1 Introduction
Osteoarthritis (OA) is characterised by degeneration of articular cartilage, synovitis, remodelling of subchondral bone, atrophy/weakness of joint muscles and is often accompanied by musculoskeletal pain and joint stiffness (Hinton et al., 2002; Arden and Nevitt, 2005). Most extensively studied is knee OA where patients present with use-dependent pain, although 43% will also have pain at rest (Creamer et al., 1998). Intra-articular anaesthetic studies and pain resolution following total joint arthroplasty in hip and knee OA implicate a peripheral drive to pain in approximately 60–80% of patients depending on the affected joint (Creamer et al., 1996; Crawford et al., 1998; Ethgen et al., 2004).
The relationship of specific structural pathologies in OA with the presence of pain is complicated. In population studies substantial discordance exists between radiographically diagnosed OA and the presence of knee pain (Hannan et al., 2000). Other imaging modalities such as magnetic resonance imaging (MRI) studies have associated subchondral bone marrow lesions with knee pain, although lesions were also observed in some asymptomatic patients (Felson et al., 2001). Given the diversity of knee structures innervated by sensory afferents it is perhaps unsurprising that other MRI based studies have also reported a significant association of pain with other pathologies including, knee effusion, synovial thickening, and tibial cartilage volume (Hill et al., 2001; Wluka et al., 2004).
Collectively, these observations demonstrate the diversity of the OA population and indicate the utility of animal models in investigating the relationship of specific elements of structural pathology typical of OA to the presence or absence of pain. Intra-articular injection of mono-iodoacetate (MIA) into the femorotibial joint space in rats induces a linear pathology with similarities to OA (Van der Kraan et al., 1989; Clarke et al., 1997; Guingamp et al., 1997; Guzman et al., 2003). Histopathological features included chondrocytic cell death (Van der Kraan et al., 1989; Clarke et al., 1997; Guingamp et al., 1997), fragmentation of cartilage and exposure of subchondral bone by day 28 (Guingamp et al., 1997; Guzman et al., 2003).
To date, an investigation of the underlying mechanisms that provoke pain-related behaviours in the MIA model (Bove et al., 2003; Kobayashi et al., 2003; Combe et al., 2004) has not been explored. More specifically, whether nerve damage/neuropathy may underpin the lack of pharmacological sensitivity to celecoxib or diclofenac but not morphine observed at later time points (Fernihough et al., 2004; Pomonis et al., 2005).
To determine the potential of a neuropathic mechanism in the MIA model of OA, we measured the expression of activating transcription factor-3 (ATF-3), a marker of neuropathy (Peters et al., 2005; Sevcik et al., 2005), in the ipsilateral lumbar (L)4 and L5 dorsal root ganglion (DRG) at time points over a 35-day period and compared it to the development of joint histopathology. In addition, we characterised behavioural profiles of MIA-induced joint hypersensitivity following acute administration of four different classes of analgesics, including those typically used to treat neuropathic pain (gabapentin and amitriptyline) at early, intermediate and established stages of model progression.
2 Methods
2.1 Animals
All procedures were conducted in accordance with the UK Home Office (Animals Scientific Procedures Act 1986) after Local Ethical Review. Male Wistar rats (Harlan, UK), 185–205g (n=400) were group housed under a 12/12h light/dark cycle with access to a standard pellet diet and water ad libitum. Prior to initiation of studies animals were randomised and assigned to treatment groups. Two separate studies were conducted.
Firstly, a temporal profile of joint histopathology and ATF-3-immunofluorescence of ipsilateral L4 and L5 DRG following intra-articular MIA, or sterile saline, was performed. In both treatment groups n=4 animals were used at each time point, 8, 14, 21, 28 and 35 days post intra-articular injection. In total n=20 animals were used per treatment.
Secondly, in a separate set of animals, a time course of acute pharmacological sensitivity of MIA-treated animals to oral amitriptyline, gabapentin, naproxen, celecoxib or vehicle (0.5% hydroxypropylmethyl cellulose, 0.1% Tween) was generated. Three dose groups of each drug were used (n=6 per group) and compared to vehicle (n=6). Separate animals were used for all treatments at each time point (14, 21 and 28 days post-MIA), as acute intervention was found to alter subsequent histopathological development of the model (unpublished observations). Intra-articular controls administered intra-articular saline (n=6 per group, per time point) were also included to confirm a significant MIA-induced weight-bearing response. In total n=90 animals were used per treatment.
2.2 Mono-iodoacetate injection
Animals were prepared for intra-articular injection of MIA (Sigma, UK) by brief anaesthesia with isoflurane (3% in O2 at 1.5l/min). Intra-articular injection of 1mg MIA in 20μl of 0.9% sterile saline or negative control (20μl of 0.9% sterile saline alone) was administered through the infra-patellar ligament into the joint space of the right knee via a Hamilton gas tight syringe (Hamilton, Switzerland). The dose of MIA was determined based on previous literature (Bove et al., 2003) and in-house data.
