If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address correspondence and reprint requests to: A.C. Hall, Centre for Integrative Physiology, School of Biomedical Sciences, University of Edinburgh, Hugh Robson Building, George Square, Edinburgh EH8 9XD, Scotland, United Kingdom. Tel: 44-(0)131-650-3263; Fax: 44-(0)131-650-2872.
Articular cartilage may experience iatrogenic injury during routine orthopaedic/arthroscopic procedures. This could cause chondrocyte death, leading to cartilage degeneration and posttraumatic osteoarthritis. In an in vitro cartilage injury model, chondrocyte death was reduced by increasing the osmolarity of normal saline (NS), the most commonly-used irrigation solution. Here, we studied the effect of hyperosmolar saline (HS) on chondrocyte viability and cartilage repair in an in vivo injury model.
Design
Cartilage injury was induced by a single scalpel cut along the patellar groove of 8 week old rats in the absence of irrigation or with either NS (300 mOsm) or HS (600 mOsm). The percentage of cell death (PCD) within the injured area was assessed using confocal microscopy. Repair from injury was evaluated by histology/immunostaining, and inflammatory response by histology, cytokine array analysis and ELISA (enzyme-linked immunosorbent assay).
Results
The PCD in saline-irrigated joints was increased compared to non-irrigated (NI) joints [PCD = 20.8% (95%CI; 14.5, 27.1); PCD = 9.14% (95%CI; 6.3, 11.9); P = 0.0017]. However, hyperosmotic saline reduced chondrocyte death compared to NS (PCD = 10.4% (95%CI; 8.5, 12.3) P = 0.0024). Repair score, type II collagen and aggrecan levels, and injury width, were significantly improved with hyperosmotic compared to NS. Mild synovitis and similar changes in serum cytokine profile occurred in all operated joints irrespective of experimental group.
Conclusions
Hyperosmotic saline significantly reduced the chondrocyte death associated with scalpel-induced injury and enhanced cartilage repair. This irrigation solution might be useful as a simple chondroprotective strategy and may also reduce unintentional cartilage injury during articular reconstructive surgery and promote integrative cartilage repair, thereby reducing the risk of posttraumatic osteoarthritis.
Chondrocytes, the cells of articular cartilage, are exclusively responsible for the turnover of the extracellular matrix (ECM) and therefore their survival is crucial for maintaining the biological and biomechanical functions of the tissue throughout life
. During arthroscopic/orthopaedic procedures, articular cartilage may be subjected to mechanical injury either by accident or design. Cartilage damage has been described as, unquestionably the most common iatrogenic lesion in arthroscopic surgery
. Iatrogenic injury can occur when cartilage is probed, trimmed, drilled or cut by surgical instruments and this may lead to articular cartilage damage and chondrocyte death increasing the risk of posttraumatic osteoarthritis
. In addition, chondrocyte death may hinder the integration between the native and repair tissue in reconstructive surgery and therefore limit the full functional and structural restoration of the joint
. It is therefore vital to maintain chondrocyte viability to limit damage and promote integration and cartilage healing following injury.
Although cartilage repair has been reported in animal models of cartilage injury, the regenerated tissue usually has inferior biomechanical properties and less durability than normal hyaline cartilage
. Improved repair of a full thickness joint surface defect in a mouse model was associated with reduced chondrocyte death while persistence of cell death was associated with cartilage degeneration
and compounds (e.g., caspase inhibitors, antioxidants) have been tested to inhibit chondrocyte death and thus the extent/severity of the cartilage damage after injury
During arthroscopic/articular surgery, the joint is normally irrigated with an artificial solution to improve visibility and provide a clear surgical field. Various fluids are currently used in clinical practice however there are concerns about the potential deleterious effect of these solutions on cartilage. Previous studies showed ultra-structural changes in the articular cartilage
. Moreover, in situ chondrocyte death following injury to human articular cartilage was significantly decreased after exposure to hyperosmotic saline (600 mOsm) compared to normal saline (NS) (300 mOsm)
Here we describe a reproducible in vivo animal model of scalpel-induced cartilage injury. Using this model, we demonstrate a chondroprotective effect of the hyperosmolar saline (HS) on chondrocyte viability and an improved repair outcome of articular cartilage following injury. We also studied the inflammatory response to the HS solution to assess its safe use.
