Abstract

Abstract

Abstract

Early changes within articular cartilage during human idiopathic osteoarthritis are poorly understood. However alterations to chondrocyte morphology occur with the development of fine cytoplasmic processes and cell clusters, potentially playing a role in cartilage degeneration. The aggrecanase ADAMTS-4 (A disintegrin and metalloproteinase with thrombospondin motifs-4) has been implicated as an important factor in cartilage degradation, so we investigated the relationship between chondrocyte morphology and levels of ADAMTS-4 in both non-degenerate and mildly osteoarthritic human cartilage.

Human femoral heads were obtained following consent from patients undergoing hip arthroplasty following femoral neck fracture. Cartilage explants of normal (grade 0; G0) and mildly osteoarthritic (grade 1; G1) cartilage were labelled with the cytoplasmic dye CMFDA (5-chloromethylfluorescein-diacetate). Explants were cryosectioned (30μm sections), and labelled for ADAMTS-4 by fluorescence immunohistochemistry. Sections were imaged with confocal microscopy, allowing the semi-quantitative analysis of ADAMTS-4 and 3D visualisation of in situ cell morphology.

With cartilage degeneration from G0 to G1, there was a decrease in the proportion of chondrocytes with normal rounded morphology (P<0.001) but an increase in the proportion of cells with processes (P<0.01) and those in clusters (P<0.001;[4(1653)]; femoral heads:cells). Although average levels of ADAMTS-4 for all cells was the same between G0 and G1 (P>0.05), a change was evident in the distribution curves for cell-specific ADAMTS-4 labelling. Cell-by-cell analysis showed that ADAMTS-4 levels were higher in chondrocytes with cytoplasmic processes compared to normal cells (P=0.044) however cells in clusters had lower levels than normal cells (P=0.003;[3(436)]). Preliminary data suggested that ADAMTS-4 levels increased with larger chondrocyte clusters.

These results suggest complex heterogeneous changes to levels of cell-associated ADAMTS-4 with early cartilage degeneration – increasing in cells with processes and initially decreasing in clusters. Increased levels of ADAMTS-4 are likely to produce focal areas of matrix weakness potentially leading to early cartilage degeneration.

The development of cytoplasmic processes from in situ chondrocytes is a characteristic feature of early osteoarthritis in human cartilage. The processes involve cytoskeletal elements and are distinct from the short primary cilia described in human chondrocytes. Vimentin is an intermediate filament playing an essential structural and signal-transduction role. We determined cellular levels and distribution of vimentin in chondrocytes of different morphologies in non-degenerate and mildly osteoarthritic cartilage.

Femoral heads were obtained after consent from patients undergoing hip arthroplasty following femoral neck fracture. Cartilage explants were graded as non-degenerate (grade 0;G0) or mildly osteoarthritic (grade 1;G1) and labelled with the cytoplasmic dye CMFDA (5-chloromethylfluorescein-diacetate) for cell shape. Explants were cryosectioned and labelled for vimentin by fluorescence immunohistochemistry. In situ chondrocyte morphology was identified by confocal microscopy as either normal (rounded/elliptical) or abnormal (with one or more cytoplasmic process of ≥2µm) and vimentin levels and distribution determined semi-quantitatively and related to chondrocyte morphology.

When all cells in G0 and G1 cartilage were compared, there was no difference between average levels of vimentin per cell (P=0.144)[6(261)];femoral heads:cells). When cells were separated on the basis of morphology, there was no difference between vimentin levels in cells with one or more cytoplasmic process compared to those of normal morphology (P>0.05;[6(261)]). However vimentin levels were much greater at the base of cytoplasmic processes compared to distant areas of the same cells (P=0.021)[5(29)]).

Although overall levels of chondrocyte vimentin do not change in these early stages of osteoarthritis, the formation and structure of these substantial chondrocyte cytoplasmic processes involves changes to its distribution. These morphological changes are similar to those occurring during chondrocyte de-differentiation to fibroblasts reported in osteoarthritis which results in the formation of mechanically-inferior fibro-cartilage. Alterations to chondrocyte vimentin distribution either directly or indirectly may play a role in cartilage degeneration.