2.3 Behavioural testing
Weight re-distribution as measured by an incapacitance meter (Linton Instruments, UK) is assumed to be a behavioural measure of musculoskeletal discomfort (Bove et al., 2003; Fernihough et al., 2004). Briefly, animals were placed into a Perspex chamber over ipsilateral and contralateral hind limb force transducer panels and given adequate time to settle. Once settled, facing forward and with both hind paws on the appropriate panel, a reading was taken. A digital readout of the mean weight exerted on each panel, over a period of 5s, was taken for each animal and measured in grams. In the current studies, all animals were habituated to the apparatus prior to initiation of the experimental period.
2.3.1 Behavioural measurements for joint histopathology and ATF-3-immunofluorescence time course
Weight-bearing measurements were taken immediately before administration of intra-articular MIA or negative control (20μl of 0.9% sterile saline) and at days 8, 14, 21, 28 and 35 by averaging over three separate readings within a given test session. Data are analysed as the percentage difference in weight distribution, (ipsilateral/contralateral value)×100. A value of 100% represented equal weight-bearing across ipsilateral and contralateral hind limbs. Data are presented as arithmetic means±standard error of the mean (SEM). Significant changes in weight-bearing asymmetry were assessed on raw data using one-way ANOVA followed by Bonferroni's post-hoc test against time-matched controls. A P value of <0.05 was set as the level of statistical significance.
2.3.2 Behavioural measurements for acute pharmacological assessment
Baseline weight-bearing measurements for pharmacological testing were taken prior to dosing on respective test days. Following pharmacological treatment animals were assessed for weight distribution at 1, 2 and 4h post-dosing on days 14, 21 and 28. Data are analysed and presented as described in Section 2.3.1.
2.4 Acute pharmacological assessment
On days 14, 21 and 28, animals were orally administered single doses of drug or vehicle (5ml/kg p.o.) (0.5% hydroxypropylmethyl cellulose, 0.1% Tween). Naproxen was dosed at 0.3, 3 and 10mg/kg p.o. (Sigma, USA) based on literature data of pharmacokinetic and pharmacodynamic modelling of anti-inflammatory effects in rat (Josa et al., 2001). Celecoxib was dosed at 1, 3 and 10mg/kg p.o. (Apin Chemicals Ltd., UK) based on extensive studies in animal models, in particular analgesic and anti-inflammatory activity in models of arthritis in the rat (Penning et al., 1997; Noguchi et al., 2005). Amitriptyline (3, 10 and 30mg/kg p.o.) (Sigma, USA) and gabapentin (10, 30 and 100mg/kg p.o.) (Apin Chemicals Ltd., UK) were dosed based on the literature of oral activity against mechanical hyperalgesia in rat models of neuropathy (Yasuda et al., 2005).
2.5 Tissue preparation – joint histopathology and DRG ATF-3-immunofluorescence
Following weight-bearing on days 8, 14, 21, 28 and 35, animals from both treatment groups (1mg MIA in 20μl of 0.9% sterile saline or negative control 20μl 0.9% sterile saline alone) were deeply anaesthetized with intra-peritoneal sodium pentobarbital (∼100mg) (Vericone Ltd., UK). Animals were perfused transcardially with 200ml heparinized saline (10IU/ml) (CP Pharmaceuticals, UK) and 250ml ice-cold fresh paraformaldehyde (PFA; 4%) (Sigma, UK) in 0.1M phosphate-buffered saline (PBS).
MIA and vehicle-treated limbs were processed for histopathological assessment as follows; limbs were taken whole and then cut mid-femur and mid-tibia. Excess skin and muscle tissue was removed and samples fixed for approximately 1 week in 10% neutral buffered formalin. They were then decalcified for 36h (changed every 2h for first 12h) in Surgipath Decalcifier II (Surgipath Europe Ltd., Venture Park, Stirling Way, Bretton, Peterborough, UK). The patella of each femorotibial joint was trimmed so that the samples could be orientated in cassettes for preparation by standard histological techniques of wax blocks for subsequent coronal (dorso-ventral) sectioning. The whole of each sample was serially sectioned. Sections at steps of 450μm were stained with haematoxylin and eosin and assessed by light microscopy.
Ipsilateral L4 and L5 DRG were extracted, post-fixed in 4% PFA overnight and cryoprotected in 20% sucrose solution for another 24h. Single ganglia were then embedded and orientated (in the plane perpendicular to the long axis of the ganglia) in tissue Tec, frozen on dry ice and stored at −80°C until sectioning.
ATF-3 expression was determined using immunofluorescence. DRG were serially sectioned (14μm) over 15 Superfrost gelatine-coated slides using a Leica 3050 cryostat. Every fourth slide for each DRG was washed in PBS and PBS with 0.2% Triton X-100. After 30-min incubation in 2% donkey serum (in PBS with 0.2% Triton X-100) at room temperature, sections were incubated in rabbit anti-ATF-3 primary antibody (1:500; Santa Cruz Biotechnologies, US) overnight at 4°C. Following three washes in PBS, sections were incubated in an Alexa Fluor 488 secondary donkey anti-rabbit antibody (1:200; Molecular Probes, US). After three further washes, sections were mounted with a fluorescent mounting reagent (DAKO, US). Tissue sections were viewed using a fluorescence microscope (Leica, UK). Digital images were taken of ATF-3-labelled cell nuclei and the total number of ATF-3-labelled cell nuclei counted.
