Chapter 1 Properties of cartilage tissue

1.1 Introduction

Joint cartilage is highly sophisticated and has been optimised by evolution. There have been considerable research interests related to the cartilage cells, chondrocytes. In the last decades these studies made cartilage the first and very successful tissue engineering treatment. (Brittberg et al. 1994)

1.2 Categorization of cartilage tissue

Cartilage tissue are categorised in three major types by their different biochemical composition and structure of their extracellular matrix (ECM). Elastic cartilage has a few cells, a small concentration of proteoglycans (PGs), and a relatively high proportion of elastin fibres. It is found in the epiglottis, small laryngeal, the external ear, auditory tube, and the small bronchi, where it is generally required to resist bending forces. Fibrocartilage also contains a small concentration of PGs, but far less elastin. The meniscus in the knee joint is made of fibrocartilaginous tissue. The third and most widespread cartilage in the human body is hyaline. It is resistant to compression or tensile forces due to the network organisation of type II collagen fibres associated with a high concentration of PGs. Hyaline cartilage can be found in the nose, the trachea, bronchi, and synovial joints. In the latter case, it is termed as articular cartilage, representing a unique type of connective tissue. Its outwards thin layer covers the articulating joint surfaces and belies a specific structure with unique mechanical properties. These two layers acting as a covering material, is fibricated by the viscous synovial fluid. The joint capsule encloses the entire joint and retains the synovial fluid. (Schulz and Bader, 2006)

1.3 Composition of articular cartilage

Articular cartilage is composed of chondrocytes and an extracellular matrix that consists of proteoglycans, collagens and water. (Darling and Athanasiou 2005) Chondrocytes contribute only between 5% of the tissue volume; the remaining 95% being composed of extracellular matrix (ECM), which is synthesised by the chondrocytes. (Mollenhauer, 2008; Buckwalter et al. 1988) The ECM of articular cartilage consists of about 60-85% water and dissolved electrolytes. The solid framework is composed of collagens (10-20%), PGs (3-10%), noncollagenous proteins and glycoproteins. (Buckwalter et al. 1997; Buckwalter et al. 1990) In articular cartilage, 95% of collagen in the ECM is comprised of collagen type II fibrils. The rest other collagen types are collagen type IX and XI and a small fraction of types III, VI, XII and XIV. (Eyre 2002) Type-I collagen forms thick fibres. Type-III forms thin fibres. Unlike these two collagens, Type-II collagen which is present in hyaline and elastic cartilages does not form fibres, and its very thin fibrils are disposed as a loose mesh that strongly interacts with the ground substance. (Montes, 1996) This collagen component in articular cartilage provides tensile stiffness and strength to the tissue and opposes the swelling capacity generated by highly negatively charged glycosaminoglycans (GAGs) of the proteoglycans (PGs). The majority (50-85%) of the overall PG content in this tissue type were presented by large molecule aggrecan. This consist of a protein backbone, the core protein, to which unbranched GAGs side chains of chondroitin sulphate (CS) and keratan sulfate (KS) are covalently attached. ( 1.1) (Watanabe et al. 1998; Schulz and Bader, 2006)

1.1. Illustration of the extracellular matrix (ECM) organization of articular cartilage (Left) and the schematic sketches (Right) of the most relevant polysaccharides of proteoglycans (PGs) in articular cartilage. The PGs consist of a strand of hyaluronic acid

(HA), to which a core protein is non-covalently attached. On the core protein, glycosaminoglycans (GAGs) such as keratan sulphate (KS) and chondroitin sulfate (CS) are covalently bound in a bottle brush fashion. (Modified from Schulz and Bader, 2006 and Mow and Wang, 1999)

1.4 Low capacity of self-repair

The aneural and avascular nature of articular cartilage, coupled with its low cellularity, contribute to both the limited rate and incomplete nature of the repair process following damage. (Heywood et al., 2004) In addition, the low mitotic potential of chondrocytes in vivo also contributes to its poor ability to undergo self-repair. (Kuroda et al., 2006) Indeed, in experimental studies on adult animals, full-thickness cartilage defects extending into the subchondral bone, have been reported to heal with the formation of fibrous tissue, which contains relatively low amounts of type II collagen and aggrecan. It is also composed of a relatively high content present in type I collagen, not present in normal adult articular cartilage and accordingly exhibits impaired mechanical integrity. (Hjertquist et al., 1971; Eyre et al., 1992)

1.5 Metabolism of articular cartilage

Joint cartilage is supplied with nutrients and oxygen by the synovial fluid diffusion facilitated by compressive cyclic loading during joint movements as a pumping function. (Mollenhauer, 2008) Chondrocytes are imbedded in ECM. Within synovial joints, oxygen supply to articular chondrocytes is very limited. The oxygen tensions are very low varying from around 6% at the joint surface to 1% in the deep layers of healthy articular cartilage. It is supposed to be even further decreased under pathological conditions, such as osteoarthritis or rheumatoid arthritis. The metabolism of chondrocytes is largely glycolytic. Oxygen-dependent energy generated by oxidative phosphorylation is just a minor contributor to the overall energy in chondrocytes. Articular chondrocytes appear to show a so-called negative Pasteur effect, whereby, glycolysis falls as O2 levels drop leading to the fall in ATP and matrix synthesis. (Gibson JS et al., 2008) A negative Pasteur effect would make chondrocytes particularly liable to suffer a shortage of energy under anoxic conditions. (Lee and Urban, 1997) Changes in O2 tension also have profound effects on cell phenotype, gene expression, and morphology, as well as response to, and production of, cytokines. Condrocytes live in hypoxic environments implies that specific factors are required to control certain genes that are responsible for glucose metabolism, energy metabolism, pH regulation, and other responses. The most important component of this hypoxic response is mediated by transcription factor hypoxia-inducible factor-1 (HIF-1), which is present in most hypoxia inducible genes. (Pfander and Gelse, 2007; Gibson JS et al., 2008) HIF-1a is necessary for anaerobic energy generation by upregulation of glycolytic enzymes and glucose transporters. (Yudoh et al. 2005) A previous study shows chondrocytes are not able to survive hypoxia in the absence of HIF-1. (Schipani et al. 2001)

Moreover, the matrix turnover in articular cartilage is extremely slow. Proteoglycan turnover is up to 25 years. Collagen half-life is estimated to range from several decades up to 400 years. No immune-competent cells (macrophages, T-cells) enter the cartilage tissue. Thus chondrocytes have to defend themselves against hostile microorganisms, leading to its immunologically privileged. (Mollenhauer, 2008)

1.6 Mechanical conditions in vivo

In vivo joint loading can result in high peak mechanical stresses (15-20 MPa) that occur over very short durations (1 s) causing cartilage compressive strains of only 1-3%. (Mollenhauer, 2008; Hodge et al., 1986) In contrast, sustained physiological stresses applied to knee joints for 5-30 min can cause compressive strains in certain knee cartilages as high as 40-45%. (Mollenhauer, 2008; Herberhold et al., 1999)

A study of the response of articular cartilage from humans to impact load showed that articular cartilage could withstand impact loads of as much as 25 MPa at strain rates from 500 to 1000 s-1 without apparent damage. Impact loads exceeding this level caused chondrocyte death or fissure in the hip or knee. (Repo RU and Finlay JB, 1977)

Chapter 2 Osteoarthritis and Treatments

2.1 Osteoarthritis, diagnosis and classification

Most cartilage defects are due to direct trauma, but may also occur in avascular necrosis, osteochondritis dissecans, and a variety of cartilage disorders. The defect may be limited to the joint surface (chondral) or involve the underlying bone (osteochondral). (NHS guidance 2006) Articular cartilage defects can progress to osteoarthritis (OA) in some patients, which is a major health problem in developed countries. (Kuroda et al. 2006; Schulz and Bader, 2006; Buckwalter, 2002; Cicuttini 1996) Symptoms may include pain, catching, locking and swelling, and may lead to degenerative changes within the joint. (NHS guidance 2006)

Arthroscopy has been used as the “gold standard” to confirmed cartilage defects. In a review of 31,516 knee arthroscopies of cartilage injury patients, the incidence of chondral lesions was 63%; the incidence of full-thickness articular cartilage lesions with exposed bone were 20% , with 5% of these occurring in patients under 40-years-old. (Marlovits, et al. 2008)

Osteoarthritis (OA) severity is commonly graded from radiographic images in accordance with the Kellgren and Lawrence scale Bilateral. (Kellgren and Bier, 1956; Kellgren and Lawrence, 1957) Osteoporosis and erosions which included narrowing of joint space were recorded separately and graded as follows: 0 = no changes; 1 = doubtful joint space narrowing; 2 = minimal change, mostly characterized by osteophytes; 3 = moderate change, characterized by multiple osteophytes and/or definite joint space narrowing; and 4 = severe change, characterized by marked joint space narrowing with bone-on-bone contact with large osteophytes. (Kellgren and Bier, 1956; Husing et al. 2003) The radiologic grade of OA was inversely associated with the joint space width (JSW). (Agnesi et al. 2008)

MRI is currently the standard method for cartilage evaluation, as it is a non-invasive, non-contact, multi-planar technique capable of producing high resolution, high contrast images in serial contiguous slices and it enables morphological assessment of the cartilage surface, thickness, volume and subchondral bone. The MRI classification of articular chondral defects are as follows: 1=Abnormal intrachondral signal with a normal chondral surface; 2=Mild surface irregularity and/or focal loss of less than 50% of the cartilage thickness; 3=Severe surface irregularity with focal loss of 50% to 100% of the cartilage thickness; 4=Complete loss of articular cartilage, with exposure of subchondral bone. (Marlovits et al. 2008) Agnesi et al. compared the radiologic grading of OA patients with the joint surface width measurements obtained from MRI images. (Agnesi et al. 2008)

