaining the tissuespecific phenotype), but also the further development of new culture strategies including bioreactor concepts (5, 7, 8). Bioreactors are well established for the cultivation of microbes or mammalian cells under monitored and controlled environmental and operational conditions (e. g. pH, temperature, oxygen tension, and nutrient supply) up to an industrial scale. However, as individual cells are mostly applied, these concepts are inapplicable to 3D tissue constructs. Furthermore, each type of tissue construct (e. g. , skin, bone, blood vessel, and cartilage) will likely require an individualized bioreactor design (7). Therefore, tissue-specific bioreactors should be designed on the basis of comprehensive understanding of biological and engineering aspects. Addi- 236 PORTNER ET AL. J. BIOSCI. BIOENG. , ionally, typical engineering aspects such as reliability, reproducibility, scalability and safety should be addressed (5, 8). Several bioreactor systems have been developed and usually the expectations are very high (5, 6). The question is, however, whether bioreactors can indeed fulfil these expectations. In this review, key technical challenges are identified and an overview of existing culture systems and bioreactors used for tissue engineering is provided. These topics have been addressed to some extent by several authors (4, 5, 7–16). Therefore, they will be discussed only briefly.

Particular focus will be given to the interaction between biological and engineering aspects. Using cartilage tissue formation as an example, it will be shown how an increased fundamental understanding of biological, biochemical, and engineering aspects can significantly improve the properties of 3D tissue constructs. IMPORTANT ASPECTS FOR BIOREACTOR DESIGN With respect to tissue engineering, bioreactors are used for different purposes such as (i) cell proliferation on a small scale (e. g. , for individual patients) and on a large scale (e. g. for allogenic therapy concepts), (ii) generation of 3D tissue constructs from isolated and proliferated cells in vitro and (iii) direct organ support devices (15). These bioreactors should enable the control of environmental conditions such as oxygen tension, pH, temperature, and shear stress and allow aseptic operation (e. g. , feeding and sampling). Furthermore, a bioreactor system should allow for automated processing steps. This is essential not only for controlled, reproducible, statistically relevant basic studies but also for the future routine manufacturing of tissues for clinical application (5, 17).

Besides these global requirements, specific key criteria for 3D tissue constructs based on cells and scaffolds have to be met, including the proliferation of cells, seeding of cells on macroporous scaffolds, nutrient (particularly oxygen) supply within the resulting tissue, and mechanical stimulation of the developing tissues (5). Proliferation of cells is the first step in establishing a tissue culture. Usually, only a small number of cells can be obtained from a biopsy specimen and expansion up to several orders of magnitude is required. The proliferation of cells is quite often accompanied by the dedifferentiation of cells.

For example, proliferating chondrocytes show a decreased expression level of collagen type II and an increased expression level of collagen type I (18, 19). Small culture dishes (e. g. , petri dishes, 12-well plates, and T-flasks) are generally used for cell expansion. As these devices allow only an increase in cell number by a factor of approximately 10, several subcultivations are required. These are considered as a major cause for the dedifferentiation of cells. Recent studies have shown that microcarrier cultures performed in wellmixed bioreactor systems can significantly improve cell expansion (20–22).

The cell seeding of scaffolds is an important step in establishing a 3D culture in a macroporous scaffold. Not only seeding at high cell densities, but also a homogenious distribution of cells within the scaffold is essential. High initial cell densities have been associated with an enhanced tissue formation including cartilage matrix production, bone mineralization and cardiac tissue structure formation (23–25). On the other hand, an inhomogeneous distribution of cells within the scaffold can significantly affect the tissue properties. Several techniques for cell seeding were discussed by Martin et al. 5). Critical issues of all bioreactor concepts involve mass transfer problems (e. g. , oxygen and nutrient supply, and removal of toxic metabolites). The size of most engineered tissues is limited as they do not have their own blood system and the cells are only supplied by diffusion (6, 26). Oxygen supply is particularly critical, as only cell layers of 100–200 µm can be supplied by diffusion (27). However, as tissue constructs should have larger dimensions, mass-transfer limitations represent one of the greatest engineering challenges.

Various studies showed that mechanical stimulation (e. g. , mechanical compression, hydrodynamic pressure, and fluid flow), which are important modulators of cell physiology, can have a positive impact on tissue formation (28), particularly in the context of musculoskeletal tissue engineering, cartilage formation, and cardiovascular tissues among others (11, 29–34). Despite accumulating evidence that mechanical stimulation can improve the properties of engineered tissues, only little is known about specific mechanical forces or the ranges of application (i. . , magnitude, frequency, continuous or intermittent, and duty cycle [5, 9]). Further studies of these factors have to be coupled with quantitative and computational analyses of physical forces experienced by cells and changes in mass transport induced by the method used. Bioreactors allow for different process strategies including the batch, fed-batch or continuous cultivation. In particular, continuous perfusion enables cultivation under constant and controlled environmental conditions (35–38). Martin et al. 5) summarized some of the effects of direct perfusion on tissue-specific properties such as growth, differentiation and mineralized matrix deposition by bone cells, proliferation of human oral keratinocytes, rates of albumin synthesis by hepatocytes, expression of cardiac-specific markers by cardiomyocytes, and GAG synthesis and accumulation by chondrocytes. On the other hand, a bioreactor system becomes more complex as additional features such as feeding pumps, vessels for fresh and spent medium, and control strategies are required (Fig. ), particularly in the case of mechanical stimulation. The bioreactor system has to be integrated into the entire cultivation scheme including biopsy, proliferation, cell seeding, tissue formation and delivery to the site of application (e. g. , hospital). In many cases, the bioreactor itself is only used for tissue formation. However, for a wholistic approach from biopsy to the implantation of tissue, the intire procedure should be coordinated to decrease the number of steps, risk of contamination, and labor costs among others.

This is particularly important with respect to the manufacture of engineered tissue constructs for clinical applications, in which good manufacturing practice (GMP) requirements have to be met (7, 8). VOL. 100, 2005 BIOREACTOR DESIGN FOR TISSUE ENGINEERING 237 FIG. 1. Set-up of flow-chamber-bioreactor system consisting of flow chamber with inserts for tissue constructs, conditioning vessel, peristaltic pump, humidifier, exhaust flask, and medium exchange bottle. CULTURE SYSTEMS AND BIOREACTORS USED IN TISSUE ENGINEERING The engineering of 3D tissue constructs include cell maintenance, proliferation and tissue formation.