Statistical analysis for ATF-3-positive cell counts was carried out on total count data using a Kruskal–Wallis and Mann–Whitney U-test. Populations of total counts in 1mg MIA-treated animals in each DRG were compared with time-matched negative controls (20μl of 0.9% sterile saline alone). A P value of <0.05 was set as the level of statistical significance for all tests.
3 Results
3.1 General
No differences in body weight gain were noted between MIA-injected animals and age-matched control groups at any time point during the studies.
3.2 Temporal profile of histopathology and DRG ATF-3-immunofluorescence
Intra-articular injection of MIA significantly decreased ipsilateral weight-bearing at every time point compared to saline controls, day 8 (P<0.001, ANOVA, post-hoc Bonferroni's, n=4 per group per time point; Fig. 1) and days 14 (P<0.01), 21 (P<0.01), 28 (P<0.01) and 35 (P<0.01). By day 8 ipsilateral weight-bearing was reduced to 50±6%. On days 14, 21, 28 and 35 weight-bearing showed marginal improvement (59±5%, 66±11%, 55±10% and 66±9%, respectively) but remained significantly different from saline controls.
Fig. 1:
Behavioural profile of joint hypersensitivity over a 35-day time course. Intra-articular administration of MIA (1 mg in 20 μl) into the rat knee joint produced a significant decrease in ipsilateral weight-bearing compared to saline controls, sustained over 35 days. Significance was determined using one-way ANOVA, post-hoc Bonferroni's ***P < 0.001, **P < 0.01 compared to saline controls. Data presented as arithmetic means ± SEM (n = 4 per group, per time point).
At all time points saline controls exhibited normal articular cartilage and subchondral bone histology. In brief, articular cartilage displayed regularly arranged chondrocytes forming small clusters and cords overlying an intact, subchondral bone plate (Fig. 2). The bone in this area was lined by numerous osteoblasts and occasional osteoclasts. The cancellous bone of the epiphysis was generally interspaced by marrow cavity filled with haematopoietic cells and other normal components including occasional fat cells.
Fig. 2:
Histological appearance of the epiphysis following intra-articular injection of 1 mg MIA (f, femoral condyle; t, tibial condyle). (a) control 8 days – the condyles of the femur and tibia show normal articular cartilage (asterisk) overlying the subchondral bone plate and cancellous bone of the epiphysis (arrowhead) which is interspaced by marrow cavity (arrow). Inset – lower power view of lateral tibial and femoral condyles showing articular surfaces (arrowhead) and epiphyseal growth plates (arrow) – 10× magnification. (b) MIA injected; 8 days – pallor of the intact articular cartilage of the femoral and tibial condyles (asterisks) with ghosting of chondrocytes and degeneration of the underlying subchondral bone plates (arrowheads). In the marrow cavity there is fibrous tissue replacing the haematopoietic marrow with occasional fragments of degenerate cartilage (arrow). (c) MIA injected; 12 days – erosion of the articular cartilage (asterisk) with marked subchondral bone degeneration and loss (arrowhead), and extensive replacement of marrow by fibrous tissue (arrow). (d) MIA injected; 14 days – pronounced ulceration of the articular cartilage and superficial subchondral bone (asterisk) with ongoing degeneration, resorption and remodelling of deeper cancellous epiphyseal bone (arrowheads). (e) MIA injected; 21 days – pronounced loss of articular cartilage and subchondral bone (asterisk) with ongoing resorption and remodelling of remaining cancellous bone of the epiphysis (arrowheads) down to the epiphyseal growth plate cartilage. Fibrous tissue is quite dense (arrow). (f) MIA injected; 28 days – profound fragmentation and loss of epiphyseal cartilage and bone tissues (arrowhead) focally down to the epiphyseal growth plate (asterisk). No 35-day image is shown, as histopathology after day 28 remained relatively consistent. Scale bar, 400 μm.
In contrast MIA-treated animals showed a time-dependent development of histopathology. At 8 days there was a diffuse pallor of the generally intact articular cartilage of the femoral and tibial condyles consistent with loss of proteoglycan seen on equivalent tissue stained with toluidine blue. Ghosting of chondrocytes indicative of chondrocyte necrosis was also apparent (Fig. 2). The underlying subchondral bone plates were degenerated with areas of bone loss/resorption and associated activation of osteoclasts/chondroclasts. Proliferating spindle cells had formed a fine fibrous stroma replacing the haematopoietic marrow in these areas. Also, the inner surface of the overlying articular cartilage displayed areas of fragmentation, and the slightly deeper epiphyseal marrow space contained occasional fragments of degenerate cartilage accompanied by superficial accumulation of activated osteoblasts. Synovial membranes were hyperplastic and accompanied by mononuclear inflammatory cell infiltrates in synovial/subsynovial tissues and intra-articular ligaments.
