INDEX Incorporation of foreign body particles following injury in immature rat articular cartilage

Nobuyuki Sato, Eisuke Sakuma*, Hong Jian Wang *, Yoshio Mabuchi*, Tsuyoshi Soji* and Nobuo Matsui
Department of Orthopedic Surgery and
*Department of Anatomy
Nagoya City University Medical School, 1 Kawasumi, Mizuho-cho,
Mizuho-ku, Nagoya City, Aichi, 467-8602, Japan

Running title: Incorporation of particles into cartilage
Key words: cartilage damage, wear debris, polystyrene latex beads, incorporation, fluid flow

Correspondence: Dr. Nobuyuki Sato: Department of Orthopedic Surgery Nagoya City University Medical School, 1 Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya City, Aichi, 467-8602, Japan

The production of particulate wear debris has recently been recognized to result in delayed mechanical failure of joint arthroplastic surgery. In endoprosthtic femoral head replacement, there is also an underlying damage to the acetabular cartilage. To investigate the influence of foreign bodies upon injured cartilage, we studied ultrastructual changes of immature rat patella cartilage following the intraarticular injection of polystyrene latex beads. The articular surfaces of the patellae of 20 and 30 day-old rats were damaged, creating a groove with maximum depth of 100 um. Following the damage, polystyrene beads (240nm diameter) suspended in saline were injected intraarticularly. Group A received the treatment on day 20, group B on day 30. The patellae in group C were damaged on day 20 and beads applied on day 30. Patella samples were taken an hour after the injection and examined by transmission electron microscope. Erythrocytes were found in the groove which was performed in injury to patella cartilage. Beads were observed entering the damaged cartilage matrix only in group , and not in the other two groups. A new layer overlying the damaged surface was seen in group C. The prerequisites for incorporation of beads in cartilage are destruction of the collagen network and interstitial fluid flow which depends on compressive loading. Our results indicate that wear debris following endoprostheses can enter damaged cartilage.

Mature articular cartilage lacks a vascular system. It is now widely accepted, through studies using radioisotope and dyes administered intraarticularly, that the nutrition of articular cartilage is mainly derived from the synovial fluid (1-4). It is also known that especially in immature rabbits, articular cartilage is nourished from both the synovial fluid and through the subchondral bone (3,4). Furthermore, movement of nutrients in cartilage is shown to depend not only on static diffusion but also on dynamic fluid flow which results from compressive loading (5-7).
In the field of orthopedic arthroplastic surgery, microscopic wear debris leads to loosening and ultimate failure of implants by disruption of the bone-implant interface. In certain cases, devices are implanted on one side of a joint, making a joint with the original cartilage on the other side. The effect of wear particles on cartilage, intact or damaged, is unknown. We studied the ultrastructual changes in immature rat patella cartilage following intraarticular administration of particles using polystyrene latex beads.
The all animals were treated with メThe Guidelines for Animal Experimentationモ of the Experimental Animal Science Center of the University. Twenty male Wister rats were separated into 4 groups (intact control, A, B and C). All were weaned on day 20 and permitted to move freely in the cage. Polystyrene latex beads (240 nm in diameter) composed of layered polystyrene and methyl methacrylate with a central void cavity were suspended in saline at a concentration of 0.02mg beads/ml. The animals were anesthetized with intraperitoneal Nembutal (pentobarbital, mg/kg). In group A, a 28G needle, bent 100 micrometers from the tip, was introduced into the right knee joint through the patellar ligament. The articular surface of the patella was scratched, resulting in a longitudinal groove with a maximum depth of 100 micrometers. The injured depth were checked after the experiment, and the too deep (over 100um) or too shallow (under 70 um) grooves were not used for observations. Following this, 0.1ml of the bead suspension was injected intraarticularly except for the intact control group. The contralateral left knee was given the beads injection only, without the chondral scratch. Group B received the same treatment on day 30. In group C, the patella was scratched on day 20 and beads injected on day 30. All animals were allowed to move for one hour after the treatments, then patella samples were taken from each rat. The specimens were fixed for 180 min in 0.05 M cacodylate buffered (pH. 7.4) 2.5% glutaraldehyde with 2 % sucrose. After the fixation, specimens were rinsed in buffer solution for 2 hours and subsequently postfixed in 0.05M cacodylate buffered 1% osmium tetroxide (pH. 7.4) with 2% sucrose for 2.5 hours. After the postfixation, the tissues were rinsed in ice-cold water for 10 min and dehydrated in a graded series of ethanol. Following two rinses in 100 % ethanol, the tissues were immersed twice in absolute propylene oxide for 15 min and then embedded in epoxy resin (8). Ultra-thin sections were prepared and electron stained, and then examined with a Hitachi H-7000 transmission electron microscope.