Methods
Operative procedure
Eight-week-old male Sprague Dawley rats were anaesthetised using 3% isoflurane. After medial para-patellar arthrotomy, the patella was dislocated laterally to expose the patellar groove
. A partial thickness cartilage defect was then induced along the groove by a single gentle pass of a fresh No.11 scalpel blade in the absence of irrigation (no irrigation NI) or in the presence of normal (0.9%) saline (300 mOsm, (NS) Baxter Healthcare Ltd., UK) or hyperosmolar saline (HS) (600 mOsm, sucrose addition to saline)
(Fig. 1) with osmolarity measured by osmometer (Vitech Scientific Ltd., UK). Joints were lavaged for 5 min before and 5 min after the induction of the cartilage injury to allow chondrocytes to respond to the altered osmotic environment
. The patella was then relocated and the wound sutured in layers with coated vicryl 6-0 (polyglactin 910, Ethicon, UK). Sham operation was performed on the contralateral joint. Rats were given the analgesic buprenorphine (0.01 mg/kg) subcutaneously and allowed unrestricted activities in standard cages. All procedures were approved by the Local Ethics committee and UK Home Office.
Fig. 1Articular cartilage injury model in the rat knee joint. (A) H&E (left) toluidine blue (middle) and CMFDA with PI (right) staining of sham (lower panel) and injured (upper panel) rat articular cartilage. Arrows indicate the injury site. (B and C) Measurements of cartilage thickness, depth, width of the defect (B) and the percentage of injury depth to cartilage thickness under different irrigation strategies (C). Data are expressed as mean ± 95% CI; N = 10. Cartilage injury was induced by a single pass of a number 11 scalpel blade (see text for details) in the absence of irrigation solution (no irrigation; NI) or in the presence of normal saline (NS, 300 mOsm) or hyperosmolar saline (HS, 600 mOsm). Scale bar in (A) = 50 μm.
For each group, ten rats were killed immediately after surgery (day 0) and five rats were killed at 1 day, 1, 2 and 8 weeks after surgery and serum obtained by cardiac puncture. Knee joints were dissected and the viability assay for in situ chondrocytes performed by incubating joints with 5-chloromethyl-fluorescein diacetate (CMFDA) and propidium iodide (PI) (1 h; both 10 μmol/L, Invitrogen, UK) to label live/dead cells respectively
. After confocal imaging, joints were decalcified and paraffin-embedded for histology and immunohistochemistry (Fig. 1).
Cartilage imaging
Consecutive axial optical sections of the fluorescently-labelled chondrocytes were acquired using confocal laser scanning microscopy (CLSM; Carl Zeiss Ltd., UK) at 10 μm intervals and combined to create a three dimensional image. For quantifying in situ chondrocyte death after injury, a region of interest (ROI) extending 200 μm on each side of the scalpel injury (x-axis) × 921 μm (y-axis) × 40 μm (z-axis) was created within the image using software [Volocity 4·0, UK; Fig. 2(A)]. Live and dead cells were identified by %voxel intensity and the percentage cell death (PCD = 100 × number of dead cells/number of dead and live cells) calculated in the ROI. Injury width was measured by LSM image software (Carl Zeiss Ltd, UK) at 50 μm interval in the y-axis using the CLSM images.
Fig. 2In situ chondrocyte death after scalpel-induced injury. (A) A ROI was created in the axial CLSM image and live cells (marked in purple) and dead cells (marked in yellow) were identified and counted based on voxel intensity. (B) Axial CLSM projections of CMFDA and PI labelled injured chondrocytes within articular cartilage that was exposed to no irrigation (NI), normal saline (NS) or hyperosmolar saline (HS) at the indicated time points. (C and D) The PCD at day 0 (C) and at later time points (D) in injured and sham-operated joints exposed to different irrigation conditions. Data are expressed as mean ± 95% CI. P values are given in text. Scale bar in (B) = 100 μm.
