Biomaterials
Volume 24, Issue 26 , November 2003, Pages 4761-4770

Chondrocyte interactions with porous titanium alloy and calcium polyphosphate substrates

V. J. D. Ciolfi^a, R. Pilliar^a, C. McCulloch^b, S. X. Wang^c, M. D. Grynpas^{a, c} and R. A. Kandel<sup>, , a, c

a</sup> Institute of Biomaterials and Bioengineering, Mount Sinai Hospital, University of Toronto, Ontario M5G 1X5, Canada
^b Faculty of Dentistry, Mount Sinai Hospital, University of Toronto, Ontario M5G 1X5, Canada
^c Department of Pathology and Laboratory Medicine and Samuel Lunenfeld Research Institute, Mount Sinai Hospital, University of Toronto, Ontario M5G 1X5, Canada

Received 25 October 2002;  accepted 13 May 2003. ; Available online 28 June 2003.

1. Abstract

Chondrocytes maintain their phenotype and form cartilagenous tissue when cultured on calcium polyphosphate (CPP) or titanium alloy (Ti alloy), porous three-dimensional materials. To understand how these materials may influence chondrocyte phenotype and matrix synthesis, the early interactions of cultured cells with CPP and titanium alloy were examined. These were compared to chondrocytes grown in monolayer culture on tissue culture polystyrene, conditions in which cultured chondrocytes dedifferentiate and do not form cartilagenous tissue. Scanning electron microscopy of cells up to 72 h in culture showed that bovine chondrocytes on CPP, Ti alloy, and polystyrene were an admixture of round and spread cells. The spread cells on CPP and titanium alloy were not entirely flattened but maintained a polygonal shape. In contrast, spread chondrocytes in monolayer culture were flatter and significantly larger, a difference that was maintained even in the absence of serum. All cells cultured on CPP and Ti alloy exhibited subcortical ring-like distribution of actin filaments whereas the flattened cells on polystyrene showed actin filaments distributed throughout the cytoplasm. Cells on CPP and Ti alloy synthesized significantly less collagen and proteoglycans than cells cultured on polystyrene at 72 h of culture. In summary the cells on the porous three-dimensional materials differed from those on polystyrene in terms of cell morphology and size, actin cytoskeleton organization, and synthesis of selected matrix macromolecules. The data suggests that CPP and titanium alloy may mediate their effect by limiting cell spreading in part by favoring the maintenance of a ring-like actin distribution.

Author Keywords: Materials; Chondrocytes; Morphology; Cytoskeleton; Proteoglycan; Collagen

2. Article Outline

1. Introduction

2. Materials and methods

2.1. Formation of calcium polyphosphate and titanium alloy substrates

2.2. Cell culture

2.3. Scanning electron microscopy

2.4. Confocal microscopy

2.5. Proteoglycan and collagen synthesis

2.6. DNA content

2.7. Quantification of cell attachment

2.8. Statistical analyses

3. Results

3.1. Disc characteristics

3.2. Percent of cells attached

3.3. Cell morphology after attachment

3.4. Extent of cell spreading on materials

3.5. Actin cytoskeleton

3.6. Proteoglycan and collagen synthesis

4. Discussion

Acknowledgements

References

3. 1. Introduction

Damage to articular cartilage can result in joint pain and limited mobility. As articular cartilage in adult humans has limited, if any, potential for repair, various methods to restore the surface of the damaged sites have been developed [1, 2, 3, 4, 5 and 6]. One procedure used currently involves transplantation of osteochondral plugs (subchondral bone and overlying articular cartilage) obtained from donor sites within the same joint, into the cartilage defect [7, 8, 9, 10, 11, 12 and 13]. Several plugs are usually required to repair an area that may be several centimeters in diameter. This treatment (`mosaicplasty') has been widely used over the past several years but has several major disadvantages including donor site morbidity, movement of implanted plugs relative to each other, the formation of repair tissue between plugs (i.e. spaces between juxtaposed round plugs), and the unpredictable integration of transplanted plugs with host tissues [3, 11, 12 and 13]. Misalignment of the grafts during placement can compromise graft viability because of formation of fibrous tissue at the perimeter of the grafts and/or abrasion of graft cartilage if protruding above the intended surface of articulation.

To circumvent some of these problems we have developed an approach that consists of generating a biphasic (i.e. cartilage and porous bone substitute) "osteochondral-type" plug in vitro [14, 15 and 16]. This approach precludes the need to harvest donor plugs. The biphasic plug is composed of cartilagenous tissue that is formed by isolated chondrocytes placed on the surface of the porous substrate (intended articulation surface). The porous substrate is generated in vitro and acts as a bone substitute [15]. The material is designed to allow both predictable cartilage-to-substrate anchorage as well as fixation to bone by bone ingrowth into the pores not filled with cartilage [17]. As the cartilage is already formed and anchored, lateral integration of the implanted cartilage to the adjacent host cartilage is possible immediately after implant insertion. There is no donor site morbidity and rather than a `mosaic' arrangement of smaller components, one plug, which can be custom shaped to fill the defect, is generated. Initial studies demonstrate that the cartilage tissue that forms in vitro is similar to native cartilage, although like other tissue-engineered cartilage it contains less collagen than the native tissue [16, 18 and 19]. Both porous calcium polyphosphate (CPP) and porous titanium alloy (Ti6Al4 V) have been utilized as substrates in these studies [16 and 20]. Although these substrates have very different compositions and surface roughness, chondrocytes cultured on the top surface of these materials will form a continuous layer of cartilagenous tissue. The cells maintain their phenotype as characterized by the synthesis of large proteoglycans and type II collagen (no type I collagen detected) [16 and 20]. These observations contrast with chondrocytes grown in monolayer culture, in which cells will dedifferentiate and lose their chondrogenic potential and will not form cartilagenous tissue [21, 22 and 23].

