Regulation of osteoarthritis by omega-3 (n-3) polyunsaturated fatty acids in a naturally occurring model of disease

Open AccessPublished:July 11, 2011DOI:https://doi.org/10.1016/j.joca.2011.06.005

      Summary

      Objective

      To examine effects of high omega-3 (n-3) polyunsaturated fatty acid (PUFA) diets on development of osteoarthritis (OA) in a spontaneous guinea pig model, and to further characterise pathogenesis in this model. Modern diets low in n-3 PUFAs have been linked with increases in inflammatory disorders, possibly including OA. However, n-3 is also thought to increases bone density, which is a possible contributing factor in OA. Therefore we aim to determine the net influence of n-3 in disease development.

      Method

      OA-prone Dunkin-Hartley (DH) Guinea pigs were compared with OA-resistant Bristol Strain-2s (BS2) each fed a standard or an n-3 diet from 10 to 30 weeks (10/group). We examined cartilage and subchondral bone pathology by histology, and biochemistry, including collagen cross-links, matrix metalloproteinases (MMPs), alkaline phosphatase, glycosaminoglycan (GAG), and denatured type II collagen.

      Results

      Dietary n-3 reduced disease in OA-prone animals. Most cartilage parameters were modified by n-3 diet towards those seen in the non-pathological BS2 strain – significantly active MMP-2, lysyl-pyridinoline and total collagen cross-links – the only exception being pro MMP-9 which was lower in the BS2, yet increased with n-3. GAG content was higher and denatured type II lower in the n-3 group. Subchondral bone parameters in the DH n-3 group also changed towards those seen in the non-pathological strain, significantly calcium:phosphate ratios and epiphyseal bone density.

      Conclusion

      Dietary n-3 PUFA reduced OA in the prone strain, and most disease markers were modified towards those of the non-OA strain, though not all significantly so. Omega-3 did not increase markers of pathology in either strain.

      Keywords

      Introduction

      In an ageing population, osteoarthritis (OA) is set to become the fourth leading cause of disability by 2020
      • Woolf A.D.
      • Pfleger B.
      Burden of major musculoskeletal conditions.
      , so with the current lack of effective treatments there is an urgent need for preventative measures.
      The high omega-6 (n-6) polyunsaturated fatty acid (PUFA) content in Western diets has been associated with a number of inflammatory disorders such as heart disease, rheumatoid arthritis and colitis. Similarly, elevated levels of n-6 have been linked to OA in bone and cartilage
      • Plumb M.S.
      • Aspden R.M.
      High levels of fat and (n-6) fatty acids in cancellous bone in osteoarthritis.
      • Wang Y.
      • Wluka A.E.
      • Hodge A.M.
      • English D.R.
      • Giles G.G.
      • O’Sullivan R.
      • et al.
      Effect of fatty acids on bone marrow lesions and knee cartilage in healthy, middle-aged subjects without clinical knee osteoarthritis.
      . “Western” diets are cited as having an n-6:n-3 ratio of between 15 and 30:1, whereas diets for which we are naturally adapted, such as those for Palaeolithic and modern hunter-gatherers, contain roughly equal levels of n-6 and n-3 (1:1–2:1)
      • Simopoulos A.P.
      Evolutionary aspects of diet, the omega-6/omega-3 ratio and genetic variation: nutritional implications for chronic diseases.
      . Thus, dietary supplementation with n-3 PUFA has been recommended to redress this imbalance.
      PUFAs are essential fatty acids and precursors to a number of important factors called eicosanoids, such as prostaglandins, thromboxanes, leukotrienes and resolvins. These mediate various processes including inflammation and bone metabolism. Conversion of PUFAs to eicosanoids is achieved by the action of cyclo-oxygenases (COX) and lipoxygenases (LOX). Omega-3 and n-6 generate different series of eicosanoids, those from n-6 are generally highly pro-inflammatory and those from n-3 somewhat less so. In particular, PGE2, derived from n-6 PUFAs, has been associated with inhibition of anabolic processes and increased proteolytic degradation of cartilage
      • Heinecke L.F.
      • Grzanna M.W.
      • Au A.Y.
      • Mochal C.A.
      • Rashmir-Raven A.
      • Frondoza C.G.
      Inhibition of cyclooxygenase-2 expression and prostaglandin E2 production in chondrocytes by avocado soybean unsaponifiables and epigallocatechin gallate.
      • Li X.
      • Ellman M.
      • Muddasani P.
      • Wang J.H.
      • Cs-Szabo G.
      • van Wijnen A.J.
      • et al.
      Prostaglandin E2 and its cognate EP receptors control human adult articular cartilage homeostasis and are linked to the pathophysiology of osteoarthritis.
      • Attur M.
      • Al-Mussawir H.E.
      • Patel J.
      • Kitay A.
      • Dave M.
      • Palmer G.
      • et al.
      Prostaglandin E2 exerts catabolic effects in osteoarthritis cartilage: evidence for signaling via the EP4 receptor.
      . Therefore, manipulation of the precursors of prostaglandins is likely to have an effect on the initiation and progression of OA. There is some evidence that n-3 PUFA can be beneficial in humans
      • Gruenwald J.
      • Petzold E.
      • Busch R.
      • Petzold H.P.
      • Graubaum H.J.
      Effect of glucosamine sulfate with or without omega-3 fatty acids in patients with osteoarthritis.
      and animals
      • Roush J.K.
      • Cross A.R.
      • Renberg W.C.
      • Dodd C.E.
      • Sixby K.A.
      • Fritsch D.A.
      • et al.
      Evaluation of the effects of dietary supplementation with fish oil omega-3 fatty acids on weight bearing in dogs with osteoarthritis.
      • Fritsch D.A.
      • Allen T.A.
      • Dodd C.E.
      • Jewell D.E.
      • Sixby K.A.
      • Leventhal P.S.
      • et al.
      A multicenter study of the effect of dietary supplementation with fish oil omega-3 fatty acids on carprofen dosage in dogs with osteoarthritis.
      , and in vitro studies have also demonstrated positive outcomes
      • Hurst S.
      • Zainal Z.
      • Caterson B.
      • Hughes C.E.
      • Harwood J.L.
      Dietary fatty acids and arthritis.
      • Zainal Z.
      • Longman A.J.
      • Hurst S.
      • Duggan K.
      • Caterson B.
      • Hughes C.E.
      • et al.
      Relative efficacies of omega-3 polyunsaturated fatty acids in reducing expression of key proteins in a model system for studying osteoarthritis.
      . On the other hand, there is considerable evidence indicating that n-3 PUFAs promote bone formation and increase bone density
      • Hogstrom M.
      • Nordstrom P.
      • Nordstrom A.
      n-3 fatty acids are positively associated with peak bone mineral density and bone accrual in healthy men: the NO2 study.
      • Watkins B.A.
      • Li Y.
      • Seifert M.F.
      Dietary ratio of n-6/n-3 PUFAs and docosahexaenoic acid: actions on bone mineral and serum biomarkers in ovariectomized rats.
      . OA is characterised by increased subchondral bone density, and therefore n-3 induced bone formation may exacerbate OA progression
      • Muraoka T.
      • Hagino H.
      • Okano T.
      • Enokida M.
      • Teshima R.
      Role of subchondral bone in osteoarthritis development: a comparative study of two strains of guinea pigs with and without spontaneously occurring osteoarthritis.
      • Young B.D.
      • Samii V.F.
      • Mattoon J.S.
      • Weisbrode S.E.
      • Bertone A.L.
      Subchondral bone density and cartilage degeneration patterns in osteoarthritic metacarpal condyles of horses.
      • Quasnichka H.L.
      • Anderson-MacKenzie J.M.
      • Bailey A.J.
      Subchondral bone and ligament changes precede cartilage degradation in guinea pig osteoarthritis.
      • Bruyere O.
      • Dardenne C.
      • Lejeune E.
      • Zegels B.
      • Pahaut A.
      • Richy F.
      • et al.
      Subchondral tibial bone mineral density predicts future joint space narrowing at the medial femoro-tibial compartment in patients with knee osteoarthritis.
      • Dequeker J.
      • Mokassa L.
      • Aerssens J.
      • Boonen S.
      Bone density and local growth factors in generalized osteoarthritis.
      . With the potential for both beneficial and detrimental effects, it is essential that the net influence of n-3 PUFAs on OA is determined before being recommended to sufferers and those at risk.
      Previous studies have established the use of the Dunkin-Hartley (DH) guinea pig as a naturally occurring model of OA
      • Bendele A.M.
      • Hulman J.F.
      Spontaneous cartilage degeneration in guinea pigs.
      • Jimenez P.A.
      • Glasson S.S.
      • Trubetskoy O.V.
      • Haimes H.B.
      Spontaneous osteoarthritis in Dunkin Hartley guinea pigs: histologic, radiologic, and biochemical changes.
      • McDougall J.J.
      • Andruski B.
      • Schuelert N.
      • Hallgrimsson B.
      • Matyas J.R.
      Unravelling the relationship between age, nociception and joint destruction in naturally occurring osteoarthritis of Dunkin Hartley guinea pigs.
      , with the Bristol Strain-2 (BS2) as an OA-resistant control
      • Quasnichka H.L.
      • Anderson-MacKenzie J.M.
      • Bailey A.J.
      Subchondral bone and ligament changes precede cartilage degradation in guinea pig osteoarthritis.
      • Anderson-MacKenzie J.M.
      • Quasnichka H.L.
      • Starr R.L.
      • Lewis E.J.
      • Billingham M.E.
      • Bailey A.J.
      Fundamental subchondral bone changes in spontaneous knee osteoarthritis.
      . Studies have also examined bone and cartilage changes in DH guinea pigs compared with other OA-resistant strains
      • Muraoka T.
      • Hagino H.
      • Okano T.
      • Enokida M.
      • Teshima R.
      Role of subchondral bone in osteoarthritis development: a comparative study of two strains of guinea pigs with and without spontaneously occurring osteoarthritis.
      • Huebner J.L.
      • Hanes M.A.
      • Beekman B.
      • TeKoppele J.M.
      • Kraus V.B.
      A comparative analysis of bone and cartilage metabolism in two strains of guinea-pig with varying degrees of naturally occurring osteoarthritis.
      . The specific aims of this study are (1) to determine how dietary supplementation with n-3 fatty acids influences the progression of OA in the disease-prone DH guinea pigs and the OA-resistant BS2; and (2) to further characterise the DH/BS2 spontaneous guinea pig model of human OA.