Normal control
In the normal cartilage of the patella of the knee joint, the articular surface was smooth (Fig. 1) and a thin electron dense layer, which seemed to be basement membrane, approximately 0.5 um thick, was observed (Fig. 2). The layer was called "lamina splendens". Just below the thin electron dense layer, elongated cells were found. The cell resembled fibroblast. They became more spherical in shape and formed groups as the size of the territorial matrix increased with the depth of the cartilage (Fig. 2).

Group A
A deep groove approximately 80 um in depth and containing erythrocytes was clearly observed. The erythrocytes were discharged into the joint cavity from a hemorrhage as a result of the needle being placed into the cavity. Slender elongated cells which resembled so-celled メsynovial B cellsモ were arranged along the surface of the groove (Fig. 3).
The thin electron dense layer so called メlamina splendenceモ was either very thin or absent. The surface of the groove was incompletely covered by two types of cells. One type resembled so-called メsynovial A cellsモ had numerous cell processes. The other type resembled so-called メsynovial B cellsモ had longer cell processes at either end.  The synovial B -like cell usually paired with the synovial A-like cell (Fig. 4).  No chondrocyte layer was found on the face of the groove but near the surface of the groove, two chondrocytes were intimately situated (Fig. 5). Occasionally, very narrow intercellular spaces existed between adjacent chondrocytes which had an interval estimated to be several nanometers. High magnification electron micrographs revealed a dotted line between plasma membranes. The periodicity of this structure corresponded to a gap junction (Fig. 6).
Chondrocytes around the matrix defects (groove) had disorganized cytoplasmic contents, displaying in situ necrosis. Some latex beads were observed within the groove, but no beads were found invading the roughened collagen network in the all cases, electron microscopically (Fig. 7). Latex beads were not observed within roughened collagen network but attached the damaged surface of the cartilage (Fig. 8).

Group B
In contrast to the findings of group A, latex beads were found in the all cases. The beads penetrated into the damaged cartilage matrix at various depths; roughly half the diameter of a particle (Fig. 9), fully buried but adjacent to the damaged surface of the groove (Fig.10), and migrating within the damaged matrix (Fig. 11). The matrix into which the beads invaded showed roughened collagen fibers. In the intact knee at day 30, the lamina splendens presented as a yet incomplete electron dense layer. It consisted of randomly orientated microfibrils with gaps between the microfibrils estimated at 4 to 7 nm (Fig 12).

Group C
Erythrocytes were seen similarly in the groove, and a new layer rich in collagen fibers was found overspreading the injured cartilage. Underlying chondrocytes presented a round shape and differed from the flat, fibroblast-like morphology seen in the superficial layer of normal cartilage. In the all cases, the latex beads were not found in the matrix at all and existed only on the outside of the new layer. The particles and debris were seen in clusters, suggesting a focal flow of articular fluid (Fig 13 ).