2.2 Non-tissue engineering treatments

Pain caused by osteoarthritis can be reduced through a number of methods. (Altman et al. 2006) These include:

· “Exercise programmes (strength and flexibility) and lifestyle changes

· Medication

· Dietary supplements

· Knee viscosupplementation

· Guidelines for viscosupplementation

· Other injections

· Custom foot orthotics

· Knee braces

· Other assisted devices (canes and walkers)

· Surgery”

Total knee replacement is most commonly performed in people over 60 years of age. (NHS guidance, 2006; Altman et al., 2006; Brittberg et al., 1994) Besides that, the most frequently used treatments include the mosaicplasty, marrow stimulation, and autologous condrocyte implantation (ACI). (Steinwachs et al., 2008) Mosaicplasty is an autologous osteochondral transplantation method through which cylindrical periosteum grafts are taken from periphery of the patellofemoral area which bears less weight, and transplanted to defective areas. This transplantation can be done with various diameters of grafts. (Haklar et al., 2008; NHS guidance, 2006) Marrow stimulation methods include arthroscopic surgery to smooth the surface of the damaged cartilage area; microfracture, drilling, abrasion. All marrow stimulation methods base on the penetration of the subchondral bone plate at the bottom of the cartilage defect. The outflowing bone marrow blood contains the mesenchymal stem cells which are stabilised by the clot formation in the defect. These pluripotent stem cells which are able to differentiate into fibrochondrocytes, result in fibrocartilage repair with varying amounts of type I, II and III collagen. (Steinwachs et al., 2008)

2.3 The tissue engineering treatment

A 1984 study in rabbits reported successful treatment of focal patellar defects with the use of autologous condrocyte implantation (ACI). One year after transplantation, newly formed cartilage-like tissue typically covered about 70 percent of the defect. (Grande et al. 1989) In 1987, Mats Brittber et al. firstly performed ACI in 23 people with deep cartilage defects in the knee. (Brittberg et al., 1994) ACI is described as a three steps procedure: cartilage cells are taken from a minor load-bearing area on the upper medial femoral condyle of the damaged knee via an arthroscopic procedure, cultivated for four to six weeks in a laboratory and then, in open surgery, introduced back into the damaged area as a liquid or mesh-like transplant; at last, a periosteal flap sutured in place to secure the transplant. ( 2.1) (Husing et al., 2008) The cell density of the cultivated cell solution is required to be 30 x 106 cells/ml, or 2 x 106 cells per cm2 in a clinical setting today. (Brittberg et al., 2003)

Genzyme Biosurgery with its product Carticel® was the first company which introduced ACT into the market and is market leader in USA. Carticel® is a classic ACT procedure using the periosteal cover. (Husing et al., 2008)

Today the periosteum is often replaced by an artificial resorbable cover such as collagen I/III and hyaluronan membrane, such as ChondroGide or Restore (De Puy, Warzaw, Indiana). (Gooding et al., 2006; Jones and Peterson, 2006) Another new method uses chondrocytes cultured on a three-dimensional, biodegradable scaffold. The scaffold, cut to the required size, is fixed into the lesion site with anchoring stitches. This method does not need the cover, thus simplifying the surgery and shorting the surgery time; opens up the possibility of arthroscopic surgery instead of open surgery which causes more tissue damage. HYALOGRAFT from Italy is one of the European market leaders. It is a cartilage substitute made of autologous chondrocytes delivered on a biocompatible tridimentsional matrix, entirely composed of a derivative of hyaluronic acid. (Marcacci et al. 2005)

2.4 Clinical follow-ups of ACI

Brittberg studied the long-term durability of ACI-treated patients, 61 patients were followed for at least five years up to 11 years post-surgery (mean 7.4 years). After two years, 50 out of 61 patients were graded good-excellent. At the five to 11 years follow-up, 51 of the 61 were graded good-excellent. The total failure rate was 16% (10/61) at mean 7.4 years. (Brittberg et al., 2003)

Since 1997 the year the FDA approved ACI, this method has been widely performed all over the world, in more than 20 000 patients. It has been reported to be effective in

improving clinical symptoms, such as pain and function. (Wakitani et al., 2008)

2.5 Randomised studies

In 2004, Knutsen et al. studied 80 patients who needed local cartilage repair because of symptomatic lesions on the femoral condyles measuring 2-10cm2. The results showed there was no significant difference in macroscopic or histological results between ACI and microfracture, and that there was no association between the histological findings and the clinical outcome at the 2-year time point. (Knutsen et al., 2004) In the same series, there were no significant differences in results at 5 years follow-up. (Knutsen et al., 2007)

In another randomised controlled study that compared mosaicplasty with ACI, there was no significant difference in the number of patients who had an excellent or good clinical outcome at 1 year (69% [29/42] and 88% [51/58], respectively). In the subgroup of patients who had repairs to lesions of the medial femoral condyle, significantly more patients who had ACI had an excellent or good outcome (88% [21/24]) compared with those who had mosaicplasty (72% [21/29]) (p < 0.032). In a large case series, the proportion of patients having an excellent or good outcome based on standardised clinical scores ranged from 79% to 92% depending on the site of mosaicplasty, at up to 10 years follow-up. (NHS guidance, 2006)

2.6 The limitation of ACI

The microfracture is a very simple and low-cost procedure whereas ACI costs almost $10 000 per patient. If ACI is not found to be more effective for improving articular cartilage repair than microfracture, the procedure will not be continued. (Wakitani et al., 2008)

There are several possible reasons which should be blamed for the limitations of the traditional ACI procedure. The cell source in ACI is the cartilage tissue taken from a minor load-bearing area on the upper medial femoral condyle of the damaged knee via an arthroscopic procedure. However, Wiseman et al. found the chondrocytes isolated from the low loaded area of the knee joint respond in a distinct manner with the chondrocytes from the high loaded area, which suggests the traditional cell source of ACI may not provide enough mechanical response and may further lead to the insufficient mechanical properties of the repaired tissue. (Wiseman et al. 2003)

As cultured in monolayer, chondrocytes undergo a process of dedifferentiation and adopt a more fibroblast-like morphology, which is accompanied by an increase in proliferation (Glowacki et al., 1983) and an altered phenotype. Type II collagen, the major protein produced by chondrocytes in articular cartilage, are down-regulated culture, while collagen types I and III are increased. (Stocks et al., 2002; Benya et al., 1978) The agregating proteoglycan aggrecan of articular cartilage, is down-regulated during dedifferentiation and replaced by proteoglycans not specific to cartilage, such as versican. (Glowacki et al., 1983; Stocks et al., 2002) Therefore, monolayer cultured chondrocytes do not express the true chondrocyte phenotype, and their ability to regenerate damaged cartilage tissue is impaired. Upon implantation, dedifferentiated cells may form a fibrous tissue expressing collagen type I that does not have the proper mechanical properties, which may lead to degradation and failure of the repair tissue. (Brodkin et al., 2004) Chondrocytes grown in conditions that support their round shape, such as plating in high-density monolayer (Kuettner et al., 1982; Watt, 1988) and seeding in 3-D gels (Benya et al., 1982) can maintain their differentiated phenotype much longer compared to cells spread in monolayer cultures.

Chapter 3 Tissue engineering strategies for articular cartilage

3.1 Introduction

Although ACI can still be considered to be one of commonly form of repair of cartilage defects, it does have a number of scientific limitations. Some of those can be resolved using a more comprehensive tissue engineered strategy which incorporates cells, scaffold materials and potentially biochemical, biomechanical and/or physical stimulation in a controlled bioreactor environment.

3.2 Cell sources

For a conventional ACI approach, chondrocytes are derived from the low loading area and then cultured in a monolayer. However, chondrocytes derived from the low load bearing area of the knee joint respond in a distinct manner with the chondrocytes from the high loaded area. Chondrocytes cultured in monolayer have a dedifferentiation phenomenon (Described in the previous chapter). In addition, the limitation of the transplant volume is always a major problem in autograft to be overcome (Kitaoka et al., 2001). Thus, potential cell sources are widely studied for the future improvement of ACI approach.

Chondrocytes from immature animals (approximately 1-6 weeks old) have been used widely in tissue engineering studies for their ability to increase matrix synthesis and to produce better mechanical properties (Darling and Athanasiou, 2005).

Kitaoka et al. examined the possibility of using hyaline cartilage of costal cartilage as a substitute to the knee joint articular cartilage. Costal cartilage cells are derived from 8-week-old male SV40 large T-antigen transgenic mice. Three mouse chondrocyte cell lines (MCC-2, MCC-5, and MCC-35) were established using cloning cylinders, which is a kind of mold. These cell lines showed chondrocytic characteristics, including formation of cartilage nodules that could be stained with alcian blue, and mRNA expression for type II collagen, type XI collagen, ALPase, osteopontin, aggrecan, and link protein (Kitaoka et al., 2001).

Animal-derived bone marrow cells, in particular from rabbit origin, have shown a highly variable chondrogenic potential (Solchaga et al., 1999). The establishment of some bone marrow stromal cell lines having the ability of diffrentiation to chondrocyte has been reported, as well as some other cell lines established from rat calvaria, mouse c-fos-induced cartilage tumor and mouse embryonic carcinoma, respectively. Each of the cell lines showed chondrocytic phenotypes (Kitaoka et al., 2001).

LVEC cells derived from EBs of human embryonic germ cells were reported to be homogenously differentiated into hyaline cartilage. Three dimensional tissue formation is achieved by encapsulating cells in synthetic hydrogels poly (ethylene glycol diacrylate) (PEGDA) followed by incubation in chondrocyte-conditioned medium (for the recipe, please see the paper) (Varghese et al., 2006).