      Materials and methods

       Animals and diets

      We used the established DH guinea pig model of spontaneous OA (Harlan UK, Oxon, UK), with the BS2 as an OA-resistant control (ASU, University of Bristol, UK). Animals were given a standard commercial diet until 10 weeks old, and then 10 animals from each strain were randomly allocated to either a defined control feed with n-6:n-3 ratio of 22:1 (typical of a “Western” diet), or a ‘high n-3’ diet with n-6:n-3 of 1.5:1 (see Table I; Teklad Custom Research Diets, UK). Each diet had 5% fat contents, contained the anti-oxidant tert-butylhydroquinone, and was vacuum packed in nitrogen at −20°C to minimise oxidation. Animals underwent no intervention other than dietary modification, and were reared with free access to food and water. After euthanasia (in accordance with UK regulations) at 30 weeks old, the hind legs were removed. All samples were stored at −80°C prior to analysis.
      Table IComposition of diets
      g/kgControl (high n-6)High n-3
      Alfalfa meal380380
      Wheat204204
      Oats146146
      Soybean meal140140
      Soybean hulls5050
      Safflower oil1810
      Canola oil128
      Corn oil10
      Fish oil17
      Coconut oil, hydrogenated107
      Mineral mix0.50.5
      Dicalcium phosphate1515
      Sodium chloride77
      Magnesium oxide11
      Vitamin mix55
      Stay-C35 (35% ascorbic acid)11
      DL-Alpha-Tocopheryl acetate (500 IU/g)0.30.3
      Choline chloride11
      Folic acid0.0040.004
      Vitamin A palmitate (500,000 U/g)0.040.04
      Vitamin D3, cholecalciferol (500,000 U/g)0.0040.004
      TBHQ (antioxidant)0.040.04
      n-6:n-322:11.5:1

       Histological disease assessment

      The right tibial plateau was removed, divided coronally through the insertion point of the collateral ligament, and the full width anterior section was fixed in neutral buffered formalin, decalcified and processed for wax histology using standard procedures. Sections were stained with toluidine blue or Safranin-O, and scored semi-quantitatively for OA severity, using a combination of the adapted Mankin scoring system and the Osteoarthritis Research Society International (OARSI) OA histopathology grading system which includes elements of subchondral bone changes
      • Anderson-MacKenzie J.M.
      • Quasnichka H.L.
      • Starr R.L.
      • Lewis E.J.
      • Billingham M.E.
      • Bailey A.J.
      Fundamental subchondral bone changes in spontaneous knee osteoarthritis.
      • Pritzker K.P.
      • Gay S.
      • Jimenez S.A.
      • Ostergaard K.
      • Pelletier J.P.
      • Revell P.A.
      • et al.
      Osteoarthritis cartilage histopathology: grading and staging.
      . Three sections were “blind” scored per animal, and scores were summed for each animal to give an overall histological score, with a possible range of 0–20.

       Cartilage biochemistry

      Cartilage from the posterior section of each tibial plateau was removed for analysis. Tissue taken from the left knee was analysed for glycosaminoglycan (GAG), and denatured and native type II collagen, whilst matrix metalloproteinase (MMP)-2 and MMP-9 activity and collagen cross-link analysis were undertaken on the right. All chemicals were purchased from Sigma unless otherwise stated.

       Cartilage extract preparation

      Freeze-dried samples were digested with tosyl-l-lysine chloromethyl ketone (TLCK)-treated bovine pancreatic α-Chymotrypsin (1 mg/ml) in 50 mM Tris buffer pH 7.5 containing inhibitors (1 mM iodoacetamide, 1 mM EDTA, 10 μg/ml pepstatin A). Digestion of denatured collagen was continued for 18 h at 34°C and the reaction stopped by the addition of tosyl-l-phenylalanine chloromethyl ketone (TPCK). The samples were centrifuged, supernatants removed and the pellets (remaining native collagen) digested with Proteinase K (1 mg/ml) in 50 mM Tris buffer pH7.5 at 56°C, with TLCK and TPCK. After 18 h, digestion was stopped by heating samples at 100°C for 15 min. Undigested material was removed, freeze-dried and weighed.

       Type II collagen

      The chymotrypsin and proteinase K extracts were assayed for type II collagen (% dry weight) and denatured type II (% type II) by inhibition ELISA using a mouse IgG monoclonal antibody to denatured type II collagen, as previously described
      • Hollander A.P.
      • Heathfield T.F.
      • Webber C.
      • Iwata Y.
      • Bourne R.
      • Rorabeck C.
      • et al.
      Increased damage to type II collagen in osteoarthritic articular cartilage detected by a new immunoassay.
      .

       GAG

      The extracts were assayed for sulphated GAGs (% dry weight) by a colorimetric assay, using dimethylmethylene blue (DMMB) as previously reported
      • Handley C.J.
      • Buttle D.J.
      Assay of proteoglycan degradation.
      .

       MMPs

      Tissue was freeze-dried, and extracted (0.1% Brij 35 in 20 mM triethanolamine; 20 μl/mg) at 4°C for 18 h. Extracts were assayed for MMP-2 and MMP-9 activity (% standard MMP-2) by gelatin zymography as previously described
      • Tarlton J.F.
      • Vickery C.J.
      • Leaper D.J.
      • Bailey A.J.
      Postsurgical wound progression monitored by temporal changes in the expression of matrix metalloproteinase-9.
      .

       Collagen analyses

      Collagen content (% dry weight) was determined by hydroxyproline analysis of a tissue hydrolysate using a continuous-flow autoanalyser (Burkard Scientific, Uxbridge, UK) as previously described
      • Knott L.
      • Tarlton J.F.
      • Bailey A.J.
      Chemistry of collagen cross-linking: biochemical changes in collagen during the partial mineralization of turkey leg tendon.
      • Bannister D.W.
      • Burns A.B.
      Adaptation of the Bergman and Loxley technique for hydroxyproline determination to the autoanalyzer and its use in determining plasma hydroxyproline in the domestic fowl.
      . Mature [hydroxylysyl-pyridinoline (HL-Pyr) and lysyl-pyridinoline (L-Pyr)] and immature hydroxyketonorleucine (HLKNL) collagen cross-links were quantified (M/M collagen) using modified amino acid analysis as described previously
      • Avery N.C.
      • Sims T.J.
      • Bailey A.J.
      Quantitative determination of collagen cross-links.
      . Lysine hydroxylation was determined using a modified amino acid analysis program as previously described
      • Knott L.
      • Tarlton J.F.
      • Bailey A.J.
      Chemistry of collagen cross-linking: biochemical changes in collagen during the partial mineralization of turkey leg tendon.
      .