To understand how these materials may influence chondrocyte phenotype and tissue formation, the attachment of the cells, the morphology and size of the attached chondrocytes and the organization of actin cytoskeleton during the first 3 days of culture on CPP, titanium alloy, and polystyrene were examined. The synthesis of the major cartilage matrix molecules, collagen and proteoglycans was also quantified. Elucidating the differences between materials that favor cartilage tissue formation and those that do not will allow us to design materials with the optimal characteristics for tissue engineering of cartilage.

4. 2. Materials and methods

2.1. Formation of calcium polyphosphate and titanium alloy substrates

The formation of calcium polyphosphate substrates has been previously described [15]. Briefly, calcium phosphate monobasic monohydrate powder (nCa(H₂PO₄)₂·H₂O) (J.T. Baker, Phillipsburg N.J.) was calcined at 500°C for 10 h to produce calcium polyphosphate [Ca(PO₃)₂]_n which was melted to an amorphous glass by heating to 1100°C for 1 h and then quenched in distilled water to form an amorphous frit. Powder (75-106 
m) was produced by ball milling this glass frit followed by screening for the desired size range (−40/+200 mesh). Porous CPP discs were produced by gravity sintering the resulting powder at 950°C for 2 h. The CPP cylinder (4 mm diameter) was cross-sectioned into 2 mm thick discs.

The titanium alloy (Ti6Al4 V) substrate was formed by sintering titanium alloy particles in the size range of 75-106 
m as described previously [20]. The discs were cleaned for 2 h in 2% Decon, rinsed 3 times in distilled water, and then placed in 28% nitric acid solution for 1 h in order to promote formation of a continuous passive oxide layer. This represents the normal practice with sintered porous-surfaced implants although it is recognized that a passive oxide layer will form spontaneously in air with this alloy. After treatment, the discs were rinsed five times with distilled water.

In order to keep the cells from spilling over the edge of the discs, both substrates (CPP and Ti alloy) were inserted into tubing (Bev-a-line, Cole Palmer Instruments Co., Vernon, IL) that served to form a `well' to hold the cells on the top of the discs. The tube-encased materials (CPP or Ti alloy) were sterilized by gamma radiation (3.5 Mrad) in preparation for cell culture.

2.2. Cell culture

Cartilage was harvested from bovine metacarpal-carpal joints under sterile conditions. The animals were approximately 9 months of age. The tissue from 2 to 6 animals (depending on the number of cells needed) were pooled together for each experiment in order to obtain sufficient cells. The chondrocytes were isolated by sequential enzymatic digestion as previously described [24]. The cells were resuspended in Ham's F-12 supplemented with 25 m
HEPES and 5% fetal bovine serum (Sigma, St. Louis, MI) and placed on the top of the discs at either 0.6×10⁶ (for confocal microscopy studies) or 2.0×10⁶ cells (for all other experiments) per disc or as a monolayer in 24 well plates (5×10⁵ cells/cm²). The cultures were maintained for various times up to 120 h under standard cell culture conditions. The medium was replaced at 48 and 96 h for the appropriate experiments.

For the serum-free experiments, the freshly isolated chondrocytes were collected into Ham's F-12 media supplemented with 2% FBS and allowed to recover for 10 min. The chondrocytes were then pelleted by centrifugation, washed three times with serum-free Ham's F-12, and seeded as described above in Ham's F-12 media only.

2.3. Scanning electron microscopy

For morphological analysis, samples were fixed in 2% glutaraldehyde for 1 h and post-fixed in osmium tetroxide for 1 h. The cells were dehydrated in an ethanol series and critical-point dried. The samples were imaged with a scanning electron microscope (Hitachi S2500, Mito City, Japan) and the images were imported into Image J program for morphometry (available at http://rsb.info.nih.gov/ij/). The scale was set using that of the SEM images. The projected area of the chondrocytes was determined by tracing the cell outline and calculating the area. Between 88 to 138 cells were evaluated/time point for cells placed on titanium alloy or CPP. For polystyrene between 20 and 45 cells were evaluated.

2.4. Confocal microscopy

At 24 hours, the cells on the materials were fixed in 4% paraformaldehyde for 10 min and then permeabilized with 0.2% Triton X-100 (Sigma). Cells were stained for actin filaments using Alexa Fluor 568 phalloidin (1:6 dilution in PBS, Molecular Probes, Eugene, OR) for 20 minutes at room temperature. Cells were washed 3 times with PBS (containing Ca²⁺ and Mg²⁺) and the titanium alloy or CPP substrates were mounted onto a glass slide with the cell-containing surface at the top. The specimens were coverslipped using Immuno Floure mounting medium (ICN Biomedical Inc., Aurora, OH). Images were obtained with a confocal laser scanning microscope (Zeiss LSM 410). In some experiments, actin filaments were depolymerized with cytochalasin D (2 
g/ml, Sigma) prior to staining as a control.

Native cartilage was harvested, fixed in 4% paraformaldehyde for 10 minutes and then frozen. Five micron frozen sections were cut and the tissue then stained and examined similarly as the cells on the materials.