Periosteum consists of two layers. Fibroblasts are from the fibrous layer and progenitor cells are from the cambium layer. Progenitor cells are supposed to be able to differentiate into chondrocytes. Emans et al. compared the chondrocyte and the periosteum cell as cell source for autologous chondrocyte implantation (ACI) on animals. The results turned out that the condrtocytes are much better for ACI procedure (Emans et al., 2006).

3.3 Scaffolds

Biomaterial scaffolds provide a critical means of controlling engineered tissue architecture and mechanical properties; allow cells attach, grow in and proliferation; allow the cell signals travelling through (Freed et al., 2006).

In many in vitro or in vivo approaches, cells are grown on biomaterial scaffolds for further research or just for implantation, where new functional tissue is formed, remodelled and integrated into the body.

The biomaterials which compose scaffolds are required to satisfy several properties. At first, the material as a support structure must possess enough mechanical strength to protect the cells contained in. Secondly, the material must have some bioactivity to accommodate cells for attachment, growth, proliferation and migration. The material should act as a vehicle for gene, protein and oxygen delivery. Furthermore, the material should be biodegradable for the new cartilage to form and replace the original structure. In this regard, the material should be non-toxic, non-inflammatory active, and also non-immunogenic. Finally, the material should be easy to handle for surgery procedure (Stoop, 2008).

3.3.1 Natural materials

Collagen-based biomaterials are widely used in today's clinical practice (for example, haemostasis and cosmetic surgery). Collagen is also be commonly used as main components in tissue engineered skin products. Several commercial autologous chondrocyte transplantation (ACT) products have used collagenous membraneas the replacement for the periosteum to close the defect, such as ChondroGide or Restore (De Puy, Warzaw, Indiana) (Cicuttini et al., 1996; Jones and Peterson, 2006) The .combination of collagen with glycosaminoglycan (GAG) in scaffolds had a positive effect on chondrocyte phenotype. Condrocytes were cultured in porous type I collagen matrices in the presence and absence of covalently attached chondroitin sulfate (CS) up to 14 days in a study (van Susante et al., 2001).

Hyaluronic acid is a non-sulphated GAG that makes up a large proportion of cartilage extracellular matrix. In its unmodified form, it has a high biocompatibility (Schulz and Bader, 2007). Matrices composed of hyaluronan have been frequently used as a carrier for chondrocytes. Facchini et al. confirms the hyaluronan derivative scaffold Hyaff ®11 as a suitable scaffold both for chondrocytes and mesenchymal stem cells for the treatment of articular cartilage defects in their study. HYALOGRAFT from Italy is one of the European market leaders for ACT. It is a cartilage substitute made of autologous chondrocytes delivered on a biocompatible tridimentsional matrix, entirely composed of a derivative of hyaluronic acid (Marcacci et al., 2005).

Fibrin plays a major role in general wound healing and specially during healing of osteochondral defects. Fibrin glue is currently used for the fixation of other chondrocyte scaffold constructs in defects. Some investigators used fibrin glues as a matrix, but owing to the exceedingly high concentrations and protein densities involved, the glue impeded rather than facilitated cell invasion and did not support a healing response (Stoop, 2008). Susante et al. found fibrin glue does not offer enough biomechanical support as a three-dimensional scaffold (van Susante et al., 1999). Another study found fibrin and poly(lactic-co-glycolic acid) hybrid scaffold promotes early chondrogenesis of articular chondrocytes in vitro. They used the natural polymer fibrin to immobilize cells and to provide homogenous cells distribution in PLGA scaffolds (Sha'ban et al., 2008).

Sugar-based natural polymers such as chitosan, alginate and agarose can be formulated as hydrogels and in some cases sponges or pads. Although these materials are extensively used in in vitro research, their role in in vivo cartilage reconstruction is still limited (Stoop, 2008). Alginate possesses a number of suitable properties as a scaffold material for cartilage tissue engineering. The mobility of alginate allows the ability of cells to be distributed throughout the scaffold before the gelling phase. Its well-characterized mechanical properties are suitable for the transmission of mechanical stimuli to cells. Furthermore, it has been proved its ability to reestablish and maintain the differentiated state of chondrocytes during long-term culture (Heywood et al., 2004). Agarose is a clear, thermoreversible hydrogel which has been applied in numerous studies in cartilage tissue engineering. This hydrogel is supportive of the chondrocyte phenotype and allows for the synthesis of a functional extracellular matrix. Agarose is neutrally charged, and forms solid gels at room temperature. The initial strength of the gel is dependent on the rate of gelling, which in turn is dependent on the ambient temperature. Gel strength is also strongly dependent on the concentration of the gel in solution. Basic science studies involving agarose gel formation have demonstrated that rapid cooling leads to a decreased, more homogeneous pore size. Increasing the gel concentration additionally decreases gel pore size and permeability. A number of studies have used agarose for the investigation of chondrocyte growth and response to mechanical stimuli (Ho MMY et al., 2003).

3.3.2 Synthetic materials

Potential synthetic material scaffolds for the tissue engineering of bone or cartilage include:

• PL (Polylactic acid)

• PGLA (Polyglycolicacid and copolymers)

• CF-PU-PLLA (Carbonfibre-Polyurethane-Poly(L-lactide)-Graft)

• CF-Polyester (Polyester-Carbonfibre)

• PU (Polyurethane)

• PLLA (Capralactone (Poly-L-Lactide/epsilon-Caprolactone)

• PLLA-PPD (Poly- L-Lactic Acid and Poly- p-Dioxanol)

• PVA-H (Polyvinylalcohol-Hydrogel)

• ß-TCP (Tricalcium phosphate)

• CDHA (Calcium-deficient hydroxyapatite) (Haasper et al., 2008)

The major advantages of the synthetic polymers are their design flexibility and avoid of disease transmission. Synthetic polymers can be easily processed into highly porous 3-dimensional scaffolds, fibres, sheets, blocks or microspheres. However, there are also disadvantages of some synthetic polymers, such as the potential increase in local pH resulting from acidic degradation products, excessive inflammatory responses and poor clearance and chronic inflammation associated with high molecular weight polymer (Stoop, 2008).

Poly(glycolic acid) (PGA), poly(lactic acid) (PLA), and poly(lactic-co-glycolic acid) (PLGA) have been investigated for use as cartilage tissue engineering scaffolds (Cima et al., 1991; Vacanti et al., 1991). Both, in vitro and in vivo studies have demonstrated these scaffold maintained the chondrocyte phenotype and the production of cartilage-specific extracellular matrix (ECM) (Barnewitz et al., 2006; Kaps et al., 2006). In addition, PLGA is used as a scaffold material for matrix-based autologous chondrocyte transplantation clinically for more than 3 years (Ossendorf et al., 2007).

3.4 Biomedical stimulation

Growth factors are proved to be able to promote the formation of new cartilage tissue in both explants and engineered constructs (Darling and Athanasiou, 2005). Insulin-like growth factor-I (IGF-I) can dramatically increase biosynthesis level of choncroctyes, especially in the presence of mechanical stimulation (Bonassar et al. 2001; Jin et al. 2003). Transforming growth factor-β1 (TGF-β1) increases biosynthesis in engineered constructs and also stimulates the cellular proliferation (Blunk et al. 2002; van der Kraan et al. 1992). Basic fibroblast growth factor (bFGF) stimulates cell proliferation (Adolphe et al. 1984) and biosynthesis (Fujimoto et al. 1999) in chondrocytes which were cultured under a variety of conditions.

3.5 Mechanical stimulation

ACI is considered a proper way for the repair of cartilage defect. However, one of the obstacles to the use of autologous chondrocytes is the limited in vitro proliferation rate of these cells.

A lot of stimulations have been found to be effective in stimulating cell proliferation and ECM synthesis, including mechanical, electrical, ultrasound (Parvizi et al., 1999; Noriega et al., 2007) and even laser (Torricelli et al., 2001) stimulation.

Mechanical forces due to body movement are experienced by articular cartilage every day. These forces include direct compression, tensile and shear forces, or the generation of hydrostatic pressure and electric gradients. Some level of these forces is beneficial to chondrocytes. (Schulz and Bader, 2007; Shieh and Athanasiou, 2007)

There are many studies which have described the design of bioreactor systems, which can apply shear forces, perfusion, tension, hydrostatic pressure, static compression, dynamic compression on cartilage explants, monolayer cultured cells or tissue engineered constructs. (Schulz and Bader, 2007)

Previous work on these bioreactor systems has demonstrated that chondrocytes are highly mechanosensitive. A summary of the key studies is provided in Table 3.1. Static compression leads to decreased levels of sulfate and proline incorporation (Sah et al., 1989; Ragan et al., 1999). Type II collagen and aggrecan gene expression increase transiently, but decrease when exposed to longer durations of static compression (Ragan et al., 1999). In contrast, dynamic compression of cartilage explants stimulates sulfate and proline incorporation, while chondrocytes em bedded in hydrogels produce more matrix and form robust constructs when cyclically compressed. (Lee and Bader, 1997; Mauck et al., 2000)

Table 3.1. Influence of the different models of mechanical stimulation on the biochemical response of articular chondrocytes.

References

Type of mechanical stimulation

Magnitude of stimulation

Cell culture type

Cell source

Response and Result

Regan et al., 1988

Static compression

Free-swelling thickness

of 1.15 mm or were compressed to 1 .0 (cut thickness), 0.7, or 0.5

mm.

1 mm thick cartilage disks

Cartilage from femoropatellar

groove of 1- to 2-week-old calves

Aggrecan and Type-II collagcn mRNA, [35SO4] and [3Hl-proline incorporation decreased with increasing compression after 24 hours.