Joints were sectioned at 5 μm intervals and stained with haematoxylin and eosin (H&E) and toluidine blue according to standard protocols. Cartilage thickness and injury depth were assessed by ImageJ software at day 0 using three non-consecutive sections (200 μm apart) and the percentage of injury depth to cartilage thickness calculated. The repair outcome was assessed using polarised light microscopy and the Wakitani scoring system
. Control non-operated, sham-operated and injured joints were assessed at the synovial insertion of medial femur, medial tibia, lateral femur and lateral tibia and the score of all four regions summed for a total joint synovitis score.
For immunostaining, sections were deparaffinised, digested with either pepsin (0.25 mg/ml, Sigma–Aldrich, UK) for type I and II collagen or chondroitinase ABC (0.25 unit/ml, Sigma–Aldrich, UK) for aggrecan, blocked in serum-free protein solution (Dako, UK) and incubated overnight with either anti-collagen I antibody (Abcam, UK), mouse anti-rat CD43 (AbD, Serotec, UK), anti-collagen II antibody (CIIC1) or anti-aggrecan antibody (12/21/1-C-6; both from the Developmental Studies Hybridoma Bank, Iowa). Sections were incubated with EnVision dual link system-HRP, and DAB substrate chromogen system used as peroxidase substrate (DakoCytomation, UK). Areas of immunostaining were measured using ImageJ software.
Cytokine array and ELISA
Relative changes in cytokine levels in serum samples from rats sacrificed 1 day and 1 week after surgery were screened and compared to control un-operated animals using the rat proteomic profiler array kit (R&D Systems, UK). Samples were pooled for each group, equal amounts of serum loaded on the blots in duplicate, and the average pixel density calculated using ImageJ. IL-1α levels were measured in serum samples from control and experimental rats using ELISA kits (R&D Systems, UK). The standard dilution series was extended to detect protein levels up to 15.6 pg/ml.
Statistical analysis
Data were analysed using SPSS v.21 (IBM, UK) and presented as means (95% CI lower limit, upper limit) where N indicated the number of animals in each group. Student's unpaired t-tests were used to compare between groups and Mann–Whitney U tests to compare histological scores. Differences were considered statistically significant at P < 0.05 and are indicated by asterisks with the actual P values given in text.
Results
Development of a reproducible in vivo partial thickness cartilage injury model
Articular cartilage injury was induced by a single pass of a scalpel blade along the patellar groove of the rat knee joint. To ensure consistency of the scalpel-induced injury, histomorphometric parameters i.e., cartilage thickness, depth and width of the cut were assessed. A partial thickness cartilage defect extending to 57.5% (95%CI, 54.3, 60.6) of cartilage thickness was induced in injured joints. The depth/width of the defect were not significantly different between the groups (Fig. 1).
Time course analysis of cell death after scalpel-induced injury
Chondrocyte death is a well-known consequence of cartilage injury
. CLSM images showed a band of cell death at the injury at day 0 [Fig. 2(A) and (B)]. A significant increase in the PCD of injured cartilage was observed within the ROI [Fig. 2(A)] in joints exposed to NS 20.8% (95%CI 14.5, 27.1) compared to non-irrigated (NI) joints 9.14% (95%CI 6.3, 11.9) or joints exposed to HS 10.4% (95%CI 8.5, 12.3 (P = 0.0017 and 0.0024 respectively; N = 10) [Fig. 2(B) and (C)].
Chondrocyte death decreased markedly at later time points following the different conditions studied however throughout, although it remained significantly higher in joints irrigated with NS compared to NI and HS irrigated joints [P = 0.0024 (D0), P = 0.04 (D1 & W1), P = 0.02, (W4 & W8) N = 5; Fig. 2(D)]. The PCD in the articular cartilage was negligible in sham-operated joints under all conditions [Fig. 2(D)].