2.5. Proteoglycan and collagen synthesis

The effect of cell attachment to the different materials on the synthesis of matrix molecules was evaluated. To estimate proteoglycan synthesis, the cells were incubated with [³⁵S]-SO₄ (4 
Ci/sample) 24 h prior to harvest. The medium was collected and proteoglycans were precipitated with cold 70% ethanol. The pellet was washed and re-suspended in guanidinium HCl. The cells were digested with papain (Sigma; 40 
g/ml in 20 m
ammonium acetate, 1 m
EDTA, and 2 m
dithiotreitol) for 72 h at 65°C. Aliquots from the papain digest and the media were counted in a
-scintillation counter, added together and normalized to DNA content. Each time point was done in quadruplicate and three independent experiments were done.

To assess collagen synthesis, cells were incubated with [¹⁴C]-proline (5 
Ci/sample) for 24 h prior to harvest. The medium was collected and the collagen was precipitated overnight at 4°C with ammonium sulphate (30% final concentration). The medium was centrifuged at 14,000g for 24 min, and washed with 70% ethanol to remove free [¹⁴C]-proline. The pellets were re-suspended in 200 
l of 10% SDS. The collagen in the cell cultures was extracted with pepsin (Sigma, 200 
g/ml in 0.5 
acetic acid) for 72 h at 4°C. The samples were neutralized with 10 
NaOH and aliquots of the pepsin extract and the medium were counted in a
-scintillation counter, combined together and normalized to DNA. To determine the percentage of [¹⁴C]-proline incorporated into collagen, an aliquot of the pepsin extract was digested overnight at 37°C with bacterial collagenase (100 U, Worthington Biochemical Corp., Lakewood, NJ). The collagenase was heat-inactivated and the digest filtered using Micron 30 filter. The filtrate was collected and counted in the
-scintillation counter. Each time point was done in quadruplicate and three independent experiments were done.

2.6. DNA content

The cultures at the appropriate time were papain digested and an aliquot was incubated with Hoechst 33258 dye as described previously [16 and 25]. The DNA was quantified using fluorimetry (excitation WAVELENGTH=365 nm and emission WAVELENGTH=458 nm). The standard curve used in the analysis was generated using calf thymus DNA (Sigma).

2.7. Quantification of cell attachment

To determine the percentage of cells that attached to the discs and polystyrene, the freshly harvested cells (2×10⁶ cells) and the cultures after 24 h were papain digested and DNA content determined as described above. To calculate the percent of cells attached, the number of cells on the material at 24 h, as determined by quantifying DNA, was divided by the number of cells placed on it initially. Seven separate experiments were done and for every experiment quadruplicate samples of each material were evaluated (n=28 for each material).

2.8. Statistical analyses

All results are presented as the mean±the standard error of the mean (SEM). Two-way analysis of variance (ANOVA) was used to determine statistical significance. All pair-wise comparisons between groups were conducted with the use of the Fisher LSD post hoc test and statistical significance was set at p
0.05.

5. 3. Results

3.1. Disc characteristics

Consistent with our previous experience, the sintering conditions selected for both the CPP and Ti alloy powders resulted in the desired volume percent porosity of approximately 35%. The pores were interconnected and resulted in a distribution of pore sizes that was nearly normally distributed with most pores in the 20-100 
m size range. Due to the different characteristics of the particles in the two preparations (CPP powders: angular and irregular; Ti alloy powders: sphere-like; Fig. 1), the resulting pore shapes after sintering were different. This difference is illustrated in the low magnification scanning electron micrographs shown in Figs. 1A and C. Notably, the three-dimensional interconnected pore network was present in both materials. Higher magnification examination of the CPP sintered structures demonstrated two additional features ( Fig. 1D). As observed previously [15] in addition to the 20-100 
m sized pores, there was also a distribution of submicron-sized pores throughout the structure and a distinctly different appearance of the surface resulting from cutting of the CPP cylindrical rods to form the discs. Further, as shown in Fig. 1D, 1-5 
m size CPP crystals formed within the particles during the annealing of the initially glassy CPP particles at 950°C. Higher magnification examination of the Ti alloy sintered discs also showed some additional features characteristic of sinter-annealed Ti6Al4 V formed during the vacuum sintering conditions employed here. Fig. 1B shows a high magnification view of a sinter neck region as well as a regular surface pattern resulting from thermal etching occurring during the high temperature sintering [26]. The striated pattern results in a fine, submicron series of ledges or plateaux related to crystallographic planes of the Ti alloy.

(78K)

Fig. 1. Scanning electron micrographs of the titanium alloy (A, B) and calcium polyphosphate (C,D) discs.

3.2. Percent of cells attached

The percentage of cells that attached to the different materials varied. There were more cells attached to CPP (66±2%) than Ti alloy (56±2%) and polystyrene (48±3%). However the difference in percent cell attachment between Ti alloy and polystyrene were not significant (p=0.07).

3.3. Cell morphology after attachment

The morphology of chondrocytes attached to the CPP, Ti alloy or tissue culture polystyrene was assessed by scanning electron microscopy at 24 and 72 h after plating. Round and spread cells were seen attached to all three materials. The spread chondrocytes on CPP and Ti alloy had a polygonal shape (Fig. 2) which was maintained even in the absence of serum (data not shown). This shape was different than the chondrocytes in monolayer culture in which the spread cells appeared flatter and larger than the polygonal cells on the Ti alloy or CPP.

(43K)

Fig. 2. Scanning electron micrographs showing the morphology of chondrocytes attached to calcium polyphosphate (CPP), titanium alloy (Ti alloy) or tissue culture polystyrene (monolayer culture, PS). The cultures were harvested at either 24 or 72 h after plating and processed for scanning electron microscopy as described under Section 2.