By 14 days, focal areas of pronounced ulceration of the articular cartilage and superficial subchondral bone had developed. Along with ongoing degeneration, resorption and remodelling of deeper cancellous epiphyseal bone. Only slight residual fibroplasia and inflammatory cell infiltration was apparent in the synovial and ligament tissues.
Across the remaining time points (21, 28 and 35 days) pronounced widespread loss of articular cartilage and subchondral bone continued with resorption and remodelling of the remaining cancellous bone of the epiphysis down to the epiphyseal growth plate. The epiphyseal changes resulting from intra-articular injection of MIA showed no site predisposition in location or severity, affecting both femoral and tibial condyles. The progression of the epiphyseal changes was also consistent between the condyles with superficial (to the articular surface) areas affected first, generally diffusely, followed by an involvement of progressively deeper tissues. Extensive replacement of the haematopoietic marrow by dense fibrous stroma was also observed, containing numerous osteoclasts, and at the bone surface numerous active osteoblasts.
In corresponding animals, ATF-3-IR was identifiable in cell nuclei of L4 and L5 DRG sections from MIA animals at all time points during the 35-day duration (Fig. 3a and b). Baseline ATF-3-labelling was identified in DRG sections from saline control animals (Fig. 3c). No ATF-3-labelling was observed following negative control studies, whereby application of the primary antibody was performed in the absence of the secondary antibody and vise versa (Fig. 3d). Following intra-articular MIA injection a significant increase in ATF-3-IR was measured in the L5 DRG on days 8 and 14 (P<0.05, Kruskal–Wallis and Mann–Whitney U-test, n=4 per group), compared with L5 DRG from saline controls (Fig. 4). No significant difference in ATF-3-IR was measured in L5 DRG between groups on days 21, 28 and 35 (P<0.05, Kruskal–Wallis and Mann–Whitney U-test, n=4 per group). ATF-3-IR was unaltered in L4 DRG between groups at any time point during the study.
Fig. 3:
(a and b) Illustrative ATF-3-IR cell nuclei in DRG sections from the same animal 14-days post intra-articular MIA injection, (a) L4 DRG section (b) L5 DRG section. (c) Illustrative L5 DRG section from a saline control animal, 7-days post intra-articular injection, demonstrating baseline ATF-3-labelling. (d) Illustrative L5 DRG section, 7-days post intra-articular injection of MIA, with histology performed in the absence of primary antibody, demonstrating background staining of secondary antibody. Arrows indicate ATF-3-IR cell nuclei. Scale bar, 100 μm.
Fig. 4:
Profile of ATF-3-IR in the L4 and L5 DRG. Following intra-articular administration of MIA, a significant increase in ATF-3-IR was measured in the L5 DRG on days 8 and 14 compared to saline controls. No significant difference in ATF-3-IR was measured in the L5 DRG between groups on days 21, 28 and 35. No significant difference in ATF-3-IR was measured in the L4 DRG between groups at any time point. Significance was determined using a Kruskal–Wallis and Mann–Whitney U-test, *P < 0.05 as compared to saline controls. Data presented as total positive cell counts (n = 4 per group, per time point).
3.3 Acute pharmacological assessment
Acute oral doses of naproxen (0.3, 3 and 10mg/kg p.o.), celecoxib (1, 3 and 10mg/kg p.o.), amitriptyline (3, 10 and 30mg/kg p.o.), gabapentin (10, 30 and 100mg/kg p.o.) or vehicle were administered on days 14, 21 and 28. Behavioural measurements were taken 1, 2 and 4h post-dose to determine any acute resolution of joint hypersensitivity. In general, resolution of weight-bearing asymmetry at all time points was maximal in amitriptyline and gabapentin-treated animals. Efficacy of both amitriptyline and gabapentin was sustained throughout the experimental protocol. All doses of amitriptyline (3, 10 and 30mg/kg p.o.) reached significance (P<0.05, ANOVA, post-hoc Bonferroni's, n=6 per group) versus vehicle controls at all time points, with the exception of the 3mg/kg group on day 28 (Fig. 5a–c). Administration of gabapentin at all three doses (10, 30 and 100mg/kg p.o.) also produced a dose-related and significant (P<0.05 ANOVA, post-hoc Bonferroni's, n=6 per group) resolution of joint hypersensitivity at all time points (Fig. 6a–c).
Fig. 5:
Acute pharmacological sensitivity to amitriptyline (3, 10 and 30 mg/kg po) compared to vehicle (5 ml/kg p.o.) (0.5% hydroxypropylmethyl cellulose, 0.1% Tween), on days 14- (a), 21- (b) and 28- (c) post-MIA. Data are analysed as percentage difference in weight distribution, (ipsilateral/contralateral value) × 100 and presented as arithmetic means ± standard error of the mean (n = 6 per group). Controls administered intra-articular saline (20 μl) are included to confirm a significant MIA-induced weight-bearing response. Significant changes in weight-bearing asymmetry were assessed on raw data using one-way ANOVA, post-hoc Bonferroni's against time-matched controls. *P < 0.05, **P < 0.01 and ***P < 0.001.