In this study, intact cartilage allowed no invasion of latex beads into the matrix and latex beads entered damaged matrix only in group B. Added to this observation, the gap junctions were found between several pairs of chondrocytes.
Dealy et al. (1994) traced expressions of m-RNA of the connexin family of the embryonic chick limb bud using in situ hybridization. They noted that expression of connexin 42 was continuous, but connexin 43 was transitorily expressed leading them to conclude that connexin 42 was involved in the differentiation of certain organs. On the other hand, Donahue et al. (10) found that connexin 43 expression was detected on the mature articular cartilage in vitro. In these two studies, only gap junctional expression of m-RNA was investigated and not the ultrastructure of gap junctions. Langille (11) reported that gap junctions appeared within 48 hours after the mandibular and frontonasal tissue of 12 day-old rat embryos were cultured, indicating that gap junctional proteins were usually present and ready to form mature structures. It is well known that once avascular tissue such as cartilage is injured, it is hard to maintain, the injured condition continues into adulthood (12) and the chrondrocytes seem unable to retain their biological activity.
The experimental animals which were used in the present study were very young and within a developing stage. Therefore, it was rather reasonable that the chondrocytes would retain their ability to maintain the cartilage matrix. To recover from damage, tissues require more information than what is needed under normal conditions, similar to what is occurring during the proliferation stage of cartilage development. Thus, we speculate that with injury, gap junctions are playing a role in the transmission of information to aid in tissue recovery.
Stockwell and Barnett (13) and Barnett and Palfrey(14) reported the relation between particle diameter and penetration depth. In these studies, silver proteinate (4 nm in diameter) penetrated 30-170 micrometers into the cartilage, whereas particles (5-10 nm in diameter) penetrated only about 2-3 micrometers. Since these reports, it has generally been believed that there are pores of about 5 nm in diameter in normal cartilage and that foreign bodies larger than 10 nm in diameter could not be transported into cartilage matrix. In the present study, the gaps between microfibrils overlying the superficial layer of intact cartilage at day 30 proved to be 4 to 7 nm as shown in Fig 12. This value is in agreement with the size of pores described by Stockwell and Barnett (13) and Barnett and Palfrey(14). It is therefore reasonable to assume that pores of approximately 5 nm do exist in the lamina splendens and that the layer acts as a barrier against foreign bodies of larger than several nanometers. Furthermore, there have been reports as to the interval between fibers in the cartilage matrix. Maroudas (15) reported that the gaps between the collagen fibers were about 100 nm. We have also observed similar intervals between fibers in the superficial layer. Since the scratching of the cartilage in our experiments caused destruction of the lamina splendens and disruption of the collagen network in the superficial layer, all specimens could theoretically incorporate latex beads into damaged cartilage. Nevertheless, actual penetration was observed only in group B. A possible explanation for this finding may be in the characteristics of nutrient flow in normal avascular cartilage. It is generally accepted that nutrients are transported to the cells by diffusion from synovial fluid. This transport is also thought to be assisted by movement of fluid in and out of cartilage in response to cyclic loading of the tissue ( pumping )(6). The latex beads in our experiments were likewise subject to these two modes of transport, diffusion and fluid flow. Honner and Thompson (4) reported using intravenous 35S that in immature rabbits articular cartilage was nourished from both the synovial fluid and subchondral bone and that in mature rabbits it was derived only from the synovial fluid. It may be argued that the lack of particulate invasion in group A is due to the presence of stronger diffusion from the subchondral bone than from the synovial fluid. However, the large particles (240 nm in diameter) used in our study may not obey the same laws of transportation as the low molecular weight solutes (sodium-35S) were used in the experiment of Honner and Thompson (4). O, Hara et al. (6) demonstrated the different transport properties of solutes with different sizes. They reported that for small solutes (urea, NaI) fluid flow due to cyclic loading of the joint did not affect the rate of solute transport and that the rate of absorption of a large solute (serum albumin, Stokes radius 3.55 nm), however, was increased by 30-100% because of cyclic loading. In relation to this, we believe that the age of the animals in our experiments is also an important factor. The twenty day-old rats in group A had just been weaned, so the patello-femoral joints of this group suffered little compressive loading, as compared with group B which was more actively mobile. From these observations, it can be stated that particulate penetration into cartilage matrix occurs only certain conditions. There must be a preexisting damage to the integrity of the collagen barrier structures, and there must be fluid flow as a result of joint motion. When patella cartilage was injured by the needle, its superficial region from the surface to the transitional layer suffered a breakdown of the three dimensional collagen network. Latex beads injected into the intraarticular space are thought to have been transported into this compromised zone passively by interstitial fluid flow (Fig 7-11).
As for group C, there appeared a new layer of matrix overlying the damaged cartilage. That layer consisted of collagen fibers arranged in parallel to the articular surface, and underlying chondrocytes showed an activated, round morphology as opposed to the flat cells of intact cartilage. Following creation of the cartilage defect, the new layer is considered to have been organized by chondrocytes in the transitional layer. It is strongly suggested that the new layer prevented the latex beads from entering cartilage. In mature animals articular cartilage is poorly repaired, whereas, in immature animals used in this study, the chondrocytes retained the potential to hypertrophy and build up a new layer of matrix.
The generation of particulate wear debris is recognized as the major problem affecting the longevity of total hip arthroplasty and endoprostheses. Since Charnley (1975) described the problem of wear debris, the biological response to particulate wear debris resulting in aseptic loosening and pathological bone resorption has been widely studied. These phenomena are thought to derive from macrophage activation secondary to the phagocytosis of wear debris. Macrophages that phagocytose particles of wear debris release mediators like prostaglandin E2, that stimulate osteoclastic bone resorption (17). As for the size of wear particles, the mean size of particles obtained during hip replacement operation has been reported to be 0.63 um (18). The rounded polyethylene particles from a simulator were primarily 0.2 um or smaller, whereas those generated in vivo ranged primarily from 0.2 to 0.8 um (19). The size of the polystyrene latex beads using in the present study are similar to the size of wear debris described by McKellop et al. (19). These particles, consisting of double layers with a central cavity, were firstly applied to examine phagocytosis of large foreign bodies by hepatocytes (20). Their microscopic appearance is easily distinguished from other naturally occurring structures already present in the cells.
Prosthetic replacement of a failed femoral head is a frequently performed procedure. The acetabular cartilage is exposed to damaging stresses, during surgery and afterward, with repeated compressive loading and abrasion by metallic or ceramic surfaces. Cruess et al. (1984) described the degeneration process of cartilage starting with early loss of proteoglycan, followed by surface damage to the cartilage, progressive degenerative changes, and growth of pannus from the articular margins. There is no doubt that cartilage injury alone like that created in this study could lead to degenerative changes, but incorporation of wear debris after the injury also could accentuate and accelerate the damage. As discussed above, articular cartilage has at least a double filtering systems. One is a filter of about 5 nm in the lamina splendens, and the other is the collagen network of the matrix itself (100 nm gap). Destruction of this filtering system could admit foreign bodies larger than 100 nm into the cartilage matrix. This could ultimately lead to failure of the homeostasis of the joint, whose components include cartilage, synovia, and synovial fluid.
The results of this study indicate that wear debris produced following endoprosthetic surgery could enter cartilage mainly by fluid flow. This particulate penetration can only occur in cartilage whose collagen structure has been damaged. Such micropathological changes could lead to failure of the joint.