3.4. Extent of cell spreading on materials

To quantify the extent of cell spreading on the different materials, cell area was determined using the scanning electron microscopy and the Image J program. Only cells that were in direct contact with materials were measured and the results are summarized in Fig. 3. At 24 h the cells on the CPP were significantly more spread than the cells on the Ti alloy (81±5 
m² and 53±2 
m² respectively, p<0.001) but the cells on both materials were smaller than the cells grown in monolayer (105±11 
m²). After 72 h the cells on the CPP showed no significant increase in cell area. However the cells on the Ti alloy were slightly more spread with a mean area of 65±4 
m² but were still less spread than cells grown in monolayer (289±16 
m²).

(7K)

Fig. 3. The chondrocytes were placed on either calcium polyphosphate (CPP), titanium alloy (Ti alloy) or tissue culture polystyrene (monolayer culture, PS) in the presence (dark bars) or absence (light bars) of 5% fetal bovine serum. The chondrocytes and the substrates were collected after 24 or 72 h and processed for scanning electron microscopy. The cell size was determined by measuring the cell area in the scanning electron micrographs as described under Section 2. The results were pooled and expressed as the mean ± SEM.

To determine whether it was serum proteins rather than the material that influenced cell spreading on the materials, the chondrocytes were plated and grown in the absence of serum. The presence of serum did affect cell spreading as cells on all materials (Ti alloy, CPP, polystyrene) were larger at 72 h than cells grown in the absence of serum. However the cells grown in monolayer under serum-free conditions were still significantly more spread than cells on the two porous substrates at both 24 and 72 h. The cells on Ti alloy and CPP in the absence of serum were of similar size.

3.5. Actin cytoskeleton

As the organization of the actin cytoskeleton has been shown to correlate with chondrocyte differentiation and gene expression [27], the distribution of actin filaments was assessed in cells attached to the different materials ( Fig. 4). The cells, whether grown on CPP or Ti alloy, showed a similar distribution of actin filaments which was different when compared to cells in monolayer culture. At 24 h the cells on CPP and Ti alloy showed a ring-like subcortical distribution of actin filaments in both the round and polygonal cells. This distribution was maintained even at 72 h (data not shown). In contrast, the spread cells in monolayer culture had a more random distribution of actin filaments that extended through the cytoplasm. The cells that remained rounded at these early time points (up to 72 h) in monolayer culture also showed a prominent subcortical distribution of actin filaments similar to cells grown on porous substrates. As a control the actin distribution in chondrocytes in native cartilage was also examined. The cells, similar to the cells on the porous substrates, showed a ring-like subcortical distribution of actin filaments suggesting that this is the preferred actin organization.

(104K)

Fig. 4. Chondrocytes were placed on either calcium polyphosphate (A), titanium alloy (B) or tissue culture polystyrene (monolayer culture, (C) and were collected after 24 h of culture. The cells were fixed, permeabilized, stained with phalloidin and examined by confocal microscopy to visualize the actin cytoskeleton. The cells on CPP and Ti alloy and the round cells on polystyrene show a ring-like subcortical distribution of actin. However the spread cells in monolayer culture had actin filaments throughout the cytoplasm and did not show a subcortical distribution. Arrowhead (
) indicates spread cells. As a control native cartilage (D) was stained with phalloidin and similar to the in vitro grown cells showed a ring-like distribution of actin.

3.6. Proteoglycan and collagen synthesis

Proteoglycan and collagen are the major matrix components of cartilage [6]. The effect of the materials on synthesis of these macromolecules was examined. Proteoglycan synthesis was determined by quantifying the amount of [³⁵S]-SO₄ incorporated into proteoglycans (Fig. 5). The cells on both CPP and Ti alloy synthesized significantly less proteoglycans than the cells in monolayer culture by 72 h and this difference was maintained at 120 h (p<0.0005 for both times).

(11K)

Fig. 5. Total proteoglycan synthesis by chondrocytes was determined by quantifying [³⁵S]-SO₄ incorporation by cells attached to either calcium polyphosphate (diagonal lines), titanium alloy (cross-hatched lines) or tissue culture polystyrene (horizontal lines) was determined 24, 72 or 120 h after the placement of the chondrocytes on the materials. Synthesis was normalized to DNA. The results from all the separate experiments were pooled and expressed as the mean ± SEM.

The chondrocytes cultured on Ti alloy synthesized significantly more proteoglycans at 24 h than cells on CPP (p<0.0005) but similar amounts to chondrocytes in monolayer culture. This difference was observed in 5 out of 7 experiments performed. By 72 h the difference in proteoglycan synthesis by cells on CPP and Ti alloy was no longer present. The cells on the porous substrates showed a significant decrease in proteoglycan synthesis over the time examined (p<0.001) whereas the cells cultured on polystyrene showed no significant difference in synthesis up to 120 h.

The chondrocytes in monolayer culture synthesized significantly more collagen than chondrocytes on CPP or Ti alloy up to 120 h. The cells on CPP and titanium alloy showed no significant difference in collagen synthesis at any time points and no change in synthesis over time (Fig. 6). Bacterial collagenase digestion of the pepsin extracts demonstrated that approximately 94±2% (n=3) of the [¹⁴C]-proline was incorporated into collagen.