Cartilage degeneration and loss are key events in the initiation and progression of osteoarthritis (OA). Changes to chondrocyte volume and morphology (in the form of cytoplasmic processes) and thus cell phenotype are implicated, as they lead to the production of a mechanically-weakened extracellular matrix. The chondrocyte cytoskeleton is intimately linked to cell volume and morphology and hence we have investigated alterations to levels and distribution of chondrocyte F-actin that occur during early OA.

The femoral heads (FH) from hip joints (N=16) were obtained with ethical permission and patient consent following femoral neck fracture. Cartilage was assessed as grade 0 (non-degenerate) and grade 1 (superficial fibrillation) using OARSI criteria. In situ chondrocyte volume and F-actin distribution were assessed using the fluorescent indicators (5-chloromethyl fluorescein diacetate (CMFDA)) and phalloidin, and imaged and quantified by confocal microscopy, Imaris^TM and ImageJ software.

There were no differences between the volume or total F-actin levels of in situ chondrocytes within the superficial zone of grade 0 (n=164 cells) compared to grade 1 (n=145) cartilage (P>0.05). However, a more detailed analysis of phalloidin labelling was performed, which demonstrated significant increases in both intense punctuate (IP) or intense areas (IA) (P<0.0001; P=0.0175 respectively). A preliminary analysis of IP and IA F-actin labelling suggested that while the former did not appear to be associated with changes to chondrocyte morphology, most of the cytoplasmic processes were associated with the presence of IA at the starting point of the protrusion.

These results demonstrate marked changes to F-actin distribution in chondrocytes in the very early stages of cartilage degeneration as occurs in OA. These subtle changes are probably an early indication of a change to the chondrocyte phenotype and thus worthy of further study as they may lead to deleterious alterations to matrix metabolism and ultimately cartilage weakening.

Articular cartilage has poor repair potential and the tissue formed is mechanically incompetent. Mesenchymal stromal cells (MSCs) show chondrogenic properties and the ability to re-grow cartilage, however a viable human model for testing cartilage regeneration and repair is lacking. Here, we describe an ex vivo pre-clinical femoral head model for studying human cartilage repair using MSCs.

Human femoral heads (FHs) were obtained following femoral neck fracture with ethical permission/patient consent and full-depth cartilage wells made using a 3mm biopsy punch. Pancreas-derived mesenchymal stromal cells (P-MSC) were prepared in culture media at ~5000 cells/20µl and added to each well and leakage prevented with fibrin sealant. After 24hrs, the sealant was removed and medium replaced with StemPro^TM chondrogenesis differentiation medium. The FHs were incubated (37^oC;5% CO₂) for 3wks, followed by a further 3wks in standard medium with 10% human serum with regular medium changes throughout. Compared to wells with medium only, A-MSCs produced a thin film across the wells which was excised en-block, fixed with 4% paraformaldehyde and frozen for cryo-sectioning.

The cell/tissue films varied in thickness ranging over 20-440µm (82±21µm; mean±SEM; N=3 FHs). The thickness of MSC films abutting the cartilage wells was variable but generally greater (15-1880µm) than across the wells, suggesting an attachment to native articular cartilage. Staining of the films using safranin O (for glycosaminoglycans; quantified using ImageJ) was variable (3±8%; mean±SEM; N=3) but in one experiment reached 20% of the adjacent cartilage. A preliminary assessment of the repair tissue gave an O'Driscoll score of 10/24 (24 is best).

These preliminary results suggest the ex vivo femoral head model has promise for studying the capacity of MSCs to repair cartilage directly in human tissue, although optimising MSCs to produce hyaline-like tissue is essential.

Supported by the CSO (TCS/17/32).

Abstract

Abstract

Abstract

Introduction

Subacromial corticosteroid injection is widely used in the treatment of Subacromial Impingement Syndrome (SIS). There is increasing interest in using ultrasound (US) to improve the accurate placement of injections. This study investigated whether the accuracy of placement of US-guided subacromial corticosteroid injections influences patients' outcome of pain and function.