Fig. 6:
Acute pharmacological sensitivity to gabapentin (10, 30 and 100 mg/kg p.o.) compared to vehicle on days 14- (a), 21- (b) and 28- (c) post-MIA. Data are analysed and presented as for
Fig. 5. *
P < 0.05, **
P < 0.01 and ***
P < 0.001.
Naproxen demonstrated efficacy at the highest dose (10mg/kg p.o.) on days 14 and 28 (P<0.01; Fig. 7a–c). In addition, the intermediate and low doses of naproxen (3 and 0.3mg/kg p.o., respectively) exhibited a small but significant effect 1h post-dose (P<0.05 ANOVA, post-hoc Bonferroni's, n=6 per group), on days 14 and 28. In contrast, celecoxib showed no significant effect on weight-bearing at any time point (P>0.05, ANOVA, post-hoc Bonferroni's, n=6 per group) (Fig. 8a–c).
Fig. 7:
Acute pharmacological sensitivity to naproxen (0.3, 3 and 10 mg/kg p.o.) compared to vehicle on days 14- (a), 21- (b) and 28- (c) post-MIA. Data are analysed and presented as for
Fig. 5. *
P < 0.05, **
P < 0.01 and ***
P < 0.001.
Acute pharmacological sensitivity to celecoxib (1, 3 and 10 mg/kg p.o.) compared to vehicle on days 14- (a), 21- (b) and 28- (c) post-MIA. Data are analysed and presented as for
Fig. 5.
4 Discussion
In the current study, intra-articular injection of MIA induced a decrease in ipsilateral hind limb weight-bearing, which was sustained over a 35-day duration. A significant increase in ATF-3-IR was measured in the L5 DRG of MIA-injected animals between days 8 and 14. Tsujino et al. (2000) have described the utility of ATF-3 as a selective marker of cell damage following axotomy in sensory and motor neurones, which is not expressed following induction of inflammation in the skin. Therefore in the current study, increased ATF-3-IR in the L5 DRG is presumed to be indicative of neuronal damage and implies that intra-articular MIA is associated with an early phase neuropathy. Clearly, there is a temporal dissociation between peak ATF-3-IR (day 14) and the sustained weight-bearing asymmetry observed throughout the remainder of the experimental protocol. This observation is not unusual. Several groups have noted a lack of correlation between ATF-3 expression and pain behaviour in models of nerve injury (Obata et al., 2003; Shortland et al., 2006). This disconnect may be due to the complex downstream network of both positive and negative regulation of genes that ATF-3 will transcriptionally modulate. Regulation of genes such as growth arrest and DNA damage gene 153 (gadd153/Chop10), heat shock protein 27 and gadd45, which are thought to play a role in neuronal regeneration and survival, has been proposed to underpin neuronal phenotype changes and pain behaviour rather than ATF-3 per se (for review, see Hai and Hartman, 2001; Shortland et al., 2006). Whilst this is an attractive theory, systems biology approaches to ATF-3 transcriptional networks are identifying a growing list of downstream gene events in other cell types, although further studies are needed to ascertain whether these genes are also regulated in neurones (Gilchrist et al., 2006).
It is probable that the increase in ATF-3-IR in the L5 DRG but not in the L4 DRG reflects an anatomically specific origin of neuropathy and an association with pathology of the joint. Retrograde axonal tracing studies in the rat have indicated that the majority of innervating cells of the knee joint reside in L3–L5, with some labelling of Thoracic 12, L1, L2 and upper sacral DRG (Widenfalk and Wiberg, 1989). Somatotopic organisation of primary afferent terminals in the superficial laminae of rat DRG, has only been described for cutaneous projections of articular nerves (Swett and Woolf, 1985). With this caveat in mind, L4 DRG are primarily associated with peroneal and tibial nerve projections and L5 DRG predominantly, tibial nerve afferents. Articular neuroanatomy of the rat knee is poorly defined, however studies of other species indicate that the posterior articular nerve, which innervates postero-medial and lateral compartments of the knee, is the principal tibial nerve subdivision with projections to the knee joint (Freeman and Wyke, 1967). Histopathological analysis of this model indicates diffuse damage in both medial and lateral compartments (current study, Guingamp et al., 1997; Janusz et al., 2001) although some authors describe a medial focus to the lesion (Bove et al., 2003; Fernihough et al., 2004). Therefore, it is likely that the observed ATF-3-IR in L5 DRG may be due to nerve damage in distal processes of the posterior articular nerve in the medial and/or lateral compartment of the knee, the cell bodies of which may reside in L5 not L4 DRG in the rat.