(7K)

Fig. 6. Total collagen synthesis by chondrocytes was determined by quantifying [¹⁴C]-proline incorporation by cells attached to either calcium polyphosphate (diagonal lines), titanium alloy (cross-hatched lines) or tissue culture polystyrene (horizontal lines) was determined 24, 72 or 120 h after the attachment of the chondrocytes to the materials. Synthesis was normalized to DNA. The results from all the separate experiments were pooled and expressed as the mean±SEM.

6. 4. Discussion

In previous studies we observed that when articular chondrocytes were grown for long periods on the surface of porous substrates, titanium alloy or CPP, the chondrocyte phenotype was maintained and the cells accumulated matrix and formed cartilagenous tissue [16, 18 and 20]. This finding contrasts with chondrocytes grown in monolayer culture on polystyrene as these cells "dedifferentiate" over time and do not form cartilage tissue [21, 22 and 23]. These different materials utilized as tools to determine if there were cellular characteristics that could be identified that might favor cartilage tissue formation. This study demonstrated that the CPP and titanium alloy had very different surfaces. Although the materials differ the cells on these porous three-dimensional substrates remained rounder and did not spread as much as cells in monolayer culture. This difference in cell morphology and cell area between the porous substrates and polystyrene was maintained even when the cells were grown in the absence of serum. The chondrocytes on the 3D substrate showed a ring-like distribution of actin, which differed from the random distribution in the flattened cells in monolayer culture. As cells on Ti alloy and CPP synthesized less collagen and proteoglycans than cells on polystyrene, this suggests that the total amount of matrix molecules produced may not be a factor influencing tissue formation, provided synthesis is not completely suppressed.

It is not clear whether it is the maintenance of the rounded shape or the organization of the actin cytoskeleton that determined whether chondrocytes remain differentiated on these porous substrates. Previous data implicates both of these mechanisms in influencing chondrocyte phenotype. In support of the role of cell shape, it has been shown that dedifferentiated chondrocytes can be induced to redifferentiate when grown in culture on agarose [23], alginate [28], or in suspension culture [29], conditions that restrict spreading and favour a rounded shape. Others have demonstrated that chondrocytes remaining round in culture, when grown in collagen gels or on alginate beads, also maintained their phenotype [30 and 31]. In these studies the organization of the actin filaments was not examined. However it has been shown that rounded chondrocytes can synthesize type I collagen and spread chondrocytes can express type II collagen mRNA and protein [32 and 33], suggesting that cell shape may not be critical in influencing chondrocyte differentiation.

A role for the organization of the actin cytoskeleton in regulating chondrocyte differentiation is supported by the findings of several studies. Dedifferentiated chondrocytes can be induced to synthesize type II collagen when the cells are incubated in the presence of cytochalasin B, an agent that disrupts actin filaments [27]. No change in cell shape was observed in that study. Others have observed that the cells with "large actin cables" (likely stress fibers) synthesized type I collagen, a marker of chondrocyte dedifferentiation [34]. More recently, Vinall and Reddi demonstrated that articular chondrocyte phenotype modulation by BMP-7 and IL-1 was dependent on the cytoskeleton [35 and 36]. As chondrocyte shape is determined in part by the organization of the actin cytoskeleton, it is likely that actin is relatively more important in influencing chondrocyte phenotype and function [36]. Interestingly the chondrocytes grown on the porous substrates in our study exhibited a ring-like subcortical distribution of actin, which is similar to that exhibited by chondrocytes in vivo as we and others have shown [37]. Notably, we observed that cells that were round in monolayer culture exhibited a similar distribution of actin filaments as cells grown on the porous materials. This observation raises the possibility that it may not be the material itself that induces a more round shape and the formation of a ring-like actin organization; rather it may be the geometry of the porous CPP and titanium alloy substrates that provides the conditions that favour the maintenance of the chondrogenic phenotype.

The cells on CPP were more spread than the cells on Ti alloy at 24 h but this difference disappeared by 72 h. Calcium phosphates (hydroxyapatite) have been shown to bind more serum proteins, such as fibronectin and vitronectin, than titanium raising the possibility that CPP, which contains calcium and phosphate but with a different Ca:P ratio than apatite, may have bound more proteins by the 24 h period [38]. This increased binding may favor spreading and could be the reason why there was greater cell spreading on CPP at 24 h compared to the Ti alloy. The differential protein adhesion would also explain why the difference in cell area was not seen when the cells were grown on Ti alloy or CPP in the absence of serum and why it was not maintained at the later time points when the cells were grown in the presence of serum. Presumably, after 24 h, there was sufficient protein adsorption on Ti alloy surfaces to facilitate cell spreading. Alternatively differences between the CPP and Ti alloy in surface roughness or other material properties may explain the differences in cell spreading [39 and 40].

It is not clear why cells grown on polystyrene synthesize more proteoglycans and collagen by 72 h when compared to cells cultured on porous substrates. Although the difference may reflect a direct effect of the materials on biosynthesis, an alternative explanation may be that the actin cytoskeleton is involved as an intermediary. Other cultured cells, such as neurosecretory cells, exhibit subcortical rings of actin filaments and these structures can act as a barrier to exocytosis [41]. In rat pituitary cells the subcortical distribution of actin filaments suppressed prolactin secretion [42] and disruption of the actin filaments increased prolactin secretion [43]. Conceivably, the presence of subcortical actin filaments in cells cultured on CPP and titanium alloy is associated with limited secretion of matrix molecules.