Chondrocytes are essential to the maintenance of articular cartilage and it is thought that chondrocyte death occurs early in septic arthritis. Understanding the causes of chondrocyte death will allow the development of chondroprotective strategies to improve long-term outcomes following septic arthritis.

We utilised a murine model of septic arthritis using intra-articular injection of 10µL of a 10⁷ concentration of S. aureus suspended in PBS. Seventy-five adult male C57/Bl6 mice were randomised to receive injection of either S. aurues 8325-4 (a wild-type of S. aurues capable of alpha toxin production), DU1090 (an isogenic mutant of 8325-4 that is identical to 8325-4 other than being incapable of producing alpha toxin) or a PBS control. Establishment of septic arthritis was confirmed through gait changes (5 mice/group), limb swelling and histological changes (10 mice/group). 10 animals from each group were sacrificed at 48 hours and the injected knee joints were dissected before being stained with CFMDA (labelling live chondrocytes green) and PI (labelling dead chondrocytes red). The samples were imaged using a confocal laser scanning microscope and the percentage of chondrocyte death was calculated.

Mice injected with S. aureus 8325-4 or DU1090 developed septic arthritis with evidence of weight loss, limb swelling and gait changes whereas these were absent in the control group. There was a significantly higher level of chondrocyte death in the group infected with 8325-4 (2.7% chondrocyte viability) when compared to DU1090 (73.9% chondrocyte viability) and PBS injected mice (95% chondrocyte viability). One-Way ANOVA revealed that the difference between each group was statistically different (p < 0.05).

Alpha toxin is the major damaging toxin in S. aurues septic arthritis. Any adverse effect of the immune system is negligible in comparison. Development of treatments counteracting the effect of alpha toxin is required.

Staphylococcus aureus is a highly virulent pathogen and implicated in approximately 50% of cases of septic arthritis. Studies investigating other S. aureus-related infections suggest that alpha-(Hla), beta-(Hlb) and gamma-(Hlg) toxins are key virulence factors, with the ‘pore-forming’ alpha-toxin considered the most potent. Here, we have assessed the influence of alpha-toxin alone on in situ chondrocyte viability. Osteochondral explants were harvested from the metacarpophalangeal joints of 3-year-old cows and cultured in Dulbecco's Modified Eagle's Medium. The flasks were then inoculated with isogenic ‘knockout’ strains of S. aureus: DU5946 (Hla+Hlb-Hlg-: alpha-toxin only strain) or DU1090 (Hla-Hlb+Hlg+: beta- and gamma-toxin only strain). Explants were incubated (37°C) and stained after 18, 24 and 40hrs with chloromethylfluorescein-di-acetate and propidium iodide, labelling living chondrocytes green and dead cells red, respectively. Axial sections were imaged by confocal microscopy and the percentage cell death determined. Alpha-toxin-producing S. aureus caused 24.8+/−3.7% chondrocyte death at 18hrs and 44.6+/−7.2% death at 24hrs. At 40hrs, there was significantly more chondrocyte death (87.4+/−3.6%;p<0.001) compared to the alpha-toxin knockout strain, which was negligible (4.1+/−1.7%; means+/−SEM; N=4 independent experiments). In this in vitro bovine cartilage explant model, whereby the effects of defined toxins were determined in isolation of a complex host immune response, in situ chondrocyte viability was dramatically and exclusively reduced by alpha-toxin. This work forms the basis for developing a rational treatment to reduce the extent of cartilage destruction during an episode of septic arthritis. IDMS was supported by Orthopaedic Research UK and The Royal College of Surgeons of Edinburgh.