       Subchondral bone biochemistry

      Subchondral bone was sampled from the posterior medial tibial plateau after removal of the articular cartilage, by cryosectioning to a depth of 450 μm, and collecting the sections. Sectioning in this manner powders the subchondral bone with minimal losses or contamination. The resulting samples were extracted for soluble proteins and the extracts assayed for MMP-2 and MMP-9 as described earlier. In addition extracts were analysed for alkaline phosphatase (ALP) activity (mU/L) by conversion of p-nitrophenylphosphate, to p-nitrophenol and monitored at 405 nm using a Konelab analyser
      • Tietz N.W.
      • Burtis C.A.
      • Duncan P.
      • Ervin K.
      • Petitclerc C.J.
      • Rinker A.D.
      • et al.
      A reference method for measurement of alkaline phosphatase activity in human serum.
      .
      Collagen quantification and cross-link analysis were undertaken on the insoluble pellet as described earlier. Mineral calcium and phosphate were quantified from aliquots of hydrolysate, using a Konelab analyser, whereby inorganic phosphate forms a complex with ammonium molybdate, measured at 340 nm
      • Daly J.A.
      • Ertingshausen G.
      Direct method for determining inorganic phosphate in serum with the “CentrifiChem”.
      , and calcium ions with arsenazo III, measured at 660 nm
      • Janssen J.W.
      • Helbing A.R.
      Arsenazo III: an improvement of the routine calcium determination in serum.
      . These values are used to give a mineral calcium:phosphate ratio.

       Dual-energy X-ray absorptiometry (DXA)

      The remaining left proximal tibial epiphysis adjacent to the subchondral was analysed by DXA using a PIXImis 1.44 (Lunar, Maddison, WI) fitted with small animal software. The bone mineral density (BMD, mg/cm2) of an area (7 mm×7 mm) immediately distal to the growth plate was measured
      • Tidswell H.K.
      • Innes J.F.
      • Avery N.C.
      • Clegg P.D.
      • Barr A.R.
      • Vaughan-Thomas A.
      • et al.
      High-intensity exercise induces structural, compositional and metabolic changes in cuboidal bones – findings from an equine athlete model.
      .

       Statistical analysis

      Results were analysed using PASW Statistics v17 (SPSS.com). Data was tested for normality using the Kolmogorov–Smirnov test and for homogeneity of variances using Levene’s test of equality of variances. Assuming normality and equal variances, a one-way ANOVA was used to test for significant differences specific for each aim: (1) model characterisation, BS2 compared to DH for n-6 control diet; (2) OA progression, n-3 compared to n-6 diet for DH strain; (3) OA promotion, n-3 compared to n-6 diet in control BS2 strain. Where data was not normal and variances not equal, log transformations were used, otherwise a Kruskal–Wallis test with Dunn’s multiple comparison tests. To evaluate multiple outcomes in the model, binomial discrete probability distribution and Chi-square tests were performed. Graphical data is presented as mean±S.E.M.

      Results

       Characterisation of the DH/BS2 model

       OA pathology

      The pathology scores were significantly lower in the BS2 strain compared to the DH strain (P=0.001), in line with previous studies verifying the BS2 strain as an OA-resistant guinea pig strain
      • Anderson-MacKenzie J.M.
      • Quasnichka H.L.
      • Starr R.L.
      • Lewis E.J.
      • Billingham M.E.
      • Bailey A.J.
      Fundamental subchondral bone changes in spontaneous knee osteoarthritis.
      . When the separate scores were considered, there was a strain effect overall (P<0.0001), with toluidine blue (P=0.0008), cartilage structure (P=0.0057) and OARSI criteria (P=0.0007) each being lower in the BS2 (Fig. 1).
      Figure thumbnail gr1
      Fig. 1Differences in individual pathology scores for each criteria as a result of diet (Std vs n-3 in the DH strain) and strain (DH vs BS2 both fed a standard diet). Closed circles, DH fed a standard diet; open circles, DH fed an n-3 diet; closed triangles, BS2 fed a standard diet. Total pathology scores for all comparisons are shown in .

       Pathogenic markers

       Cartilage

      The GAG and type II collagen (total or denatured) contents of the cartilage were similar between strains [Fig. 2(A and B)]. However the post-translational modifications of the collagen were markedly different between OA-prone DH and non-prone BS2 strains, with significantly lower levels of lysyl hydroxylation and higher levels of the cross-link L-Pyr, associated with a mineralising matrix, in the DH strain [P=0.01 and P<0.0001 respectively; Fig. 3(A and B)].
      Figure thumbnail gr2
      Fig. 2Differences in cartilage composition as a result of Strain (DH vs BS2) or Diet (standard diet black bars, n-3 diet grey bars, BS2/n-3 unfilled grey bars). (A) GAG (DH std n=9, DH n-3 n=10, BS2 std n=9, BS2 n-3 n=8); (B) Denatured type II collagen (DH std n=9, DH n-3 n=9, BS2 std n=10, BS2 n-3 n=10). For Diet ##P=0.003.
      Figure thumbnail gr3
      Fig. 3Differences in biochemical parameters of cartilage as a result of Strain (DH vs BS2) or Diet (standard diet black bars, n-3 diet grey bars, BS2/n-3 unfilled grey bars). (A) Lysyl hydroxylation of collagen (DH std n=9, DH n-3 n=8, BS2 std n=9, BS2 n-3 n=9); (B) Lysyl-pyridinoline (DH std n=9, DH n-3 n=8, BS2 std n=8, BS2 n-3 n=7): (C) Pro MMP-9 (DH std n=9, DH n-3 n=10, BS2 std n=10, BS2 n-3 n=10); (D) Pro MMP-2 (DH std n=9, DH n-3 n=10, BS2 std n=10, BS2 n-3 n=10); (E) MMP-2 activation (DH std n=9, DH n-3 n=10, BS2 std n=10, BS2 n-3 n=10). For diet #P<0.05, ##P<0.01. For strain ∗∗P<0.01, ∗∗∗P<0.001.
      Differences were also notable in mediators of matrix metabolism – the DH strain having significantly higher cartilage pro MMP-9 and pro MMP-2 levels than the control BS2 Strain [P=0.012 and P<0.0001 respectively, Fig. 3(C and D)], and the levels of pro MMP-2 were positively correlated with the total histological changes of the joint surface (r=0.576; P=0.003) and for each individual parameter and overall (P<0.05). The BS2s had very low levels of pro MMP-2, with no detectable active MMP-2 [Fig. 3(E)].