Cells on CPP and Ti alloy synthesized measurable amounts of proteoglycans and collagen within 24 h of plating. Although the cells synthesized similar amounts of collagen, cells cultured on CPP produced less proteoglycans than the cells on titanium alloy during the first 24 h. There are at least three possible explanations for this observation. First, the difference in the extent of cell spreading on the two materials may have influenced proteoglycan synthesis. However, this explanation is considered less likely as cells attached to tissue culture plastic were significantly more spread than cells on the Ti alloy and there was no difference in the amount of proteoglycans synthesized per cell on either of these two materials at 24 h. Secondly, the cells on the CPP may have synthesized less proteoglycans because increased calcium levels in the medium. Within the first 24 h after immersion in the media there is dissolution of particles of CPP from the surfaces, which could potentially result in higher calcium levels. (This degradation results in negligible change in sample geometry as significant loss of CPP, approaching 50%, occurs only by 12 months as shown in the in vivo experiments [17]). Calcium can inhibit proteoglycan synthesis by chondrocytes as has been described previously by others [44]. In support of this mechanism, no difference in proteoglycan synthesis was detected between cells grown on CPP or Ti alloy at longer times, ie greater than 24 h. Thirdly, the difference in proteoglycan synthesis may be due to a direct effect of the substrates themselves on the cells [45, 46 and 47]. For example, chondrocytes grown on polyglycolic acid (PGA) exhibit enhanced proteoglycan synthesis compared to cells grown on collagen matrices [47]. Titanium surface roughness has been shown to influence proteoglycan and collagen synthesis by growth plate chondrocytes [45]. Finally it is possible that the CPP bound more of a proteoglycan stimulatory protein from the serum than the titanium [38]. Further study is required to determine which one of these mechanisms is responsible for influencing proteoglycan synthesis. Nevertheless the difference in proteoglycan synthesis by cells on these two materials did not affect their ability to form cartilagenous tissue as shown previously [16 and 20].

In conclusion, chondrocytes attached to 3D-porous substrates, Ti alloy and CPP, remained less spread than cells in monolayer culture and exhibited a ring-like, subcortical organization of actin filaments. The presence of these features may be involved in favouring tissue formation by the cells. The data also suggests that these substrates may exert their effect by limiting cell spreading in part by favoring the maintenance of the ring-like actin distribution. Further studies are ongoing to evaluate how these porous materials induce these responses in chondrocytes. Knowing this it should be possible ultimately to synthesize novel materials that would incorporate these characteristics and enhance cartilage tissue formation.

7. Acknowledgements

Supported by the National Science and Engineering Research Council, The Arthritis Society of Canada, the Canadian Arthritis Network and the Canadian Institutes of Health Research. We thank Hanje Chen and Tajinder Bhardwaj for technical assistance.

8. References

1. G. Bentley, Articular tissue grafts. Ann Rheum Dis 51 (1992), pp. 292-296. Abstract-EMBASE | Abstract-MEDLINE  

2. J.S. Temenoff and A.G. Mikos, Review: tissue engineering for regeneration of articular cartilage. Biomaterials 21 (2000), pp. 431-440. SummaryPlus | Full Text + Links | PDF (119 K)

3. J.E. Gilbert, Current treatment options for the restoration of articular cartilage. Am J Knee Surg 11 (1998), pp. 42-46. Abstract-MEDLINE  

4. S.R. Frenkel and P. Di Cesare, Degradation and repair of articular cartilage. Front in Biosci 4 (1999), pp. 671-685.

5. E.B. Hunziker, Articular cartilage repair: are the intrinsic biological constraints undermining this process insuperable?. Osteoarthritis Cart 7 (1999), pp. 15-28. Abstract | PDF (434 K)

6. Buckwalter JA, Rosenberg LC, Hunziker EB. Articular cartilage: composition, structure, response to injury, and methods of facilitating repair. In: Ewing JW editor. Articular cartilage and knee joint function: basic science and arthroscopy. New York: Raven Press Ltd.; 1990.

7. L. Hangody, G. Kish, Z. Karpati, I. Udvarhelyi, I. Szigeti and M. Bely, Mosaicplasty for the treatment of articular cartilage defects: application in clinical practice. Orthopedics 21 (1998), pp. 751-756. Abstract-MEDLINE | Abstract-EMBASE  

8. L. Hangody, P. Feczko, L. Bartha, G. Bodo and G. Kish, Mosaicplasty for the treatment of articular defects of the knee and ankle. Clin Orthop 391 (2001), pp. S328-S336. Abstract-MEDLINE | Abstract-EMBASE   | Full Text via CrossRef

9. L. Hangody, G. Kish, L. Modis, I. Szerb, L. Gaspar, Z. Dioszegi and Z. Kendik, Mosaicplasty for the treatment of osteochondritis dissecans of the talus: two to seven year results in 36 patients. Foot Ankle Int 22 (2001), pp. 552-558. Abstract-EMBASE | Abstract-MEDLINE  

10. J.G. Lane, W.L. Tontz, S.T. Ball, J.B. Massie, A.C. Chen, W.C. Bae, M.E. Amiel, R.L. Sah and D. Amiel, A Morphologic, biochemical, and biomechanical assessment of short-term effects of osteochondral autograft plug transfer in an animal model. Arthroscopy 17 (2001), pp. 856-863. Abstract-MEDLINE | Abstract-EMBASE  

11. S. Pearce, M.B. Hurtig, R. Clarnette, M. Kalra and A. Miniaci, Investigation of two techniques for optimizing joint surface congruency with mosaic arthroplasty. J Arthroscopy 17 (2001), pp. 50-55. Abstract | PDF (1040 K)

12. M. Hurtig, P. Evans, R. Clarnette, S. Pearce and K.M. Singh, Mosaic arthroplasty for resurfacing the equine third carpal bone. Vet Surg 30 (2001), pp. 228-239. Abstract-MEDLINE  

13. M. Hurtig et al., The effect of graft size and number on the outcome of mosaic arthroplasty resurfacing: An experimental model in sheep. Proc Arthroscopy Assoc NA 18 (2001), pp. 16-17.