We have demonstrated that toxins produced by Staphylococcus aureus, a common infective agent in septic arthritis (SA), cause rapid in situ chondrocyte death. Here, we have compared the sensitivity of chondrocytes within the superficial and deep zones (SZ, DZ) of cartilage to the same toxins. Culture medium containing the toxins produced by S. aureus strain 8325-4, which include alpha-, beta-, and gamma-toxin, was prepared. Cartilage explants free of subchondral bone were taken from the metacarpophalangeal joints of 3-year-old cows, and incubated (37°C) with the toxins. Explants were stained after 6hrs with chloromethylfluorescein-di-acetate and propidium iodide, labelling living chondrocytes green and dead cells red, respectively. Full-thickness coronal sections were imaged by confocal microscopy and the percentage cell death within the SZ (100μm from articular surface) and DZ (100μm from subchondral bone interface) determined. Both zones were incubated with the same toxin culture medium for the same time period. At 0hrs, chondrocytes within all zones were >98% viable. However, after incubation with toxin-containing culture medium for 6hrs, 71.9+/−11.2% of the SZ cells were dead compared to only 47.4+/−6.7% in the DZ (p=0.03;data are means+/−SEM;N=4). These results suggest that SZ chondrocytes are considerably more sensitive to S. aureus toxins than those within deeper zones. As SZ chondrocytes are close to the synovial fluid harbouring bacterial toxins, these data emphasise the need to remove bacteria and their products aggressively as part of the treatment of SA. IDMS was supported by Orthopaedic Research UK and The Royal College of Surgeons of Edinburgh.

The internal fixation of osteochondral fragments in fractures normally utilizes intra-articular screws inserted through a pilot hole drilled into cartilage/bone. This trauma causes cartilage injury leading to chondrocyte death. We have quantified the cell death following cartilage drilling and identified irrigation conditions that can protect chondrocytes. Articular cartilage of bovine metacarpophalangeal joints of 3yr-old cows was irrigated in the presence/absence of saline of various compositions. Holes were then made using a standard 1.5mm drill (Ortho Solutions Ltd.) at 18,000 rpm through the articular cartilage into bone. Osteochondral explants were then harvested and cultured in Dulbecco's Modified Eagle's Medium containing chloromethylfluorescein-di-acetate and propidium iodide (10uM each), to label living chondrocytes green and dead cells red, respectively. Axial images were taken by confocal microscopy and the width of the zone of cell death (ZCD) around the hole determined. With no irrigation, new drills caused a ZCD of 171±25um, which was increased when drills used 50+ times were tested (279±31um;p=0.03). With saline irrigation, the ZCD was reduced for old drills (150±6um;p=0.016) but not for new drills (124±8um) suggesting the heating effect of the old drills caused additional chondrocyte death. However for new drills, the ZCD was further reduced significantly to 82±7um when the osmolarity of the saline irrigation solution was raised to 480mOsm using sucrose. Data are mean±s.e.m., from at least 5 separate experiments each with a minimum of 3 replicates. The results demonstrate a chondroprotective effect of raising the osmolarity of saline used during drilling of cartilage which could be clinically beneficial.

The cartilage diseases such as osteoarthritis and chondral injuries are considered irreversible and the result of recent treatments remains not optimal. One of the reasons is due to the poor understanding of chondrocyte behaviours. To understand more about cartilage, we designed a series of novel experiments. First, a total joint of bovine metatarsophalanges was isolated as our novel model. We chose it because the configuration and the healing potential were similar to human, and many variables of large animal studies could be controlled in laboratory. The model not only provided a good ex vivo platform for cartilage researches but also connected in vitro cellular studies and in vivo animal studies. To mimic joint movement a special driving machine was designed. To characterise the novel model viabilities of chondrocytes and contents of sulphated glycosaminoglycan (GAGs) in extracellular matrixes were measured every seven days. The preliminary results revealed the viabilities of chondrocytes remained above 80% alive in the middle zone after four-weeks culture. The GAGs contents decreased after this culturing period. The experiments still carry on going to compare the static and dynamic models which joint movement could be a determinative factor to the viability of chondrocytes. Cellular treatment is the recent mainstream for cartilage diseases. If advanced knowledge in chondrocyte behaviours could be obtained from this model, development of optimal treatment will be possible in the future.