An overview of culture systems and bioreactors used for these purposes is shown in Fig. 2. Cell maintenance and proliferation is usually performed in culture systems developed for the monolayer culture of adherent cells (T-flasks, petri dishes, multi well plates). These systems enable sterile handling procedures and are easy to use, disposable and low-cost (39). Moreover, they require individual handling as for example in medium exchange and cell seeding; their usefulness is limited when large quantities are required (7).

This can be overcome to some extent using sophisticated robotics (5). Furthermore, controlling environmental parameters including pH, pO2, and temperature is generally impossible. A further drawback is the limited increase in cell number (approximately 10–20 times during cultivation). The generation of a large number of cells requires several enzymatic subcultivation steps accompanied by an increase in passage number and cell dedifferentiation. In recent studies, small well-mixed bioreactors (e. g. shake flasks, stirred vessels, and super spinner) have been suggested for cell proliferation, in which the cells are grown on microcarriers (20–22). These systems were used in the cultivation of encapsulated cells (18, 19) or neural stem cells in single-cell suspension culture (40). In fixed bed and fluidized bed bioreactors, cells are immobilized in macroporous carriers. The carriers are arranged in a column either packed (fixed bed) or floating (fluidized bed). The column is permanently perfused with a conditioned medium from a medium reservoir, mostly in a circulation oop. These types of reactor are very efficient for the long-term cultivation of mammalian cells for the production of biopharmaceuticals (e. g. , monoclonal antibodies, recom- binant drugs including tPA, and EPO) or recombinant retroviruses for gene therapy (41–43). With respect to tissue engineering they have been investigated for several applications including the cultivation of liver cells as an extracorporal liver device (4, 10, 37), proliferation of stem cells (44–46), cultivation of cardiovascular cells (11) and cartilage cells (9).

As most of the scaffolds have large, interconnected pores, during seeding, cells are distributed quite uniformly. During cultivation, medium flow through a construct enhances the mass transfer of substrates, particularly oxygen to immobilized cells when interconnected cell free pores are available. In membrane bioreactors including hollow-fiber reactors (47), the miniPerm system (48) or the tecnomouse (49), cells are cultivated at tissuelike densities in a compartment, which contains one or several types of membrane for nutrient and oxygen supply and removal of toxic metabolites.

Hollow-fiber systems are widely used in the production of biopharmaceuticals including monoclonal antibodies. Several examples of modified membrane bioreactors exist for the 3D culture of tissue cells including hepatocytes (4, 50–54), skin cells (55) or other human cells (49, 56). Most of the culture systems and bioreactors discussed so far were first developed for the cultivation of mammalian cells and adapted to the engineering of 3D tissue constructs. However, besides some exceptions, they can hardly be used in the generation of implantable tissue constructs.

Each type of tissue intended for implantation (e. g. , skin, heart valve, blood vessel, and cartilage) requires a different geometric structure and a specific bioreactor design, particularly when mechanical stimulation has to be provided. Finally, engineered constructs have to be transferred from the culture system to the patient after cultivation. Therefore, several tissue-specific culture systems and bioreactors have been suggested. One of the most prominent culture systems is the rotat- 238 PORTNER ET AL. J. BIOSCI. BIOENG. , FIG. 2. Cell culture systems used in tissue engineering. ng-wall vessel (23), in which a construct remains in a state of free-fall through the medium with a low shear stress and a high mass transfer rate. This system has a wide range of practical applications (5, 7). A multipurpose culture system was introduced by Minuth et al. (35) for perfusion cultures under organotypic conditions. Several tissue carriers can be placed inside a perfusion container. Depending on the type of tissue-specific cell supports can be selected. A perfused flow-chamber bioreactor with a new concept for aeration has been introduced recently (20), in which tissue-specific inserts for various types of tissue (e. . , cartilage, skin, and bone) can be applied. Besides these examples of multipurpose bioreactors, numerous of tissue-specific culture systems have been suggested. These are well reviewed with respect to specific tissues (5, 6, 8–11, 14, 15, 26, 50, 57–59). Most of them are custom-made; only very few have been commercialized. MODELLING OF BIOREACTOR SYSTEMS FOR TISSUE ENGINEERING The appropriate molecular and macroscopic architecture of 3D tissue constructs is essential to producing a phenotypically appropriate tissue (6). However, in many cases this is not realized.

The solution may lie in providing appropriate physiological conditions during cultivation. In many cases, the culture systems and bioreactors were not optimized in this respect. Parameters such as perfusion rate, flow conditions, shear stress, and compression magnitude were varied, quite often by a trial-and-error approach. Furthermore, different conditions have to be examined accurately with regard to their effect. For example, hydrostatic pressure applied during cartilage culture can lead to an improved mass transfer of small and large molecules into the cartilage matrix, but can also induce a mechanical stimulation of embedded cells.

These two effects have to be examined separately. Particularly, when if forces (e. g. , hydrostatic pressure, shear stress induced by perfusion, shear stress or gliding forces induced by mechanical impact, and pulsation flow) are applied, the exact local conditions experienced by the cells have to be determined. Therefore, experimental studies should be supported by simulation methods such as the computational fluid dynamics (CFD) or finite-element approach. Several examples underline the potential of an integrated study of mechanical and biomechanical factors that control the functional development of tissue engineered constructs (60–66).

This approach will significantly improve bioreactor design in the near future. As an attempt to evaluate and compare different tissue engineering reactors, Omasa et al. (67) applied the analytic hierarchy process (AHP). For evaluating a reactor for a bioartificial liver (BAL) support system, five criteria were identified (namely, safety, scalability, cell growth environment, mimicking native liver functions and handling). On the basis of these criteria, six different types of BAL systems were successfully ranked using AHP.

