Requirement for protein kinase R in interleukin-1alpha-stimulated effects in cartilage

Abstract

Interleukin-1 (IL-1) has pleiotropic effects in cartilage. The interferon-induced, double stranded RNA-activated protein kinase PKR that phosphorylates eukaryotic initiation factor 2 (eIF2) alpha has been implicated in cytokine effects in chondrocytes. A compound was recently identified that potently suppresses PKR autophosphorylation (IC50 approximately 200 etaM) and partially restores PKR-inhibited translation in a cell-free system with significant effect in the nanomolar range. The objectives of this study were to exploit this potent PKR inhibitor to assess whether PKR kinase activity is required for catabolic and proinflammatory effects of IL-1alpha in cartilage and to determine whether IL-1alpha causes an increase in eIF2alpha phosphorylation that is antagonized by the PKR inhibitor. Cartilage explants were incubated with the PKR inhibitor and IL-1alpha. Culture media were assessed for sulfated glycosaminoglycan as an indicator of proteoglycan degradation and for prostaglandin E(2). Cartilage extracts were analyzed by Western blot for cyclooxygenase-2 and phosphorylated signaling molecules. Nanomolar concentrations of the PKR inhibitor suppressed proteoglycan degradation and cyclooxygenase-2 accumulation in IL-1alpha-activated cartilage. The PKR inhibitor stimulated or inhibited PGE(2) production with a biphasic dose response relationship. IL-1alpha increased the phosphorylation of both PKR and eIF2alpha, and nanomolar concentrations of PKR inhibitor suppressed the IL-1alpha-induced changes in phosphorylation. The results strongly support PKR involvement in pathways activated by IL-1alpha in chondrocytes.

Fig. 1

The PKR inhibitor suppresses IL-1α-stimulated cartilage proteoglycan degradation. Cartilage disks (n = 8) were placed individually in 200 µL DMEM supplemented with 10 µg/ml BSA in the wells of a 96-well plate. Disks were preincubated for 2 hours with the PKR inhibitor (PKRi) at the indicated concentration. IL-1α (10 ηg/ml) was added (black bars) or not (open bar), and incubation was continued for 24 hours. Proteoglycan degradation was assessed by assaying aliquots of culture media for GAG. Values are reported as mean + S. E. M. Significance of difference of IL-1α-stimulated GAG release in PKR inhibitor-treated disks relative to non-inhibited disks: ***, p < 0.001; ns, not significant.

Fig. 2

The PKR inhibitor reduces the accumulation of COX-2 in IL-1α-activated cartilage. Cartilage disks (n = 8) were preincubated for 2 hours with the indicated concentration of the PKR inhibitor in DMEM supplemented with BSA. IL-1α was added to selected wells, and incubation was continued for 24 hours. Protein extracts were prepared from cartilage disks for analysis by Western blotting using antibody to COX-2 with anti-β-actin to monitor loading and transfer. COX-2 accumulation was dependent on IL-1α and inhibited by the PKR inhibitor.

Fig. 3

The PKR inhibitor has biphasic dose-dependent effects on IL-1α-stimulated PGE₂ production. Cartilage disks (n = 8) were preincubated for 2 hours with the PKR inhibitor at the indicated concentration. IL-1α (10 ηg/ml) was added (black bars) or not (open bar not visible), and incubation was continued for 24 hours. Harvested culture media were assayed for PGE₂ by ELISA. Significance of difference of IL-1α-stimulated PGE₂ release in PKR inhibitor-treated disks relative to non-inhibited disks: *, p < 0.05; ns, not significant.

Fig. 4

The PKR inhibitor prevents the IL-1α-stimulated increase in phosphorylation of PKR and eIF2α in cartilage. (A) Cartilage disks were preincubated for 2 hours with or without 20 ηM PKR inhibitor. IL-1α (10 ηg/ml) or 3 mM DTT was added as indicated, and the incubation was continued for 30 minutes. Protein extracts of the disks were analyzed by Western blot for phosphorylated eIF2α, phosphorylated PKR, and β-actin. Signals for phosphorylated proteins were normalized to signals obtained using antibody to β-actin. Fold-stimulation shown below the Western blot for each phosphorylated protein is calculated as the ratio of each normalized signal relative to the normalized signal for the extract from unstimulated cartilage disks shown in lane 1. (B) Disks were preincubated for 2 hours with the indicated concentrations of PKR inhibitor before the addition of IL-1α to selected wells. The incubation was continued for 24 hours. Phosphorylated and total eIF2α in disk extracts were analyzed by Western blot, and the signals for the phosphorylated form were normalized to signals for total eIF2α. Fold-stimulation is determined as in the legend for Fig. 4B. IL-1α caused a rapid and prolonged increase in eIF2α phosphorylation that was inhibited in disks treated with the PKR inhibitor.

Fig. 4

The PKR inhibitor prevents the IL-1α-stimulated increase in phosphorylation of PKR and eIF2α in cartilage. (A) Cartilage disks were preincubated for 2 hours with or without 20 ηM PKR inhibitor. IL-1α (10 ηg/ml) or 3 mM DTT was added as indicated, and the incubation was continued for 30 minutes. Protein extracts of the disks were analyzed by Western blot for phosphorylated eIF2α, phosphorylated PKR, and β-actin. Signals for phosphorylated proteins were normalized to signals obtained using antibody to β-actin. Fold-stimulation shown below the Western blot for each phosphorylated protein is calculated as the ratio of each normalized signal relative to the normalized signal for the extract from unstimulated cartilage disks shown in lane 1. (B) Disks were preincubated for 2 hours with the indicated concentrations of PKR inhibitor before the addition of IL-1α to selected wells. The incubation was continued for 24 hours. Phosphorylated and total eIF2α in disk extracts were analyzed by Western blot, and the signals for the phosphorylated form were normalized to signals for total eIF2α. Fold-stimulation is determined as in the legend for Fig. 4B. IL-1α caused a rapid and prolonged increase in eIF2α phosphorylation that was inhibited in disks treated with the PKR inhibitor.