Potential mechanisms of nerve injury may be identified in the histopathology of the model. In this study (day 8) and others (day 7), subchondral bone and bone marrow pathology is associated with increased osteoclastic activity and large reactive osteoblasts in adjacent trabeculae (Guzman et al., 2003; Morenko et al., 2004). By day 14 sclerosis of the tibial subchondral plateau is observed and extensive replacement of bone marrow beneath damaged cartilage. Bone is densely innervated with Aδ and C-fibres in regions of maximal load and greatest bone turnover such as proximal and distal heads as well as marrow and periosteum (Mach et al., 2002). Therefore, osteoclast-induced injury and mechanical compression may be priority mechanisms for nerve damage. In this respect, similar mechanisms have been proposed to account for bone cancer pain (Thurlimann and Stoutz, 1996; Luger et al., 2001; Mantyh and Hunt, 2004) where modification of osteoclast maturation and activity results in diminished movement evoked pain behaviour (Luger et al., 2001).
Acute, pharmacological, characterisation of analgesics in this model has previously been reported versus several behavioural endpoints. In general, early stages of the model (day 14) are characterised by sensitivity to non-steroidal anti-inflammatory drugs (NSAIDs) such as cyclo-oxygenase-2 (COX-2) selective and non-selective drugs (Bove et al., 2003). However, as the pathology progresses, Pomonis et al. (2005) have demonstrated that at day 21 the model is not acutely sensitive to COX-2 inhibition, although morphine does show dose-dependent inhibition of weight-bearing asymmetry. Similarly, Fernihough et al. (2004) have reported positive effects of NSAIDs at day 3, but not at day 14 and day 28, versus mechanical hyperalgesia. Broadly our results are consistent with these previous findings. NSAID efficacy was greatest at day 14 although relatively weak compared to amitriptyline and gabapentin throughout the time course. The time-point at which this model loses NSAID efficacy may depend on dose of MIA used. Temporal evolution and severity of histopathology is highly dose-dependent (Guingamp et al., 1997). Pharmacology studies of Bove et al. (2003); Pomonis et al. (2005) and the current study employ 1mg MIA, whereas Fernihough et al. (2004) used 2mg MIA. Therefore an accelerated pathology may underpin a lack of sensitivity to NSAIDs at day 14 in the 2mg MIA model when compared to Bove et al. (2003) and the results of this study.
Throughout the time course of the present study gabapentin and amitriptyline exhibited a robust and significant reversal of weight-bearing asymmetry. Gabapentin a ligand of α2δ calcium channel subunit and the tricyclic antidepressant amitriptyline, which acts to non-selectively inhibit the reuptake of 5-HT and noradrenaline, are effective in reducing experimental and clinical neuropathic pain (Rosner et al., 1996; Xiao and Bennett, 1996; Bomholt et al., 2005). Of relevance to this study, gabapentin has been demonstrated to have both central and peripheral sites of action. Spinal administration of gabapentin reduces nociceptive behaviour in knee joint inflammation, whilst peripheral exposure is associated with inhibition of spontaneous firing in injured nerves and reduced mechano-sensitivity of knee afferents (Chapman et al., 1998; Lu and Westlund, 1999; Hanesch et al., 2003). The observed efficacy of gabapentin in the current study may be a summation of both central and peripheral effects. Nerve injury observed as an increase in ATF-3-IR in L5 DRG at days 8 and 14 may result in spontaneous nerve firing and/or central sensitisation, which is modulated by gabapentin. Alternatively, gabapentin may directly reduce the mechanosensitivity of primary afferent nerves of the knee, normalising weight-bearing asymmetry through an, as yet, unknown mechanism (Hanesch et al., 2003).
In contrast, spinal cord modulation of nociceptive transmission is likely to be the predominant action of amitriptyline via descending serotonergic and noradrenergic pathways of the rostral ventromedial medulla and dorsolateral pontomesencephalic tegmentum. In the rat formalin test, efficacy observed with local administration of amitriptyline dosed with formalin has been cited as supporting a peripheral mode of action (Sawynok et al., 1999). Although concentrations at sensory afferent terminals during localised administration may exceed that achieved systemically (Sawynok et al., 1999). Of direct relevance to this study, amitriptyline has been demonstrated to reduce mechanical and thermal hyperalgesia after nerve injury (Bomholt et al., 2005).
In conclusion, this study demonstrates a role for nerve injury in the pain-related behaviour in a putative model of OA. The absence of inflammatory cell infiltrates after day 8 and lack of efficacy of NSAIDs beyond initial phases of the model would not support a role of inflammation in maintaining pain behaviour beyond this time point. Extrapolating these findings to patients with OA will depend on the face validity of the MIA model and in particular the bone pathology that may initiate the nerve damage observed in the current study.
References
Arden N, Nevitt MC. Osteoarthritis: epidemiology.
Best Pract Res Clin Rheumatol. 2005;20:3-25.
Bomholt SF, Mikkelsen JD, Blackburn-Munro G. Antinociceptive effects of the antidepressants amitriptyline, duloxetine, mirtazapine and citalopram in animal models of acute, persistent and neuropathic pain.
Neuropharmacology. 2005;48:252-263.