The authors thank Professor Damon C. Herbert (Department of Cellular and Structural Biology, The Health Science Center at San Antonio, University of Texas) for their scientific advice on this study.

1) Barnett, C. H. and A. J. Palfrey (1965) Absorption into the rabbit articular cartilage. J Anat 99, 2: 365-375.

2) Brower, T. D., Y. Akahoshi, and P. Orlic (1962) The diffusion of dyes through articular cartilage in vivo. J Bone and Joint Surg 44-A, No. 3: 456-463.

3Charnley, J. and D. K. Harry (1975) Rate of wear in total hip replacement. Clin Orthop 112: 170-179.

4) Cruess, R. L., D. C. Kwok, P. N. Duc, M. A. Lecavalier, and G-T. Dang (1984) The response of articular cartilage to weight-bearing against metal. J Bone and Joint Surg 66-B, No. 4: 592-597.

5) Dealy, CN, Beyer, EC and Kosher, RA 1994 Expression patterns of mRNAs for the gap junction proteins connexin 43 and connexin 42 suggest their involvement in chick limb morphogenesis and specification of the arterial vasculature. Developmental Dynamics, 199: 156-167

6) Donahue, HJ, Guilak, F, Vander Molen, MA, McLeod, KJ, Rubin, CT, Grande, DA and Brink, PR 1995 Chondrocytes isolated from mature articular cartilage retain the capacity to form functional gap junctions. J. Bone Mineral Res., 10: 1359-1364

7) Garcia, A. M., E. H. Frank, P. E. Grimshaw and A. J. Grodzinsky (1996) Contributions of fluid convection and electrical migration to transport in cartilage: Relevance to loading. Arch Biochem Biophys 333, No. 2: 317-325.

8) Hodge, J. A., and B. McKibbin (1969) The nutrition of mature and immature cartilage in rabbits. J Bone and Joint Surg 51-B, No. 1: 140-147.

9) Honner, R., F.R.A.C.S. and R.C. Thompson (1971) The nutritional pathways of articular cartilage: An autoradiographic study in rabbits using 35 S injected intravenously. J Bone and Joint Surg 53-A, No. 4: 742-748.

10) Kanai, M., Y. Murata, Y. Mabuchi, N. Kawahashi, M. Tanaka, T. Ogawa, M. Doi, T. Soji, and D. C. Herbert (1996) In vivo uptake of lecithin-coated polystyrene beads by rat hepatocytes and sinusoidal endothelial cells. Anat Rec 244: 175-181.

11) Langille, RM 1994 Chondrogenic differentiation in cultures of embryonic rat mesenchyme. Micro. Res. Techn., 28:455-469

12) Luft, J.H. 1961. Improvement in epoxy resin embedding methods. J. Biophys. Biochem. Cytol. 9: 409-414.