The process is a valuable tool for decision making in tissue engineering, not only for BAL systems but for tissue engineering reactors in general. INTERACTION BETWEEN BIOLOGICAL AND ENGINEERING ASPECTS IN GENERATION OF ARTIFICIAL CARTILAGE The following considerations are intended to demonstrate how the interdisciplinary application of biological and engineering knowledge can significantly improve the properties of tissue engineered 3D cartilage constructs. Severe health problems can arise from a lack or damage of hyaline cartilage in joints, particularly in knee joints (68).

The methods of tissue engineering enable the generation of cartilage tissue in vitro and provide new strategies to restoring damaged cartilage (9, 36, 58, 69). Whereas many approaches to generating tissue-engineered cartilage are based on either alginate encapsulation or the use of bioresorbable scaffolds (9, VOL. 100, 2005 BIOREACTOR DESIGN FOR TISSUE ENGINEERING 239 19, 58, 59, 69, 70), recently, a new concept involving the use of a bone substitute carrier covered with a layer of tissue-engineered cartilage ithout scaffolds has been presented (17). This concept consists of four steps: (i) explanted articular chondrocytes are expanded in a monolayer culture and then seeded on a solid carrier to form a primer layer; (ii) simultaneously, chondrocytes are suspended in alginate gel and cultivated for two weeks to allow the formation of a pericellular matrix; (iii) redifferentiated chondrocytes are then eluted out of the alginate gel and sedimented on the primer layer; (iv) the biphasic constructs are further cultivated for three weeks.

Below, mainly different aspects of step iv are addressed. A similar concept was suggested by Waldman et al. (71). A critical parameter during the cultivation of cartilage tissue is oxygen supply. The effect of oxygen during the in vitro cultivation of chondrocytes is poorly understood, and therefore it is presently a controversial issue (72). In several studies, chondrocytes were immobilized in alginate beads and cultivated under different oxygen concentrations in the gas phase (for review see Ref. 72).

O’Driscoll et al. (73) observed a limited collagen type II production at very high (90% O2) and very low (1–5% O2) oxygen concentrations. Domm et al. (19) showed a stimulatory effect of a decreased oxygen tension (5% O2) on matrix production. In a recent study of Malda et al. (18), the pellets of chondrocytes were suspended in a stirred bioreactor under different oxygen concentrations. They observed an increased production of glycosaminoglycan at 5% and 1% O2 (v/v) in comparison with aeration at 21% O2 (air).

The increased glycosaminoglycan production is accompanied by a decrease in collagen type I level. On the other hand, several studies of chondrocytes embedded in a 3D matrix or a scaffold have demonstrated an enhanced matrix formation, particularly proteoglycan synthesis under more aerobic conditions (74, 75). The main difference between the applied methodologies (alginate and pellet culture vs. cartilage generation in 3D scaffolds) can be observed in oxygen gradients in the vicinity and within the formed cartilage (38, 76).

In the case of alginate and pellet culture, oxygen gradients at the surface of constructs can be neglected, and only oxygen limitations within the constructs are likely. For the cultivation of 3D scaffolds, even more significant oxygen gradients within the formed matrix are expected. Another attempt to improve tissue-engineered cartilage is the application of mechanical force during the cultivation to produce a phenotypically appropriate tissue. Four main types of force are currently used in cartilage cultivation: hydrostatic pressure, direct compression, and high- and low-shear fluid environments.

All these forces have been integrated into culturing devices used as bioreactors for articular cartilage. The individual effects have been reviewed by Darling and Athansiou (9). Our work started from an engineering view point, that is, we invastigated the oxygen transport in culture systems for the generation of cartilage pellets and within the pellet itself theoretically. Two culture strategies were modelled, a static culture system with the supply of oxygen through the medium by diffusion and a perfused culture system with an xygen-enriched medium. For a static culture system, the oxygen supply of a cartilage pellet from the medium, which is not flowing (static medium layer, Fig. 3A) was calculated with the following assumptions: (i) the medium is in contact with the atmosphere via a gas-liquid interface and the oxygen concentration at the surface is in equilibrium with the atmosphere. (ii) The oxygen concentration cR in the static medium layer can be described by one-dimensional diffusion (D= diffusion coefficient in the medium). ?cR ? cR (1) ———- = D ? ———–? x2 ? t (iii) The oxygen concentration cP within a pellet can be described by Eq. 2. X ? qmax cP dcP d2cP (2) ———- = Deff ? ———– – ——————- ? —————V dt ks + cP dx2 where X is the cell density, qmax the maximal cell-specific oxygen uptake rate, V the volume of the cartilage and kS the Monod constant. (iv) The mass transfer of oxygen from the medium to the pellet is only by diffusion. At the interface, oxygen concentration and slope are identical on both sides.

At the pellet wall, the slope of oxygen concentration is zero. Figure 3 shows the calculated oxygen concentration profile as a function of length scale and time with the following parameters: a cell density of 2 ? 106 cells per pellet, a maximal cell specific oxygen uptake rate of qmax = 10–13 mol cell–1 h–1, kS = 1075 ? 10–5 mol l–1 and Deff = 2. 16 ? 10–5 cm2 s–1, which is 80% of the diffusion coefficient in the medium (38). As shown in Fig. 3A, oxygen concentration in the vicinity of the pellet decreases very rapidly to zero within a very short time.

This effect is mainly due to the very low diffusive transport of oxygen through the boundary layer between the cartilage pellet and the medium. Alternatively, a perfused system was modelled, in which an oxygen enriched medium flows over the cartilage and the thickness of the boundary layer over the cartilage should be significantly decreased (Fig. 3B). The oxygen concentration cm in the flowing medium as a function of time and length xm can be described by Eq. 3. dcm d2cm 4 (3) ———– = D ? ———— + u ? ————– ? (c0 – cm) dt dxm2 dP ? where U is the flow velocity, dP the pellet diameter, and c0 the oxygen concentration at xm = 0. The results of these simulations are summarised in Fig. 3B for a flow rate of 3 ml min–1. These calculations show that the mass transfer resistance in the boundary layer between the cartilage and the medium can be significantly decreased, but still oxygen concentration within the pellet decreases very rapidly. For both systems, a severe oxygen limitation has to be expected within the tissue matrix. The main reason for this limitation is the very low rate of transport by diffusion.

