The neuropathic phenotype of the K/BxN transgenic mouse with spontaneous arthritis: pain, nerve sprouting and joint remodeling

Abstract

The adult K/BxN transgenic mouse develops spontaneous autoimmune arthritis with joint remodeling and profound bone loss. We report that both males and females display a severe sustained tactile allodynia which is reduced by gabapentin but not the potent cyclooxygenase inhibitor ketorolac. In dorsal horn, males and females show increased GFAP⁺ astrocytic cells; however, only males demonstrate an increase in Iba1⁺ microglia. In dorsal root ganglia (DRG), there is an increase in CGRP⁺, TH⁺, and Iba1⁺ (macrophage) labeling, but no increase in ATF3⁺ cells. At the ankle there is increased CGRP⁺, TH⁺, and GAP-43⁺ fiber synovial innervation. Thus, based on the changes in dorsal horn, DRG and peripheral innervation, we suggest that the adult K/BxN transgenic arthritic mice display a neuropathic phenotype, an assertion consistent with the analgesic pharmacology seen in this animal. These results indicate the relevance of this model to our understanding of the nociceptive processing which underlies the chronic pain state that evolves secondary to persistent joint inflammation.

Figure 1

Male and female K/BxN mice develop destructive arthritis and mechanical allodynia. Male and female WT and K/BxN mice (n = 9/group) were followed from 4 to 16 weeks of age scored for arthritis by visual inspection (A). Groups of mice (n = 5–7/group) were sacrificed at 16 weeks and the ankles were histologically examined. Both male and female K/BxN mice had severe bone erosion and cartilage loss (B–G). The same mice in (A) were assessed for tactile allodynia by von Frey testing (H). Groups of 16-week-old mice (n = 6/group) were tested for their responses to ketorolac or gabapentin by withdrawal threshold 60 min after drug delivery (I). Data are shown as mean ± SEM. *p < 0.01, ***p < 0.001 by strain two-way ANOVA for repeated measures and Bonferroni post hoc test.

Figure 2

Differences in trabecular bone in WT and K/BxN mice. Representative images of tridimensional reconstructions at different sites and quantitative analyses of trabecular bone parameters (A–C). MicroCT analysis was performed on the tibiae of 16-week-old mice. Female and male K/BxN mice show a decrease of BMD, BV/TV and Tb.N compared with wild type (WT) in the tibiae (A), calcanea (B) and tali (C), both in males and females, except for Tb.N of talus in females (C). Data are shown as mean $\pm$ SEM of 6–9 mice/group. *p < 0.05 by strain, ^&p < 0.05 by sex by two-way ANOVA and Bonferroni post hoc test.

Figure 3

K/BxN male and female spinal cord glial activation. Spinal cords were harvested from 16-week-old male and female WT and K/BxN mice (n = 6–9 mice /group) and the lumber regions were incubated with antibodies against GFAP (A) and Iba1 (B). Representative staining images which were used for the Iba1 and GFAP quantification are shown (bar is 50 µm). (C) Schematic of Rexed’s laminae I–II (in red) and four circular (blue) areas where the optical density (OD) was analyzed. The OD of an area (blue circle) in the deep region of the same slice was subtracted as the background control. (D) GFAP immunofluorescence was increased in both the male and female mice compared to WT. (E) Iba1 staining was increased in the male K/BxN mice compared to WT. *p < 0.05 strain, ^&p < 0.05 by sex one-way ANOVA with Bonferroni post hoc test data are represented as mean ± SEM.

Figure 4

K/BxN male and female have increased CGRP⁺ and TH⁺ neurons in the DRG. The L4 and L5 DRG were harvested from 16-week-old male and female WT and K/BxN mice (n = 7–8 mice /group) and stained for DAPI (blue), NeuN (green), TH (red) and CGRP (red). Representative staining images are shown (bar is 50 µM), which were used for quantification of CGRP (A) and TH (B). The mean percentages ± SEM of NeuN positive cells that also stained for CGRP (C) and TH (D) are shown. *p < 0.05; one-way ANOVA with Bonferroni post hoc test. There were no differences between sexes of the same strain (p > 0.05).

Figure 5

Arthritis is followed by increase of macrophage (Iba1) with lack of ATF3 expressing neurons in the DRG of K/BxN mice. The L4 and L5 DRG were harvested from 16-week-old male and female WT and K/BxN mice (n = 6 mice /group) and stained for DAPI (blue), NeuN (green), Iba1 (red) and ATF3 (red). Representative staining images are shown (bar is 50 µM), which were used for quantification of Iba1 (A) and ATF3 (B). The mean percentages ± SEM of NeuN positive cells that also stained for Iba1 (C) and ATF3 (D) are shown. *p < 0.05 one-way ANOVA with Bonferroni post hoc test. There were no differences between sexes of the same strain (p > 0.05). (D) No immunoreactivity for ATF3 was found in the DRG of WT or K/BxN; however, L5 nerve axotomy induced an increase of ATF3 in at least 50% of neurons on the DRG. *p < 0.05 one-way ANOVA with Bonferroni post hoc test. N.D. not detected.

Figure 6

Increase in sensory CGRP⁺, TH⁺ and GAP-43 nerve fibers in arthritic ankles. (A) Hematoxylin and eosin staining of a K/BxN ankle joint at 16 weeks (bar is 400 µm). The square illustrates the synovial region from which the confocal images were obtained (B; bar is 100 µm). Representative confocal images of calcitonin gene-related peptide (CGRP; cyan, C, D), tyrosine hydroxylase (TH; purple, E, F) and growth-associated protein (GAP-43, marker of fibers undergo regeneration; yellow, G, H) in ankle-joint sections (20 μm thick) from WT (C, E, G) and K/BxN (D, F, H) mice (bar is 100 µm). A significantly greater density of CGRP⁺, TH⁺ and GAP-43⁺ sprouted nerve fibers are observed in the synovia of K/BxN as compared to those observed in WT mice (I, J, K). Bars represent the mean ± SEM of 6–9 mice/group. *p < 0.05 one-way ANOVA with Bonferroni post hoc test.

Figure 7

Dense bundles of CGRP⁺ fibers near the ankle/calcaneus of K/BxN mice. Confocal images of dense fibers in female (A) and three males (B–D) 16-week-old K/BxN mice are shown (bar is 100 µm). DAPI-labeled nuclei (blue; A–D), and CGRP (white; A–D), TH (purple; E–H) and GAP-43 (yellow; I–L) immunoreactive nerve fibers in the ankle joint sections show that the densest bundles were formed by CGRP⁺ nerve fibers in the synovia of arthritic mice.