The nitric oxide donor glyceryl trinitrate increases subchondral bone sclerosis and cartilage degeneration following ovine meniscectomy

OsteoArthritis and Cartilage (2004) 12, 974e981 ª 2004 OsteoArthritis Research Society International. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.joca.2004.08.012

International Cartilage Repair Society

The nitric oxide donor glyceryl trinitrate increases subchondral bone sclerosis and cartilage degeneration following ovine meniscectomy1 Martin A. Cake Ph.D.y*, Richard A. Read Ph.D.y, Richard C. Appleyard Ph.D.z, Su-Yang Hwa M.D.x and Peter Ghosh Ph.D.k y School of Veterinary and Biomedical Sciences, Murdoch University, Perth, WA 6105, Australia z Orthopaedic Research Institute, St. George Hospital (University of New South Wales), NSW, Australia x Department of Orthopaedics, Tri-Services General Hospital, National Defence Medical Centre, Taipei, Taiwan k Institute of Bone and Joint Research, Royal North Shore Hospital (University of Sydney), NSW, Australia Summary Aim: To examine the effect of glyceryl trinitrate (GTN), a nitric oxide (NO) donor compound, on the concurrent progression of cartilage and subchondral bone changes in an ovine meniscectomy model of osteoarthritis (OA). Methods: Bilateral lateral meniscectomy (MX) was performed on 12 ewes to induce OA. Six were treated with topical GTN (0.7 mg/kg twice weekly) (MX C GTN). Six other sheep formed non-operated controls (NOC). After sacrifice at six months, the subchondral bone density (BMD) of the lateral and medial femoral condyles (LFC, MFC) and tibial plateau (LTP, MTP) was assessed by DEXA. Dynamic biomechanical testing was performed across the MTP and LTP. Histological sections from each region were scored qualitatively and the thickness of the subchondral bone plate (SCB) was determined by image analysis. Results: MX C GTN displayed significantly greater SCB thickness relative to MX in the LFC (mean increase C88% and C42%, respectively) and the MFC. SCB BMD was 10e12% greater in MX C GTN relative to MX in the LFC, LTP and MTP. MX C GTN sheep also showed greater increases in some histopathology variables, greater central erosion of the LTP, and changes in dynamic stiffness (decreased) and phase lag (increased) in the outer zone of the LTP. Conclusions: Treatment with GTN significantly increased subchondral bone thickness and density during subchondral remodelling following meniscectomy. In addition, it slightly but significantly worsened degeneration of cartilage structure and function. These results suggest that clinical use of GTN may accelerate both cartilage degeneration and subchondral bone sclerosis if used in the presence of OA, and demonstrate that NO has the potential be an important mediator of the subchondral bone changes seen in OA. ª 2004 OsteoArthritis Research Society International. Published by Elsevier Ltd. All rights reserved. Key words: Glyceryl trinitrate, Nitric oxide, Osteoarthritis, Subchondral bone.

osteoarthritic cartilage damage through altered metabolism of the underlying subchondral bone (SCB). In bone, constitutive local production of NO appears to support the growth and activity of both osteoblasts and osteoclasts8. However, higher levels of NO, such as those induced by cytokine stimulation, inhibit osteoclast function and differentiation of osteoclasts from precursor cell types, as well as inhibiting osteoblast maturation and activity8e10. Induced production of NO may represent a protective mechanism against cytokine-stimulated bone resorption in pathological conditions11. While the in vitro effects of NO in bone are somewhat variable11, its dominant role in vivo appears to be tonic inhibitory restraint of bone resorption. This is supported by the observation that in rats, the iNOS inhibitor aminoguanidine potentiates loss of bone mineral density12. Wimalawansa et al. have demonstrated that topical use of an NO donor (glyceryl trinitrate) alleviates both ovariectomy-induced and prednisolone-induced bone loss in rats13,14. A pilot trial in humans has also shown that glyceryl trinitrate effectively counteracts bone loss after ovariectomy in women15. Glyceryl trinitrate (GTN) is a lipophilic organic nitrate rapidly cleaved in vivo to inorganic nitrite, then nitric oxide.

Introduction Nitric oxide (NO) is a highly reactive, gaseous free radical that has been implicated in the pathogenesis of osteoarthritis (OA). Synovial fluid levels of nitric oxide are elevated in arthritic joints1, and osteoarthritic chondrocytes show elevated expression of the generative enzyme inducible nitric oxide synthase (iNOS)2. Nitric oxide is thought to induce a range of disturbances in chondrocyte metabolism, including suppression of proteoglycan (PG) synthesis3, increased matrix degradation4, and chondrocyte apoptosis5. Animal studies using iNOS knockout models and iNOS inhibiting agents have demonstrated a clear role for NO in the progression of arthritis6,7. In addition to a direct influence on chondrocytes, it is also possible that NO may influence the progression of 1

This study was supported by an Australian Research Council Small Grant. MAC was supported by the Murdoch University Veterinary Trust and Boehringer Ingelheim Vetmedica. *Address correspondence and reprint requests to: Dr Martin Cake. School of Veterinary and Biomedical Sciences, Murdoch University, Perth, WA 6105, Australia. Tel: 61-8-9360-2175; Fax: 61-8-9310-4144; E-mail: [email protected] Received 28 March 2004; revision accepted 26 August 2004.