Similar conclusions have been drawn elsewhere (38, 76). A flow with an enriched medium, as in the perfused culture system, can overcome this problem only to some extent by decreasing the thckness of the boundary layer between the tissue matrix and the medium. These findings lead to the question of whether it is possible to predict the properties of cartilage generated in these bioreactor systems on the basis 240 PORTNER ET AL. J. BIOSCI. BIOENG. , FIG. 3. Modelling of oxygen profiles within tissue-engineered cartilage. A) Calculated oxygen profile in static culture system for cultivation of artificial cartilage pellets. (B) Calculated oxygen profile in perfusion culture system for cultivation of artificial cartilage pellets. of the above conclusions. This will be addressed below by comparing experimental results from different cultivation systems. The methods will be introduced only briefly. For details, see Ref. 17. Threedimensional cartilage-carrier constructs were produced according to the protocol described above from chondrocytes of an adult mini pig.

The final step of this protocol was performed either in bioreactors (in a flow-chamber or under intermittent hydrostatic pressure) or in 12-well plates as the control (Fig. 4). The flow-chamber bioreactor was specially designed for the generation of three-dimensional cartilagecarrier constructs. A specific feature of the flow chamber is a very thin medium layer for improved oxygen supply and a counter current flow of the medium and gas (20). Another bioreactor system was designed in which an intermittent hydrostatic pressure is applied during the cultivation.

It enables the cultivation of seven cartilage-carrier constructs in parallel. The main results of the experiments are summarized in Table 1. Experiments at different oxygen concentrations (21% and 10% v/v O2) were performed in 12-well plates as the control and the constructs were compared qualitatively and quantitatively. The appearance of the cartilage obtained under decreased oxygen tension seemed to be closer to the native cartilage with respect to the shape of cells, distribution of cells within the matrix, and smoothness of the surface among others.

The thickness of the cartilage formed by free swelling was always in the same range as that of the native cartilage (approximately 1 mm). For both oxygen concentrations the content of glycosaminoglycan (GAG) was similar, but still significantly lower than that in the native cartilage (data not shown). Qualitatively, the most stable attachment of the cartilage on top of the carrier was found for 10% O2 (v/v). Furthermore, the cultivated cartilage contained a large content of collagen type II (data not shown). From these observations, a lower oxygen concentration in the gas phase is recommended.

Experiments in the flow-chamber bioreactor showed a significantly higher matrix thickness but a lower content of GAG then cultures in 12-well plates as the control. The appearance of the cartilage obtained in the bioreactor seemed to be closer to the native cartilage with respect to the shape of cells, distribution of cells within the matrix, and smoothness of the surface among others. The cartilage obtained from 12-well plates showed an inhomogeneous distribution of cells, a more uneven surface and holes within the matrix.

Another important requirement for a successful implantation is the consistency of the cartilage. In the case of bioreactor cultures, the compaction of the cartilage and the attachment between the cartilage and carrier were very good, indicating that the construct is appropriate for implantation. In contrast, the cartilage-carrier constructs cultivated in 12-well plates were soft and the attachment between the cartilage and carrier was not sufficient. In this case, the cartilage tended to slip off the carrier with only a slight mechanical impact.

Furthermore, the cultivated cartilage should VOL. 100, 2005 BIOREACTOR DESIGN FOR TISSUE ENGINEERING 241 FIG. 4. Culture systems used for generation of three-dimensional cartilage carrier constructs. contain a significant content of collagen type II. This was confirmed qualitatively by immunohistological analysis. In the third set of experiments cartilage-carrier constructs were cultivated in a bioreactor, in which intermittent hydrostatic pressure was applied during the last week of cultivation (total time, 5 weeks).

A total pressure of 3 ? 105 Pa was applied eight times per hour for 2. 5 min during this week. Constructs cultivated under an intermittent hydrostatic pressure load and a decreased oxygen tension (10% O2 v/v) showed a significantly higher matrix thickness and a higher GAG content than those cultivated in 12-well plates as the control (Table 1). The compaction of matrix as well as attachment between the cartilage and carrier was evaluated qualitatively and found to be much better in the bioreactor ultures with intermittent hydrostatic pressure loading than in constructs cultivated without loading in 12-well plates. In a small animal trial, constructs were implanted into mini pigs. One pig recieved constructs cultivated with pressure loading and at 21% O2 v/v and anotther pig received constructs cutlivated with pressure loading and at 10% O2 v/v. After surgery, none of the animals showed restrictions in movement. In both pigs, the cultivated cartilage integrated into the surrounding native cartilage without the formation of gaps.

The immunohistological analysis revealed that explants cultivated at 10% O2 v/v showed a higher collagen type II content than those cultivated at 21% v/v O2 (Fig. 5). The explants cultivated at 10% O2 v/v indicated a hyaline like cartilage, whereas those cultivated at 21% O2 TABLE 1. Biochemical data of cartilage-carrier constructs cultivated in 12-well plates, in flow-chamber bioreactor, under intermittent hydrostatic pressure, and for native cartilage (glycosaminoglycane, GAG) Matrix GAG thickness (µg) (mm) 12-well plate 4 1. 2 ± 0. 7 162± 103. 12-well plate 4 1. 3 ± 0. 6 155. 9 ± 74. 4 Flow-chamber bioreactor 6 1. 6 ± 0. 2 142. 7 ± 9. 7 12-well plate (control) 6 1. 1 ± 0. 1 238. 9 ± 22. 1 Static culture under intermittent hydrostatic pressure 3 1. 1 ± 0. 7 466. 53 ± 7. 1 12-well plate (control) 3 0. 7 ± 0. 0 268. 8 ± 74 Native cartilage 1. 0 ± 0. 0 1680. 1 ± 211. 8 The cultivation principle consists of the following steps (17): (i) chondrocytes are isolated from the articular cartilage of an adult mini-pig, expanded in a monolayer culture and then seeded on a solid ceramic carrier (diameter 4. mm) to form a primer layer; (ii) simultaneously, chondrocytes are suspended in alginate gel and cultivated for two weeks to allow the formation of a pericellular matrix; (iii) redifferentiated chondrocytes are then recovered from the alginate gel and sedimented on the primer layer; (iv) the biphasic constructs are further cultivated for three weeks to allow free-swelling of the cartilage matrix. This step was performed either in a 12-well plate, a flow-chamber bioreactor (flow rate, 0. 5 ml/min) or under intermittent hydrostatic pressure (total pressure of 3 ? 105 Pa, 8? 2. min per hour, applied hourly during the last week of cultivation). Three sets of experiments were performed: (i) cultivation of cells in passage 4 in 12-well plates at 21% and 10% oxygen tensions in the headspace; (ii) cultivation of cells in passage 6 in a flow-chamber bioreactor and in 12-well plates (control) at 21% oxygen tension; (iii) cultivation of cells in passage 3 in a bioreactor under intermittent hydrostatic pressure and in 12-well plates (control) at 10 % oxygen tension. Native cartilage from an adult mini-pig; diameter, 4. 5 mm; average of five samples. Culture system Passage no.