The effect of irrigation solutions on the repair outcome following cartilage injury
One week after injury, the defects generated under the three different irrigation conditions were filled with lightly toluidine blue-stained matrix and spindle-shape cells, however an area of hypo-cellularity was noticed around the injury in joints irrigated with NS [Fig. 3(A)]. After 2 weeks, chondrocyte-like cells and darker stained matrix were observed in NI and HS-irrigated joints [Fig. 3(A)]. By 8 weeks, the defects in these two groups were almost completely filled with well-developed hyaline-like cartilage [Fig. 3(A) and (B)] except the superficial layer of the cut that exhibited a minor reduction in toluidine blue staining and hypo-cellularity [Fig. 3(A)]. In the NS-irrigated joints, the hypo-cellular area around the cut persisted for up to 8 weeks after surgery and the defects were filled with less developed repair tissue with decreased toluidine blue staining, few spindle-shaped cells and less organised collagen [Fig. 3(A) & (B)]. Wakitani repair scores were significantly lower (implicating better repair) in NI and HS-irrigated joints compared to joints irrigated with NS [P = 0.0047 and 0.002 respectively; N = 5; Fig. 3(C)]. No signs of cartilage degeneration in terms of cellular disorganization, surface fibrillation or loss of metachromasia were observed in the articular cartilage adjacent to, or distant from, the injury in all experimental groups [Fig. 3(A)]. No pathological changes were observed in bone, muscles, ligaments or menisci following exposure to different irrigation protocols and for up to 8 weeks post-surgery (data not shown).
Fig. 3Histological assessment of the changes to articular cartilage following scalpel-induced injury. (A) H&E (left) and toluidine blue (right) staining of sections from injured and sham-operated rat knee joints at the indicated time points. Arrows indicate areas of hypo-cellularity. (B) Polarized light microscopy of sections from sham and injured joints 8 weeks after injury. Arrows identify the site of injury and asterisks indicate the area of poor collagen organization. (C) Wakitani repair score at the 8 week time point. Significantly improved repair score was observed in non-irrigated joints (NI) and joints irrigated with hyperosmolar saline (HS) as compared to joints irrigated with normal saline (NS). Data are expressed as mean ± 95% CI; N = 5, P values are given in text. Scale bar in (A) and (B) = 50 μm.
Immunohistochemical staining for type II collagen and aggrecan showed strong staining in the injury site in NI and HS-irrigated joints 8 weeks post-injury indicating the presence of type II collagen and aggrecan in the reparative tissue [Fig. 4(A) and (B)]. However, cartilage injured while irrigated with NS showed pale staining for both type II collagen and aggrecan suggesting less developed cartilaginous repair tissue [Fig. 4(A) and (B)]. The area of reduced staining for type II collagen and aggrecan was significantly higher in the NS group compared to the NI and the HS groups [P = 0.0001 for all comparisons shown; N = 5; Fig. 4(C) and (D)]. Articular cartilage sections from all experimental groups exhibited no staining for type I collagen (data not shown).
Fig. 4Collagen and aggrecan immunoreactivity. Immunostaining for type II collagen (A) and aggrecan (B) in sham and injured articular cartilage 8 weeks after surgery. Injured articular cartilage was exposed to no irrigation (NI), normal saline (NS) or hyperosmolar saline (HS) before and after injury. Arrows identify the site of injury and asterisks indicate the area of reduced staining. (C and D) The area of reduced type II collagen (C) and aggrecan (D) in injured cartilage at 8 week time point. Data are expressed as mean ± 95% CI. P values are given in text. Scale bar in (A) and (B) = 50 μm.
A significant reduction in injury width was observed in the CLSM images of NI and HS irrigated joints but not in the NS irrigated joints at 8 weeks compared to the injury width immediately after surgery (P = 0.03 and 0.009 respectively; Fig. 5(A) and (B)). A significant reduction was also noted in the width of the defect of joints exposed to HS compared to NS 8 weeks after surgery (P = 0.013) [Fig. 5(B)].
Fig. 5Quantitative measurements of the width of the injury in the axial view. (A) CLSM projections of CMFDA and PI labelled injured articular cartilage that was exposed to no irrigation (NI), normal saline (NS) or hyperosmolar saline (HS) at Day 0 and 8 weeks after injury. (B) The width of the defect at day 0 and week 8. Data are expressed as mean ± 95% CI. P values are given in text. Scale bar in (A) = 50 μm.
revealed no significant difference between irrigation conditions on synovial membrane pathology following surgery at any time point [Fig. 6(A) and (B)]. Synovitis progressed similarly in all operated joints (sham and injured) for up to 8 weeks post-surgery. Synovitis marked by an increase in cellular density and thickening of the synovial lining was mainly observed at the 1 week time point and then progressively decreased and nearly resolved by 8 weeks after surgery [Fig. 6(A) and (B)]. CD43 labelled cells were detected in the synovial membrane under all three conditions at day 1 but there were very few at the 1W time point and none at 8W. There was no difference between the three conditions at any time point (data not shown).