In macroscopically-normal and early degenerate human articular cartilage, chondrocytes often exhibit increased volume and abnormal morphology with cytoplasmic processes. With further degeneration, chondrocyte clusters are a characteristic feature. These changes can influence matrix metabolism leading to matrix loss and predisposition to osteoarthritis (OA). Here, we report that articular chondrocytes cultured in a weak 3D agarose gel develop some of the morphological changes observed in degenerate cartilage. Cells were isolated from bovine metacarpal-phalangeal joints using collagenase. Gels were prepared with agarose (2% or 0.2% (v/v)) and cultured for 7 days (Dulbecco's modified Eagle's medium;37superscriptC;pH7.4;fetal calf serum (FCS; 1–10%)). Cells were fluorescently-labelled and volume/morphology examined by confocal microscopy. After one week of culture, chondrocytes in 2% gels (10% FCS) were mostly spheroidal; only 18.5±1% cells exhibited fine processes and 42.5±0.1% formed small clusters. However, in weak (0.2%) gels 66.9±1.3% (P=0.011) of chondrocytes had processes with lengths 7–63μm and 80.8±0.2% (P<.005) formed large clusters. In the weak gel after seven days, increasing FCS concentration markedly elevated the %age of chondrocytes in clusters from 31.1±0.2% in 1% FCS, to 87.3±0.27% in 10% FCS (P⊖.05). (Data from a minimum of three separate experiments at each condition with at least three replicates). These results suggest some similarities between the morphological changes to chondrocytes with OA development and those observed in weak agarose gels. The increased prevalence of abnormal chondrocytes with raised FCS concentration suggests that action of e.g. growth factors on chondrocytes is a more potent controller of cell shape than the strength of agarose.

Staphylococcus aureus is the most common bacterial isolate in septic arthritis. From studies on isolated cartilage cells, the ‘pore-forming’ alpha and gamma toxins are considered the most virulent factors. However, understanding the response of in situ chondrocytes is important in order to identify new treatments to reduce the extent of cartilage damage during, and following, episodes of septic arthritis. Animal models can give useful information; however the interpretation of data can be complex because of the strong immune response. Thus, to clarify the role of S. aureus toxins on in situ chondrocytes we have developed a bovine cartilage explant model.

Metacarpophalangeal joints, from 3-year-old cows, were opened under sterile conditions within 6hrs of slaughter and cartilage explants harvested. Explants were placed into flasks containing Dulbecco's Modified Eagle Medium (DMEM). Aspirates from a patient with septic arthritis of the hip, containing S. aureus, were compared to negative aspirates (no bacterial growth) from a patient with an inflamed knee joint (controls).

The explants were incubated at 37 degrees Celsius and stained after 18, 24 and 40hrs with the fluorescent probes chloromethylfluorescein di-acetate and propidium iodide (10 micromolar each) to label living chondrocytes green and dead cells red respectively. Following imaging of cartilage by confocal laser scanning microscopy, the percentage cell death at each time point was obtained using Volocity 4 software.

There was no detectable change in chondrocyte viability (<1% cell death) over 40hrs incubation with the negative aspirate. However, for the aspirate from a patient positive for S. aureus, there was a rapid increase in cell death between 18 and 24hrs (0.2 +/− 0.3% to 23 +/− 5% cell death respectively) and almost complete cell death at 40hrs (80 +/− 12%; data are means +/− s.d; n=4).

These results show that a strain of S. aureus capable of manifesting clinical disease exerts a potent effect on in situ chondrocytes. In the absence of an immune response, chondrocyte death was purely the result of the bacteria and their products. This bovine cartilage explant model could therefore be useful for studying the effects of S. aureus on chondrocyte behaviour and, ultimately, cartilage integrity.

Saline (0.9%) is typically used to rinse joints during osteo-articular surgery. It is not unusual for cartilage to then be exposed to the air of the operating theatre for 1-2hrs, which can lead to chondrocyte death. We have compared the survival of in situ chondrocytes within bovine cartilage which has been rinsed in various solutions or simply drained of synovial fluid (SF) and then allowed to dry, to identify approaches that could reduce chondrocyte death arising from cartilage drying.