Aggrecan and type-II collagcn mRNA increased after initial 0.5 hours, but decreased after 4-24 hours.

Sah et al., 1989

Static mechanical compressions for 12 hours;

One time compression for 2 hours then released;

Repetitive compression-release

Cyclic compression-release

Oscillatory strain

0.5-1.3mm

Calf articular cartilage explants

Articular cartilage of Young calves ( 1-2 week old)

↓ [3H]-proline after static compression

No response after the release of a compression

↑ 35SO4, ↑ [3H]-proline after repetitive compression, cyclic compression and oscillatory strain

Smith et al., 1995

Fluid-induced shear using a cone viscometer for periods of 24, 48, and 72 hours

1.6 Pa

Monolayer culture

Human and Bovine adult knee articular cartilage

At 48 and 72 hours

↑GAG

After 48 hours

↑ PGE

↑Metalloproteinase

Lee and Bader, 1997

Dynamic compression at 0.3-3 Hz for 48 hours

15% compressive strain

Cell seeded agarose cylinders

Articular cartilage of metacarpo-

phalangeal joints of 18-month-old steers

↑GAG at 1 Hz

↑ [3H]-thymidine at 0.3Hz,1Hz and 3 Hz

↓ [3H]-proline at all frequencies

Fujisawa et al., 1999

Cyclic tension force. High frequency (30 cycles/min, 1 s duration, 1 s interval),

middle frequency (1 cycle/2 min, I s duration and 119 s interval),

low frequency

(1 cycle/4 min, 1 s duration and 239 s interval)

5 or 15 kPa

Monolayer culture

A human chondrosarcoma-derived chon-

drocyte cell line (HCS-2/8)

↓DNA synthesis and ↓ [3H]proline

after high frequency CTF at 5 and 15kPa.

The mRNA levels of IL-1 and MMP-9 increased 3-6 h after high

frequency CTF at 5 and 15 kPa.

The mRNA levels of TIMP-1 showed no remarkable change at either level of

high frequency CTF.

Mauck et al., 2000

Dynamic compression at 1 Hz, 3×(1 hour on, 1 hour off)/day, 5 days/week for 4 weeks

10 percent strain

Cell-seeded agarose disks

Cartilage from carpometacarpal joint of 3-5 month bovine calves

↑ Equilibrium aggregate modulus

Smith et al., 2000

Intermittent hydrostatic pressure at a frequency of 1 Hz for periods of 2, 4, 8, 12, and 24 hrs.

10 MPa

High-density monolayers or formed aggregates

Adult bovine articular chondrocytes

Type II collagen mRNA increased at 4 and 8 hours, subsequently decreased at 12 hours.

↑Aggrecan mRNA at all periods

Jin et al., 2001

Shear loading at 0.01-1.0 Hz

1-3% sinusoidal shear

strain

Cartilage disks (3 mm in diameter by 1-1.1 mm thick)

Cartilage from femoropatellar

groove of 1- to 2-week-old calves

↑ [3H]-proline and

↑ 35SO4 when shear loading of 3% amplitude at all frequencies tested (0.01, 0.1, and 1.0 Hz)

Chowdhury et al. 2003

Continuous dynamic compression at 1 Hz for 1.5, 3, 6, 12, 24 or 48 hours;

Intermittent compression for 1.5, 3, 6 or 12 hours compression with equivalent

unstrained periods for a total period of 48 h

15% compressive strain

Cell seeded agarose cylinders

Articular cartilage of metacarpo-

phalangeal joints of 18-month-old steers

↑35SO4 incorporation for all test conditions

↑ [3H]thymidine for the majority of test conditions

↓NO for all test conditions

Chapter 4 Electrical stimulation

4.1 Cartilage electrokinetics

The charge density within the cartilage tissue is caused by fixed negative charges including the sulfate (SO-) and carboxyl (COO-) groups attached to the chondroitin sulfate chains, the sulfate group attached to the keratan sulfate chains, and the carboxyl group attached to the hyaluronan chains. (Schmidt-Rohlfing et al., 2002) ( 1.1)

During the past decades, investigators have found the electromechanical interaction play an important role in the physiology of cartilage and the other connective tissues. As early as 1974, Lotke et al. discovered the strain-related potential can be generated on human cartilage. Several hypotheses about the origin of the strain-related potentials were discussed in their paper, which include the streaming potential and the Donnan potential. (Lotke et al. 1974) Since then, a lot of works have been done on the study of electromechanical transduction phenomena on cartilage tissue either mathematically or experimentally.

The early mathematical model on cartilage e.g. poroelastic theory successfully predicted the creep and stress-relaxation behaviours of the articular cartilage. This theory considered the cartilage has both solid and liquid phases and derived the interaction between the two phases mathematically. Lai and co-workers extended it into a triphasic theory which incorporates an ion phase. (Lai et al., 1991) Considering the concentration of the cation in the external solution is required to balance the negative charges on the GAG chains of the PG molecules in the internal solution, Lai et al. derived the relationship of the ion concentrations between the external solution and the internal solution. This work also followed the Donnan equilibrium theory. Furthermore, the ions generated the chemical potential, which is considered to yield the electrical tensor. Thus the electrical system can be combined into the tensor system. The link between ion concentrations and the deformation of the cartilage tissue thus was established. (Lai et al., 1991)

Several hypotheses are used to explain the electrokinetic phenomena, including streaming potentials, diffusion potentials and piezoelectricity. (Schmidt-Rohlfing et al., 2002)

Streaming potentials refer to electrical phenomena which are generated by fluid flow over a charged surface under an open circuit condition. (Mow and Wang 1999) As the fluid flows through the specimen, the positive ions will be attached to the fixed negative charge. To maintain electroneutrality within the specimen, an electrical potential must be established against the direction of water flow to restrict the loss of counter ions (for example, Na+) from the tissue. This streaming potential is given by:

Where

∆ψ = the streaming potential = Permeability coefficient

∆p = the applied pressure =The tissue's apparent charge density

= Faraday's constant = Conductivity of the cartilage material

For the inhomogeneous nature of the cartilage tissue, there is a diffusion potential theory. Because of the nature or deformation induced inhomogeneous fixed charge density (FCD) distribution in the tissue, the FCD gradient will lead to gradients of ion concentrations, with cations and anions having opposite directions for the gradients. As a consequence, there exists a gradient of electric potential inside the tissue, i.e., the diffusion potential, and it is caused by the tendency of the ions to diffuse from a region of higher concentration to a region of lower concentration. The relation between these ion concentrations are given by:

and

where c1, c2, cF, c*, and are cation concentration, anion concentration, fixed charge density, external bathing solution salt concentration, and the ratio of activity coefficients of salt, respectively. (Lai et al., 2000)

In addition, the electrical potential caused by piezoelectricity is thought to be too small to generate the electrical potential detected. (Grodzinsky et al., 1978)

Except for these theoretical studies, mechanically induced electrical potentials have been observed by experimentally testing circular plugs of articular cartilage subjected to compressive loading. (Grodzinsky et al., 1978; Lotke et al. 1974) Schmidt-Rohlfing et al. further investigated the electromechanical phenomena on the whole joints, which preserved the joint as a whole as well as protect the cartilage against further damage. (Schmidt-Rohlfing et al., 2002)

4.2 Electrical stimulation: literature review

Electrical stimulation has long been used as a tool to promote connective tissue healing. A recently study reported the electrical stimulation successfully attenuated knee OA symptoms. (Farr et al. 2006) Considering the electromechanical interaction phenomena within articular cartilage tissue, electrical stimulation is supposed to be as effective as mechanical stimulation to up-regulate chondrocyte matrix and products in vitro. Many studies have proved this hypothesis. (Table 4.1)

Table 4.1. Literature review about the previous studies of electrical stimulation effect on different chondrocytes models.

References

Type of electrical signal

Intensity of stimulation

Cell culture type

Cell Source

Response and Result

Brighton et al., 1976

Capacitive coupled fields (Constant)

1500 V/cm

Explant

Growth plate of ostochondral junction of 21 day male Sprague-Dawley rat

↑ 35SO4

↑ [3H]-thymidine

↑45Ca

Rodan et al., 1978

Capacitive coupled fields (Constant)

1166 V/cm at 5 Hz

Monolayer culture

Epiphyseal growth plate chondrocytes from chick embryos

↑ [3H]-thymidine

Brighton et al., 1984

Capacitive coupled fields (AC)

250V

peak to peak (37 µA/ cm2) at 60 kHz

Cell pellets

Articular cartilage of Young calves (3 - 6 weeks)

↑ 35SO4

↑ [3H]-thymidine

Brighton and Townsend 1986

Capacitive coupled fields (AC)

500 V

peak to peak at 60 kHz

Monolayer culture

Growth plate of ostochondral junction of new born calf (2-5 days)

↑ [3H]-thymidine

↑cAMP

Armstrong et al., 1988

Capacitive coupled fields (AC)

50, 75, 1000V p-p (15 to 30 mV/cm) at 60 kHz

Monolayer culture

Metacarpal growth plate cartilage of young calves (1-8 weeks)

↑35SO4

No change or decrease in [3H]-thymidine

Okihana and Shimomura, 1988

Direct stimulation (DC)

1 µA in the chamber

Monolayer culture

Articular cartilage of costochondral junction of young rabbits' ribs

↑ [3H]-thymidine

No change or decrease in 35SO4

Brighton et al., 1989

Capacitive coupled fields (AC)

7,20,50, and 126 mV/cm2 at 60 kHz

Monolayer culture

Epiphyseal growth plate of new born calf

↑ [3H]-thymidine

No change or decrease in 35SO4

MacGinitie et al., 1994

Direct stimulation (AC)