O2-tension in the gas phase (%) 21 10 21 21 10 10 242 PORTNER ET AL. J. BIOSCI. BIOENG. , further investigations showed that due to the very low mass transfer between the static culture medium and the gas phase, oxygen concentration in the culture medium did not increase during intervals with a higher pressure. Therefore, other factors that explain the significantly improved matrix properties of constructs cultivated under intermittend hydrostatic pressure compared with static cultures in 12-well plates should be identified. Probably, the mechanical stimulation of chondrocytes is the main effect.

However, to clarify this, further biochemical and cell biological studies are required to determine, which cell biological effects are responsible for the observed phenomena. CONCLUSIONS Bioreactors for the generation of 3D tissue constructs can provide a better process control by taking into account different demands of cells during cultivation. Furthermore, they can provide the technical means to perform controlled studies aimed at understanding specific biological, chemical or physical effects. Moreover, bioreactors enable a safe and reproducible production of tissue constructs.

An overall comparison of different culture methods shows clearly the advantages of bioreactor culture. Not only can the properties of cultivated 3D tissue constructs be improved, aspects such as safety of operation argue for the use of bioreactor systems. Furthermore, the bioreactor can be used to study effects such as shear flow and/or hydrostatic pressure on the generation of tissues. For future clinical applications, the bioreactor system should be an advantageous method in terms of low contamination risk, ease of handling and scalability.

To date the goals and expectations of bioreactor development have been fulfilled only to some extent, as bioreactor design in tissue engineering is very complex and still at an early stage of development. In the future, a very intimate collaboration between engineers and biologists will lead to an increased fundamental understanding of complex issues that can have an impact on tissue formation in bioreactors. On one hand, devices are required with a well-described microenvironment of cells for fundamental studies. On the other hand, a ransition from a laboratory scale to an industrial scale will require a high adaptability of specialized bioreactors in a standardized production process. These advances will aid in ensuring that tissue engineering fulfil the expectations for revolutionizing medical care. ACKNOWLEDGMENTS We would like to thank Drs. Klaus Baumbach, Frank Feyerabend, Jan-Philipp Petersen and Jens Schroder for scientific input and help in animal trials, Sven Cammerer and Katja Schmid for modelling the cartilage reactor as well as Katharina Braun, Ditte Siemesgeluss and Richard Getto for technical support.

The financial support of Biomet Deutschland GmbH, Berlin under the BMBF grant no. 03N4012 and the city of Hamburg within the “Qualtitatsoffensive Tissue Engineering” is gratefully acknowledged. FIG. 5. Images of immunohistological staining for collagen type II of implanted cartilage-carrier construct (mini-pig) after explanation and decalcification of carrier. Cartilage-carrier constructs were cultivated prior to implantation under loading with intermittent hydrostatic pressure in a bioreactor aerated with 21% O2 (a, b) and 10% O2 (c, d). Time of observation for the animals: 8 weeks.

Bars indicate 500 µm. 1, Native cartilage; 2, newly formed cartilage; 3, bone; 4, extracted carrier after decalcification. The protocol for the generation of cartilagecarrier constructs is described in Table 1. v/v showed unsatisfactory results. The above considerations underline the idea that engineering knowledge can help improve cultivation systems and applied strategies for tissue engineering. On the other hand, it becomes obvious, that both engineering and fundamental studies on cell biology are required to further clarify the observed effects.

The theoretical simulations indicate that even under ideal conditions (no mass transfer limitation in the fluid phase) a severe oxygen limitation within the engineered tissue should be expected. If oxygen supply would be the limiting factor during cartilage formation, a bioreactor system (flow chamber) with an improved oxygen supply should lead to a better quality of the engineered cartilage. On the other hand, lower oxygen concentrations in the gas phase seem to improve some matrix properties.

From the results discussed above, these discrepancies can be solved only to some extent. With respect to important biochemical properties, particularly the content of GAG, the constructs from the flow chamber bioreactor showed significantly lower values than those from the 12-well plates, probably due to a higher, detrimental oxygen concentration in the matrix. On the other hand, other matrix properties, particularly the attachment between the cartilage and carrier was better for constructs from the flow chamber than for those from 12well plates.

This can be due to a better oxygen supply within the matrix close to the surface of the carrier. The best results were obtained from constructs cultivated under intermittent hydrostatic pressure and a decreased oxygen concentration in the gas phase. Initially, this phenomenon is difficult to understand. As the partial pressure of oxygen in the gas phase depends on total pressure, a higher pressure should even increase oxygen concentration significantly, leading to even worse matrix properties. However, VOL. 100, 2005 BIOREACTOR DESIGN FOR TISSUE ENGINEERING 243

NOMENCLATURE c0 : oxygen concentration in the flowing medium layer at xm = 0 m, mol l–1 cm : oxygen concentration in flowing medium layer, mol l–1 cR : oxygen concentration in static medium layer, mol l–1 cP : oxygen concentration in cartilage matrix, mol l–1 D : diffusion coefficient in medium, m2 s–1 Deff : diffusion coefficient in cartilage matrix, m2 s–1 dP : pellet diameter, m kS : Monod constant, mol l–1 qmax : maximal cell-specific oxygen uptake rate, mol cell–1 h–1 u : flow velocity, m2 s–1 V : volume of cartilage, l X : cell density, cells l–1 x : length, m xm : length in medium, m REFERENCES 1.

