Characterisation of a peripheral neuropathic component of the rat monoiodoacetate model of osteoarthritis

Abstract

Joint degeneration observed in the rat monoiodoacetate (MIA) model of osteoarthritis shares many histological features with the clinical condition. The accompanying pain phenotype has seen the model widely used to investigate the pathophysiology of osteoarthritis pain, and for preclinical screening of analgesic compounds. We have investigated the pathophysiological sequellae of MIA used at low (1 mg) or high (2 mg) dose. Intra-articular 2 mg MIA induced expression of ATF-3, a sensitive marker for peripheral neuron stress/injury, in small and large diameter DRG cell profiles principally at levels L4 and 5 (levels predominated by neurones innervating the hindpaw) rather than L3. At the 7 day timepoint, ATF-3 signal was significantly smaller in 1 mg MIA treated animals than in the 2 mg treated group. 2 mg, but not 1 mg, intra-articular MIA was also associated with a significant reduction in intra-epidermal nerve fibre density in plantar hindpaw skin, and produced spinal cord dorsal and ventral horn microgliosis. The 2 mg treatment evoked mechanical pain-related hypersensitivity of the hindpaw that was significantly greater than the 1 mg treatment. MIA treatment produced weight bearing asymmetry and cold hypersensitivity which was similar at both doses. Additionally, while pregabalin significantly reduced deep dorsal horn evoked neuronal responses in animals treated with 2 mg MIA, this effect was much reduced or absent in the 1 mg or sham treated groups. These data demonstrate that intra-articular 2 mg MIA not only produces joint degeneration, but also evokes significant axonal injury to DRG cells including those innervating targets outside of the knee joint such as hindpaw skin. This significant neuropathic component needs to be taken into account when interpreting studies using this model, particularly at doses greater than 1 mg MIA.

Figure 1. Behavioural and articular histological assessment of MIA animals.

A – withdrawals in response to 1 g, 6 g or 8 g von frey hair or acetone applied to the plantar hindpaw 3, 7 or 14 days after intra-articular saline, 2 mg or 1 mg MIA injection (Kruskal Wallis test of: timecourse, P<0.05 +, P<0.01 ++, P<0.001 +++ vs preinjection baseline; inter-group comparison P<0.05 *, P<0.01 ** n = 12 animals/group). B – incapacitance testing of 2 mg or 1 mg MIA treated animals at 3, 7 and 14d after treatment (2-way ANOVA with Bonferroni's post test, n = 12 animals/group. P>0.05 at all time points). C – 2 mg MIA treated rat knee sagittal sections from day 14, stained with toluidine blue to visualise cartilage proteoglycan content. Fem = femoral condyl, Tib = tibial condyl. Ant = anterior aspect of knee, Post = posterior aspect. D – 1 mg MIA rat knee sections from day 14. E – saline injected contralateral control knee from day 14.

Figure 2. Cell size distribution for BIIITubulin (green bars, left y axis) and ATF-3 (red bars, right y axis) expressing profiles in DRG L3, L4 and L5 in sham or 2 mg MIA animals 3, 7 and 14 days after injection.

n = 9 sham, 7 at day 3, 12 at day 7, 7 at day 14.

Figure 3. Approximation of total DRG ATF-3 expression.

A, B, C - DRG profiles show nuclear expression of ATF-3 7 days after 2 mg (A) or 1 mg (B) MIA, but not 14 days after saline sham injection (C). D - Estimated fraction of DRG profiles expressing ATF-3 in 2 mg MIA-treated animals (One-way ANOVA with Dunnett's multiple comparison test, n = 9×14d sham, 7×3d, 12×7d MIA, 7×14d MIA, P<0.01 **, P<0.05 *). E - Comparison of ATF-3 expression in 2 mg vs 1 mg MIA injected groups at 7 day timepoint (1 way ANOVA with Bonferroni's multiple comparison test, n = 9×7d sham, 12×2 mg MIA, 12×1 mg MIA, P<0.01 ** P<0.001 ***).

Figure 4. Quantification of intrapepidermal nerve fibre density in plantar hindpaw following MIA treatment.

A, B - sections of naïve (A) and MIA (B) plantar hindpaw skin containing PGP9.5 immunoreactive intraepidermal nerve fibres (arrows). C – quantification of IENFs at the 7 and 14 day timepoints (paired t-test of ipsi vs contralateral density, n = 8×7d, 4×14d, P<0.05). IENF density reduction is not seen in 1 mg MIA animals.

Figure 5. Quantification of microglial activation in spinal cord following MIA treatment.

A, B - expression of the microglial marker iba1 in L4 spinal cord in 2 mg (A) and 1 mg (B) MIA groups. C, D – Quantification of microglia with effector morphology in dorsal (C) and ventral (D) horn (1-way ANOVA, n = 4×7d sham, 11×2 mg MIA, P<0.05 *). E – Comparison of dorsal horn microgliosis in 2 mg vs 1 mg MIA groups at the 7d timepoint (1-way ANOVA, n = 4×7d sham, 11×2 mg MIA, 8×1 mg MIA, P<0.05 *).

Figure 6. Effect of pregabalin treatment (10 mg/kg s.c.) on evoked responses of deep dorsal horn WDR neurones.

A - Electrically evoked responses in the Aß, A∂ and C fibre range, as well as PD (post-discharge/repetitive firing) expressed as % of predose baseline in 2 mg MIA, 1 mg MIA or sham injected animals. B - Mechanically evoked responses expressed as % of predose baseline in 2 mg MIA, 1 mg MIA or sham injected animals. C - Thermally evoked responses expressed as % of predose baseline in 2 mg MIA, 1 mg MIA or sham injected animals. Note the significantly greater effectiveness of pregabalin in the 2 mg treated group (normalized post-drug responses compared using Kruskal Wallis test with Dunn's posttest. n = 14d sham×9, 1 mg MIA×8, 2 mg MIA×9. P<0.05 *, P<0.01, **). The sham and 2 mg data is included for comparison, but has previously been published in a different form .