10 - 30 mA/ cm2

at 1 to 10 kHz

Explant

Articular cartilage of young calves (1 - 2 weeks)

↑35S-methionine incorporation

Chao et al., 2000

Direct stimulation (DC)

0.003 to 0.1 mA/cm2

Monolayer culture

Articular cartilage of calves (4 - 6 months)

↑Galavanotaxis

↑ Galvanotropism

Szasz et al., 2003

Direct stimulation (AC)

25mA/cm2at 0.01 to 10 kHz

3D chondrocyte/agarose model

Articular cartilage of Newborn calves

↑ 35SO4

↑ [3H]-hydroxyproline

↓ [3H]-thymidine

Wang et al., 2004

Capacitive coupled stimulation (AC)

20mV/cm, 300µA/cm2 at 60kHz

Explant

Articular cartilage of foetal tissue

↑ Aggrecan mRNA

↑ Collagen II mRNA

Brighton et al., 2006

Capacitive coupled stimulation (AC)

20mV/cm, 300µA/cm2 at 60kHz

Explant

Articular cartilage of mature cows (18‑ 30months )

↑ Aggrecan mRNA

↑ 35SO4

↑ Collagen II mRNA

↑ Hydroxyproline

Bone growth at the epiphyseal plate is sensitive to many stimuli. The early researches of electrical stimulation on chondrocytes were mostly done on epiphyseal growth plate cartilage and costochondral junction cartilage in order to find a way to improve bone fracture healing. Considering the cell growth is very active in epithyseal plate, many studies have found the electrical stimulation is effective in stimulating the cell proliferation and matrix regeneration of epithyseal plate. Later, several studies found the electrical stimulation are effective on articular cartilage as well, although the mechanism and the efficacy need to be further proved and studied.

4.3 Electrical stimulation: different stimulation systems

The electrical stimulation can be generated through direct stimulation, capacitively coupled stimulation, or undirectly by electromagnetic fields (EMF).

Direct stimulation: The direct electrical stimulation is provided by the direct passage of current through a cell suspension or monolayer, which is typically located within a rectangular chamber. The chamber is then connected via agar or agarose salt bridges to two other containers with the electrolyte solution in them. These two containers are connected to the power source by electrode leads. Salt bridge and the electrolyte solution are used to provide electrolytes for conductance. (Okihana and Shimomura, 1988; MacGinitie et al., 1994; Chao et al., 2000) ( 4.1)

Capacitive coupling: Electrode leads are connected to a couple of electrode plates to provide an electrical field. Different kinds of signals of this electrical field can be provided by the power source. If the cultured cells or cartilage explants are put in this electrical field, the ions in the cells or explants will be affected and move along the electrical field line because of the potential difference between the two electrode plates. (Brighton et al., 1976) ( 4.2)

Electromagnetic stimulation: Except for the direct stimulation or capacitive coupled electrical stimulation described in Table 3.1, the PEMFs (pulsed electromagnetic fields) can also be effective in stimulating bone or cartilage tissues. While PEMFs are presumed to induce the ion current in vivo, it can be considered as the relatively optimal electrical stimulation with uniform current and flexible stimulating direction.

The relationship between an electrical field and a magnetic field is described as follows. A moving charge or current creates a magnetic field in the surrounding space in addition to its electric field. The magnetic field exerts a force on any other moving charge or current present in the field. (Young et al. 2008)

The magnetic field is generated by two wire coils, separated by a prescribed distance. Electricity passes through the coils. The cultured cells or cartilage explants are put at a distance midway between the coils, where the magnetic field is uniform. ( 4.3)

The EMFs are proved be able to promote bone repair after fracture. Pulsing electromagnetic fields (PEMFs) of differing waveforms and frequencies have been applied successfully in the treatment of congenital and acquired pseudoarthroses and of non-union fractures (2-6). (Elliott JP et al., 1988)

Several researchers are trying to find the basic mechanisms of the bone healing induced by EMF. They examined the chondrocytes response induced by EMF on different models.

Elliot et al. investigated the relationship of electromagnetic vectors to the plane of bovine cell culture plates. Cells were derived from radiocarpal joints of 2 to 5 year old bovines. The results showed that the chondrocyte proliferation in medium containing 3% serum with PEMF coils in a vertical orientation was significantly decreased (p < 0.001) when compared with unstimulated controls while the horizontal stimulation appears to have no significant effect on chondrocyte proliferation. (Elliott et al., 1988)

In the study of Pezzetti et al., the data show that PEMFs induce an increase in the proliferation of human chondrocytes, measured by 3H-thymidine incorporation. The chondrocyte phenotype is examined by RT-RNA PCR of the type II collagen, which is a specific marker of the chondrocyte phenotype. (Pezzetti et al., 1999)

Ciombor DM et al. firstly reported chondrogenic response to an EMF in a developing system of endochondral bone formation. The exposure to EMF of certain configurations can regulate chondrocyte differentiation from a mesenchymal progenitor CD rat cell pool. The subsequent extracellular matrix synthesis and maturation in that model (mesenchymal progenitor cell pool) recapitulates the cell biology of endochondral ossification. (Ciombor et al., 2002)

A recent study reported the EMF increased matrix GAG synthesis in undigested, untreated explants derived from young calves but has no such effect on the adult tissue or damaged tissue. In another sets of experiment on trypsin treated samples, the EMF showed no stimulatory effect of proteoglycan biosynthesis on matrix depleted cartilage. (Bobacz et al., 2006)

4.4 Introduction for the present study, aims and objectives

However, most of previous studies about electrical stimulation were experimented under alternating electrical stimulation. Few experiments about electrical stimulation effect on cell proliferation and ECM synthesis has been done by direct current stimulation although some studies were about DC electrical stimulation effect of galavanotaxis and galvanotropism on chondrocytes. (Chao et al., 2000) The current proposed study will employ DC stimulation because direct current is unidirectional and its influence at specific magnitudes was more easily ascertained while alternating current varies its magnitude and direction in each cycle and the effect of specific magnitudes of current or voltage on chondrocytes cannot be determined. (Akanji, 2008)

In addition, no research on 3D chondrocyte/agarose construct has been reported except for a conference paper (Szaz et al., 2003). MacGinitie et al. noted that the responses to electrical stimulations were affected by different positions in which chondrocytes were derived. This may be because either the cell activity or chondrocyte-matrix interactions. Chondrocyte seeded agarose constructs can provide a relatively homogeneous cell population and homogeneous matrix properties for study of chondrocyte response to electrical stimulation. 3D chondrocyte/agarose structure can eliminate the affecting factor of the extracellular matrix. (MacGinitie et al. 1994)

A recent PhD study in our host lab have shown that DC electrical stimulation was relatively ineffective in providing enhanced stimulation of chondrocyte seeded agarose constructs, in terms of cell proliferation and matrix synthesis, using either mRNA for selected genes (collagen II and aggrecan) or radioisotope labeling techniques. In addition, it had only minimal effect on enhancing intracellular Ca2+ signalling. (Akanji, 2008) This is inconsistent with the previous studies which have reported intermittent or alternating electric fields can significantly up-regulation of chondrocyte matrix. This indicates the stimulation protocol can be improved by the application of intermittent electrical fields to our 3D cell model systems.

The present study tests the hypothesis that intermittent DC stimulation provides an up-regulation of the bio-synthesis activity of chondrocytes seeded in 3D agarose constructs.

The objectives include:

* To modify the existing electrical stimulation system to enable it to accommodate intermittent stimulation to chondrocyte seeded agarose constructs.

* To utilise a range of methodologies to assess chondrocyte viability, GAG and DNA content in agarose constructs and the synthesis of NO within the conditioned medium

* To optimize the protocol of the intermittent stimulation in terms of chondrocyte activity.

Chapter 5 Material and Methods

5.1 Chondrocyte medium (DMEM + 16% FCS)

Chondrocytes medium, namely DMEM + 16% FCS, was prepared by Dulbecco's modified Eagle's medium (DMEM) supplemented with 16% (v/v) fetal calf serum (FCS), 2 mM L-glutamine, 5 mg/mL penicillin, 5 mg/mL streptomycin, 20mM Hepes buffer, and 0.85 mM L-ascorbic acid (DMEM plus 16% FCS, all from Sigma Chemical Co., Poole, UK). Dulbecco's Modified Eagle's Medium (DMEM) contains inorganic salts, vitamins, glucose and amino acids. All of the materials were thoroughly mixed and then filtered through a 0.22μm cellulose acetate filter (Millipore, Watford, UK) for sterility. (Lee and Knight, 2004)

5.2 Chondrocyte isolation

Full-depth slices of cartilage are removed under aseptic conditions from the metacarpophalangeal joints of mature steers aged from 18 to 24 months. Each hoof was scrubbed and then immersed in 70% industrial methylated spirit (IMS) for 15minutes or more. The joint capsule was opened and then the joint was disarticulated using a sterile scalpel blade in a class II flow cabinet. Full thickness cartilage explants were excised from the proximal surface of the joint and placed in a Petri dish containing sterile Earle's Balanced Salt Solution (EBSS, Sigma Aldrich, Poole, UK). ( 5.1) The EBSS was aspirated and the explants were chopped into approximately 1 - 2mm3 pieces (Chowdhury et al., 2001).