Langer, R. : Tissue engineering. Mol. Ther. , 1, 12–15 (2000). 2. Langer, R. and Vacanti, J. P. : Tissue engineering. Science, 260, 920–926 (1993). 3. Lysaght, M. J. and Loughlin, J. A. : Demographic scope and economic magnitude of contemporary organ replacement therapies. ASAIO J. , 46, 515–521 (2000). 4. Chamuleau, R. A. : Artificial liver support in the third millennium. Artif. Cells Blood Substit. Immobil. Biotechnol. , 31, 117–126 (2003). 5. Martin, I. , Wendt, D. , and Heberer, M. : The role of bioreactors in tissue engineering. Trends Biotechnol. , 22, 80–86 (2004). 6. Griffith, L. G. and Naughton, G. Tissue engineering — current challenges and expanding opportunities. Science, 295, 1009–1014 (2002). 7. Ratcliffe, A. and Niklason, L. E. : Bioreactors and bioprocessing for tissue engineering. Ann. N. Y. Acad. Sci. , 961, 210– 215 (2002). 8. Naughton, G. K. : From lab bench to market: critical issues in tissue engineering. Ann. N. Y. Acad. Sci. , 961, 372–385 (2002). 9. Darling, E. M. and Athanasiou, K. A. : Articular cartilage bioprocesses and bioreactors. Tissue Eng. , 9, 9–26 (2003). 10. Allen, J. W. and Bhatia, S. N. : Improving the next generation of bioartificial liver devices. Semin.

Cell Dev. Biol. , 13, 447–454 (2002). 11. Barron, V. , Lyons, E. , Stenson-Cox, C. , McHugh, P. E. , and Pandit, A. : Bioreactors for cardiovascular cell and tissue growth: a review. Ann. Biomed. Eng. , 31, 1017–1030 (2003). 12. Godbey, W. T. and Atala, A. : In vitro systems for tissue engineering. Ann. N. Y. Acad. Sci. , 961, 10–26 (2002). 13. Morsi, Y. S. , Birchall, I. E. , and Rosenfeldt, F. L. : Artificial aortic valves: an overview. Int. J. Artif. Org. , 27, 445–451 (2004). 14. Park, K. D. , Kwon, I. K. , and Kim, Y. H. : Tissue engineering of urinary organs. Yonsei Med. J. , 41, 780–788 (2000). 5. Shachar, M. and Cohen, S. : Cardiac tissue engineering, ex vivo: design principles in biomaterials and bioreactors. Heart Fail Rev. , 8, 271–276 (2003). 16. Vunjak-Novakovic, G. : The fundamentals of tissue engineering: scaffolds and bioreactors. Novartis Found Symp. , 249, 34–46 (2003). 17. Nagel-Heyer, S. , Goepfert, Ch. , Morlock, M. M. , and Portner, R. : Relationship between gross morphological and biochemical data of tissue engineered cartilage-carrier-constructs. Biotechnol. Lett. , 27, 187–192 (2005). 18. Malda, J. , van Blitterswijk, C. A. , van Geffen, M. , Martens, D. E. , Tramper, J. and Riesle, J. : Low oxygen tension stimulates redifferentiation of dedifferentiated adult human nasal chondrocytes. Osteoarthritis Cartilage, 12, 306–313 (2004). 19. Domm, C. , Schunke, M. , Christesen, K. , and Kurz, B. : Redifferentiation of dedifferentiated bovine articular chondrocytes in alginate culture under low oxygen tension. Osteoarthritis Cartilage, 10, 13–22 (2002). 20. Nagel-Heyer, S. , Leist, Ch. , Lunse, S. , Goepfert, C. , and Portner, R. : From biopsy to cartilage-carrier constructs by using microcarrier cultures as sub-process, p. 139. In Proceedings of 19th ESACT meeting.

European Society for Animal Cell Technology (ESACT), Harrogate, UK (2005). 21. Malda, J. , van den Brink, P. , Meeuwse, P. , Grojec, M. , Martens, D. E. , Tramper, J. , Riesle, J. , and van Blitterswijk, C. A. : Effect of oxygen tension on adult articular chondrocytes in microcarrier bioreactor culture. Tissue Eng. , 10, 987– 994 (2004). 22. Bardouille, C. , Lehmann, J. , Heimann, P. , and Jockusch, H. : Growth and differentiation of permanent and secondary mouse myogenic cell lines on microcarriers. Appl. Microbiol. Biotechnol. , 55, 556–562 (2001). 23. Freed, L. E. , Langer, R. , Martin, I. , Pellis, N.

R. , and Vunjak-Novakovic, G. : Tissue engineering of cartilage in space. Proc. Natl. Acad. Sci. USA, 94, 13885–13890 (1997). 24. Holy, C. E. , Shoichet, M. S. , and Davies, J. E. : Engineering three-dimensional bone tissue in vitro using biodegradable scaffolds: investigating initial cell-seeding density and culture period. J. Biomed. Mater. Res. , 51, 376–382 (2000). 25. Carrier, R. L. , Papadaki, M. , Rupnick, M. , Schoen, F. J. , Bursac, N. , Langer, R. , Freed, L. E. , and Vunjak-Novakovic, G. : Cardiac tissue engineering: cell seeding, cultivation parameters, and tissue construct characterisation.

Biotechnol. Bioeng. , 64, 580–589 (1999). 26. Kannan, R. Y. , Salacinski, H. J. , Sales, K. , Butler, P. , and Seifalian, A. M. : The roles of tissue engineering and vascularisation in the development of micro-vascular networks: a review. Biomaterials, 26, 1857–1875 (2005). 27. Fassnacht, D. and Portner, R. : Experimental and theoretical considerations on oxygen supply for animal cell growth in fixed bed reactors. J. Biotechnol. , 72, 169–184 (1999). 28. Butler, D. L. , Goldstein, S. A. , and Guilak, F. : Functional tissue engineering: the role of biomechanics. J. Biomech. Eng. , 122, 570–575 (2000). 29.