       Subchondral bone

      The collagen content was significantly higher in the DH animals compared to BS2’s (P=0.04; Table II), and was also positively correlated with histological score of the cartilage (r=0.403, P=0.034). Again the levels of lysyl hydroxylation were low in DH animals on the standard n-6 diet compared to control BS2s, although this was not significant [Fig. 4(A)]. DH subchondral bone had significantly more total cross-links [P=0.006, Fig. 4(B)] compared to BS2s.
      Table IISummary of results
      DHBS2
      Standard dietn-3 dietStandard dietn-3 diet
      Total path. Score
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      8.97 (6.17, 11.8)4.56 (0.83, 8.28)#2.20 (0.05, 4.34)∗∗∗1.78 (0.67, 4.24)
      Cartilage
      Type II collagen
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      55.7 (47.9, 63.4)58.7 (50.7, 66.6)59.2 (53.7, 64.8)56.9 (46.4, 67.4)
      Denatured type II1.33 (1.01, 1.65)1.16 (0.85, 1.48)1.39 (1.11, 1.68)1.23 (0.88, 1.58)
      GAG
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      29.3 (27.0, 31.7)35.7 (32.3, 39.1)##30.1 (27.4, 32.7)30.3 (26.5, 34.1)
      Pro MMP-2
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      0.858 (0.671, 1.04)0.769 (0.670, 0.865)0.289 (0.212, 0.366)∗∗∗0.312 (0.245, 0.37)
      Activated MMP-2
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      19.3 (16.1, 22.6)11.6 (8.0, 15.2)##00
      ProMMP-91.50 (0.98, 2.02)1.93 (1.5, 2.25)1.03 (0.85, 1.2)∗∗1.38 (1.08, 1.68)
      Hydroxylysine
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      21.2 (20.1, 22.2)22.6 (20.9, 24.3)24.0 (22.0, 26.0)∗∗22.8 (21.0, 24.5)
      HL-Pyr1.1 (1.0, 1.27)1.03 (0.96, 1.1)1.28 (1.09, 1.46)1.32 (1.26, 1.49)
      L-Pyr
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      0.228 (0.203, 0.251)0.200 (0.178, 0.221)#0.103 (0.076, 0.131)∗∗∗0.107 (0.070, 0.144)
      Total cross-links1.92 (1.75, 2.08)1.68 (1.57, 1.78)##2.05 (1.84, 2.26)2.13 (1.96, 2.30)
      Subchondral bone
      Collagen
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      13.4 (12.3, 14.6)13.0 (12.49, 13.48)11.7 (10.4, 13.1)∗12.7 (11.6, 13.7)
      Ca:P
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      2.73 (2.37, 3.09)3.36 (2.97, 3.75)#4.39 (3.43, 5.35)∗∗4.95 (4.11, 5.79)
      ALP
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      1.10 (0.85, 1.35)0.952 (0.873, 1.03)0.779 (0.58, 0.98)∗0.800 (0.600, 1.00)
      Pro MMP-9
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      3.07 (2.28, 3.86)7.22 (2.14, 12.3)9.83 (5.65, 14.01)∗∗∗12.4 (8.86, 15.9)
      Hydroxylysine
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      14.9 (13.7, 16.1)16.8 (14.1, 19.4)16.1 (14.7, 17.6)16.9 (14.6, 19.2)
      HLKNL
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      1.08 (0.959, 1.21)0.976 (0.889, 1.06)0.927 (0.831, 1.02)∗0.955 (0.840, 1.07)
      HL-Pyr
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      0.453 (0.391, 0.514)0.425 (0.366, 0.484)0.356 (0.312, 0.4)∗∗0.393 (0.361, 0.425)
      L-Pyr
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      0.097 (0.070, 0.124)0.091 (0.075, 0.107)0.086 (0.065, 0.106)0.087 (0.058, 0.117)
      Total cross-links
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      1.82 (1.64, 1.99)1.67 (1.52, 1.82)1.52 (1.40, 1.64)∗∗1.63 (1.52, 1.75)
      Adjacent BMD
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      519 (507, 530)495 (479, 520)#465 (448, 483)∗∗∗461.1 (440, 482)
      Mean (95% CI); #P<0.05, ##P<0.01, ###P<0.001 comparing diet for DH; ∗P<0.05, ∗P<0.01, ∗∗∗P<0.001 comparing strain (n-6 diet).
      Those parameters for which the n-3 diet modified DH values towards those of the non-OA strain.
      Figure thumbnail gr4
      Fig. 4Differences in biochemical parameters of subchondral bone as a result of Strain (DH vs BS2) or Diet (standard diet black bars, n-3 diet grey bars, BS2/n-3 unfilled grey bars). (A) Lysyl hydroxylation of collagen (DH std n=10, DH n-3 n=10, BS2 std n=10, BS2 n-3 n=8); (B) Total collagen cross-links: (C) Ca:PO4 ratio (n=10); (D) Alkaline phosphatise (DH std n=10, DH n-3 n=10, BS2 std n=8, BS2 n-3 n=10); (E) Pro MMP-9 (DH std n=10, DH n-3 n=10, BS2 std n=8, BS2 n-3 n=8). For diet #P<0.05. For strain ∗P<0.05, ∗∗P<0.01, ∗∗∗P<0.001.
      The Ca:P was lower in the DH than the control guinea pig subchondral bone [P=0.002, Fig. 4(C)]. Although we did not directly measure the BMD of the subchondral bone region, there was significantly greater BMD in the adjacent proximal tibial epiphysis in the DH (P=0.0001; Table II).
      Levels of ALP, a marker of osteoblast activity, were significantly higher in DH bone extracts compared to the BS2 strain [P=0.04; Fig. 4(D)]. Pro MMP-9 levels were more than 2-fold lower in the DH animals [P=0.001; Fig. 4(E)] and were negatively correlated with histological score (r=−0.540, P=0.004). There was no detectable pro or active MMP-2 in the BS2 extracts, and below accurately quantifiable levels of active MMP-2 in the DH animals only.

       Effect of n-3 on disease progression in OA-developing DH strain

       OA pathology

      The n-3 diet halved the average histological scores in the DH animals (P=0.048) by comparison with the standard diet. When the separate scores were considered, there was a diet effect overall (P=0.0083), with toluidine blue (P=0.059), cartilage structure (P=0.015) and OARSI criteria (P=0.046) each being lower in the n-3 (Fig. 1).

       Pathogenic markers

       Cartilage

      The GAG content of the cartilage significantly increased (P=0.003) with n-3 compared to the control diet in the DH strain [Fig. 2(A)]. Levels of denatured type II collagen were reduced slightly with the n-3 diet in both strains [Fig. 2(B)], although comparison between strains showed no clear association with pathology. High n-3 diet increased lysyl hydroxylation in the OA-prone DH towards those seen in the non-prone control BS2 [P=0.1, Fig. 3(A)], and this was reflected in significantly decreased levels of L-Pyr [P=0.05, Fig. 3(B)]. The total level of cross-links was also lower in the n-3 fed DH (P=0.008, Table II). Levels of pro MMP-9 were elevated, although not significantly [Fig. 3(C)], whilst total MMP-2 (NS) and the percentage active MMP-2 (P<0.002) levels were decreased [Fig. 3(D and E)] for the n-3 diet group.

       Subchondral bone

      The collagen biochemistry parameters measured in the OA-prone strain all changed towards control BS2 levels as a result of the n-3 diet. However, most of these failed to reach significance in this study (Table II). The Ca:P ratio increased in the DH with the n-3 diet [P=0.015, Fig. 4(C)].The BMD of the adjacent proximal epiphyseal area also decreased with n-3 (P=0.03, Table II). MMP-9 expression increased [P=0.07, Fig. 4(E)], and there was a trend towards increased levels of collagen lysine hydroxylation [P=0.1, Fig. 4(A)], decreased HLKNL (P=0.12) and decreased total collagen cross-links (P=0.17) with the n-3 diet (Table II).
      Examining the overall data, of the measures of cartilage and bone pathogenic markers examined in the DH on a high n-3 or a standard diet, 16 changed in the direction of the OA-resistant BS2 as a result of the n-3 diet (significant in 7). Of the remainder, two (denatured type II and total cross-links) showed no difference for breed, and two (cartilage pro MMP-9 and HL-Pyr) showed a non-significant change in the opposite direction (Table II). Comparing the direction of dietary changes with respect to the model, a binomial discrete probability distribution demonstrated a probability of P=0.0046 that these occurred by chance, and Chi-square test showed a probability of P=0.001 that the data distribution was equal.

       Effect of n-3 in promoting OA in control BS2 strain

      The n-3 diet had little apparent effect on control BS2 animals in histopathological OA scores, and fewer effects, compared with DH, on biochemical parameters investigated in this study. This is in line with n-3 modulating both disease parameters in OA and baseline physiological levels of bone and cartilage mediators. The only significant change with the high n-3 diet in the BS2 was raised levels of pro MMP-9 in cartilage [P=0.033, Fig. 3(C)].

      Discussion

      The principle outcomes from this study were (1) that n-3 PUFAs reduce OA pathology in this naturally occurring model, and (2) that a number of physiological markers of disease confirmed this at a mechanistic level. Also, there was no evidence that n-3 contributed to an increase in OA in either strain. Furthermore, we were able to confirm the utility of this model of OA and demonstrated a number of changes in metabolism and composition not previously reported which may give further insight into the model, and into early OA changes in human disease.