Fig. 6Assessment of the inflammatory response following different irrigation conditions. (A) Representative H&E stained histological sections of the synovial membrane from injured knee joints exposed to no irrigation (NI), normal saline (NS) or hyperosmolar saline (HS) at 1 day, 1 week and 8 weeks after surgery. (B) Synovitis score in different irrigation conditions at 1 day, 1, 2 and 8 weeks post-surgery. (C) Cytokine array of serum samples obtained from control un-operated and experimental rats 1 day after injury. (D) IL-1α ELISA of serum samples from control and experimental rats at 1 day, 1, 2 and 8 weeks following cartilage injury. Data are expressed as mean ± 95% CI. N = 5. Scale bar in (A) = 200 μm.
To investigate whether inducing cartilage injury under different irrigation conditions could affect the production of inflammatory cytokines, a cytokine array was performed on serum samples from un-operated controls and experimental animals 1 day and 1 week after surgery. Interleukin 1 alpha (IL-1α), monokine induced by gamma interferon (MIG/CXCL9), Macrophage inflammatory protein-1α (MIP-1α/CCL3), Macrophage Inflammatory Protein-3 (MIP3A/CCL20, thymus chemokine (CXCL7) and vascular endothelial growth factor (VEGF) were higher in all injured animals irrespective of the irrigation condition compared to control animals [Fig. 6(C)]. Serum levels of IL-1α significantly increased 1 day following joint injury in all experimental groups. Peak concentrations of serum IL-1α in all operated animals occurred at 1 week post-injury. The elevation of serum concentrations of IL-1α was gradually reduced over subsequent time points to nearly normal levels 8 weeks post-operative. No significant differences were found between experimental animals in different irrigation groups at any of the time points [Fig. 6(D)].
Discussion
Articular surgery is routinely performed using irrigation solutions with an osmolarity lower than that of synovial fluid, potentially rendering chondrocytes more sensitive to mechanical injury
to an in vivo animal model and provides evidence that a hyperosmolar saline for joint irrigation during surgery gives chondroprotection against mechanical insult and promotes cartilage repair with no deleterious effect on joint tissues.
With the in vivo model developed here, injury was induced by a single pass of a scalpel under its own weight over the articular cartilage. This created a reproducible partial thickness cartilage defect and the induction of a reproducible zone of chondrocyte death around the cut edge (Fig. 1). This permitted the study of the effect of increasing osmolarity on cell viability limiting any variation in mechanical injury that could potentially affect the extent of chondrocyte death.
Exposing articular cartilage to NS during injury significantly increased in situ chondrocyte death following acute mechanical injury compared to NI joints (Fig. 2). This latter protective effect was probably due to the presence of a thin film of synovial fluid covering the joint surface. During arthroscopy, this film is removed by the irrigating solution and during open orthopaedic surgery e.g., the reconstruction of the joint surface of intra-articular fractures, it will be washed away by the fluid used to prevent cartilage drying.
Raising the osmolarity of the irrigation solution significantly reduced the extent of cell death to levels comparable to NI joints (Fig. 2). Thus, even temporarily disturbing the osmotic balance of the chondrocytes using NS, may pre-dispose chondrocytes to greater cell death during scalpel injury whereas raising the osmolarity provided significant chondroprotection. The precise mechanisms by which medium osmolarity regulates cell death/survival following injury are still to be elucidated. Osmotic challenges modulate cell volume through ion transport and cytoskeleton actin restructuring
The elevated cell death in joints exposed to NS during cartilage injury was associated with less developed repair tissue that exhibited hypo-cellularity and low immunoreactivity to type II collagen and aggrecan [Fig. 2, Fig. 3]. Differences in cell death patterns have been observed between animals with normal and impaired tissue healing as cell death was markedly increased, and persistent in animals with healing deficiency
. It is therefore likely that extensive and prolonged cell death could negatively influence cartilage repair by reducing the viable cell population around the injury.