Pronase solution was prepared by 700Units/mL pronase (BDH Ltd., Poole, UK) dissolved in DMEM + 16% FCS and then filtered for sterility. Collagenase solution was prepared by 100Units/mL type IX collagenase (Sigma-Aldrich, UK). The diced explant tissue was incubated in 10mL pronase solution at 37°C for 1hour on rollers, thereafter was incubated in the collagenase solution for 16 to 18hours at 37°C on rollers. Following the incubation, the undigested tissue was allowed to settle down. The supernatant was passed through 70mm pore size cell sieve (Falcon, Oxford, UK), then centrifuged at 2000 rpm for 5 minutes yielding a cell pellet. The pellet was washed twice with DMEM + 16% FCS and was finally resuspended in 10mL of fresh DMEM + 16% FCS. (Lee and Knight, 2004)

5.3 Cell count

20μL of the cell suspension was aspirated and then mixed thoroughly with 20µL of 0.4% (w/v) trypan blue in saline solution (Sigma, Poole, UK). A small volume of the mixture was put in the Neubauer haemocytometer. Cell number and the viability were counted in the 25 square counting grid. This small grid has a 1mm2 surface area which provides an equivalent volume of 0.1mm3.

The total cell number in the total suspension=

Number of cells in 25 squares

× 2 (dilution to trypan)

× 104 (µm → mL conversion)

× (n) (total volume suspension)

5.4 Preparation of cell-agrarose constructs

After counting the cells, the cell solution was diluted in DMEM + 20% FCS to a concentration of 8 × 106, which is the twice times of the desired final concentration. Agarose was weighed up and dissolved in EBSS to make up a solution at a concentration of 8% [0.08g agarose per mL. Agarose solution was autoclaved and then placed in 37℃ incubator to cool down. Thereafter, the equal volume of the cell suspension and the agarose solution was added to the agarose solution, placed on rollers for approximately 10minutes to ensure its thorough mixed, to achieve a 4% (w/v) agarose solution at a final concentration of 4x106cells/mL (Lee and Knight, 2004). The cell-agarose solution was aliquoted into a Perspex mould designed to produce cubic 5 x5x5mm constructs. The Perspex mould consists of 12 inner cores and a hoop mould which enclose them. Each inner core has a cubic square gap in it. ( 5.2) The moulds were then placed in Petri dishes and incubated at 4°C for 15minutes to allow gelation of the agarose. After this, all the inner cores with cubic constructs were transferred into a 12 well plate (Nunc, Roskilde, Denmark) and cultured for 24hours in DMEM + 16% FCS at 37°C and 5% CO2, prior to experimentation. (Akanji, 2008)

5.5 Preparation of salt bridge

3% agarose salt bridges are prepared by dissolving agarose (type VII, Sigma, Poole, UK) in 0.01 M KCl (cell culture grade, Sigma, Poole, UK) solution in a clinical autoclave. The melted agarose solution is added into the U-shaped glass tubes (internal diameter 5 mm) and allowed to solidify.

5.6 The working stimulation rig and the electrical stimulation protocol

Our established electrical stimulation rig contains a power supply (Biorad PowerPac basic, Biorad Laboratories Inc., UK), a variable resistor, an ammeter and a Perspex tank. The tank contained two electrode wells and four separated chambers in each of which our 3D construct was well fitted. The chambers were connected to each other by “U” shape salt bridge. ( 5.3) The tank was connected to power supply by electrodes and wires in the way illustrated by the circuit diagram of 5.4. The electrodes were made of 99.9% platinum and were sufficiently resistant. The rig design permitted a continuous flow of current through the appropriate component parts.

The constructs were transferred into the tank of the stimulation rig. 1.4 mL of DMEM + 16% FCS was added into each chamber ( 5.3). Then the stimulation chamber was placed in the 37°C incubator with 5% CO2. The power supply (Biorad PowerPac basic, Biorad Laboratories Inc., UK) was connected to the chamber by electrodes and wires in the way illustrated by the circuit diagram of 5.4. The electrical current through the constructs and the chamber was monitored by an ammeter. ( 5.5) Various magnitude of current can be provided by the power source. For a 4 mA/cm2 current density and due to the 5 × 5 mm transverse section of the constructs, the current through the construct and the chamber should be set as 1 mA. Similarly, for an 8 mA/cm2 current density, the electrical current should be 2 mA. For a 2 mA/cm2 current density, the electrical current should be 0.25 mA.

After constructs were transferred into the stimulation rig, five different protocols of various intermittent electrical stimulations were employed separately over a 24 h culture period on the stimulation group. These five protocols were illustrated schematically in the 5.6. The total stimulation time of each protocol was 4 hours. The electrodes in the other stimulation tank were kept unconnected, and thus the cells in this tank served as an unstimulated control. Several other samples were kept in 12 well plate and cultured with 3 mL DMEM + 16% FCS for each sample. They were incubated as the same situation as the stimulation group and control group, served as the free swelling group. ( 5.5)

5.7 Viability test in chondrocyte seeded agrose chonstruct

Each specimen was bisected from top to base, as oriented during culture, and incubated at 37°C for 30-40 min in 0.6 mL of DMEM + 16% FCS containing calcein AM and ethidium homodimer 1 (5µM eachfrom Molecular Probes, Cambridge, UK). The cut surfaces of the specimen were placed onto a coverslip and visualized with an inverted microscope (TE200; Nikon Instruments, Melville, NY) and a xenon fluorescent light source. The viability of cells within each specimen was recorded by a systematic sampling procedure, as illustrated in 5.7. Sampling was performed in a continuous series running from the top to the base of the specimen and the number of viable (green) and nonviable (red) cells were recorded. The 0.5 × 0.5 mm sample area was standardized by the width and half the depth of a 1 cm2 eyepiece graticule as viewed under the ×20 objective ( 4.7). The live-dead cell counts were subsequently converted to values of viable percentage (Heywood et al., 2003).

5.8 Biochemical analysis

After the electrical stimulation, all the constructs and the corresponding medium were removed and frozen at -20 °C.

5.8.1 Construct digestion:

Digest buffer was prepared by 500 mL PBS supplemented with 0.788g Cysteine hydrochloride and 0.403g EDTA, which was made up to pH 6.0 with NaOH and then freeze in 20 mL aliquots (all from Sigma Chemical Co., Poole, UK). Each construct was put in a small bijou tube. 1 mL digest buffer was added into each construct. Thereafter, constructs were melted in 70 °C incubator for 1hour then mixed vortex and placed in 37 °C incubator for cooling down. 10µL agarase solution (1000Units/mL, Sigma, Poole, UK) and 5µL of papain suspension (560Units/mL, Sigma, Poole, UK) were added to each construct. After this, all the constructs were digested overnight in the 37 °C incubator for agarase digestion. The next morning, all the constructs were digested for a further 1 hour in the 60 °C incubator for papain digestion. At last, digested samples were stored at -20 °C. (Akanji, 2008)

5.8.2 Assessment of glycosaminoglycan production

Glycosaminoglycan (GAG) content was assessed by the dimethylmethylene blue (DMB) assay technique. The DMB bound sulphated GAGs causes a metachromatic shift in the absorbance from 600 to 535 nm under the light source of the spectrophotometer. The absorbance is measured at 595 nm wavelength. To determine absolute GAG values, a set of standard solutions were prepared using stock solution of bovine chondroitin-4-sulphate standard (1 mg/mL, Sigma, Poole,UK). Standard solutions were prepared by diluting the stock solution in the distilled water, to yield concentration ranging between 0 and 50 µg/mL. 40 µL of each standard solution and samples (including digested constructs and the medium) were added into 96-well plate. 250 µL DMB reagent (Sigma, Poole, UK) were added to all wells. The absorbance of the bound dye was measured in Ascent spectrophotometer with associated software. Absolute value of GAG in each sample well was determined using a standard curve, generated from the absorbance measurement of the GAG standard solutions. A typical GAG standard curve is illustrated in 5.8.

5.8.3 Fluorimetric DNA assay:

DNA content was assessed using the fluorimetric dye Hoechst 33258 (Sigma, Poole, UK), which is a DNA specific dye that binds to the adenine-thymidine base pairs emitting fluorescence at a wavelength of 460 nm.

DNA digest buffer is the same with the construct digest buffer described in the previous part. The saline sodium citrate (SSC) buffer (20×) was prepared by 500 mL distilled water supplemented with 87.65 g sodium chloride and 44.1 g trisodium citrate, which was made up to pH 7.0 with 1 M NaOH or 1 M HCL (all from Sigma Chemical Co., Poole, UK). Then “(2×) SSC/DNA buffer” was made by the SSC buffer (20×) diluted 10 times with the DNA digest buffer rather than the distilled water. The Hoechst / digest buffer solution was made by the stock of Hoechst 33258 diluted one thousand times using the “(2×) SSC/DNA buffer”.

To determine absolute DNA values, a set of standards were prepared using calf thymus DNA (1mg/mL, Sigma, Poole, UK). Standards were prepared by serial dilutions in the “(2×) SSC/DNA buffer”, to yield concentrations ranging between 0 and 100µg/mL. 100µL of each dilution of the standards was aliquoted in triplicate into separate wells of a 96 well plate and 100µL of the digested samples were added to the other wells in duplicate. Thereafter, 100µL of the Hoechst / digest buffer solution was added to each well. The fluorescence of the bound dye was measured in a fluorimeter (Fluostar Galaxy, BMG Labtechnologies, Ayelsbury, UK) with associated software. Absolute values of DNA in each sample well was determined using a standard curve, generated from the fluorescence measurements of the DNA standards. A typical standard curve is illustrated in 5.9. (Akanji, 2008)

5.8.4 No assay

Nitric Oxide (NO) production in conditioned medium was determined by quantifying its derived product, nitrite, using a spectrophotometric method based upon the Griess reaction. Content was assessed using the Greiss reagen, which is made up of 0.5% sulphanilamide, 0.05% naphthyl ethylenediamine dihydrochloride, 2.5% H3PO4. The absorbance is measured at 540 nm.