14. M.J. Filiaggi, J.D. Wells, S.A.F. Peel, R.M. Pilliar, M.D. Grynpas and R.A. Kandel, Condensed calcium phosphates for soft tissue and bone repair/regeneration. Bioceramics 11 (1998), pp. 341-344.

15. R.M. Pilliar, M.J. Filiaggi, J.D. Wells, M.D. Grynpas and R.A. Kandel, Porous calcium polyphosphate scaffolds for bone substitute applications: in vitro characterization. Biomaterials 22 (2001), pp. 963-972. SummaryPlus | Full Text + Links | PDF (450 K)

16. S.D. Waldman, M.D. Grynpas, R.M. Pilliar and R.A. Kandel, Characterization of cartilaginous tissue formed on calcium polyphosphate substrates in vitro. J Biomed Mater Res 62 (2002), pp. 323-330. Abstract-MEDLINE | Abstract-EMBASE   | Full Text via CrossRef

17. M.D. Grynpas, R.M. Pilliar, R.A. Kandel et al., Porous calcium polyphosphate scaffolds for bone substitute applications: in vivo characterization. Biomaterials 23 (2002), pp. 2063-2070. SummaryPlus | Full Text + Links | PDF (736 K)

18. S.D. Waldman, C.G. Spiteri, M.D. Grynpas, R.M. Pilliar and R.A. Kandel, Long-term intermittent shear deformation improves the quality of cartilagenous tissue formed in vitro. J Orthop Res 20 (2003), pp. 590-596. SummaryPlus | Full Text + Links | PDF (315 K)

19. G. Vunjak-Novakovic, I. Matrin, B. Obradovic, S. Treppo, A.J. Grodzinsky, R. Langer and L.E. Freed, Bioreactor cultivation conditions modulate the composition and mechanical properties of tissue-engineered cartilage. J Orthop Res 17 (1999), pp. 1130-1138.

20. T. Bhardwaj, R.M. Pilliar, M.D. Grynpas and R.A. Kandel, Effect of material geometry on cartilaginous tissue formation in vitro. J Biomed Mater Res 57 (2001), pp. 190-199. Abstract-MEDLINE | Abstract-EMBASE   | Full Text via CrossRef

21. H.G. Coon, Clonal stability and phenotypic expression of chick cartilage cells in vitro. Proc Natl Acad Sci USA 55 (1966), pp. 66-73. Abstract-MEDLINE  

22. S. Chacko, J. Abbot, S. Holtzer and H. Holtzer, The loss of phenotypic traits by differentiated cells. VI. Behavior of the progeny of a single chondrocyte. J Exp Med 130 (1969), pp. 417-442. Abstract-MEDLINE  

23. P.D. Benya and J.D. Shaffer, Dedifferentiated chondrocytes reexpress the differentiated collagen phenotype when cultured in agarose gels. Cell 30 (1982), pp. 215-224. Abstract

24. J. Boyle, B. Luan, T.F. Cruz and R.A. Kandel, Characterization of proteoglycan accumulation during formation of cartilagenous tissue in vitro. Osteoarthritis Cart 3 (1995), pp. 117-125. Abstract-EMBASE | Abstract-MEDLINE  

25. Y-.J. Kim, R.L.Y. Sah, J-.Y.H. Doong and A.J. Grodinsky, Fluorometric assay of DNA in cartilage explants using Hoechst 33258. Anal Biochem 174 (1998), pp. 168-176.

26. R.M. Pilliar, Porous-surfaced metallic implants for orthopaedic applications. J Biomed Mater Res 21 A1 (1987), pp. 1-33. Abstract-MEDLINE | Abstract-EMBASE | Abstract-Compendex  

27. P.D. Benya, P.D. Brown and S.R. Padilla, Microfilament modification by dihydrocytochalasin b causes retinoic acid-modulated chondrocytes to reexpress the differentiated collagen phenotype without a change in shape. J Cell Biol 106 (1988), pp. 161-170. Abstract-EMBASE | Abstract-MEDLINE   | Full Text via CrossRef

28. J. Bonaventure, N. Kadhom, L. Cohen-Solal, K.H. Ng, J. Bourguignon, C. Lasselin and P. Freisinger, Reexpression of cartilage-specific genes by dedifferentiated human articular chondrocytes cultured in alginate beads. Exp Cell Res 212 (1994), pp. 97-104. Abstract

29. M. Solursh, Formation of cartilage tissue in vitro. J Cell Biochem 45 (1991), pp. 258-260. Abstract-MEDLINE | Abstract-EMBASE  

30. T. Kimura, N. Yasui, S. Ohsawa and K. Ono, Chondrocytes embedded in collagen gels maintain cartilage phenotype during long-term cultures. Clin Orthop 186 (1983), pp. 231-239.