       Characterisation of the DH/BS2 model

      The pathology scores show clear and spontaneous development of OA in the DH guinea pigs, with little apparent pathology in the BS2 strain at 30 weeks. Characterisation of this model of spontaneous OA, with BS2 as a non-disease prone control, provides valuable information on changes in composition and metabolism of the cartilage and subchondral bone in early OA. However, it should be acknowledged that OA does develop in much older BS2 guinea pigs, and this strain may more accurately be described as late onset.
      This is the first study to report levels of collagen cross-linking in guinea pig joints. Cartilage collagen of the DH animals was found to have a reduced lysyl hydroxylation compared with control BS2s, which was reflected in higher levels of the less hydroxylated mature collagen cross-link, L-Pyr. Pokharna et al reported no difference in overall Pyr levels (L-Pyr plus HL-Pyr) in a rabbit model of OA, in accordance with our findings
      • Pokharna H.K.
      • Monnier V.
      • Boja B.
      • Moskowitz R.W.
      Lysyl oxidase and Maillard reaction-mediated crosslinks in aging and osteoarthritic rabbit cartilage.
      . However, Bank et al reported elevated levels of hydroxylysine and HL-Pyr in degenerated human articular cartilage
      • Bank R.A.
      • Verzijl N.
      • Lafeber F.P.
      • Tekoppele J.M.
      Putative role of lysyl hydroxylation and pyridinoline cross-linking during adolescence in the occurrence of osteoarthritis at old age.
      . However, these changes were associated with individuals with no apparent OA, unlike the present study.
      We identified higher levels of both pro MMP-9 and pro MMP-2 in the DH cartilage compared with BS2, and that active MMP-2 was only detected in OA DH cartilage, in accordance with studies in other species
      • Clegg P.D.
      • Carter S.D.
      Matrix metalloproteinase-2 and -9 are activated in joint diseases.
      . Elevated catabolism/metabolism, as shown by increased levels and activation of MMPs is associated with OA progression
      • Masuhara K.
      • Nakai T.
      • Yamaguchi K.
      • Yamasaki S.
      • Sasaguri Y.
      Significant increases in serum and plasma concentrations of matrix metalloproteinases 3 and 9 in patients with rapidly destructive osteoarthritis of the hip.
      • Kanyama M.
      • Kuboki T.
      • Kojima S.
      • Fujisawa T.
      • Hattori T.
      • Takigawa M.
      • et al.
      Matrix metalloproteinases and tissue inhibitors of metalloproteinases in synovial fluids of patients with temporomandibular joint osteoarthritis.
      • Hulejova H.
      • Baresova V.
      • Klezl Z.
      • Polanska M.
      • Adam M.
      • Senolt L.
      Increased level of cytokines and matrix metalloproteinases in osteoarthritic subchondral bone.
      . MMP-9 is linked with inflammation, and MMP-2 with collagen turnover. As both are degradative proteases, their association with OA may be expected. However, as both have a role in tissue repair their relationship with OA is likely to be complex.
      In subchondral bone, we identified elevated levels of ALP in DH compared with BS2, consistent with previous studies of OA in-vitro
      • Couchourel D.
      • Aubry I.
      • Delalandre A.
      • Lavigne M.
      • Martel-Pelletier J.
      • Pelletier J.P.
      • et al.
      Altered mineralization of human osteoarthritic osteoblasts is attributable to abnormal type I collagen production.
      • Sanchez C.
      • Deberg M.A.
      • Bellahcene A.
      • Castronovo V.
      • Msika P.
      • Delcour J.P.
      • et al.
      Phenotypic characterization of osteoblasts from the sclerotic zones of osteoarthritic subchondral bone.
      and in-vivo
      • Huebner J.L.
      • Hanes M.A.
      • Beekman B.
      • TeKoppele J.M.
      • Kraus V.B.
      A comparative analysis of bone and cartilage metabolism in two strains of guinea-pig with varying degrees of naturally occurring osteoarthritis.
      • Dieppe P.
      • Cushnaghan J.
      • Young P.
      • Kirwan J.
      Prediction of the progression of joint space narrowing in osteoarthritis of the knee by bone scintigraphy.
      • Mansell J.P.
      • Tarlton J.F.
      • Bailey A.J.
      Biochemical evidence for altered subchondral bone collagen metabolism in osteoarthritis of the hip.
      , demonstrating increased subchondral bone metabolism in guinea pig OA. Pro and active MMP-2 were barely detectable in DH bone, and absent in the control BS2 strain. Previous studies reported elevated MMP-2 and enhanced activation in OA bone, most notably in the proximal epiphysis
      • Mansell J.P.
      • Tarlton J.F.
      • Bailey A.J.
      Expression of gelatinases within the trabecular bone compartment of ovariectomized and parathyroidectomized adult female rats.
      . However MMP-9 was found to be lower in the OA-developing DH strain than the control strain. MMP-9 is produced by osteoclasts, and is elevated during active bone resorption
      • Logar D.B.
      • Komadina R.
      • Prezelj J.
      • Ostanek B.
      • Trost Z.
      • Marc J.
      Expression of bone resorption genes in osteoarthritis and in osteoporosis.
      • Teti A.
      • Migliaccio S.
      • Taranta A.
      • Bernardini S.
      • De Rossi G.
      • Luciani M.
      • et al.
      Mechanisms of osteoclast dysfunction in human osteopetrosis: abnormal osteoclastogenesis and lack of osteoclast-specific adhesion structures.
      . Reduced MMP-9 in the OA-developing strain is in line with subchondral bone sclerosis in OA
      • 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.
      , and corroborates the higher levels of bone formation (ALP) seen in DH. However, other studies have shown a link between OA and raised MMP-9. This contradiction may relate to their bone sampling site being outside the subchondral compartment
      • Hulejova H.
      • Baresova V.
      • Klezl Z.
      • Polanska M.
      • Adam M.
      • Senolt L.
      Increased level of cytokines and matrix metalloproteinases in osteoarthritic subchondral bone.
      , or disease being more advanced
      • Logar D.B.
      • Komadina R.
      • Prezelj J.
      • Ostanek B.
      • Trost Z.
      • Marc J.
      Expression of bone resorption genes in osteoarthritis and in osteoporosis.
      .
      There were elevated levels of individual and total collagen cross-links in the DH subchondral bone compared with the control strain. There have been no other reports to the author’s knowledge of collagen cross-linking in the subchondral bone in any species, although some have analysed trabecular bone immediately proximal to the subchondral plate
      • Mansell J.P.
      • Tarlton J.F.
      • Bailey A.J.
      Expression of gelatinases within the trabecular bone compartment of ovariectomized and parathyroidectomized adult female rats.
      . Raised levels of cross-links are indicative of a stiffer subchondral bone, in line with the previous findings indicating sclerosis in the OA-developing strain
      • Anderson-MacKenzie J.M.
      • Quasnichka H.L.
      • Starr R.L.
      • Lewis E.J.
      • Billingham M.E.
      • Bailey A.J.
      Fundamental subchondral bone changes in spontaneous knee osteoarthritis.
      .
      Guinea pig subchondral bone has an unusually high Ca:P ratio (mean 3.86). Although this has not been reported before in guinea pigs, in other species this is generally between 1.7 and 2, increasing with maturity of hydroxyapatite. High Ca:P ratios have been reported in OA subchondral bone
      • Kouri J.B.
      • Aguilera J.M.
      • Reyes J.
      • Lozoya K.A.
      • Gonzalez S.
      Apoptotic chondrocytes from osteoarthrotic human articular cartilage and abnormal calcification of subchondral bone.
      • Sato M.
      • Wada M.
      • Miyoshi N.
      • Imamura Y.
      • Noriki S.
      • Uchida K.
      • et al.
      Hydroxyapatite maturity in the calcified cartilage and underlying subchondral bone of guinea pigs with spontaneous osteoarthritis: analysis by Fourier transform infrared microspectroscopy.
      but in this study it was the control BS2 strain which exhibited the highest ratios. High Ca:P ratios could be indicative of the presence of amorphous, calcium rich deposits
      • Chichocki T.
      • Gonsior B.
      • Hofert M.
      • Jarczyk L.
      • Raith B.
      • Rokita E.
      • et al.
      Measurements of mineralization process in the femur growth plate and rib cartilage of the mouse using pixe in combination with a proton microprobe.
      , and matrix vesicles associated with cartilage mineralisation accumulating large amounts of calcium
      • Ali S.Y.
      • Rees J.A.
      • Scotchford C.A.
      Microcrystal deposition in cartilage and in osteoarthritis.
      • Ali S.Y.
      • Wisby A.
      • Evans L.
      • Craig-Gray J.
      The sequence of calcium and phosphorus accumulation by matrix vesicles.
      . The low Ca:P in the OA-prone DH strain may result from formation of new, less mature bone, in line with raised ALP.

