Deficiency of NADPH oxidase components p47phox and gp91phox caused granulomatous synovitis and increased connective tissue destruction in experimental arthritis models

Abstract

Recent studies indicated that the nicotinamide dinucleotide phosphate oxidase (NADPH) oxidase-derived oxygen radicals plays a deleterious role in arthritis. To study this in more detail, gonarthritis was induced in NADPH oxidase-deficient mice. Mice received an intraarticular injection of either zymosan, to elicit an irritant-induced inflammation, or poly-L-lysine coupled lysozyme, to evoke an immune-complex mediated inflammation in passively immunized mice. In contrast to wild-type mice, arthritis elicited in both p47phox(-/-) and gp91(-/-) mice showed more severe joint inflammation, which developed into a granulomatous synovitis. Treatment with either Zileuton or cobra venom factor showed that the chemokines LTB4 and complement C3 were not the driving force behind the aggravated inflammation in these mice. Arthritic NADPH oxidase-deficient mice showed irreversible cartilage damage as judged by the enhanced aggrecan VDIPEN expression, and chondrocyte death. Furthermore, only in the absence of NADPH oxidase-derived oxygen radicals, the arthritic joints showed osteoclast-like cells, tartrate-resistant acid phosphatase (TRAP)-positive/multinucleated cells, extensive bone erosion, and osteolysis. The enhanced synovial gene expression of tumor necrosis factor-alpha, interleukin-1alpha, matrix metalloproteinase (MMP)-3, MMP-9 and receptor activator of NF-kappaB ligand (RANKL) might contribute to the aggravated arthritis in the NADPH oxidase-deficient mice. This showed that the involvement of NADPH oxidase in arthritis is probably far more complex and that oxygen radicals might also be important in controlling disease severity, and reducing joint inflammation and connective tissue damage.

Figure 1.

Zymosan-induced joint swelling in WT and p47phox KO mice. A single dose of zymosan (180 μg sonicated and sterile yeast particles) was injected intraarticularly into the right (R) knee joint. The left (L) knee joint was not injected and served as an intraanimal control. Joint swelling was measured at the indicated times after zymosan injection by the ^99mTechnetium uptake method (as described in Materials and Methods). Right knee joints with a R/L ratio of more than 1.1 are regarded swollen. Data are the mean ± SD of joint swelling measured in 12 animals at day 2, 7 animals at day 3, and 6 animals at day 7. At all timepoints measured the values of the p47phox^−/− mice were significantly different (P < 0.0001, by Student’s t-test) as compared to WT mice.

Figure 2.

Histological evaluation of joint pathology in WT (A) and NADPH-OX deficient (B to H) mice, 7 days after intraarticular zymosan injection. Frontal knee-joint sections (×50) stained with safranin-O: A, WT mouse; B, p47phox KO mouse, arrowheads indicate cartilage proteoglycan depletion, open arrows indicate sites of bone erosion. Zymosan caused granulomatous synovitis in p47phox^−/− mice (C, original magnification, ×100) with a large number of PMNs as identified with NIMP-R14 antibodies (D, representative figure of 6 mice). May-Grunwald Giemsa staining of a cytospin of exudate cells obtained from the joint cavity of p47phox^−/− mice showed the presence of PMNs (E, original magnification, ×400). Exudate cells are in close proximity to the cartilage surface in p47phox^−/− mice (F, original magnification, ×400). Glycosaminoglycan depletion as detected by loss of safranin-O staining and chondrocyte death in articular cartilage of patellae and condyles from p47phox^−/− mice (G and H, original magnification, ×200). C, cartilage; F, femur; GP, growth plate; M, meniscus; P, patella, S, synovium; T, tibia.

Figure 3.

Immunolocalization of the neo-epitope VDIPEN in cartilage of WT (A and C) and p47phox KO (B and D) mice at day 7 of zymosan-induced arthritis. VDIPEN staining (black grains) was detected in cartilage at both the patella-femur junction and in the condyles of p47phox^−/− mice only. Representative figure of three experiments (total number of p47phox^−/− mice evaluated is 18 and an equal number of WT mice). Original magnification, ×100. Sections were counterstained with Orange G.

Figure 4.

X-ray photographs of whole knee-joints from WT (A and B) and p47phox KO (C and D) mice taken 7 days after zymosan injection. B and D: Mice treated with Zileuton (Table 2) ▶ . Note the loss (osteolysis) of trabecular and cortical bone in the femur of p47phox^−/− mice. Representative figure of two experiments; first experiment consisted of five intercrossed p47phox^−/− mice and four WT mice, second experiment consisted of 12 backcrossed p47phox^−/− mice (N15) and 12 C57Bl/6 mice.

Figure 5.

Saggital plane of whole knee-joint sections stained for TRAP and counterstained with hematoxylin of WT (A) and p47phox^−/− (B) mice. Black arrowhead indicates the site of cortical bone erosion, white arrowhead indicates the areas with large numbers of TRAP-positive multinucleated cells in p47phox^−/− mice. Note the overall loss of trabecular bone structure in the femur. Representative figure of 4 backcrossed p47phox^−/− mice (N15) and 4 C57Bl/6 animals. Original magnification, ×50.

Figure 6.

Detection of osteoclast-like cells at sites of bone erosion at day 7 of zymosan-induced arthritis in p47phox^−/− mice. Frontal knee-joint sections were developed for TRAP. TRAP-positive multinucleated cells were identified in close contact with cortical bone present in the bone matrix invading tissue (A) and in the synovium (B and C). Representative figure of 7 intercrossed p47phox^−/− mice. A and B: Original magnification, ×200; C, ×400.

Figure 7.

Gene expression in inflamed synovia of WT and p47phox KO mice at day 7 of zymosan-induced arthritis. Synovia were dissected and total RNA was extracted and subjected to RT-PCR using specific oligonucleotide primers for the indicated genes. Equal amounts of mRNA (normalized by their GAPDH content) were used and the expression in the inflamed synovia was set against the normal contralateral joint (Δ PCR cycles). Data represent the mean mRNA expression of 3 intercrossed p47phox^−/− and 3 WT mice.

Figure 8.

Expression of genes involved in the apoptotic process in the inflamed synovia of WT and p47phox-deficient mice. Total mRNA obtained from non-inflamed (pooled samples of 2 mice) and inflamed synovia (pooled samples of mice) were subjected to multiplex PCR. The included positive apoptosis gene set (M) showed the migration from top to bottom of GAPDH (532 bp), BCL-xL (371 bp), caspase-3 (320 bp), Bax (272 bp), Bcl-2 (235 bp), and BCL-xS (183 bp). The PCR product of caspase-3 was very faint in all of the samples.

Figure 9.

Immunolocalization of PARP in synovium of WT (A) and p47phox^−/− (B) mice at day 7 of zymosan-induced arthritis. Comparable positive staining was seen in both mouse strains. Figure is representative of 15 intercrossed p47phox^−/− and 18 WT mice. No immunoreactivity was found in the non-arthritic control knees (not shown). Original magnification, ×200. B, bone; JC, joint cavity.