974

975

Osteoarthritis and Cartilage Vol. 12, No. 12 As an efficient nitric oxide donor, it has been used for many years as a vasodilator and symptomatic treatment for angina pectoris, hypertension, and congestive heart failure. On the basis of the above studies, together with the observation that nitrate use is associated epidemiologically with increased hip and heel bone mineral density (BMD)16, it has been proposed that GTN may represent an effective treatment for osteoporosis and other forms of osteopenia17. However, given the known deleterious effects of NO on chondrocytes, and the protective effect of iNOS inhibitors in experimental arthritis7, it is reasonable to question the safety of nitrates in joint tissues. Our previous studies have shown that intermittent, topical administration of GTN to normal sheep induced significant structural and biomechanical alterations within femoro-tibial articular cartilage18. This is supported by recent epidemiological evidence of an association between nitrate use and hip OA in elderly women19. Furthermore, high BMD is itself a risk factor for OA20. Given also that subchondral bone sclerosis is a histological hallmark of OA, and is suspected by some authors to be central to its pathogenesis21, it is possible that nitrate use may increase OA risk by increasing subchondral bone mass19. The aim of this study was to examine the effect of GTN treatment on the concurrent progression of articular cartilage and subchondral bone changes in the well-characterised ovine meniscectomy model of OA22,23.

Method ANIMAL TREATMENTS

Eighteen aged (7 years old) Merino ewes, selected for uniformity of size and conformation, were purchased from a single source. Twelve sheep were subjected to bilateral lateral meniscectomy to induce osteoarthritis, as described previously22,23. Operated animals received cefazolin sodium, flunixin meglumine, and slow-release fentanyl patches in the peri-operative period. After a brief recovery period, all sheep were housed on irrigated pasture for the remainder of the trial period. Beginning 2 weeks after surgery, six of the meniscectomised animals were treated with 2% GTN ointment [Nitro-Bid
, Hoescht, Australia], applied twice weekly (0.7 mg GTN/kg body weight) to the hairless region on the medial aspect of the hindlimb. This dosage regime was selected to provide a similar weight-adjusted weekly dosage (though at lower frequency) to that used in the rat studies of Wimalawansa et al.13,14. The treatment groups (n Z 6) were therefore: non-operated control (NOC), meniscectomised (MX), and GTN-treated meniscectomised (MX C GTN). All animal procedures were approved by the Murdoch University Animal Ethics Committee (AEC R779/00). HISTOPATHOLOGICAL ASSESSMENT

Animals were euthanased after 6 months (24e26 weeks) of treatment, by intravenous injection of pentobarbitone. One hindleg (random left or right) was used for DEXA, biomechanical testing and tibial plateau histology, while the other was used for assessment of gross pathology (cartilage erosions and osteophyte development) and femoral condyle histology, as described previously18. Decalcified histological sections were prepared from midcondyle, medio-lateral osteochondral slices of the medial and lateral tibial plateau (MTP, LTP) and medial and lateral femoral condyles (MFC, LFC), and stained with toluidine blue and Masson’s trichrome23. Semi-quantitative histopathological grading was performed according to a published

modified Mankin’s scoring system24. In each joint region, four zones were scored: inner, middle, outer, and outer marginal zones. Scoring was done by a single observer, according to a six-point scale (structure, cellularity, chondrocyte cloning, territorial toluidine blue staining, interterritorial toluidine blue staining, calcified cartilage structure), and a mean aggregate score determined as the average of these four zones. In addition, calcified cartilage vascularity was measured as the number of blood vessels invading the length of the calcified cartilage layer, excluding regions of osteophytic remodeling25. Subchondral bone plate thickness was measured using computer-assisted image analysis as described previously18. Image analysis software [ImagePro Plus v3.0.1; Media Cybernetics, USA] containing an algorithm for radial thickness measurement was used to determine the mean distance (mm) between lines delimiting the most advanced tidemark and limit of subchondral bone (intertrabecular spaces !500 mm), after dividing the cartilage visually into three equal arc segments termed the outer, middle, and inner zones. In addition, the cross-sectional area (mm2) of osteophytic remodeling was determined by image analysis.

DYNAMIC BIOMECHANICAL TESTING

The biomechanical properties of tibial plateau articular cartilage were assessed using a hand-held indentation device26, as described previously27. Briefly, testing was completed at each of 18 locations according to a 3 ! 3 array marked on each condyle. At each location, the indentation probe was pressed against the surface with constant pressure whilst a flat cylindrical non-porous indenter (0.7 mm diameter) applied a sinusoidal oscillation (20 Hz) to the cartilage surface. Dynamic stiffness and phase lag were determined directly whilst dynamic shear modulus (G*), was calculated by adjusting for thickness using the theory of Hayes et al.28. On completion of biomechanical testing, articular cartilage thickness was determined at each location by needle penetration29. Results of biomechanical testing at individual locations are pooled for reporting (but not statistical analysis) into subsets corresponding to the equivalent outer (O), middle (M), and inner (I) zones of the lateral (L) or medial (M) plateau (n Z 6 zones: MO, MM, MI; LO, LM, LI), as described previously18. BONE DENSITOMETRY

Subchondral bone mineral density was determined in the LTP, MTP, LFC, and MFC using a Hologic QDR 4500-W absorptiometer [Hologic Inc., USA]30. A bed of rice of constant thickness (5 cm) was used both to position the specimen and to simulate soft-tissue density. Scanning of subchondral sites was conducted such that the weightbearing articular surface was parallel to the beam. Bone density (gm/cm2) was determined as the mean of three 5 ! 3 mm regions-of-interest (ROI), in the inner, middle and outer subchondral zones of the tibial plateau, and three 3 ! 3 mm ROI in the corresponding zones of the femoral condyles. ASSAY OF BONE MARKERS