31. H.J. Hauselmann, M.B. Aydelotte, B.L. Schumacher, K.E. Kuettner, S.H. Gitelis and J-.M.A. Thonar, Synthesis and turnover of proteoglycans by human and bovine adult articular chondrocytes cultured in alginate beads. Matrix 12 (1992), pp. 116-129. Abstract-MEDLINE | Abstract-EMBASE  

32. K. Von der Mark, V. Gauss, H. von der Mark and P. Muller, Relationship between cell shape and type of collagen synthesized as chondrocytes lose their cartilage phenotype in culture. Nature 267 (1977), pp. 531-532. Abstract-MEDLINE  

33. F. Mallein-Gerin, F. Ruggiero and R. Garrone, Proteoglycan core protein and type II collagen gene expressions are not correlated with cell shape changes during low density chondrocyte cultures. Differentiation 43 (1990), pp. 204-211. Abstract-MEDLINE | Abstract-EMBASE  

34. F. Mallein-Gerin, R. Garrone and M. van der Rest, Proteoglycan and collagen synthesis are correlated with actin organization in dedifferentiating chondrocytes. Eur J Cell Biol 56 (1991), pp. 364-373. Abstract-MEDLINE | Abstract-EMBASE  

35. R.L. Vinall, S.H. Lo and A.H. Reddi, Regulation of articular chondrocyte phenotype by bone morphogenetic protein 7, interleukin 1, and cellular context is dependent on the cytoskeleton. Exp Cell Res 272 (2002), pp. 32-44. Abstract | Abstract + References | PDF (1102 K)

36. R.L. Vinall and A.H. Reddi, The effect of BMP on the expression of cytoskeletal proteins and its potential relevance. J Bone Jt Surg 83 S1 (2001), pp. 63-69.

37. E. Langelier, R. Suetterlin, C.D. Hoemann, U. Aebi and M.D. Buschmann, The chondrocyte cytoskeleton in mature articular cartilage: structure and distribution of actin, tubulin, and vimentin filaments. J Histochem Cytochem 48 (2000), pp. 1307-1320. Abstract-MEDLINE | Abstract-Elsevier BIOBASE | Abstract-EMBASE  

38. K.L. Kilpadi, P.L. Chang and S.L. Bellis, Hydroxylapatite binds more serum proteins, purified integrins, and osteoblast precursor cells than titanium or steel. J Biomed Mater Res 57 (2001), pp. 258-267. Abstract-MEDLINE | Abstract-EMBASE   | Full Text via CrossRef

39. Z. Schwartz, J.Y. Martin, D.D. Dean, J. Simpson, D.L. Cochran and B.D. Boyan, Effect of titanium surface roughness on chondrocyte proliferation, matrix production, and differentiation depends on the state of cell maturation. J Biomed Mater Res 30 (1996), pp. 145-155. Abstract-MEDLINE | Abstract-Compendex | Abstract-EMBASE   | Full Text via CrossRef

40. B.D. Boyan, T.W. Hummert, D.D. Dean and Z. Schwartz, Role of material surfaces in regulating bone and cartilage cell response. Biomaterials 17 (1996), pp. 137-146. SummaryPlus | Full Text + Links | PDF (1343 K)

41. J-.M. Trifaro, S.D. Rose and M.G. Marcu, Scinderin, a Ca²⁺ dependant actin filament severing protein that controls cortical actin network dynamics during secretion. Neurochem Res 25 (2000), pp. 133-144. Abstract-MEDLINE | Abstract-Elsevier BIOBASE | Abstract-EMBASE  

42. M.E. Carbajal and M.L. Vitale, The cortical actin cytoskeleton of lactotropes as an intracellular target for the control of prolactin secretion. Endocrinology 138 (1977), pp. 5374-5384.

43. M. Yoneda, T. Nishizaki, K. Tasaka, H. Kurachi and A. Miyake, Changes in actin network during calcium-induced exocytosis in permeabilized GH3 cells: calcium directly regulates F-actin disassembly. J Endocrin 166 (2000), pp. 677-687. Abstract-MEDLINE | Abstract-EMBASE | Abstract-Elsevier BIOBASE  

44. A. Beit-Or, Z. Nevo, M. Kalina and Y. Eilam, Decrease in the basal levels of cytosolic free calcium in chondrocytes during aging in culture: possible role as differentiation-signal. J Cell Phys 144 (1990), pp. 197-203. Abstract-EMBASE | Abstract-MEDLINE  

45. J. Hambleton, Z. Schwartz, A. Khare, S.W. Windeler, M. Luna, B.P. Brooks, D.D. Dean and B.D. Boyan, Culture surfaces coated with various implant materials affect chondrocyte growth and metabolism. J Orthop Res 12 (1994), pp. 542-552. Abstract-MEDLINE | Abstract-EMBASE | Abstract-Compendex  

46. S. Nehrer, H.A. Breinan, A. Ramappa, G. Young, S. Shortkroff, L.K. Louie, C.B. Sledge, I.V. Yannas and M. Spector, Matrix collagen type and pore size influence behavior of seeded canine chondrocytes. Biomaterials 18 (1997), pp. 769-776. SummaryPlus | Full Text + Links | PDF (1517 K)

47. D.A. Grande, C. Halberstadt, G. Naughton, R. Schwartz and R. Manjo, Evaluation of matrix scaffolds for tissue engineering of articular cartilage grafts. J Biomed Mater Res 34 (1997), pp. 211-220. Abstract-MEDLINE | Abstract-EMBASE | Abstract-Compendex   | Full Text via CrossRef

Corresponding author. Dept. of Pathology and Laboratory Medicine, Mount Sinai Hospital, 600 University Avenue, Ste 600, , Toronto, Ontario, , Canada M5G 1X5. Tel.: [phone]; fax: [phone]

---