NADPH oxidase is involved in post-ischemic brain inflammation

Abstract

Nicotinamide adenine dinucleotide phosphate oxidase (NOX) is widely expressed in brain tissue including neurons, glia, and endothelia in neurovascular units. It is a major source of oxidants in the post-ischemic brain and significantly contributes to ischemic brain damage. Inflammation occurs after brain ischemia and is known to be associated with post-ischemic oxidative stress. Post-ischemic inflammation also causes progressive brain injury. In this study we investigated the role of NOX2 in post-ischemic cerebral inflammation using a transient middle cerebral artery occlusion model in mice. We demonstrate that mice with NOX2 subunit gp91(phox) knockout (gp91 KO) showed 35-44% less brain infarction at 1 and 3 days of reperfusion compared with wild-type (WT) mice. Minocycline further reduced brain damage in the gp91 KO mice at 3 days of reperfusion. The gp91 KO mice exhibited less severe post-ischemic inflammation in the brain, as evidenced by reduced microglial activation and decreased upregulation of inflammation mediators, including interleukin-1β (IL-1β), tumor necrosis factor-α, inducible nitric oxide synthases, CC-chemokine ligand 2, and CC-chemokine ligand 3. Finally, we demonstrated that an intraventricular injection of IL-1β enhanced ischemia- and reperfusion-mediated brain damage in the WT mice (double the infarction volume), whereas, it failed to aggravate brain infarction in the gp91 KO mice. Taken together, these results demonstrate the involvement of NOX2 in post-ischemic neuroinflammation and that NOX2 inhibition provides neuroprotection against inflammatory cytokine-mediated brain damage.

Fig. 1

Decreased brain infarct volume after I/R in mice with NOX2 inhibition or minocycline treatment. Infarct volume at 24 or 72 h of reperfusion after MCAO in WT mice, gp91 KO mice, minocycline-treated WT mice (Mino), and minocycline-treated KO mice (Mino+KO) (#p < 0.05 vs. WT mice at 24 h of reperfusion, *p < 0.05 vs. WT mice at 72 h of reperfusion, **p < 0.05 vs. gp91 KO or minocycline-treated mice at 72 h of reperfusion) (n = 3–5).

Fig. 2

Cleaved spectrin in WT and gp91 KO mice. (A) Western blot analysis of cleaved spectrin at 3 days of reperfusion. Significant cleaved spectrin products were detected in the WT mice, but not in the gp91 KO mice. Con, contralateral; Ipsi, ipsilateral. (B) Summarized data from panel A. *p < 0.05 vs. WT cleaved spectrin (150 kDa). Summarized data are presented as a ratio of (cleaved spectrin: ipsilateral/contralateral)/(uncleaved spectrin: ipsilateral/contralateral) (n = 4–5).

Fig. 3

3-NT expression in WT, apocynin-treated (Apo), and gp91 KO mice. (A) Western blot analysis of 3-NT at 24 h of reperfusion. Significant 3-NT was detected in the WT mice, but not in the apocynin-treated or the gp91 KO mice. Con, contralateral; Ipsi, ipsilateral. (B) Summarized data from A. *p < 0.05 vs. WT mice. Summarized data are presented as a ratio of 3-NT/β-actin, which are normalized to the contralateral side. (C) Western blot analysis of cleaved 3-NT at 3 days of reperfusion. Genetic ablation of NOX2 inhibition reduced 3-NT upregulation. (D) Summarized data from C. *p < 0.05 vs. WT mice (n = 4–5).

Fig. 4

NOX2 inhibition reduced microglia activation after I/R. At 24 h of reperfusion, more activated microglia were detected in the WT ipsilateral hemisphere than in the gp91 KO mice. (A) WT contralateral; (B) WT ipsilateral; (C) gp91 KO contralateral; (D) gp91 KO ipsilateral. Insets in A and B, magnification of one typical microglia. Inset in D, negative control. Scale bar = 50 μm. Images shown are representative of 4–5 experiments. (E) After I/R, microglia in KO ischemic side had more longer processes than in the WT ipsilateral hemisphere (*p < 0.05 vs. WT ipsilateral) (n = 4–5).

Fig. 5

Changes in IBA-1 expression after cerebral I/R. (A), (C) and (E) Western blot analyses of IBA-1 expression after I/R. I/R induced IBA-1 upregulation at 1 and 3 days of reperfusion, but not at 3 h of reperfusion. Con, contralateral; Ipsi, ipsilateral. (B), (D) and (F) Summarized data from (A), (C) and (E), respectively. Numbers are presented as a ratio of IBA-1/β-actin, which are normalized to the contralateral side (*p < 0.05 vs. WT mice).

Fig. 6

Inflammation factor gene transcription after cerebral ischemia in WT or gp91 KO mice. (A) gp91 ablation reduced upregulation of TNFα transcription at 3 days of reperfusion (n = 8–12). (B) gp91 ablation reduced upregulation of iNOS transcription at 3 days of reperfusion (n = 8–9). (C) gp91 ablation reduced upregulation of CCL2 transcription at 3 days of reperfusion (n = 8–9). (D) gp91 ablation reduced upregulation of CCL3 transcription at 1 and 3 days of reperfusion (n = 9–11) (*p < 0.05 vs. WT mice). Summarized data are presented as a ratio of ipsilateral side/contralateral side.

Fig. 7

NOX2 activity is required for the generation of and detrimental effect of IL-1β. (A) gp91 ablation reduced upregulation of IL-1β gene transcription at 1 and 3 days of reperfusion (n = 8–10). (B) Western blot analysis of IL-1β at 72 h of reperfusion. Significant IL-1β overexpression was detected in the WT mice, but not in the gp91 KO mice. Con, contralateral; Ipsi, ipsilateral. (C) Summarized data from (B) (*p < 0.05 vs. WT mice). Summarized data are presented as a ratio of *p < 0.05 vs. WT mice, and as a ratio of ipsilateral side/contralateral side (n = 4–5). (D) NOX2 ablation abolished IL-1β-mediated brain damage (n = 9) (*p < 0.05 compared with IL-1β treatment).