Bove SE, Calcaterra SL, Brooker RM, Huber CM, Guzman RE, Juneau PL, et al. Weight bearing as a measure of disease progression and efficacy of anti-inflammatory compounds in a model of monosodium iodoacetate-induced osteoarthritis.
Osteoarthritis Cartilage. 2003;11:821-830.
Chapman V, Suzuki R, Chamarette HLC, Rygh LJ, Dickenson AH. Effects of systemic carbamazepine and gabapentin on spinal neuronal responses in spinal nerve ligated rats.
Pain. 1998;75:261-272.
Clarke KA, Heitmeyer SA, Smith AG, Taiwo YO. Gait analysis in a rat model of osteoarthritis.
Physiol Behav. 1997;62:951-954.
Combe R, Bramwell S, Field MJ. The monosodium iodoacetate model of osteoarthritis: a model of chronic nociceptive pain in rats?
Neurosci Lett. 2004;370:236-240.
Crawford RW, Gie GA, Ling RS, Murray DW. Diagnostic value of intra-articular anaesthetic in primary osteoarthritis of the hip.
J Bone Joint Surg Br. 1998;80:279-281.
Creamer P, Hunt M, Dieppe P. Pain mechanisms in osteoarthritis of the knee: effect of intraarticular anesthetic.
J Rheumatol. 1996;23:1031-1036.
Creamer P, Lethbridge-Cejku M, Hochberg MC. Where does it hurt? Pain localization in osteoarthritis of the knee.
Osteoarthritis Cartilage. 1998;6:318-323.
Ethgen O, Bruyere O, Richy F, Dardennes C, Reginster JY. Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature.
J Bone Joint Surg Am. 2004;86:963-974.
Felson DT, Chaisson CE, Hill CL, Totterman SM, Gale ME, Skinner KM, et al. The association of bone marrow lesions with pain in knee osteoarthritis.
Ann Intern Med. 2001;134:541-549.
Fernihough J, Gentry C, Malcangio M, Fox A, Rediske J, Pellas T, et al. Pain related behaviour in two models of osteoarthritis in the rat knee.
Pain. 2004;112:83-93.
Freeman MAR, Wyke B. The innervation of the knee joint An anatomical and histological study in the cat.
J Anat. 1967;101:505-532.
Gilchrist M, Thorsson V, Li B, Rust AG, Korb M, Kennedy K, et al. Systems biology approaches identify ATF3 as a negative regulator of Toll-like receptor 4.
Nature. 2006;441:173-178.
Guingamp C, Gegout-Pottie P, Philippe L, Terlain B, Netter P, Gillet P. Mono-iodoacetate-induced experimental osteoarthritis: a dose–response study of loss of mobility, morphology, and biochemistry.
Arthritis Rheum. 1997;40:1670-1679.
Guzman RE, Evans MG, Bove S, Morenko B, Kilgore K. Mono-iodoacetate-induced histologic changes in subchondral bone and articular cartilage of rat femorotibial joints: An animal model of osteoarthritis.
Toxicol Pathol. 2003;31:619-624.
Hai T, Hartman MG. The molecular biology and nomenclature of the activating transcription factor/cAMP responsive element binding family of transcription factors: activating transcription factor proteins and homeostasis.
Gene. 2001;273:1-11.
Hanesch U, Pawlak M, McDougall JJ. Gabapentin reduces the mechanosensitivity of fine afferent nerve fibres in normal and inflamed rat knee joints.
Pain. 2003;104:363-366.
Hannan MT, Felson DT, Pincus T. Analysis of the discordance between radiographic changes and knee pain in osteoarthritis of the knee.
J Rheumatol. 2000;27:1513-1517.
Hill CL, Gale DG, Chaisson CE, Skinner K, Kazis L, Gale ME, et al. Knee effusions, popliteal cysts, and synovial thickening: association with knee pain in osteoarthritis.
J Rheumatol. 2001;28:1330-1337.
Hinton R, Moody RL, Davis AW, Thomas SF. Osteoarthritis: diagnosis and therapeutic considerations.
Am Fam Physician. 2002;65:841-848.
Janusz MJ, Hookin EB, Heitmeyer SA, Woessner JF, Freemont AJ, Hoyland JA, et al. Moderation of iodoacetate-induced experimental osteoarthritis in rats by matrix metalloproteinase inhibitors.
Osteoarthritis Cartilage. 2001;9:751-760.
Josa M, Urizar JP, Rapado J, Dios-Vieitez C, Castaneda-Hernandez G, Flores-Murrieta F, et al. Pharmacokinetic/pharmacodynamic modeling of antipyretic and anti-inflammatory effects of naproxen in the rat.
J Pharmacol Exp Ther. 2001;297:198-205.
Kobayashi K, Imaizumi R, Sumichika H, Tanaka H, Goda M, Fukunari A, Komatsu H. Sodium iodoacetate-induced experimental osteoarthritis and associated pain model in rats.
J Vet Med Sci. 2003;65:1195-1199.
Lu Y, Westlund KN. Gabapentin attenuates nociceptive behaviours in acute arthritis model in rats.