Osteocalcin was assayed in serum collected prior to euthanasia, by a commercial pathology laboratory utilising competitive radioimmunoassay. Briefly, 125I-bovine

976 osteocalcin tracer, rabbit anti-bovine osteocalcin serum, and normal rabbit serum were combined with the test sample or bovine osteocalcin standard and incubated for 48 h at 4(C. Antibody complexes were precipitated by addition of an anti-rabbit gamma globulin precipitating sera and counted by routine gamma counting. Antisera raised against bovine osteocalcin have been shown to cross-react with both human and ovine osteocalcin31. Pyridinoline and deoxypyridinoline were assayed in urine collected at necropsy, using a published high performance liquid chromatography (HPLC) method32. Assayed levels were corrected for urine creatinine concentration, as determined by modified Jaffe reaction33. STATISTICAL ANALYSIS

All data presented express mean G standard deviation, except for graphical figures which present mean G standard error. Statistical comparisons were generated using specialist software [Statview 5.0, SAS Institute Inc., USA], using analysis of variance (ANOVA) to analyse variance across groups, and Fisher’s protected least significant difference (Fisher’s PLSD) test to compare means between groups. When multiple zones or testing locations were compared simultaneously to examine differences across an entire region, zone (inner, middle, or outer) or location was included as a second independent variable. A significance level of P Z 0.05 was used throughout.

Results BODY WEIGHT

Meniscectomised sheep lost an average of 5.6 kg body weight during the month following surgery. GTN treatment had no influence on body weight at any time point. No side effects or behavioural differences were observed in GTNtreated animals. CARTILAGE DATA

As previously described, meniscectomy was associated with gross cartilage erosion and prominent osteophyte development24. Gross changes were mostly limited to the meniscectomised lateral compartment. GTN treatment had no effect on subjective scores for cartilage erosion or osteophyte development (Table I). Aggregate histopathology scores (Table I) were increased in all regions in meniscectomised animals, especially the lateral compartment. Aggregate scores did not differ between treated and non-treated meniscectomised sheep. However, MX C GTN animals were found to have significantly greater scores relative to MX sheep for structural damage (1.96 vs 1.17, P Z 0.033; NOC 0.65) and loss of interterritorial staining (1.17 vs 0.58, P Z 0.003; NOC 0.25) in the MFC, and greater scores for chondrocyte cloning in the LFC (3.37 vs 2.67, P Z 0.015; NOC 0.33). After meniscectomy, scores for both calcified cartilage pathology and vessel number (Table I) were increased in all regions (P ! 0.0001 for both variables), especially the lateral compartment. MX C GTN sheep had a greater number of vessels invading calcified cartilage in the LFC, relative to MX animals (P Z 0.031). However, scores for both calcified cartilage pathology (1.25 vs 1.96, P Z 0.023; NOC 1.21) and vessel number (Table I, P Z 0.010) were lower than MX sheep in the MTP.

M. A. Cake et al.: NO donor accelerates OA bone sclerosis Table I Effect of meniscectomy (MX) and glyceryl trinitrate (GTN) on histopathological variables in the lateral and medial femoral condyles (LFC, MFC) and tibial plateau (LTP, MTP). NOC

MX

MX C GTN

Cartilage gross 2.67 G 2.42 11.33 G 1.97*** 11.50 G 2.68*** score (max. Z 16) Osteophyte gross 0.17 G 0.41 9.00 G 2.28*** 10.5 G 2.43*** score (max. Z 12) Mean aggregate histopathology score (max. Z 30) LFC 3.21 G 0.29 14.83 G 2.18*** 15.12 G 4.02*** MFC 2.55 G 0.60 6.96 G 1.86*** 8.00 G 1.70*** LTP 3.31 G 0.43 16.75 G 1.67*** 16.21 G 2.76*** MTP 2.33 G 1.72 8.87 G 2.47** 9.46 G 3.97*** Number of blood vessels penetrating calcified cartilage zone LFC 0.83 G 0.41 4.00 G 1.79 8.167 G 6.56**y MFC 1.17 G 1.60 2.83 G 1.33 2.33 G 2.42 LTP 1.33 G 1.21 4.50 G 2.07* 5.67 G 1.75** MTP 2.17 G 1.60 5.17 G 2.93* 2.00 G 1.67y Osteophyte cross-sectional area (mm2) LFC e 25.12 G 10.71 MFC e 2.00 G 1.90 LTP e 12.13 G 3.86 MTP e 2.25 G 1.62

40.40 G 14.28a 6.97 G 5.11y 13.28 G 6.56 1.91 G 2.02

The symbol * differs from NOC (*P ! 0.05, **P ! 0.005, ***P ! 0.0005); yMX C GTN differs from MX (P ! 0.05); a (P Z 0.06 vs MX)

After meniscectomy, cartilage thickness decreased in the inner regions of both tibial condyles and increased in the outer regions (Fig. 1). MX C GTN had significantly thinner cartilage in the eroded middle regions of the LTP relative to MX (P Z 0.0034). Dynamic shear modulus (G*) was significantly lower in all meniscectomised groups (P ! 0.0001), primarily due to changes in the LTP. G* was significantly lower in MX C GTN sheep relative to MX in the outer locations of the LTP (P Z 0.039). G* data obtained in the middle locations of the LTP were found to have a high degree of error due to deep cartilage erosion in some animals exposing much stiffer subchondral tissues. Phase lag values of MX C GTN sheep were increased overall relative to MX (12.38 G 3.26 vs 11.46 G 2.38; P Z 0.018), but were not significantly different from those of NOC sheep (12.08 G 2.30). However, phase lag values of MX C GTN sheep did differ from NOC in some zones, notably a significant increase in the outer LTP (P Z 0.032) and a significant decrease in the outer MTP (P Z 0.017). The latter change was also observed in MX animals. SUBCHONDRAL BONE DATA

