Increased expression of IKK isoforms in human OA meniscus tissue and differential responses to inflammatory stimuli in meniscal cells.
The characteristics of IKK isoform expressions in human normal and OA menisci were investigated by immunohistochemical and gene expression analyses. In terms of the protein expressions of all four IKK isoforms, the OA menisci exhibited a significantly higher percentage of immunostained cells compared with normal menisci (Fig. 1A). Regarding gene expressions, IKKα, IKKβ, and IKKε were significantly upregulated in human OA meniscal cells in the presence of IL1β, TNFα, or LPS (Fig. 1B). On the other hand, in human normal meniscal cells, only gene expressions of IKKβ and IKKε were significantly increased in response to IL1β, TNFα, or LPS stimulation, while TNFα stimulation led to increased gene expressions of IKKα and IKKγ (Fig. 1C).
IKKε knockdown specifically attenuates meniscal degeneration–related genes in human OA meniscal cells.
The effects of IKK knockdown were analyzed using siRNA-transfected human OA meniscal cells stimulated with IL-1β. Figure 2A illustrates the impact of IKK knockdown on the gene expressions of other IKKs. Specifically, the gene expressions of IKKα and IKKβ were reciprocally downregulated by siRNA, whereas IKKε knockdown did not affect the gene expressions of other IKKs.
Regarding the expressions of genes associated with meniscal degeneration, IKKα knockdown led to significant decreases in the gene expressions of MMP1 and p65, while increasing those of ADAMTS4 and ADAMTS5. IKKβ knockdown resulted in reduced gene expressions of IL6, MMP3, RUNX2, and p65, along with increased gene expressions of MMP13, ADAMTS5, and ENPP1. Importantly, IKKε knockdown notably decreased the gene expressions of IL6, MMP1, MMP3, ADAMTS5, RUNX2, ENPP1, and p65 (Fig. 2B).
IKKε overexpression notably promotes meniscal degeneration–related gene expression in human OA meniscal cells, with reversal by the IKKε inhibitors amlexanox and BAY-985.
The effects of IKKε overexpression and IKKε inhibitors on meniscal degeneration–related gene expressions were assessed in human OA meniscal cells transfected with an IKKε recombinant adenoviral vector (Fig. 3A). In the absence of IL-1β, IKKε overexpression significantly increased the gene expressions of IL6, MMP1, MMP3, MMP13, ADAMTS4, ADAMTS5, ENPP1, and p65 (Fig. 3B). In the presence of IL-1β, IKKε overexpression led to significant increases in the gene expressions of IL6, MMP13, ADAMTS4, ADAMTS5, RUNX2, ENPP1, and p65, which were significantly attenuated by treatment with either of the IKKε inhibitors, amlexanox or BAY-985 (Fig. 3C).
IKKε overexpression activates the NF-κB pathway by increasing IκBα phosphorylation in the presence of IL-1β.
The role of IKKε in the NF-κB pathway was assessed by examining the phosphorylation levels of IκBα and p65 in human OA meniscal cells overexpressing IKKε. Stimulation with IL-1β induced the phosphorylation of IκBα and p65 (Fig. 4A). In the absence of IL-1β stimulation, the level of IκBα phosphorylation was not increased by IKKε overexpression compared to the Ad-GFP control (Fig. 4B). By contrast, in the presence of IL-1β, IKKε overexpression significantly increased the phosphorylation of IκBα, accompanied by increases in the levels of IκBα, p65, and p-p65 (Fig. 4C). The elevation of p-IκBα was significantly reversed by treatment with either amlexanox or BAY-985.
Intraarticular injection of either IKKε inhibitor, amlexanox or BAY-985, reduces meniscal degeneration in a mouse model of early OA.
The in vivo therapeutic effects of intraarticular injection of amlexanox or BAY-985 on meniscus degradation were evaluated using a surgical OA mouse model. In this model, OA is induced by meniscus destabilization, leading to abnormal mechanical loading-induced meniscus damage and subsequent cartilage damage. To detect early changes in meniscus degradation before cartilage degeneration, mice were sacrificed 4 weeks after surgery. This model resulted in a significantly higher percentage of IKKε-positive cells in the meniscus compared with sham-operated mice (Fig. 5A).
Amlexanox solution, BAY-985 solution, or saline as a vehicle control was injected into the knee joint space of mice, starting on the day of surgery and continuing every 5 days for 4 weeks. The histopathological scores in the anterior and posterior menisci were significantly lower in the groups treated with amlexanox or BAY-985 compared with the saline group (Fig. 5B). OARSI scores showed only mild cartilage damage and this was not significantly changed by the drug injections. Immunofluorescence analysis of p-IκBα in the total menisci of mice revealed that the percentage of p-IκBα–positive cells was significantly decreased by treatment with amlexanox or BAY-985 compared with saline (Fig. 5C).
IKKε knockout does not exacerbate meniscal degeneration in an early OA mouse model.
Finally, the in vivo causal effects of IKKε deletion on meniscus degradation were investigated using IKKε knockout mice after MMTL and MCL transection. There was no difference in the whole-body phenotype of knockout and wild-type mice (Fig. 6A). Body weights did not show significant differences between IKKε knockout mice and wild-type mice at 2, 4, 8, or 12 weeks of age. The development and structure of meniscus and other joint tissues was also normal in the mutant mice. The histopathological scores in the anterior and posterior menisci at 4 weeks after surgery were significantly lower in IKKε knockout mice compared with wild-type mice, with no significant change in early cartilage degradation assessed by OARSI scores (Fig. 6B).