       Effect of n-3 on disease progression in OA-developing DH strain

      Few experimental studies have investigated the effects of n-3 PUFAs on OA, although there is clinical evidence that increasing dietary n-3 relative to n-6 may be beneficial in terms of symptom management in humans
      • Hurst S.
      • Zainal Z.
      • Caterson B.
      • Hughes C.E.
      • Harwood J.L.
      Dietary fatty acids and arthritis.
      and other species
      • Roush J.K.
      • Cross A.R.
      • Renberg W.C.
      • Dodd C.E.
      • Sixby K.A.
      • Fritsch D.A.
      • et al.
      Evaluation of the effects of dietary supplementation with fish oil omega-3 fatty acids on weight bearing in dogs with osteoarthritis.
      . Not all studies however conclude that dietary n-3 PUFA supplementation is of benefit, in the treatment of OA
      • Rosenbaum C.C.
      • O’Mathuna D.P.
      • Chavez M.
      • Shields K.
      Antioxidants and antiinflammatory dietary supplements for osteoarthritis and rheumatoid arthritis.
      . It is possible that OA could be aggravated due to the potential for increased bone formation which is linked to the progression of OA. This study is the first to look at both cartilage and subchondral bone changes with increased dietary n-3 PUFAs.
      In terms of gross pathology, the high n-3 diet significantly and substantially (by 50%) reduced OA associated changes in the DH strain, with no significant effect on the control strain. Having established this improvement in OA pathology, we went on to demonstrate parallel changes in a number of compositional and metabolic markers of cartilage and bone function.
      Increased cartilage GAG content, reduced denatured type II collagen (NS), and reduced pro and activated MMP-2 in DH guinea pigs fed an n-3 diet are all indicative of disease attenuation. Others, such as L-Pyr and lysine hydroxylation change to be more in line with the control animals. Only cartilage MMP-9 changed in the opposite direction to that seen in the control animals, though not significantly. However, as noted, the relationship between MMP-9, cartilage degradation and regeneration is likely to be complex. Omega-3 supplemented DH guinea pigs had fewer calcifying cartilage lesions identified as part of the Mankin grading, corroborated by reduction in markers of mineralisation, collagen lysyl hydroxylation and L-Pyr cross-links, each shifting towards those seen in the BS2s. Taken together, the analyses of the biochemical parameters demonstrate an n-3 treatment effect, and provide evidence of particular mechanisms of disease, but that these should be explored further to confirm links to n-3 diet.
      Evidence that n-3 modifies some elements only in the disease-prone animals, whilst others are also moderated in the control strain, supports the contention that n-3 regulates both the disease process and the underlying physiology of cartilage and subchondral bone in the absence of overt disease. As there is evidence of OA in very old BS2 guinea pigs, it could be speculated that those elements modified in the “later onset” BS2 as well as the DH represent the earliest stages of OA pathology, such that denatured type II collagen, collagen hydroxylation (in cartilage and bone), MMP-9 (also in both) and Ca:P, are early events, whereas changes in GAG content, collagen crosslinking, MMP-2 and ALP are part of established disease in th DH animals. Thus n-3 has the potential to prevent or delay the onset of disease as well as to reduce its progression.
      Though no previous in vivo studies have examined cartilage matrix changes with n-3, in vitro studies using bovine chondrocytes identified reduced expression for cartilage-degrading proteinases, cyclooxygenase-2 and inflammatory cytokines in the presence of n-3, in particular eicosapentanoic acid
      • Zainal Z.
      • Longman A.J.
      • Hurst S.
      • Duggan K.
      • Caterson B.
      • Hughes C.E.
      • et al.
      Relative efficacies of omega-3 polyunsaturated fatty acids in reducing expression of key proteins in a model system for studying osteoarthritis.
      .
      Reports on the effect of PUFAs on bone are more numerous, with n-3 PUFAs generally reducing bone loss, by decreasing bone resorption
      • Sun D.
      • Krishnan A.
      • Zaman K.
      • Lawrence R.
      • Bhattacharya A.
      • Fernandes G.
      Dietary n-3 fatty acids decrease osteoclastogenesis and loss of bone mass in ovariectomized mice.
      , and/or increasing bone formation
      • Watkins B.A.
      • Li Y.
      • Lippman H.E.
      • Feng S.
      Modulatory effect of omega-3 polyunsaturated fatty acids on osteoblast function and bone metabolism.
      • Watkins B.A.
      • Li Y.
      • Lippman H.E.
      • Seifert M.F.
      Omega-3 polyunsaturated fatty acids and skeletal health.
      . However specific effects on subchondral bone are less well documented, and extrapolation of data from other skeletal sites may be misleading as sclerotic subchondral bone present in OA is dissimilar to other bone sites and disease states
      • Quasnichka H.L.
      • Anderson-MacKenzie J.M.
      • Bailey A.J.
      Subchondral bone and ligament changes precede cartilage degradation in guinea pig osteoarthritis.
      • Anderson-MacKenzie J.M.
      • Quasnichka H.L.
      • Starr R.L.
      • Lewis E.J.
      • Billingham M.E.
      • Bailey A.J.
      Fundamental subchondral bone changes in spontaneous knee osteoarthritis.
      • 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.
      • Bailey A.J.
      • Mansell J.P.
      • Sims T.J.
      • Banse X.
      Biochemical and mechanical properties of subchondral bone in osteoarthritis.
      . Increase in subchondral bone formation as a result of n-3 is potentially detrimental in the progression of OA. However, in the OA-prone DH guinea pigs, n-3 acts to modify markers indicating reduced subchondral bone deposition, including increased pro MMP-9, increased Ca:P ratio, decreased collagen content and decreased ALP levels. Supplementation with n-3 also directed collagen cross-linking and lysine hydroxylation towards control BS2 levels in DH subchondral bone.
      Supplementation with n-3 reduced the BMD of the adjacent proximal epiphyseal bone, again towards control levels. This is contrary to expectations as n-3 is generally reported to increase bone density
      • Watkins B.A.
      • Li Y.
      • Seifert M.F.
      Dietary ratio of n-6/n-3 PUFAs and docosahexaenoic acid: actions on bone mineral and serum biomarkers in ovariectomized rats.
      , but as increased density of bone underlying cartilage is thought to contribute to OA
      • Dequeker J.
      • Mokassa L.
      • Aerssens J.
      • Boonen S.
      Bone density and local growth factors in generalized osteoarthritis.
      , this may demonstrate a further protective effect of n-3.
      There is circumstantial evidence for a link between human OA and a high n-6:n-3 ratio. High n-6, particularly arachidonic acid, have been reported in bone from OA patients
      • Plumb M.S.
      • Aspden R.M.
      High levels of fat and (n-6) fatty acids in cancellous bone in osteoarthritis.
      , and an association between OA and subchondral osteoblastic production of PGE2, IL-6, and COX-2 levels has been identified, as well as an elevated response to exogenous PGE2
      • Massicotte F.
      • Lajeunesse D.
      • Benderdour M.
      • Pelletier J.P.
      • Hilal G.
      • Duval N.
      • et al.
      Can altered production of interleukin-1beta, interleukin-6, transforming growth factor-beta and prostaglandin E(2) by isolated human subchondral osteoblasts identify two subgroups of osteoarthritic patients.
      .
      Therefore supplementation of n-3 PUFAs returns the balance in n-3 and n-6 to a more ‘normal’ functional state, reducing signs of OA in both cartilage and subchondral bone.

       Effect of n-3 on OA promotion in control BS2 strain

      There is little evidence from this study for any detrimental effects of n-3 supplementation in the control OA-resistant strain. This suggests that there are no ill effects of n-3 on OA development in disease free individuals.

       Limitations of the study

      The principle limitations of this study were the small quantities of tissues available for analysis and the large number of analyses attempted, such that many measures were at the lower end of their detectable range. A group size of 10 was adequate for determining differences in pathology with diet, which was the primary outcome of this study and for detecting differences in strain. However, there were a number of biochemical measures that did not reach significance when comparing diets within the OA-developing strain. It should also be noted that the development of OA in this model is at an early age by comparison with human disease, and occurs on an accelerated timescale.

      Conclusion

      This study demonstrates clear benefits of n-3 supplementation in reducing the signs of OA in a naturally occurring model of disease. Furthermore, there was no sign that increased n-3 would lead to disease in the OA free strain. We have also further characterised this model, and identified fundamental differences in cartilage and bone biology associated with OA. We propose that a high n-3 diet has the potential to reduce signs of OA in both cartilage and subchondral bone. Further studies are needed to determine the influence of n-3 on established disease, and to confirm these effects in human OA.

      Author contributions

      We declare that all authors listed contributed to the acquisition of data, drafting, critical revision and final approval of this manuscript, in line with Osteoarthritis and Cartilage guidelines. Dr John Tarlton ( [email protected] ) takes responsibility for the integrity of this work.

      Role of funding source

      This work was funded by Arthritis Research UK (ARUK, 17249). Other than their support and encouragement, ARUK had no involvement in the design or execution of the study, in the preparation of the manuscript or the decision to submit this work for publication.

      Conflict of interest

      The Authors have no conflict of interest relating to this manuscript.