After meniscectomy, the SCB (Fig. 2) of MX animals was significantly thicker in the outer zone of the LFC (P Z 0.035) and LTP (P Z 0.008), while thinning of the SCB was observed in the outer zone of the MFC (P Z 0.034). MX C GTN animals showed significantly greater increase in SCP thickness relative to MX in the middle zone of the LFC (P Z 0.024), and reduced loss of bone thickness in the outer zone of the MFC (P Z 0.049). Osteophyte size (Table I) was greatest in the femoral condyles. MX C GTN sheep showed greater osteophyte growth relative to MX ewes in the MFC (P Z 0.026), with a similar trend in the LFC (P Z 0.062). After meniscectomy, subchondral BMD (Fig. 3) was increased in the LFC (P Z 0.002) and LTP (P Z 0.018), and decreased in the MFC (P Z 0.0003) and MTP

977

Osteoarthritis and Cartilage Vol. 12, No. 12

Shear modulus

NOC

1.2

MX

G* (MPa)

1

MX+GTN

Discussion

0.8

* 0.6

† *

**

0.4

**

0.2 0

Phase (degrees)

†

† *

14

Phase lag

†

12

**

*

10 8 6 4 2 0

Thickness

1.8

Thickness (mm)

1.6

*

**

1.4

*

1.2 1

0.8 0.6

and deoxypyridinoline compared to MX animals, which mirrored the changes seen in serum osteocalcin.

* *

0.4

*

† *

*

† *

0.2 0 OUTER MIDDLE INNER LATERAL TIBIA

ALL

ALL

INNER MIDDLE OUTER MEDIAL TIBIA

Fig. 1. The effect of meniscectomy (MX) and glyceryl trinitrate (GTN) on articular cartilage shear modulus (G*, MPa G SE), phase lag (degrees G SE) and cartilage thickness (mm G SE) at testing locations in outer, middle, and inner zones of the tibial plateaux. Central columns represent the mean of all test locations in the lateral (left) and medial (right) tibial plateau. The symbol * differs significantly from NOC (P ! 0.05); yMX C GTN differs from MX (P ! 0.05).

(P ! 0.0001). MX C GTN sheep showed a greater increase in subchondral BMD in the LFC (P Z 0.019) and LTP (P Z 0.011), and reduced loss of BMD in the MTP (P Z 0.023), relative to MX. When compared by zone, these differences were most prominent in the middle zone of the LFC, and the outer zones of the LTP and MTP. BONE MARKERS

At the time of sacrifice, serum osteocalcin levels (Fig. 4) were significantly increased in MX sheep relative to NOC animals (P Z 0.020). Levels in MX C GTN animals were significantly lower relative to MX (P Z 0.004), and did not differ from those of NOC sheep. Urinary levels of pyridinoline and deoxypyridinoline (Fig. 4) did not differ significantly between treatment groups, though MX C GTN sheep showed a tendency towards lower levels of pyridinoline

In this trial, treatment with GTN was found to significantly affect the structural alterations seen in cartilage and bone following meniscectomy. While there was no grossly visible difference in cartilage pathology or osteochondral remodelling, other measures suggest that the cartilage of GTNtreated animals was degraded to a greater degree compared to untreated animals. MX C GTN animals showed thinner cartilage at central testing points in the LTP, suggesting greater severity of articular cartilage erosion. Though aggregate scores were not significantly different in MX C GTN sheep, histopathological scoring revealed greater structural degeneration and loss of PG in the MFC, and a greater degree of chondrocyte cloning in the LFC relative to MX animals. The cartilage of GTN-treated animals was also biomechanically inferior, with an overall increase in phase lag, and a greater reduction in dynamic shear modulus (G*) in the outer zones of the LTP, relative to MX sheep. These locations (LO) correspond to areas previously protected by the meniscus prior to surgery, and were found to have the greatest alterations following meniscectomy in both this study and in the topographical analyses of Appleyard et al.27. The dramatic loss of cartilage dynamic stiffness in OA is thought to be due to fibrillation and rupture of collagen of the collagen meshwork, and loss of PG-supplied collagen pre-stress34,35. Increased cartilage phase lag has also been reported in experimental models of OA, and is thought to reflect increased frictional energy dissipation due to loosening of collagenePG interactions, as well as proteolytic degradation of PGs27,36. Though phase lag was not altered overall in MX C GTN sheep relative to NOC, a significant increase was observed in the outer zones of the LTP. Together, both the structural and biomechanical results suggest that topical treatment with GTN worsened cartilage degeneration following lateral meniscectomy. However, the lack of a consistent increase in phase lag in the untreated MX group was unexpected, and may cast doubt on the significance of this particular variable. As has been shown previously in this model23,37, significant remodelling of subchondral bone was evident 6 months after the induction of osteoarthritis by meniscectomy. The subchondral bone plate was thickened in the LFC but thinner in the outer MFC of meniscectomised animals, reflecting the altered pattern of loading in the meniscusdeficient joint, whereby external femoral rotation and slight varus displacement of the stifle increases loading in the lateral compartment. Changes in the tibial plateau were concentrated in the outer zone of the LTP, the region normally protected by the meniscus in the intact joint. Changes in peri-articular bone mineral density showed a similar pattern, with increased BMD in the lateral compartment and relative osteopenia in the medial side of the joint. The observation that BMD changes were more prominent than altered SCP thickness is interesting, as the classic concept of subchondral bone changes in OA involves a sclerotic but hypomineralised matrix38. Karvonen et al. used a similar DEXA method to show a significant decrease in peri-articular BMD in patients with early knee OA39. Early loss of subchondral bone volume and density is also common in unilateral surgical animal models of OA, due to decreased loading of the joint40,41. The use of a bilateral