The partial thickness cartilage defect generated here showed spontaneous healing with more developed repair tissue in non- and HS irrigated joints compared to NS-irrigated joints [Fig. 3, Fig. 4]. Type I collagen was not detected, suggesting negligible fibrocartilage was produced during repair
Immunohistochemical study of collagen types I and II and procollagen IIA in human cartilage repair tissue following autologous chondrocyte implantation.
. In contrast, aggrecan and type II collagen were observed in the repair tissue of all injured cartilage with decreased intensity of immunoreactivity in NS irrigated joints (Fig. 4). This, taken with the reduction of injury width with time (Fig. 5), suggests that articular cartilage may possess limited intrinsic repair potential which can be promoted by minimizing chondrocyte death following mechanical insult. In partial thickness cartilage defects, there is no access to the mesenchymal stem cells residing in the bone marrow spaces
and it is therefore unlikely that they have contributed to the repair of these lesions. However, progenitor cells have been identified in the synovial membrane
. Here, a single relatively small cartilage defect, was induced by a scalpel which allowed us to investigate the repair response. Although cartilage repair was more evident in NI and HS-irrigated joints (Fig. 5), tissue recovery was not complete as the defects were still detectable 8 weeks after injury with some hypo-cellularity and slightly lower toluidine blue staining and type II collagen immunoreactivity compared to surrounding cartilage (Fig. 3, Fig. 4).
Previous reports have demonstrated degenerative changes in injured cartilage
, however these were not observed here under any of the experimental conditions or time points analysed. However, it should be stressed that the pathogenic process of posttraumatic osteoarthritis development strongly depends on the severity of the mechanical trauma and associated tissue damage
. The scalpel-induced injury in this model caused minimal disruption of ECM and a limited zone of cell death that probably did not reach the critical threshold required to initiate cartilage degeneration.
Investigating the local intra-articular and systemic inflammatory response to hyperosmotic saline was essential as a first step to determine its safe use in clinical practice. Mild synovial cellular infiltration and alterations to the serum cytokine profile were observed in all the joints following surgery with no difference between experimental groups (Fig. 6), suggesting that the inflammatory reaction was induced solely by the surgical intervention. No pathological changes were observed in the surrounding muscles, bone, ligaments or menisci and therefore it is unlikely that raising saline osmolarity would have an adverse effect during surgery.
This study provides a safe translational strategy for reducing the extent of chondrocyte death associated with mechanical injury and promoting cartilage repair simply by increasing the osmolarity of the irrigation solution used during arthroscopic or open articular surgery.
Author contributions
Concept and design: NE, SH, HS, ACH.
Collection, analysis and interpretation of data: NE, ACH.
Drafting of manuscript: NE, ACH.
Critical revision of the manuscript for intellectual content; NE, SH, HS, AKA, ACH.
Final approval of the manuscript; NE, SH, HS, AKA, ACH.
NE: Noha Eltawil.
SH: Sarah Howie.
HS: Hamish Simpson.
AKA: Anish Amin.
ACH: Andrew Hall.
Role of the funding source
Funding for this study was provided by Arthritis Research (U.K.) grant number 19665. The funding body had no role in the study design, collection, analysis and interpretation of data, nor in the writing of the manuscript and in the decision to submit the manuscript for publication.
Competing interests
Details of these are included in the attached ICMJE forms submitted by each author.
Acknowledgements
We thank Dr Anisha Kubasik-Thayil of the IMPACT imaging facility for assistance with polarised light microscopy. This work was funded by Arthritis Research UK (Grant No. 19665). The CIIC1 and 12/21/1-C-6 antibodies were obtained from the Developmental studies Hybridoma Bank developed under the auspices of the NICHD and maintained by the University of Iowa, Department of Biology, Iowa City, IA 52242.
References
Buckwalter J.A.
Martin J.A.
Brown T.D.
Perspectives on chondrocyte mechanobiology and osteoarthritis.
Immunohistochemical study of collagen types I and II and procollagen IIA in human cartilage repair tissue following autologous chondrocyte implantation.
01374-0&id=gr1.jpg){kind=link}
01374-0&id=gr2.jpg){kind=link}
01374-0&id=gr3.jpg){kind=link}
01374-0&id=gr4.jpg){kind=link}
01374-0&id=gr5.jpg){kind=link}
01374-0&id=gr6.jpg){kind=link}