      Acknowledgements

      We would like to thank Arthritis Research UK for supporting this work. We would also like to acknowledge Mr Trevor Sims for his technical assistance, Laura Halliday and Isabel Mark for performing the DXA measurements, and to thank the late Dr Mike Billingham for his guidance with the guinea pig model.

      References

        • Woolf A.D.
        • Pfleger B.
        Burden of major musculoskeletal conditions.
        Bull World Health Organ. 2003; 81: 646-656
        • Plumb M.S.
        • Aspden R.M.
        High levels of fat and (n-6) fatty acids in cancellous bone in osteoarthritis.
        Lipids Health Dis. 2004; 3: 12
        • Wang Y.
        • Wluka A.E.
        • Hodge A.M.
        • English D.R.
        • Giles G.G.
        • O’Sullivan R.
        • et al.
        Effect of fatty acids on bone marrow lesions and knee cartilage in healthy, middle-aged subjects without clinical knee osteoarthritis.
        Osteoarthritis Cartilage. 2008; 16: 579-583
        • Simopoulos A.P.
        Evolutionary aspects of diet, the omega-6/omega-3 ratio and genetic variation: nutritional implications for chronic diseases.
        Biomed Pharmacother. 2006; 60: 502-507
        • Heinecke L.F.
        • Grzanna M.W.
        • Au A.Y.
        • Mochal C.A.
        • Rashmir-Raven A.
        • Frondoza C.G.
        Inhibition of cyclooxygenase-2 expression and prostaglandin E2 production in chondrocytes by avocado soybean unsaponifiables and epigallocatechin gallate.
        Osteoarthritis Cartilage. 2010; 18: 220-227
        • Li X.
        • Ellman M.
        • Muddasani P.
        • Wang J.H.
        • Cs-Szabo G.
        • van Wijnen A.J.
        • et al.
        Prostaglandin E2 and its cognate EP receptors control human adult articular cartilage homeostasis and are linked to the pathophysiology of osteoarthritis.
        Arthritis Rheum. 2009; 60: 513-523
        • Attur M.
        • Al-Mussawir H.E.
        • Patel J.
        • Kitay A.
        • Dave M.
        • Palmer G.
        • et al.
        Prostaglandin E2 exerts catabolic effects in osteoarthritis cartilage: evidence for signaling via the EP4 receptor.
        J Immunol. 2008; 181: 5082-5088
        • Gruenwald J.
        • Petzold E.
        • Busch R.
        • Petzold H.P.
        • Graubaum H.J.
        Effect of glucosamine sulfate with or without omega-3 fatty acids in patients with osteoarthritis.
        Adv Ther. 2009; 26: 858-871
        • Roush J.K.
        • Cross A.R.
        • Renberg W.C.
        • Dodd C.E.
        • Sixby K.A.
        • Fritsch D.A.
        • et al.
        Evaluation of the effects of dietary supplementation with fish oil omega-3 fatty acids on weight bearing in dogs with osteoarthritis.
        J Am Vet Med Assoc. 2010; 236: 67-73
        • Fritsch D.A.
        • Allen T.A.
        • Dodd C.E.
        • Jewell D.E.
        • Sixby K.A.
        • Leventhal P.S.
        • et al.
        A multicenter study of the effect of dietary supplementation with fish oil omega-3 fatty acids on carprofen dosage in dogs with osteoarthritis.
        J Am Vet Med Assoc. 2010; 236: 535-539
        • Hurst S.
        • Zainal Z.
        • Caterson B.
        • Hughes C.E.
        • Harwood J.L.
        Dietary fatty acids and arthritis.
        Prostaglandins Leukot Essent Fatty Acids. 2010; 82: 315-318
        • Zainal Z.
        • Longman A.J.
        • Hurst S.
        • Duggan K.
        • Caterson B.
        • Hughes C.E.
        • et al.
        Relative efficacies of omega-3 polyunsaturated fatty acids in reducing expression of key proteins in a model system for studying osteoarthritis.
        Osteoarthritis Cartilage. 2009; 17: 896-905
        • Hogstrom M.
        • Nordstrom P.
        • Nordstrom A.
        n-3 fatty acids are positively associated with peak bone mineral density and bone accrual in healthy men: the NO2 study.
        Am J Clin Nutr. 2007; 85: 803-807
        • Watkins B.A.
        • Li Y.
        • Seifert M.F.
        Dietary ratio of n-6/n-3 PUFAs and docosahexaenoic acid: actions on bone mineral and serum biomarkers in ovariectomized rats.
        J Nutr Biochem. 2006; 17: 282-289
        • Muraoka T.
        • Hagino H.
        • Okano T.
        • Enokida M.
        • Teshima R.
        Role of subchondral bone in osteoarthritis development: a comparative study of two strains of guinea pigs with and without spontaneously occurring osteoarthritis.
        Arthritis Rheum. 2007; 56: 3366-3374
        • Young B.D.
        • Samii V.F.
        • Mattoon J.S.
        • Weisbrode S.E.
        • Bertone A.L.
        Subchondral bone density and cartilage degeneration patterns in osteoarthritic metacarpal condyles of horses.
        Am J Vet Res. 2007; 68: 841-849
        • Quasnichka H.L.
        • Anderson-MacKenzie J.M.
        • Bailey A.J.
        Subchondral bone and ligament changes precede cartilage degradation in guinea pig osteoarthritis.
        Biorheology. 2006; 43: 389-397
        • Bruyere O.
        • Dardenne C.
        • Lejeune E.
        • Zegels B.
        • Pahaut A.
        • Richy F.
        • et al.
        Subchondral tibial bone mineral density predicts future joint space narrowing at the medial femoro-tibial compartment in patients with knee osteoarthritis.
        Bone. 2003; 32: 541-545
        • Dequeker J.
        • Mokassa L.
        • Aerssens J.
        • Boonen S.
        Bone density and local growth factors in generalized osteoarthritis.
        Microsc Res Tech. 1997; 37: 358-371
        • Bendele A.M.
        • Hulman J.F.
        Spontaneous cartilage degeneration in guinea pigs.
        Arthritis Rheum. 1988; 31: 561-565
        • Jimenez P.A.
        • Glasson S.S.
        • Trubetskoy O.V.
        • Haimes H.B.
        Spontaneous osteoarthritis in Dunkin Hartley guinea pigs: histologic, radiologic, and biochemical changes.
        Lab Anim Sci. 1997; 47: 598-601
        • McDougall J.J.
        • Andruski B.
        • Schuelert N.
        • Hallgrimsson B.
        • Matyas J.R.
        Unravelling the relationship between age, nociception and joint destruction in naturally occurring osteoarthritis of Dunkin Hartley guinea pigs.
        Pain. 2009; 141: 222-232
        • Anderson-MacKenzie J.M.
        • Quasnichka H.L.
        • Starr R.L.
        • Lewis E.J.
        • Billingham M.E.
        • Bailey A.J.
        Fundamental subchondral bone changes in spontaneous knee osteoarthritis.
        Int J Biochem Cell Biol. 2005; 37: 224-236
        • Huebner J.L.
        • Hanes M.A.
        • Beekman B.
        • TeKoppele J.M.
        • Kraus V.B.
        A comparative analysis of bone and cartilage metabolism in two strains of guinea-pig with varying degrees of naturally occurring osteoarthritis.
        Osteoarthritis Cartilage. 2002; 10: 758-767
        • Pritzker K.P.
        • Gay S.
        • Jimenez S.A.
        • Ostergaard K.
        • Pelletier J.P.
        • Revell P.A.
        • et al.
        Osteoarthritis cartilage histopathology: grading and staging.
        Osteoarthritis Cartilage. 2006; 14: 13-29
        • Hollander A.P.
        • Heathfield T.F.
        • Webber C.
        • Iwata Y.
        • Bourne R.
        • Rorabeck C.
        • et al.
        Increased damage to type II collagen in osteoarthritic articular cartilage detected by a new immunoassay.
        J Clin Invest. 1994; 93: 1722-1732
        • Handley C.J.
        • Buttle D.J.
        Assay of proteoglycan degradation.
        Methods Enzymol. 1995; 248: 47-58
        • Tarlton J.F.
        • Vickery C.J.
        • Leaper D.J.
        • Bailey A.J.
        Postsurgical wound progression monitored by temporal changes in the expression of matrix metalloproteinase-9.
        Br J Dermatol. 1997; 137: 506-516
        • Knott L.
        • Tarlton J.F.
        • Bailey A.J.
        Chemistry of collagen cross-linking: biochemical changes in collagen during the partial mineralization of turkey leg tendon.