978

M. A. Cake et al.: NO donor accelerates OA bone sclerosis Outer

Middle

2.2

Subchondral bone plate thickness (mm)

2

1.8 1.6 1.4 1.2

Inner

Inner

Middle

Outer 2.2

*†

LFC

MFC

*

1.8 1.6 1.4 1.2

1

1

*

0.8 0.6 0.4 0.2

†

* *

0.8 0.6 0.4 0.2 0

0

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

2

LTP

MTP

2 1.8 1.6 1.4 1.2

*

1

*

0.8 0.6 0.4 0.2 0

Outer

Middle

Inner

Inner

Middle

Outer

Fig. 2. Subchondral bone plate thickness (mm G SE) by zone in the lateral and medial femoral condyles (LFC, MFC) and tibial plateau (LTP, MTP) of NOC (,), MX ( ), and MX C GTN sheep (-). The symbol * differs significantly from NOC (P ! 0.05); yMX C GTN differs from MX (P ! 0.05).

meniscectomy model in this study prevented unloading through shift of weight to the contralateral limb, though compensatory gait changes may still have reduced peak loads on the stifle joints. Joint loading was also affected by the slight (w11%) loss of body weight recorded following meniscectomy; this is likely to be the result of post-surgical stress and/or an effect of subclinical lameness on grazing behaviour23. Results in this surgical OA model show that it is possible (as in the LTP in this study) for a significant increase in subchondral BMD to occur, accompanied by only minor thickening of the subchondral plate. However, the concurrent loss of bone in the medial compartment of the

same joint suggests that primarily mechanical, rather than biochemical changes were responsible for these subchondral bone changes; this model may therefore not directly reflect the subchondral changes of naturally-occurring OA. Glyceryl trinitrate treatment significantly influenced the subchondral bone remodelling processes associated with meniscectomy. GTN-treated animals showed a greater increase in subchondral bone plate thickness in the middle (in-contact) zone of the LFC, and reduced loss of thickness in the outer zone of the MFC, relative to MX sheep (Fig. 2). Similarly, treated animals showed a greater increase in subchondral bone density in the LFC and LTP, and less

Fig. 3. Boxplots of subchondral bone mineral density (BMD, gm/cm2) in the lateral and medial femoral condyles (LFC, MFC) and tibial plateau (LTP, MTP) of NOC (,), MX ( ), and MX C GTN sheep ( ). Box indicates 25the75th percentile, line indicates 10the90th percentile, line indicates median, dot indicates mean. The symbol * differs significantly from NOC (P ! 0.05); yMX C GTN differs from MX (P ! 0.05).

979

Osteoarthritis and Cartilage Vol. 12, No. 12

Serum osteocalcin (ng/ml)

16

*

14 12 10 8 6 4 2 0

†

Urine pyridinoline (pmol/ g creatinine)

OC

0.012

Pyr

DPyr

0.25

0.01

0.2

0.008

0.15

0.006

0.1

0.004

0.05

0.002

0

Urine deoxypyridinoline (pmol/ g creatinine)

0.3

18

0

Fig. 4. Serum osteocalcin (left, ng/ml G SE), urinary pyridinoline (middle, pmol/mg creatinine G SE), and urine deoxypyridinoline (right, pmol/mg creatinine G SE) in NOC (,), MX ( ), and MX C GTN sheep (-). The symbol * differs significantly from NOC (P Z 0.011); y MX C GTN differs from MX (P Z 0.0023).

loss of BMD in the MTP (Fig. 3). Induction of endogenous NO production in bone is thought to be a protective response against cytokine-stimulated bone resorption in pathological conditions9,11. NO is known to inhibit both osteoclast function and osteoclast differentiation from precursor cell types10. GTN, as an exogenous donor of nitric oxide, appears to have had a similar bone-promoting effect in this study. It has been proposed by Fazzalari et al. that a different set of mediators might predominate in the remodelling processes of OA subchondral bone, compared to those operative during normal homeostatic turnover, such as the osteoprotegerin/receptor activator of nuclear factor kappa-b ligand (OPG/RANKL) axis42. Hilal et al. reported an alteration of subchondral bone metabolism in OA patients, suggesting that local release of inflammatory mediators may be at least as important in subchondral bone remodelling as mechanical stimuli43. The pronounced influence of exogenous NO in this study suggests that endogenously generated nitric oxide has the potential to be a prime mediator of the subchondral bone plate sclerosis that typically occurs in OA joints. However, the observed responses may again relate to mechanical factors peculiar to this surgical model; NO has been shown to potentiate mechanically induced bone formation in rat vertebrae44. GTN treatment also appeared to promote osteophytic remodelling, increasing the cross-sectional area of marginal osteophytes especially in the femoral condyles. Reports of reduced osteophyte growth in arthritic iNOS-deficient mice45 and in osteoarthritic dogs treated with a selective iNOS inhibitor46 support this association. Tidemark alterations and neovascularisation of the calcified cartilage layer were prominent following meniscectomy in this trial. Possibly due to the advanced age of the subjects, multiple tidemarks were present in nearly all sheep, including control animals. However, blood vessels were seen to invade the calcified cartilage in greater number and to a greater depth in meniscectomised animals, particularly in the lateral joint compartment. GTN treatment significantly altered the pattern of calcified cartilage neovascularisation following meniscectomy. An increased NO is known to promote angiogenesis in other tissues, possibly by feedback interaction with fibroblast growth factor-2 (FGF-2)47. The GTN-induced increase and decrease in vessel number observed in the LFC and