Carver, S. E. and Heath, C. A. : Increasing extracellular matrix production in regenerating cartilage with intermittent physiological pressure. Biotechnol. Bioeng. , 62, 166–174 (1999). 30. Carver, S. E. and Heath, C. A. : Influence of intermittent pressure, fluid flow, and mixing in the regenerative properties of articular chondrocytes. Biotechnol. Bioeng. , 65, 274–281 (1999). 31. Hall, A. C. , Urban, J. P. G. , and Gehl, K. A. : The effects of hydrostatic pressure on matrix synthesis in articular cartilage. J. Othop. Res. , 9, 1–10 (1991). 32. Jagodzinski, M. , Cebotari, S. , Tudorache, I. , Zeichen, J. Hankemeier, S. , Krettek, C. , van Griensven, M. , and Mertsching, H. : Tissue engineering of long bones with a vascular matrix in a bioreactor. Orthopade, 33, 1394–1400 (2004). 33. Seidel, J. O. , Pei, M. , Gray, M. L. , Langer, R. , Freed, L. E. , and Vunjak-Novakovic, G. : Long-term culture of tissue engineered cartilage in a perfused chamber with mechanical stimulation. Biorheology, 41, 445–458 (2004). 34. Yu, X. , Botchwey, E. A. , Levine, E. M. , Pollack, S. R. , and Laurencin, C. T. : Bioreactor-based bone tissue engineering: 244 PORTNER ET AL. J. BIOSCI. BIOENG. , 35. 36. 37. 38. 39. 40. 41. 42. 3. 44. 45. 46. 47. 48. 49. 50. 51. the influence of dynamic flow on osteoblast phenotypic expression and matrix mineralization. Proc. Natl. Acad. Sci. USA, 101, 11203–11208 (2004). Minuth, W. W. , Stockl, G. , Kloth, S. , and Dermietzel, R. : Construction of an apparatus for perfusion cell cultures which enables in vitro experiments under organotypic conditions. Eur. J. Cell. Biol. , 57, 132–137 (1992). Sittinger, M. , Schultz, O. , Keyser, G. , Minuth, W. W. , and Burmester, G. R. : Artificial tissues in perfusion culture. Int. J. Artif. Organs, 20, 57–62 (1997). Fassnacht, D. , Rossing, S. , Stange, J. and Portner, R. : Long-term cultivation of immortalised mouse hepatocytes in a high cell density fixed bed reactor. Biotechnol. Tech. , 12, 25–30 (1998). Nehring, D. , Adamietz, P. , Meenen, N. M. , and Portner, R. : Perfusion cultures and modelling of oxygen uptake with three-dimensional chondrocyte pellets. Biotechnol. Tech. , 13, 701–706 (1999). Morgan, J. R. and Yarmush, M. L. : Tissue engineering methods and protocols. Humanda Press, Totowa, NJ (1999). Sen, A. , Kallos, M. S. , and Behie, L. A. : New tissue dissociation protocol for scaled-up production of neural stem cells in suspension bioreactors.

Tissue Eng. , 10, 904–913 (2004). Fassnacht, D. , Rossing, S. , Singh, R. , Al-Rubeai, M. , and Portner, R. : Influence of bcl-2 expression on antibody productivity in high cell density hybridoma culture systems. Cytotechnology, 30, 95–105 (1999). Fussenegger, M. , Fassnacht. , D. , Schwartz, R. , Zanghi, J. A. , Graf, M. , Bailey, J. E. , and Portner, R. : Regulated overexpression of the survival factor bcl-2 in CHO cells increases viable cell density in batch culture and decreases DNA release in extended fixed-bed cultivation. Cytotechnology, 32, 45–61 (2000). Nehring, D. , Gonzales, R. Czermak, P. , and Portner, R. : Mathematical model of a membrane filtration process using ceramic membranes to increase retroviral pseudotype vector titer. J. Membr. Sci. , 237, 25–38 (2004). Noll, T. , Jelinek, N. , Schmid, S. , Biselli, M. , and Wandrey, C. : Cultivation of hematopoietic stem and progenitor cells: biochemical engineering aspects. Adv. Biochem. Eng. Biotechnol. , 74, 111–128 (2002). Cabrita, G. J. , Ferreira, B. S. , da Silva, C. L. , Goncalves, R. , Almeida-Porada, G. , and Cabral, J. M. : Hematopoietic stem cells: from the bone to the bioreactor. Trends Biotechnol. 21, 233–240 (2003). Schubert, H. , Garrn, I. , Berthold, A. , Knauf, W. U. , Reufi, B. , Fietz, T. , and Gross, U. M. : Culture of haematopoietic cells in a 3-D bioreactor made of Al2O3 or apatite foam. J. Mater. Sci. Mater. Med. , 15, 331–334 (2004). Davis, J. M. and Hanak, J. A. : Hollow-fiber cell culture. Methods Mol. Biol. , 75, 77–89 (1997). Weichert, H. , Falkenberg, F. W. , Krane, M. , Behn, I. , Hommel, U. , and Nagels, H. O. : Cultivation of animal cells in a new modular minifermenter, p. 907–913. In Beuvery, E. C. , Griffiths, J. B. , and Zeijlemaker, W. P. (ed. , Animal cell technology: developments towards the 21st century. Kluwer Academic Publishers, The Netherlands (1995). Nagel, A. , Effenberger, E. , Koch, S. , Lubbe, L. , and Marx, U. : Human cancer and primary cell culture in the new hybrid bioreactor system tecnomouse, p. 296–298. In Spier, R. E. , Griffiths, J. B. , and Berthold, W. (ed. ), Animal cell technology: products of today, prospects for tomorrow. ButterworthHeinemann, Oxford (1994). De Bartolo, L. and Bader, A. : Review of a flat membrane bioreactor as a bioartificial liver. Ann. Transplant. , 6, 40–46 (2001). Jasmund, I. and Bader, A. Bioreactor developments for tissue engineering applications by the example of the bioartificial liver. Adv. Biochem. Eng. Biotechnol. , 74, 99–109 (2002). 52. Kulig, K. M. and Vacanti, J. P. : Hepatic tissue engineering. Transpl. Immunol. , 12, 303–310 (2004). 53. Gerlach, J. C. : Development of a hybrid liver support system: a review. Int. J. Artif. Organs, 19, 645–654 (1996). 54. Sielaff, T. D. , Hu, M. Y. , Amiot, B. , Rollins, M. D. , Rao, S. , McGuire, B. , Bloomer, J. R. , Hu, W. S. , and Cerra, F. B. : Gel-entrapment bioartificial liver therapy in galactosamine hepatitis. J. Surg. Res. , 59, 179–184 (1995). 55.