        Biochem J. 1997; 322: 535-542
        • Bannister D.W.
        • Burns A.B.
        Adaptation of the Bergman and Loxley technique for hydroxyproline determination to the autoanalyzer and its use in determining plasma hydroxyproline in the domestic fowl.
        Analyst. 1970; 95: 596-600
        • Avery N.C.
        • Sims T.J.
        • Bailey A.J.
        Quantitative determination of collagen cross-links.
        Methods Mol Biol. 2009; 522: 103-121
        • Tietz N.W.
        • Burtis C.A.
        • Duncan P.
        • Ervin K.
        • Petitclerc C.J.
        • Rinker A.D.
        • et al.
        A reference method for measurement of alkaline phosphatase activity in human serum.
        Clin Chem. 1983; 29: 751-761
        • Daly J.A.
        • Ertingshausen G.
        Direct method for determining inorganic phosphate in serum with the “CentrifiChem”.
        Clin Chem. 1972; 18: 263-265
        • Janssen J.W.
        • Helbing A.R.
        Arsenazo III: an improvement of the routine calcium determination in serum.
        Eur J Clin Chem Clin Biochem. 1991; 29: 197-201
        • Tidswell H.K.
        • Innes J.F.
        • Avery N.C.
        • Clegg P.D.
        • Barr A.R.
        • Vaughan-Thomas A.
        • et al.
        High-intensity exercise induces structural, compositional and metabolic changes in cuboidal bones – findings from an equine athlete model.
        Bone. 2008; 43: 724-733
        • Pokharna H.K.
        • Monnier V.
        • Boja B.
        • Moskowitz R.W.
        Lysyl oxidase and Maillard reaction-mediated crosslinks in aging and osteoarthritic rabbit cartilage.
        J Orthop Res. 1995; 13: 13-21
        • Bank R.A.
        • Verzijl N.
        • Lafeber F.P.
        • Tekoppele J.M.
        Putative role of lysyl hydroxylation and pyridinoline cross-linking during adolescence in the occurrence of osteoarthritis at old age.
        Osteoarthritis Cartilage. 2002; 10: 127-134
        • Clegg P.D.
        • Carter S.D.
        Matrix metalloproteinase-2 and -9 are activated in joint diseases.
        Equine Vet J. 1999; 31: 324-330
        • Masuhara K.
        • Nakai T.
        • Yamaguchi K.
        • Yamasaki S.
        • Sasaguri Y.
        Significant increases in serum and plasma concentrations of matrix metalloproteinases 3 and 9 in patients with rapidly destructive osteoarthritis of the hip.
        Arthritis Rheum. 2002; 46: 2625-2631
        • Kanyama M.
        • Kuboki T.
        • Kojima S.
        • Fujisawa T.
        • Hattori T.
        • Takigawa M.
        • et al.
        Matrix metalloproteinases and tissue inhibitors of metalloproteinases in synovial fluids of patients with temporomandibular joint osteoarthritis.
        J Orofac Pain. 2000; 14: 20-30
        • Hulejova H.
        • Baresova V.
        • Klezl Z.
        • Polanska M.
        • Adam M.
        • Senolt L.
        Increased level of cytokines and matrix metalloproteinases in osteoarthritic subchondral bone.
        Cytokine. 2007; 38: 151-156
        • Couchourel D.
        • Aubry I.
        • Delalandre A.
        • Lavigne M.
        • Martel-Pelletier J.
        • Pelletier J.P.
        • et al.
        Altered mineralization of human osteoarthritic osteoblasts is attributable to abnormal type I collagen production.
        Arthritis Rheum. 2009; 60: 1438-1450
        • Sanchez C.
        • Deberg M.A.
        • Bellahcene A.
        • Castronovo V.
        • Msika P.
        • Delcour J.P.
        • et al.
        Phenotypic characterization of osteoblasts from the sclerotic zones of osteoarthritic subchondral bone.
        Arthritis Rheum. 2008; 58: 442-455
        • Dieppe P.
        • Cushnaghan J.
        • Young P.
        • Kirwan J.
        Prediction of the progression of joint space narrowing in osteoarthritis of the knee by bone scintigraphy.
        Ann Rheum Dis. 1993; 52: 557-563
        • Mansell J.P.
        • Tarlton J.F.
        • Bailey A.J.
        Biochemical evidence for altered subchondral bone collagen metabolism in osteoarthritis of the hip.
        Br J Rheumatol. 1997; 36: 16-19
        • Mansell J.P.
        • Tarlton J.F.
        • Bailey A.J.
        Expression of gelatinases within the trabecular bone compartment of ovariectomized and parathyroidectomized adult female rats.
        Bone. 1997; 20: 533-538
        • Logar D.B.
        • Komadina R.
        • Prezelj J.
        • Ostanek B.
        • Trost Z.
        • Marc J.
        Expression of bone resorption genes in osteoarthritis and in osteoporosis.
        J Bone Miner Metab. 2007; 25: 219-225
        • Teti A.
        • Migliaccio S.
        • Taranta A.
        • Bernardini S.
        • De Rossi G.
        • Luciani M.
        • et al.
        Mechanisms of osteoclast dysfunction in human osteopetrosis: abnormal osteoclastogenesis and lack of osteoclast-specific adhesion structures.
        J Bone Miner Res. 1999; 14: 2107-2117
        • 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: 891-899
        • Kouri J.B.
        • Aguilera J.M.
        • Reyes J.
        • Lozoya K.A.
        • Gonzalez S.
        Apoptotic chondrocytes from osteoarthrotic human articular cartilage and abnormal calcification of subchondral bone.
        J Rheumatol. 2000; 27: 1005-1019
        • Sato M.
        • Wada M.
        • Miyoshi N.
        • Imamura Y.
        • Noriki S.
        • Uchida K.
        • et al.
        Hydroxyapatite maturity in the calcified cartilage and underlying subchondral bone of guinea pigs with spontaneous osteoarthritis: analysis by Fourier transform infrared microspectroscopy.
        Acta Histochemica Et Cytochemica. 2004; 37: 101-107
        • Chichocki T.
        • Gonsior B.
        • Hofert M.
        • Jarczyk L.
        • Raith B.
        • Rokita E.
        • et al.
        Measurements of mineralization process in the femur growth plate and rib cartilage of the mouse using pixe in combination with a proton microprobe.
        Histochemistry. 1988; 89: 99-104
        • Ali S.Y.
        • Rees J.A.
        • Scotchford C.A.
        Microcrystal deposition in cartilage and in osteoarthritis.
        Bone Miner. 1992; 17: 115-118
        • Ali S.Y.
        • Wisby A.
        • Evans L.
        • Craig-Gray J.
        The sequence of calcium and phosphorus accumulation by matrix vesicles.
        Calcif Tissue Res. 1977; : 490-493
        • Rosenbaum C.C.
        • O’Mathuna D.P.
        • Chavez M.
        • Shields K.
        Antioxidants and antiinflammatory dietary supplements for osteoarthritis and rheumatoid arthritis.
        Altern Ther Health Med. 2010; 16: 32-40
        • Sun D.
        • Krishnan A.
        • Zaman K.
        • Lawrence R.
        • Bhattacharya A.
        • Fernandes G.
        Dietary n-3 fatty acids decrease osteoclastogenesis and loss of bone mass in ovariectomized mice.
        J Bone Miner Res. 2003; 18: 1206-1216
        • Watkins B.A.
        • Li Y.
        • Lippman H.E.
        • Feng S.
        Modulatory effect of omega-3 polyunsaturated fatty acids on osteoblast function and bone metabolism.
        Prostaglandins Leukot Essent Fatty Acids. 2003; 68: 387-398
        • Watkins B.A.
        • Li Y.
        • Lippman H.E.
        • Seifert M.F.
        Omega-3 polyunsaturated fatty acids and skeletal health.
        Exp Biol Med (Maywood). 2001; 226: 485-497
        • Bailey A.J.
        • Mansell J.P.
        • Sims T.J.
        • Banse X.
        Biochemical and mechanical properties of subchondral bone in osteoarthritis.
        Biorheology. 2004; 41: 349-358
        • Massicotte F.
        • Lajeunesse D.
        • Benderdour M.
        • Pelletier J.P.
        • Hilal G.
        • Duval N.
        • et al.
        Can altered production of interleukin-1beta, interleukin-6, transforming growth factor-beta and prostaglandin E(2) by isolated human subchondral osteoblasts identify two subgroups of osteoarthritic patients.
        Osteoarthritis Cartilage. 2002; 10: 491-500