MTP, respectively (Table I), mirrored local subchondral bone responses, suggesting that the effect of NO was dependent on other metabolic or mechanical lesions extant in this model. Serum osteocalcin levels (a marker of bone formation) were significantly increased in meniscectomised sheep. This suggests an increase in bone formation (or bone turnover) in these animals, and was seen despite a slight loss of body weight. Stress is unlikely to have caused this elevation, as it has been shown in sheep that while shortterm stresses such as handling or transport do not influence osteocalcin levels, long-term corticosteroid administration reduces plasma levels31. Human studies have shown both elevated43,48 and reduced49,50 osteocalcin levels in OA patients. Serum osteocalcin levels of GTN-treated sheep were significantly lower than those of MX animals, and similar to NOC sheep. Urinary collagen cross-links (markers of bone resorption) showed a similar (but not statistically significant) trend, with GTN reversing a meniscectomyassociated increase in pyridinoline and deoxypyridinoline excretion. This suggests that GTN treatment attenuated an increase in bone turnover following meniscectomy. This finding was in contrast to both human and rat trials, in which topical GTN treatment did not affect the ovariectomyinduced increase in serum osteocalcin or deoxypyridinoline levels15,51,52. Previous topographical studies of the cartilage and bone changes in this OA model23,27,30 have shown that the observed structural alterations reflect altered joint kinematics consequent to loss of the important stabilizing and loaddistributing roles of the meniscus. Bone changes particularly mirror altered loading patterns, with a relative increase in bone formation in the middle and outer zones of the LFC and LTP, and relative bone loss in the outer MFC and MTP23,30. In this study, GTN treatment significantly influenced post-meniscectomy changes in SCB thickness and/ or density in four of these six regions (middle LFC, outer LTP, outer MFC, outer MTP). However, in less affected regions (e.g., middle MFC) GTN treatment had no apparent influence. Thus, GTN appears to have accentuated postmeniscectomy bone formation whilst simultaneously reducing post-meniscectomy bone loss, without a global effect on SCB thickness or density. Cartilage lesions induced by meniscectomy show a slightly different distribution, with net

980 thickening of cartilage predominant except at sites of focal erosive lesions23,27,30. Appleyard et al.27 found that postmeniscectomy alterations in cartilage composition and biomechanical integrity were most marked in the outer and middle zones of the LTP, sites previously protected by the meniscus. Thus, the effect of GTN on G* and phase lag in the outer LTP, and increased loss of cartilage in the middle LTP, are highly significant within the context of the lesions induced by this model and are again consistent with a local worsening of key pathology rather than a clear global influence on joint tissue structure. In summary, intermittent topical treatment with GTN, an efficient NO donor compound, was found to subtly but significantly worsen cartilage degeneration following ovine meniscectomy. This is consistent with the known adverse effects of NO in cartilage in other models, and complements findings of altered cartilage structure and function in GTNtreated normal sheep in a similar previous study18. Secondly, GTN treatment significantly increased subchondral bone thickness and density during subchondral remodelling following lateral meniscectomy, whilst peripheral markers suggested a reduction in bone turnover. It has been proposed that the local bone sclerosis accompanying OA results in a stiffer, less compliant subchondral plate, thereby increasing compressive forces on the overlying chondrocytes and accelerating cartilage degeneration21. The results of this study provide evidence that clinical use of NO donor compounds such as GTN may accentuate bone sclerosis and contribute to disease progression if used in the presence of OA. Furthermore, the changes observed here suggest that nitric oxide has the potential be an important mediator of the subchondral bone changes seen in OA.

Acknowledgments The authors would like to sincerely thank Ms Susan Smith for her competent preparation of the histological sections; Ms Joanna Makovey and Prof. Philip Sambrook (Department of Rheumatology, Royal North Shore Hospital, Sydney) for generous use of the Hologic QDR 4500W densitometer; Graeme Worth (Sir Charles Gairdner Hospital, Perth) for his help with osteocalcin assays; and Nick Avery (Collagen Research Group, University of Bristol) for performing collagen cross-link analyses.

References 1. Farrell A, Blake D, Palmer R, Moncada S. Increased concentration of nitrite in synovial fluid and serum samples suggest increased nitric oxide synthesis in rheumatic diseases. Ann Rheum Dis 1992;51: 1219e22. 2. Melchiorri C, Meliconi R, Frizziero L, Silvestri T, Pulsatelli L, Mazzetti I, et al. Enhanced and coordinated in vivo expression of inflammatory cytokines and nitric oxide synthase by chondrocytes from patients with osteoarthritis. Arthritis Rheum 1998;41:2165e74. 3. Taskiran D, Stefanovic-Racic M, Georgescu H, Evans C. Nitric oxide mediates suppression of cartilage proteoglycan synthesis by interleukin-1. Biochem Biophys Res Commun 1994;200:142e8. 4. Murrell G, Jang D, Williams R. Nitric oxide activates metalloprotease enzymes in articular cartilage. Biochem Biophys Res Commun 1995;206:15e21.