To determine absolute the absolute nitrite values, sodium nitrite (NaNO2) standard solutions were made raging from 0 nM to 50 nM by diluting the stock 0.1M NaNO2 solution in DMEM + 16% FCS. 50 µL of each standard solution and conditioned medium were added into 96-well plate. 50 µL of the Greiss reagents were added to all wells. The absorbance of the bound dye was measured in Ascent spectrophotometer with associated software. Absolute value of nitride in each sample well was determined using a standard curve, generated from the absorbance measurement of the NaNO2 standard solutions. A typical standard curve is illustrated in 5.10. (Fermer et al., 2005)

5.9 Statistical analysis

Significant differences were assessed using two-tailed Student's t-test using Microsoft Office Excel 2007 (Redmond,WA) and P<0.05 was used to indicate statistical significance. (Cassino et al., 2007)

Chapter 6 Results

6.1 The results of the stimulation protocol I

Under Protocol I, stimulated constructs were subjected to a current density of 4 mA/cm2. The electrical stimulation signal was designed as two hours on, two hours off, anther two hours on, with 18 hours recovery time. Protocol I were repeated for three times. At the first two times, all the chondrocyte seeded agrose chonstructs were taken out for the viability test described in Chapter 4.

6.1.1 Viability test

For the very first experiment under protocol I, the viabilities of both of the control group and the stimulation group were observed lower than 50%.

For the second experiment under protocol I, the viabilities of constructs from both of the control group and the stimulation group were observed more than 97% ( 6.1). The average cell density ranged from 31 to 35 per field of view, observed under ×20 objective lens for these first two experiments.

For the third experiment under protocol I, only one control construct was taken out for viability test. However, the result was not consistent with the previous two practices. It was only averagely 8.7 per field of view, ranging from 4 to 12 per field of view. The viability was counted as 76%. All the other constructs were digested for biochemical assay.

6.1.2 Effect on DNA content

The effects of intermittent stimulation under protocol I on DNA content are illustrated in Table 6.1 and 6.2. The DNA content ranged between 16.85 and 26.94 µg across all the samples, which was not consistent with the data (5.9 to 6.4 µg) of the previous study in our host lab. The difference between the control group and the stimulation group was not significant. (p > 0.05).

Proximity to electrodes

DNA content (µg)

Control

Stimulation

Free swelling

1

22.69

16.85

22.60

2

18.17

19.48

25.25

3

26.94

20.00

NA

4

NA

19.81

NA

Mean value ± SD

19.03 ± 1.47

22.6 ± 4.38

23.92 ± 1.87

Table 6.1. The DNA content of the cell seeded construct in each position of the control group, the stimulation group and the free swelling group after being subjected to the stimulation of protocol I. 1, 2, 3, 4 represent the location of the constructs within the tank with position 1 closest to the negative electrode and position 4 closest to the positive electrode. The numbers 1, 2, 3, 4 in the following s and tables are all in this order.

6.2. DNA content in chondrocytes seeded in 4% agarose construct after being subjected to intermittent stimulation under protocol I. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group; F represents free-swelling group. Two-tailed Student's t-test indicate none statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.1.3 Effect on glycosaminoglycan (GAG) content

The effects of intermittent stimulation under protocol I on GAG content are illustrated in Table 6.2 and 6.3. To assess the influence of applied intermittent stimulation on GAG content, the GAG content in each construct and the GAG content dissolved in the conditioned medium of it were added together to obtain the total GAG content of each sample. To eliminate the effect of the variance of cell number and DNA content in each construct, the total GAG content of each sample was normalised to the DNA content, which means the mass of the GAG was divided by the mass of the DNA in each construct. However, under the protocol I, the difference between the mean of the stimulated and control samples was not statistically significant (p > 0.05).

Proximity to electrodes

GAG (µg) in construct

GAG (µg) in medium

Total GAG (µg)

Total GAG (µg) / DNA (µg)

Control

Stimulation

Free swelling

Control

Stimulation

Free swelling

Control

Stimulation

Free swelling

Control

Stimulation

Free swelling

1

15.41

17.39

18.9

7.59

6.4

5.82

26.04

26.34

36.36

1.15

1.56

1.61

2

16.45

13.589

19.57

8.38

5.41

7.91

28.17

21.15

43.29

1.55

1.09

1.71

3

20.2

13.48

NA

8.43

7.13

NA

32

23.45

NA

1.19

1.17

NA

4

NA

15.56

NA

NA

6.14

NA

NA

24.15

NA

NA

1.22

NA

Mean value ± SD

NA

NA

NA

NA

NA

NA

NA

NA

NA

1.30 ± 0.22

1.26 ± 0.21

1.66 ± 0.07

Table 6.2. GAG (µg) in construct, GAG (µg) in medium, Total GAG (µg), Total GAG (µg) / DNA (µg) of the sample in each position of the control group, stimulation group and the free swelling group.

6.3. GAG content in chondrocytes seeded in 4% agarose construct after being subjected to intermittent stimulation under protocol I. The GAG content was normalised by DNA content of each sample. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group; F represents free-swelling group. Two-tailed Student's t-test indicate none statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.1.4 Effect on NO release

The effects of intermittent stimulation under protocol I on NO release are illustrated in 6.4. The NO release between control group and stimulation group were compared by the concentration of NO in the conditioned medium after the stimulation protocol was applied. The difference between the control group and the stimulation group was not significant. (p > 0.05).

6.4. Nitrite release (C) by chondrocytes seeded in 4% agarose and subjected to intermittent stimulation under protocol I. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group;. Two-tailed Student's t-test indicates no statistically significant difference between the the stimulated group and the unstimulated control group exists (p > 0.05).

6.2 The results of the stimulation protocol II

Under protocol II, stimulated constructs were subjected to a current density of 4 mA/cm2. The electrical stimulation signal was designed as one hour on, one hour off, repeated for four times to keep the total stimulation time as four hours, plus a recovery period to keep the total culture time as 24 hours.

6.2.1 Effect on DNA content

The effects of intermittent stimulation under protocol II on DNA content are illustrated in Table 6.3 and 6.5. The DNA content ranged between 3.83 and 5.69 µg across all the samples. The difference between the control group and the stimulation group was not significant. (p > 0.05).

Proximity to electrodes

DNA content (µg)

Control

Stimulation

Free swelling

1

4.86

4.02

4.42

2

4.87

4.48

3.83

3

4.72

5.69

4.19

4

4.5

4.42

NA

Mean value ± SD

4.65 ± 0.72

4.74 ± 0.17

4.15 ± 0.30

Table 6.3. The DNA content of the cell seeded construct in each position of the control group, the stimulation group and the free swelling group after being subjected to the stimulation of protocol II. 1, 2, 3, 4 represent the location of the constructs within the tank with position 1 closest to the negative electrode and position 4 closest to the positive electrode. The numbers 1, 2, 3, 4 in the following s and tables are all in this order.

6.5. DNA content in chondrocytes seeded in 4% agarose construct after being subjected to intermittent stimulation under protocol II. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group; F represents free-swelling group. Two-tailed Student's t-test indicate none statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.2.2 Effect on glycosaminoglycan (GAG) content

The effects of intermittent stimulation under protocol II on GAG content are illustrated in 6.6, under the protocol II, the difference between the mean of the stimulated and control samples was not statistically significant (p > 0.05).

6.6. GAG content in chondrocytes seeded in 4% agarose construct after being subjected to intermittent stimulation under protocol II. The GAG content was normalised by DNA content of each sample. Error bars represent means and standard deviation of the four samples within one stimulation tank namely either in control group or in stimulation group. Two-tailed Student's t-test indicate none statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.2.3 Effect on NO release

The effects of intermittent stimulation under protocol II on NO release are illustrated in 6.7. The NO release between control group and stimulation group were compared by the concentration of NO in the conditioned medium after the stimulation protocol was applied. The difference between the control group and the stimulation group was not significant (p > 0.05), which is consistent with the result of protocol I.

6.7. Nitrite release by chondrocytes seeded in 4% agarose and subjected to intermittent stimulation under protocol II. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group. Two-tailed Student's t-test indicates no statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.3 The results of the stimulation protocol III

Under protocol III, stimulated constructs were still subjected to a current density of 4 mA/cm2. The electrical stimulation signal was designed as half hour on, half hour off, repeated for eight times to keep the total stimulation time as four hours, plus a recovery period to keep the total culture time as 24 hours. The effects of intermittent stimulation under protocol III on DNA content were illustrated respectively in 6.8, 6.9 and 6.10. Two-tailed Student's t-test indicate none statistically significant difference of the DNA content, GAG content and NO release exists between the stimulated group and the unstimulated control group (p > 0.05).

6.8. DNA content in chondrocytes seeded in 4% agarose construct after being subjected to intermittent stimulation under protocol III. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group. Two-tailed Student's t-test indicate none statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.9. GAG content in chondrocytes seeded in 4% agarose construct after being subjected to intermittent stimulation under protocol III. The GAG content was normalised by DNA content of each sample. Error bars represent means and standard deviation of the four samples within one stimulation tank namely either in control group or in stimulation group. Two-tailed Student's t-test indicate none statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.10. Nitrite release by chondrocytes seeded in 4% agarose and subjected to intermittent stimulation under protocol III. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group. Two-tailed Student's t-test indicates no statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.4 The results of the stimulation protocol IV

Under protocol III, stimulated constructs were subjected to a current density of 8 mA/cm2. The electrical stimulation signal was designed as half hour on, half hour off, repeated for eight times to keep the total stimulation time as four hours, plus a recovery period to keep the total culture time as 24 hours. Protocol IV was repeated once. In the first experiment, one sample of position 4 from stimulation group was taken out for cell viability test. The result showed all the cells were dead. In the second experiment, one sample of position 4 from stimulation group and another sample of the same position from control group were taken out for cell viability test. The result showed all the cells were dead after the stimulation under protocol IV while the cell viability was still more than 70% in the control group ( 6.11). Then the rest of the samples were taken out for biochemical assay. Although the cell viabilities between the stimulated sample and control sample were quite different as described above, the biomedical assay still showed no significant difference of the GAG content exists between the control group and the stimulation group. ( 6. 12)

6.11. The illustration of the viability test of one stimulation sample and one control sample. In the second experiment, one sample from stimulation group and another sample from control group were taken out for cell viability test. The result showed all the cells were dead after the stimulation under protocol IV while the cell viability was still more than 70% in the control group.