J Pharmacol Exp Ther. 1999;290:214-219.
Luger NM, Honore P, Sabino MA, Schwei MJ, Rogers SD, Mach DB, et al. Osteoprotegerin diminishes advanced bone cancer pain.
Cancer Res. 2001;61:4038-4047.
Mach DB, Rogers SD, Sabino MC, Luger NM, Schwei MJ, Pomonis JD, et al. Origins of skeletal pain: sensory and sympathetic innervation of the mouse femur.
Neuroscience. 2002;113:115-166.
Mantyh PW, Hunt SP. Mechanisms that generate and maintain bone cancer pain.
Novartis Found Symp. 2004;260:221-238. discussion 238–40, 277–279.
Morenko BJ, Bove SE, Chen L, Guzman RE, Juneau P, Bocan TM, et al. In vivo micro computed tomography of subchondral bone in the rat after intra-articular administration of monosodium iodoacetate.
Contemp Top Lab Anim Sci. 2004;43:39-43.
Noguchi M, Kimoto A, Kobayashi S, Yoshino T, Miyata K, Sasamata M. Effect of celecoxib, a cyclo-oxygenase-2 inhibitor, on the pathophysiology of adjuvant arthritis in rat.
Eur J Pharmacol. 2005;513:229-235.
Obata K, Yamanaka H, Fukuoka T, Yi D, Tokunaga A, Hashimoto N, et al. Contribution of injured and uninjured dorsal root ganglion neurons to pain behaviour and the changes in gene expression following chronic constriction injury of the sciatic nerve in rats.
Pain. 2003;101:65-77.
Penning TD, Talley JJ, Bertenshaw SR, Carter JS, Collins PW, Docter S, et al. Synthesis and biological evaluation of the 1,5-diarylpyrazole class of cyclo-oxygenase-2 inhibitors: identification of 4-[5-(4-methylphenyl)-3-(trifluoromethyl)-1
H-pyrazol-1-yl]benzenesulfonamide (SC-58635, celecoxib).
J Med Chem. 1997;40:1347-1365.
Peters CM, Ghilardi JR, Keyser CP, Kubota K, Lindsay TH, Luger NM, et al. Tumor-induced injury of primary sensory nerve fibers in bone cancer pain.
Exp Neurol. 2005;193:85-100.
Pomonis JDM, Boulet JM, Gottshall SL, Phillips S, Rani S, Bunton T, et al. Development and pharmacological characterisation of a rat model of osteoarthritis pain.
Pain. 2005;114:339-346.
Rosner H, Rubin L, Kestenbaum. Gabapentin adjuctive therapy in neuropathic pain states.
Clin J Pain. 1996;2:56-58.
Sawynok J, Reid A, Esser MJ. Peripheral antinoceptive action of amitriptyline in the rat formalin test: involvement of adenosine.
Pain. 1999;80:45-55.
Sevcik MA, Ghilardi JR, Peters CM, Lindsay TH, Halvorson KG, Jonas BM, et al. Anti-NGF therapy profoundly reduces bone cancer pain and the accompanying increase in markers of peripheral and central sensitization.
Pain. 2005;115:128-141.
Shortland PJ, Baytug B, Krzyzanowska A, McMahon SB, Priestley JV, Averill S. ATF3 expression in L4 dorsal root ganglion neurones after L5 spinal nerve transaction.
Eur J Neurosci. 2006;23:365-373.
Swett JE, Woolf CJ. The somatotopic organisation of the primary afferent terminals in the superficial laminae of the dorsal horn of the rat spinal cord.
J Comp Neurol. 1985;231:66-77.
Thurlimann B, Stoutz ND. Causes and treatment of bone pain of malignant origin.
Drugs. 1996;51:383-398.
Tsujino H, Kondo E, Fukuoka T, Dai Y, Tokunaga A, Miki K, et al. Activating transcription factor 3 (ATF3) induction by axotomy in sensory and motoneurons: a novel neuronal marker of nerve injury.
Mol Cell Neurosci. 2000;15:170-182.
Van der Kraan PM, Vitters EL, van de Putte LBA, van den Berg WB. Development of osteoarthritic lesions in mice by “metabolic” and mechanical” alterations in the knee joints.
Am J Pathol. 1989;135:1001-1014.
Widenfalk B, Wiberg M. Origin of sympathetic and sensory innervation of the knee joint.
Anat Embryol. 1989;180:317-323.
Wluka AE, Wolfe R, Stuckey R, Cicuttini FM. How does tibial cartilage volume relate to symptoms in subjects with knee osteoarthritis?
Ann Rheum Dis. 2004;63:264-268.
Xiao WH, Bennett GL. Gabapentin has an anti-nociceptive effect mediated via a spinal site of action in a rat model of painful peripheral neuropathy.
Analgesia. 1996;2:267-273.
Yasuda T, Miki S, Yoshinaga N, Senba E. Effects of amitriptyline and gabapentin on bilateral hyperalgesia observed in an animal model of unilateral axotomy.
Pain. 2005;155:161-170.