M. A. Cake et al.: NO donor accelerates OA bone sclerosis 5. Hashimoto S, Takahashi K, Amiel D, Coutts R, Lotz M. Chondrocyte apoptosis and nitric oxide production during experimentally-induced osteoarthritis. Arthritis Rheum 1998;41:1266e74. 6. van de Loo F, Arntz O, van Enckevort F, van Lent P, van den Berg W. Reduced cartilage proteoglycan loss during zymosan-induced gonarthritis in NOS2-deficient mice and in anti-interleukin-1-treated wild-type mice with unabated joint inflammation. Arthritis Rheum 1998;41:634e46. 7. Pelletier J, Lascau-Coman V, Jovanovic D, Fernandes J, Manning P, Connor JR, et al. Selective inhibition of inducible nitric oxide synthase in experimental osteoarthritis is associated with reduction in tissue levels of catabolic factors. J Rheumatol 1999;26:2002e14. 8. Evans D, Ralston S. Nitric oxide and bone. J Bone Miner Res 1996;11:300e5. 9. Hukkanen M, Hughes F, Buttery L, Gross S, Evans T, Seddon S, et al. Cytokine-stimulated expression of inducible nitric oxide synthase by mouse, rat, and human osteoblast-like cells and its functional role in osteoblast metabolic activity. Endocrinology 1995;136: 5445e53. 10. Brandi M, Hukkanen M, Umeda T, Moradi-Bidhendi N, Bianchi S, Gross S, et al. Bidirectional regulation of osteoclast function by nitric oxide synthase isoforms. Proc Natl Acad Sci U S A 1995;92:2954e8. 11. Lowik C, Nibbering P, van de Ruit M, Papapoulos S. Inducible production of nitric oxide in osteoblast-like cells and in fetal mouse bone explants is associated with suppression of osteoclastic bone resorption. J Clin Invest 1994;93:1465e72. 12. Kasten T, Collin-Osdoby P, Patel N, Krukowski M, Misko T, Settle S, et al. Potentiation of osteoclast bone-resorption activity by inhibition of nitric oxide synthase. Proc Natl Acad Sci U S A 1994;91:3569e73. 13. Wimalawansa S, De Marco G, Gangula P, Yallampalli C. Nitric oxide donor alleviates ovariectomy-induced bone loss. Bone 1996;18:301e4. 14. Wimalawansa S, Chapa M, Yallampalli C, Zhang R, Simmons D. Prevention of corticosteroid-induced bone loss with nitric oxide donor nitroglycerin in male rats. Bone 1997;21:275e80. 15. Wimalawansa S. Nitroglyerin therapy is as efficacious as standard estrogen replacement therapy (Premarin) in prevention of oophorectomy-induced bone loss: a human pilot clinical study. J Bone Miner Res 2000; 15:2240e4. 16. Jamal S, Browner W, Bauer D, Cummings S. Intermittent use of nitrates increases bone mineral density: the study of osteoporotic fractures. J Bone Miner Res 1998;13:1755e9. 17. Wimalawansa S. Prevention and treatment of osteoporosis: efficacy of combination of hormone replacement therapy with other antiresorptive agents. J Clin Densitom 2000;3:187e201. 18. Cake M, Appleyard R, Read R, Ghosh P, Swain M, Murrell G. Topical administration of the nitric oxide donor glyceryl trinitrate modifies the structural and biomechanical properties of ovine articular cartilage. Osteoarthritis Cartilage 2003;11:872e8. 19. Lane NE, Williams EN, Hung YY, Hochberg MC, Cummings SR, Nevitt MC. Association of nitrate use with risk of new radiographic features of hip osteoarthritis in elderly white women: the study of osteoporotic fractures. Arthritis Rheum 2003;49: 752e8.