13) Malony, W. J., R. L. Smith, T. P. Schmalzried, J. Chiba, D. Huene, and H. Rubash (1995) Isolation and characterization of wear particles generated in patients who have had failure of a hip arthroplasty without cement. J Bone and Joint Surg 77-A, No. 9: 1301-1310.

14) Maroudas, A., P. Bullough, S. A. V. Swanson, and M. A. R. Freeman (1968) The permeability of articular cartilage. J Bone and Joint Surg 50-B, No. 1: 166-177.

15) Maroudas, A (1970) Distribution and diffusion of solutes in articular cartilage. Biophys J 10: 365-379.

16) Maroudas, A. (1976) Transport of solutes through cartilage: permeability to large molecules. J Anat 122, 2: 335-347.

17) Mckellop, H.A., P. Campbell, S.H. Park, T.P. Schmalzried, P. Grigoris, H.C. Amstutz, and A. Sarmiento 1995 The origin of submicron polyethylene wear debris in total hip arthroplasty. Clin. Orthop., 311: 3-20.

18) Michell, N and Shepard, N 1976 The resurfacing of adult rabbit articular cartilage by multiple perforations through the subchondral bone. J. Bone Joint Surg., 58-A: 230-233

19) Murray, D. W., and N. Rushton (1990) Macrophages stimulate bone resorption when they phagocytose particles. J Bone and Joint Surg 72-B, No. 6: 988-992.

20) O,Hara, B. P., J. P. G. Urban, and A. Maroudas (1990) Influence of cyclic loading on the nutrition of articular cartilage. Ann Rheum Dis 49: 536-539.

21) Stockwell, R. A., and C. H. Barnett (1964) Changes in permeability of articular cartilage with age. Nature, Lond 201: 835-836.

Figure 1: Semi-thin section of normal control. The surface of the articular cartilage is smooth and fibroblastic cells are present.

Figure 2: Electron micrograph of a normal control. An thin electron dense layer is clearly observed (between arrowheads). Fibroblastic-like chondrocytes are present in several layers of the cartilage matrix.

Figure 3: Semi-thin sections of an experimental animal. A deep groove is observed (arrow) and around the groove, a non-cellular amorphous layer is can be found (*) along with flat-shaped cells arranged on the surface of the groove (arrows in inset)

Figure 4: Electron micrograph of the lining cells of the groove. Note a pair of synovial A (A) and a B (B)cell.

Figure 5: Cartilage near the surface of groove. The chondrocyte layer is absent and was replaced by a dense collagen network. Amorphous areas which are presumed to come from the chondrocytes were found (arrow).
Figure 6: Gap junctional connections between two cells.
Inset: High magnification of rectangle A. The interval of the intercellular space was estimated to be 7 nm. A dotted electron dense line which was between the trilamellar structure of the junction (arrow) were observed.

Figure 7: In the groove (cartilage defect), a few erythrocytes was seen. Chondrocytes around the groove had disorganized cytoplasmic contents, suggesting in situ necrosis. Latex beads were found in the groove (arrow), but no invasion into matrix fibers was seen. (A group). Inset: A high power photograph of a rectangular part.

Figure 8: The cartilage surface which was injured of A group. Latex beads were not observed within roughened collagen network but attached the damaged surface of the cartilage.

Figure 9: The injured surface of B group. A half sinking particle was found into the matrix fiber of the transitional layer (arrow). Chondrocytes in not superficial but transitional layer was seen below the injured surface. Inset: High magnification of the arrow

Figure 10: Whole body of a particle was buried (arrow), but a part of it was facing on the scratched surface (B group).

Figure 11: A bead of B group was completely penetrating into roughened collagen fibers (arrow).

Figure 12: Lamina splendens from control at 30 day. Incomplete high density layer was seen on the surface of patella cartilage. Normal chondrocyte in the superficial layer presented flat like fibroblast. Inset: Note the cross section of fine collagen fibrils of lamina splendens (arrow). The intervals of fine fibers (arrowheads) were ranged from 4 to 7 nm.

Figure 13: A new layer which had collagen fibers was found overspreading injured cartilage (arrowhead). Note underlying chondrocytes present round shape, suggesting ones of the transitional layer following the operation. Cluster of latex beads was attached on the surface of the new layer (C group).

Nagoya City University
Medical School