Prenosil, J. E. and Villeneuve, P. E. : Automated production of cultured epidermal autografts and sub-confluent epidermal autografts in a computer controlled bioreactor. Biotechnol. Bioeng. , 59, 679–683 (1998). 56. Marx, U. , Matthes, H. , Nagel, A. , and Baehr, R. V. : Application of a hollow fiber membrane cell culture system in medicine. Am. Biotechnol. Lab. , 11, 26 (1993). 57. Ratcliffe, A. : Tissue engineering of vascular grafts. Matrix Biol. , 19, 353–357 (2000). 58. Risbud, M. V. and Sittinger, M. : Tissue engineering: advances in in vitro cartilage generation. Trends Biotechnol. , 20, 351– 356 (2002). 59. Minuth, W. W. Strehl, R. , and Schumacher, K. : Tissue engineering — from cell biology to artificial organs. WileyVCH Verlag, Weinheim (2005). 60. Lima, E. G. , Mauck, R. L. , Shelley, H. H. , Park, S. , Ng, K. W. , Ateshian, G. A. , and Hung, C. T. : Functional tissue engineering of chondral and osteochondral constructs. Biorheology, 41, 577–590 (2004). 61. Raimondi, M. T. , Boschetti, F. , Falcone, L. , Fiore, G. B. , Remuzzi, A. , Marinoni, E. , Marazzi, M. , and Pietrabiss, R. : Mechanobiology of engineered cartilage cultured under a quantified fluid-dynamic environment. Biomech. Model. Mechanobiol. , 1, 69–82 (2002). 62. Connelly, J.

T. , Vanderploeg, E. J. , and Levenston, M. E. : The influence of cyclic tension amplitude on chondrocyte matrix synthesis: experimental and finite element analyses. Biorheology, 41, 377–387 (2004). 63. Sengers, B. G. , Oomens, C. W. J. , and Baaijens, F. P. T. : An integrated finite-element approach to mechanics, transport and biosynthesis in tissue engineering. J. Biomech. Eng. , 126, 83–91 (2004). 64. Williams, K. A. , Saini, S. , and Wick, T. M. : Computational fluid dynamics modelling of steady-state momentum and mass transport in a bioreactor for cartilage tissue engineering. Biotechnol. Prog. , 18, 951–963 (2002). 65.

Mauck, R. L. , Hung, C. T. , and Ateshian, G. A. : Modeling of neutral solute transport in a dynamically loaded porous permeable gel: implications for articular cartilage biosynthesis and tissue engineering. J. Biomech. Eng. , 125, 602–614 (2003). 66. Begley, C. M. and Kleis, S. J. : The fluid dynamic and shear environment in the NASA/JSC rotating-wall perfused-vessel bioreactor. Biotechnol. Bioeng. , 70, 32–40 (2000). 67. Omasa, T. , Kishimoto, M. , Kawase, M. , and Yagi, K. : An attempt at decision making in tissue engineering: reactor evaluation using the analytic hierarchy process (AHP). Biochem. Eng. J. , 20, 173–179 (2004). 8. Petersen, J. P. , Rucker, A. , von Stechow, D. , Adamietz, P. , Portner, R. , Rueger, J. M. , and Meenen, N. M. : Present and future therapies of articular cartilage defects. Eur. J. Trauma, 1, 1–10 (2003). 69. Freed, L. E. , Martin, I. , and Vunjak-Novakovic, G. : Frontiers in tissue engineering. In vitro modulation of chondrogenesis. Clin. Orthop. , 367 (Suppl. ), 46–58 (1999). 70. Buschmann, M. D,, Gluzband, Y. A. , Grodzinsky, A. J. , Kimura, J. H. , and Hunziker, E. B. : Chondrocytes in agarose culture synthesize a mechanically functional extracellular matrix. J. Orthop. Res. , 10, 745–758 (1992). 71. Waldman, S. D. Grynpas, M. D. , Pilliar, R. M. , and Kandel, VOL. 100, 2005 BIOREACTOR DESIGN FOR TISSUE ENGINEERING 245 R. A. : Characterization of cartilaginous tissue formed on calcium polyphosphate substrate in vitro. J. Biomed. Mater. Res. , 62, 323–330 (2002). 72. Malda, J. , Martens, D. E. , Tramper, J. , van Blitterswijk, C. A. , and Riesle, J. : Cartilage tissue engineering: controversy in the effect of oxygen. Crit. Rev. Biotechnol. , 23, 175– 194 (2003). 73. O’Driscoll, S. W. , Fitzsimmons, J. S. , and Commisso, C. N. : Role of oxygen tension during cartilage formation by periosteum. J. Orthop. Res. , 15, 682–687 (1997). 74.

Obradovic, B. , Carrier, R. L. , Vunjak-Novakovic, G. , and Freed, L. E. : Gas exchange is essential for bioreactor cultivation of tissue engineered cartilage. Biotechnol. Bioeng. , 63, 197–205 (1999). 75. Ysart, G. E. and Mason, R. M. : Responses of articular cartilage explant cultures to different oxygen tensions. Biochim. Biophys. Acta, 1221, 15–20 (1994). 76. Malda, J. , Rouwkema, J. , Martens, D. E. , Le Comte, E. P. , Kooy, F. K. , Tramper, J. , van Blitterswijk, C. A. , and Riesle, J. : Oxygen gradients in tissue-engineered PEGT/PBT cartilaginous constructs: measurement and modelling. Biotechnol. Bioeng. , 86, 9–18 (2004).