Osteoarthritis and Cartilage Vol. 12, No. 12 20. Dequeker J, Aerssens J, Luyten FP. Osteoarthritis and osteoporosis: clinical and research evidence of inverse relationship. Aging Clin Exp Res 2003;15:426e39. 21. Radin E, Rose R. Role of subchondral bone in the initiation and progression of cartilage damage. Clin Orthop 1986;213:241e8. 22. Little C, Ghosh P, Bellenger C. Topographic variation in biglycan and decorin synthesis by articular cartilage in the early stages of osteoarthritis: an experimental study in sheep. J Orthop Res 1996;14:433e44. 23. Cake M, Read R, Guillou B, Ghosh P. Modification of articular cartilage and subchondral bone pathology in an ovine meniscectomy model of osteoarthritis by avocado and soya unsaponifiables (ASU). Osteoarthritis Cartilage 2000;8:404e11. 24. Little C, Smith S, Ghosh P, Bellenger C. Histomorphological and immunohistochemical evaluation of joint changes in a model of osteoarthritis induced by lateral meniscectomy in sheep. J Rheumatol 1997;24: 2199e209. 25. Hwa S, Smith M, Burkhardt D, Ghosh P. Oestrogen depletion provokes significant vascular and matrix changes in the calcified cartilage and subchondral bone of the femoraletibial joints of aged ewes (abstract). Trans Orthop Res Soc 2001;abstract 0373. 26. Appleyard R, Swain M, Khanna S, Murrell G. The accuracy and reliability of a novel handheld dynamic indentation probe for analysing articular cartilage. Phys Med Biol 2001;46:541e50. 27. Appleyard R, Burkhardt D, Ghosh P, Read R, Cake M, Swain M, et al. Topographical analysis of the structural, biochemical, and dynamic biomechanical properties of cartilage in an ovine model of osteoarthritis. Osteoarthritis Cartilage 2003;11:65e77. 28. Hayes W, Keer L, Herrmann G, Mockros L. A mathematical analysis for indentation tests of articular cartilage. J Biomech 1972;5:541e51. 29. Swann A, Seedhom B. Improved techniques for measuring the indentation and thickness of articular cartilage. Proc Inst Mech Eng [H] 1989;203:143e50. 30. Hwa S, Burkhardt D, Little C, Ghosh P. The effects of orally administered diacerein on cartilage and subchondral bone in an ovine model of osteoarthritis. J Rheumatol 2001;28:825e34. 31. Corlett S, Couch M, Care A, Sykes A. Measurement of plasma osteocalcin in sheep: assessment of circadian variation, the effects of age and nutritional status and the response to perturbation of the adrenocortical axis. Exp Physiol 1990;75:515e27. 32. Sims T, Avery N, Bailey A. Quantitative determination of collagen crosslinks. Methods Mol Biol 2000;139: 11e26. 33. Heinega˚rd D, Tiderstro¨m G. Determination of serum creatinine by direct colorimetric method. Clin Chim Acta 1973;43:305e10. 34. Zhu W, Mow V, Koob T, Eyre D. Viscoelastic shear properties of articular cartilage and the effects of glycosidase treatments. J Orthop Res 1993;11:771e81. 35. Le Roux M, Arokoski J, Vail T, Guilak F, Hyttinen M, Kiviranta I, et al. Simultaneous changes in the mechanical properties, quantitative collagen organization, and proteoglycan concentration of articular cartilage following meniscectomy. J Orthop Res 2000;18:383e92. 36. Setton L, Mow V, Howell D. Mechanical behavior of articular cartilage in shear is altered by transection of the anterior cruciate ligament. J Orthop Res 1995;13: 473e82.

981 37. Armstrong S, Read R, Ghosh P. The effects of intraarticular hyaluronan on cartilage and subchondral bone changes in an ovine model of early osteoarthritis. J Rheumatol 1994;21:680e8. 38. Grynpas M, Alpert B, Katz I, Lieberman I, Pritzker K. Subchondral bone in osteoarthritis. Calcif Tissue Int 1991;49:20e6. 39. Karvonen R, Miller P, Nelson D, Granda J, Ferna´ndezMadrid F. Periarticular osteoporosis in osteoarthritis of the knee. J Rheumatol 1998;25:2187e94. 40. Dedrick D, Goldstein S, Brandt K, O’Connor B, Goulet R, Albrecht M. A longitudinal study of subchondral plate and trabecular bone in cruciate deficient dogs with osteoarthritis followed up for 54 months. Arthritis Rheum 1993;36:1460e7. 41. Messner K, Fahlgren A, Ross I, Andersson B. Simultaneous changes in bone mineral density and articular cartilage in a rabbit meniscectomy model of knee osteoarthritis. Osteoarthritis Cartilage 2000;8: 197e206. 42. Fazzalari N, Kuliwaba J, Atkins G, Forwood M, Findlay D. The ratio of messenger RNA levels of receptor activator of nuclear factor kappa b ligand to osteoprotegerin correlates with bone remodelling indices in normal human cancellous bone but not in osteoarthritis. J Bone Miner Res 2001;16:1015e27. 43. Hilal G, Martel-Pelletier J, Pelletier J-P, Ranger P, Lajeunesse D. Osteoblast-like cells from human subchondral osteoarthritic bone demonstrate an altered phenotype in vitro: possible role in subchondral bone sclerosis. Arthritis Rheum 1998;41:891e9. 44. Chow J, Fox S, Lean J, Chambers T. Role of nitric oxide and prostaglandins in mechanically produced bone formation. J Bone Miner Res 1998;13:1039e44. 45. van den Berg W, van de Loo F, Joosten L, Arntz O. Animal models of arthritis in NOS2-deficient mice. Osteoarthritis Cartilage 1999;7:413e5. 46. Pelletier J, Jovanovic D, Fernandes J, Manning P, Connor JR, Currie MG, et al. Reduced progression of experimental osteoarthritis in vivo by selective inhibition of inducible nitric oxide synthase. Arthritis Rheum 1998;41:1275e86. 47. Ziche M, Parenti A, Ledda F, Dell’Era P, Granger H, Maggi C, et al. Nitric oxide promotes proliferation and plasminogen activator production by coronary venular endothelium through endogenous bFGF. Circ Res 1997;80:845e52. 48. Franck H, Ittel T. Serum osteocalcin levels in patients with psoriatic arthritis: an extended report. Rheumatol Int 2000;19:161e4. 49. Sowers M, Lachance L, Jamadar D, Hochberg M, Hollis B, Crutchfield M, et al. The associations of bone mineral density and bone turnover markers with osteoarthritis of the hand and knee in pre- and perimenopausal women. Arthritis Rheum 1999;42: 483e9. 50. Stewart A, Black A, Robins S, Reid D. Bone density and bone turnover in patients with osteoarthritis and osteoporosis. J Rheumatol 1999;26:622e6. 51. Wimalawansa S. Restoration of ovariectomy-induced osteopenia by nitroglycerin. Calcif Tissue Int 2000;66: 56e60. 52. Hukkanen M, Platts LA, Lawes T, Girgis SI, Konttinen YT, Goodship AE, et al. Effect of nitric oxide donor nitroglycerin on bone mineral density in a rat model of estrogen deficiency-induced osteopenia. Bone 2003; 32:142e9.