Metacarpophalangeal joints from 3yr-old cows were opened under aseptic conditions. The joints were then (a) rinsed with saline (Baxter's Healthcare, Newbury), (b) rinsed with saline+glucose (20mM; both 300mOsm) or (c) drained of SF, and allowed to dry at room temperature. Full depth cartilage explants were taken after 2hrs, placed into Dulbecco's modified Eagle's medium and incubated with CMFDA (5-chloromethyl-fluorescein diacetate; 10microM) and propidium iodide (10microM) for the identification/quantification of living and dead cells respectively by confocal scanning laser microscopy and image analysis.

After 2hrs, the appearance and properties of the cartilage of the drying joints were clearly different. Saline-rinsed cartilage was dark purple and appeared dull with the cartilage difficult to sample. However when the rinsing solution was saline+glucose, or when joints were drained of SF, the cartilage was almost identical to the freshly-opened joint with a pearly-blue, shiny appearance, and cartilage sampling was easy.

Chondrocyte death was markedly increased in saline rinsed/dried joints after 2hrs (21±9% cell death). In contrast, there was no significant (P>0.05) death in saline+glucose rinsed/dried (2±1%) or SF-drained joints (3±2%;means±s.e.m.;n=5). The loss of cartilage wet weight over 2hrs (time=0 taken as 100%) was almost identical between cartilage rinsed in saline (73.6±1.6%), saline + glucose (78.6±1.1%) or SF (75.0±0.2%; data means±s.d.;n=2).

These results suggest that it was not the loss of water per se during cartilage drying that was the key determinant of chondrocyte viability. As chondrocytes are normally anaerobic, the rise in cartilage pO2 which occurs during exposure to air could have a deleterious effect on cell viability however the presence of glucose or SF protects through an anti-oxidant effect.

Intra-articular screw fixation is indicated for internal fixation of large osteochondral fragments secondary to trauma or osteochondritis dissecans. During surgery, orthopaedic drills are used to prepare a hole through which the screw can pass. Previous work has shown that mechanical injury to articular cartilage results in a zone of cell death adjacent to the traumatised articular cartilage (1). Here, we characterise and quantify the margin of in situ chondrocyte death surrounding drill holes and screws (standard cortical and headless compression designs) placed in mature bovine articular cartilage to model the orthopaedic procedure.

Drill holes (1mm) were made through the articular cartilage and bone of intact bovine metacarpophalangeal joints obtained from 3-yr old cows within 12hrs of slaughter. Osteochondral explants (∼1cm square and 2-3mm thick) encompassing the drilled holes in articular cartilage and subchondral bone were harvested using a chisel. Explants were then incubated in Dulbecco's modified Eagle's medium for 45mins with CMFDA (5-chloromethylfluorescein diacetate) and PI (propidium iodide; both at 10micromolar) to identify/quantify living and dead in situ chondrocytes respectively in a consecutive series of axial optical sections using confocal scanning laser microscopy (CLSM).

The drill holes through cartilage appeared to have clearly defined edges with no macroscopic evidence of cartilage splitting. However visualisation of fluorescently-labelled in situ chondrocytes by CLSM demonstrated clear cell death around the periphery of the drilled hole which was 166±19 micrometers in width. This increased with a larger diameter (1.5mm) drill to 450±151 micrometers (all data are means±s.e.m.; n=3). Preliminary experiments indicated that the margin of chondrocyte death around a 1.5mm hole was dramatically increased further by the insertion of screws into pre-drilled holes.

These results suggest that the mechanical trauma associated with cartilage drilling and the insertion of intra-articular screws occurs with marked death of in situ chondrocytes extending into normal cartilage beyond the area occupied by the screw. As chondrocytes are not replaced in mature cartilage, their loss around the hole/screw will mean that the extracellular matrix is not maintained, inevitably leading to cartilage failure.