6.12. GAG content in chondrocytes seeded in 4% agarose construct after being subjected to intermittent stimulation under protocol IV. The GAG content was normalised by DNA content of each sample. Error bars represent means and standard deviation of the four samples within one stimulation tank namely either in control group or in stimulation group. Two-tailed Student's t-test indicate none statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.13. Nitrite release by chondrocytes seeded in 4% agarose and subjected to intermittent stimulation under protocol IV. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group. Two-tailed Student's t-test indicates no statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.5 The results of the stimulation protocol V

Under protocol V, stimulated constructs were subjected to a current density of 2mA/cm2. The electrical stimulation signal was designed as half hour on, half hour off, repeated for eight times to keep the total stimulation time as four hours, plus a recovery period to keep the total culture time as 24 hours. The effects of intermittent stimulation under protocol V on DNA content, GAG content, and NO release were illustrated respectively in 6.14, 6.15 and 6.16. Two-tailed Student's t-test indicate none statistically significant difference of the DNA content and GAG content exists between the stimulated group and the unstimulated control group (p > 0.05). However, the release of NO in the stimulation group was significantly less than in the control group (p < 0.05).

6.14. DNA content in chondrocytes seeded in 4% agarose construct after being subjected to intermittent stimulation under protocol V. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group. Two-tailed Student's t-test indicate none statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.15. GAG content in chondrocytes seeded in 4% agarose construct after being subjected to intermittent stimulation under protocol V. The GAG content was normalised by DNA content of each sample. Error bars represent means and standard deviation of the four samples within one stimulation tank namely either in control group or in stimulation group. Two-tailed Student's t-test indicate none statistically significant difference exists between the stimulated group and the unstimulated control group (p > 0.05).

6.16. Nitrite release by chondrocytes seeded in 4% agarose and subjected to intermittent stimulation under protocol V. Error bars represent means and standard deviation of the four samples within one stimulation tank, namely the four samples either in control group or in stimulation group. C represents control group; S represents stimulation group. Two-tailed unpaired Student's t-test indicates statistically significant difference exists between the stimulated group and the unstimulated control group (p < 0.05).

Chapter 7 Discussion

The aim of this study was to test the hypothesis that intermittent DC stimulation provides an up-regulation of the bio-synthesis activity of chondrocytes seeded in 3D agarose constructs. In order to test this hypothesis, chondrocytes were isolated from the mature steer metacarpophalangeal joints and then seeded cubic 5 x5x5mm constructs, which were then incorporated into the electrical stimulation rig established by Akanji in our host lab. (Akanji, 2008) Five separate different intermittent stimulation protocols were successfully applied to the constructs in the stimulation tank. The stimulation time was controlled by manually operate the control panel of the power source. Different current densities were supplied by the power source and monitored by an ammeter. Subsequently, the GAG content, DNA content and NO release were tested by the biochemical assay. The results between stimulation group and control group were compared. The results are limited by the available time schedule. Accordingly, the experiments under most of the protocols were only carried on once. The lack of repeat would impair the reliability of the results. However, the results were generally consistent with the previous Phd program done by Akanji in our host lab.

Three major markers of chondrocyte metabolism are proteoglycan synthesis, cell division, and protein synthesis. Proteoglycan synthesis was measured by GAG assay in our experiments. GAG assay was used in the previous work in our host lab to analyse the effect of dynamic compressive stress on bovine chondrocytes. GAG synthesis was prominently improved at 1 Hz frequency. The radioactive 35SO4 incorporation is another way to detect GAG synthesis. It was not used in the present study because of the rule for radioactive materials in our lab. The cell division and can be assessed by the [3H]-thymidine incorporation during s-phase of the cell cycle (Chowdry et al., 2003). The protein synthesis can be assessed by the [3H]-hydroxyproline (Brighton et al., 2006) or [35S]-methionine incorporation respectively (Macginitie et al., 1994). For the same reason, these two kinds of experiment were not carried on in present study. Capacitive coupled alternating current (AC) electrical stimulation has been proved effective in improving 35SO4 incorporation and [3H]-thymidine incorporation of chondrodytes from both growth plate chondrocytes (Brighton et al., 1976) and articular cartilage chondrocytes (Brighton et al., 1984, 2006). Szasz et al. has proved that direct AC stimulation improved 35SO4 incorporation but decreased the [3H]-thymidine (Szasz et al., 2003). Most of the studies showed positive result which means electrical stimulation can improve chondrocyte biosynthesis. Converse to most of previous experiments, the result of Akanji's study in our host lab showed that the continuous direct current (DC) did not influence 35SO4 and [3H]-thymidine uptake by chondrocytes in a 3D agarose system.

In the light of the success of intermittent mechanical stimulation (chowdhury et al., 2003) and the intermittent capacitively coupled electrical stimulation (Wang et al., 2004; Brighton et al., 2006) in improving condrocyte bio-synthesis, the present study shift the continuous DC stimulation into the intermittent stimulation based on Akanji's stimulation rig. Protocol III (half hour interval, 4 mA/cm2) was the only protocol under which the GAG content in stimulated group was slightly higher than the control group. However, according to the statistical analysis, the applied intermittent DC electrical stimulation with the current density of 4 mA/cm2 or 2 mA/cm2 did not alter the absolute DNA amounts and the GAG content of the stimulation group from control levels in all the applied stimulation protocols.

Specially, when the current density of the electrical stimulation was added up to 8 mA/cm2 under protocol IV (half hour interval, 8 mA/cm2), the viability test showed all the cells were died in the sample taken out from stimulation group while the cell viability was good in control group. The biochemical assay done on the rest of the samples showed a small reduction in stimulated constructs compared to controls ( 6.12), although the differences were not statistically significant (p > 0.05). That implies the current density up to 8 mA/cm2 can impair the viability. In addition, the GAG assay might not be enough to reveal the real viability and the biosynthesis condition.

NO is a key mediator of catabolic processes in chondrocytes. Physical stimulation has previously been reported to influence NO production in condrocytes, which suggests that NO plays a key role in mechanotransduction (Lee et al., 1998). In addition, dynamic compression is proved to be able to inhibit the synthesis of nitric oxide and PGE2 by IL-1b-stimulated chondrocytes cultured in agarose constructs. None of the previous studies have been done to examine the electrical stimulation effect on NO production. Our experiments examined the NO production after intermittent electrical stimulation. According to the statistical analysis, the applied intermittent DC electrical stimulation with the current density of 4 mA/cm2 or 8 mA/cm2 did not alter the NO production of the stimulation group from control levels in all the applied stimulation protocols I II III IV. However, under protocol IV (half hour interval, 2 mA/cm2), NO production of the stimulation group was significantly reduced. Considering the high standard deviation of each group and the lack of experiment repeat, the result needs to be further proved.

Chondrocyte seeded agarose constructs in our study provided a homogeneous cell population and homogeneous matrix properties for study of chondrocyte response to electrical stimulation. Electrical current can pass through the construct cultured in DMEM + 16% FCS homogeneously thus it can affect each cell equally. 3D chondrocyte/agarose structure can also eliminate the affecting factor of the extracellular matrix (MacGinitie et al. 1994).

In summary, the hypothesis that intermittent DC stimulation provides an up-regulation of the bio-synthesis activity of chondrocytes seeded in 3D agarose constructs is tested to be a null hypothesis up until to now. The intermittent DC stimulation had no significant effect on DNA product and GAG product while the effect on NO product needs to be further proved. The present work contributes to the human knowledge of the electromechanics within cartilage and the effect of the electrical stimulation on chondrocyte biosynthesis, providing valuable experience for the further study in future.

To find an optimal protocol to upgrade the biosynthesis of chondrocytes, the future work needs to be done on a further change of the stimulation time and the current density. The present study indicated the current density should be controlled lower than 8 mA/cm2. In addition, Okihana and Shimomura's study suggests that the constant 1µA DC stimulation up to seven days can significantly upgrade the proteoglycans and DNA synthesis of chondrocytes from growth plate, while the 10µA DC stimulation devitalise the chondrocytes (Okihana and Shimomura, 1988). This suggests the future work can be attempted on longer stimulation time and lower current density. As has been reported previously, once a cell has been stimulated to

enter the cell cycle, it will proceed through the complete cycle, which is believed to take several days for a bovine chondrocyte (Chowdry et al., 2003). This suggests future work can be attempted on peak to peak stimulation with a long interval of recovery time.

As described above, GAG assay and DNA assay might not be sufficient to reveal the biosynthesis condition in molecular level. The isotopic 35SO4 and [3H]-thymidine incorporation need to be examined in the future work. In addition, the influence of the DC intermittent stimulation might only be limited in the level of gene expression. As we know, collagen II and aggrecan are the phenotype marks of articular cartilage. The future work could investigate the mRNA expression levels of Collagen II and Aggrecan by the RT-PCR technique.