Determination of optimal dose and evaluation of drug safety. The peptide sequences used in this study, including the Nrf2 peptide (NP) and the corresponding control peptide (CP), are shown in Figure 1A, and a schematic representation of the experimental design is provided in Figure 1B. Next, we assessed the dose-dependent effects of NP on alcohol-induced endothelial and microvascular cell damage using annexin V staining as a marker of vascular apoptosis. To identify the optimal NP concentration, mice subjected to a 4-week alcohol regimen received daily NP injections (0–150 μg) for 2 consecutive weeks after alcohol exposure. Brain sections were immunostained for annexin V, and annexin V–positive cells were quantified across treatment and control groups. Annexin V–positive cells were markedly reduced at the NP dose of 100 μg per mouse (~4 mg/kg in 100 μL) compared with other doses, indicating decreased apoptotic activity (P < 0.0001; Figure 1, C and D). In parallel, ROS-associated lipid peroxidation was assessed by measurement of malondialdehyde (MDA) levels in the brains of NP-treated animals. Among the tested doses, the 100 μg NP treatment produced the most pronounced neuroprotective effect, showing a significant reduction in MDA levels compared with other doses (P < 0.0001; Figure 1E). This dose was well tolerated, as indicated by stable body weight, normal grooming and activity, and unaltered serum biochemical markers of oxidative stress and cell death. Together, these results demonstrate both the safety and efficacy of NP at the selected in vivo concentration.

Alcohol impairs the expression of Nrf2 while Nrf2 peptide treatments activate it. To test our hypothesis that activation of Nrf2 through NP treatment would suppress BBB dysfunction, vascular permeability, and leukocyte transmigration into the brain following alcohol ingestion, we treated mice daily with a single subcutaneous dose of the NP (100 μg/100 μL). First, activation of Nrf2 by the NP was analyzed using immunofluorescence and Western blotting. We specifically analyzed the expression of Nrf2 in the brain microvessels (Figure 1F). Immunostaining of brain tissue sections showed Nrf2 predominantly in the cytoplasm, while phosphorylated Nrf2 (p-Nrf2) was localized mainly in the nucleus, indicating activation and nuclear translocation (Figure 1G). Next, we analyzed the expression level of Nrf2 and p-Nrf2 in the frontal brain tissue section following alcohol ingestion using immunofluorescence. A 2-way ANOVA revealed significant main effects of diet and treatment on Nrf2 (F(3,36) = 28.10, P < 0.0001) and p-Nrf2 (F(3,36) = 27.72, P < 0.0001). Post hoc comparisons indicated that both Nrf2 and p-Nrf2 levels were significantly reduced in ethanol diet–fed (ED-fed) mice compared with the control diet (CD) mice (P < 0.0001; Figure 1, H–J). Importantly, NP treatment activated Nrf2 signaling, leading to increased Nrf2 and p-Nrf2 expression in ED+NP mice compared with ED+CP mice (P < 0.0001; Figure 1, H–J).

Further, Western blot analysis confirmed these findings, showing significantly reduced Nrf2 and p-Nrf2 expression in ED+CP WT group compared with CD+CP WT group animals (P < 0.0001; Figure 2, A–C). Notably, NP treatment caused an increased expression of both Nrf2 and p-Nrf2 in the ED+NP WT group compared with the ED+CP group (~3.5-fold for both Nrf2 and p-Nrf2, P < 0.0001). Here, we used Nrf2−/− mice to validate the Nrf2 expression and activation of Nrf2 by NP treatment. Consistent with the genotype, neither Nrf2 nor p-Nrf2 was detected in Nrf2−/− mice (Figure 2, A–C); however, a very faint signal in Nrf2–/– samples likely reflects nonspecific binding or residual nonfunctional protein, consistent with basal background levels.

Nrf2 peptide improves expression of antioxidant proteins in alcohol ingestion. (A–C) Western blot analysis of Nrf2, p-Nrf2, and β-actin from mouse cortex tissue lysate of CD- and ED-fed animals in WT and Nrf2-knockout animals (Nrf2–/–) with CP or NP treatment. Bar graphs (B and C) with dot plot show ratio densitometry of Nrf2 and p-Nrf2 to β-actin (n = 8 per group). (D) mRNA expression level of p-Nrf2 using qPCR from the cortex of CD-fed animals with CP or NP and ED-fed animals with CP or NP (n = 8 per group). (E) Western blot analysis of Keap1 and β-actin from mouse cortex tissue lysate of CD- and ED-fed animals in WT and Nrf2-knockout animals with CP or NP treatment. Bar graph with dot plot shows ratio densitometry of Keap1 to β-actin (n = 8 per group). (F and G) Immunofluorescence staining of the antioxidant protein HO-1 colocalized with GLUT1 (microvessel marker) and DAPI (nucleus) in intact brain microvessels of mice fed with CD or ED with CP or NP treatments. Yellow arrows show expression of HO-1. Scale bars: white, 50 μm; yellow, 10 μm. (G) Quantification of HO-1 staining was analyzed using ImageJ software (n = 10 per group). (H–K) Western blot analysis of antioxidant proteins HO-1, GPx1, and GSTm1 and β-actin from mouse frontal cortex tissue lysate of CD- and ED-fed animals in WT and Nrf2-knockout animals with CP or NP treatment. Bar graphs with dot plot show ratio densitometry of HO-1 (I), GPx1 (J), and GSTm1 (K) to β-actin (n = 12 per group). All values are expressed as mean ± SD. Statistical analysis was performed by ANOVA (2-way for D and G, 3-way for B, C, E, I, and J) followed by Bonferroni’s post hoc test. P < 0.05 was considered statistically significant. @P < 0.05, @@P < 0.01, @@P < 0.001 vs. representative groups in WT animals. Exact P values are shown between the compared groups.

Similarly, the expression level of Nrf2 mRNA by RT-qPCR was reduced to half in ED+CP animals compared with CD+CP control animals (~2.0-fold, F(3,28) = 99.07, P < 0.0001). NP treatment significantly increased the expression of Nrf2 mRNA in the ED+NP group compared with ED+CP as we observed in the protein expression (~1.5-fold, P < 0.0001; Figure 2D). When we analyzed Keap1 protein expression, we observed a significant reduction in the ED+CP group compared with the CD+CP controls (F(7,56) = 27.61, P < 0.0001; Figure 2E), whereas NP treatment significantly increased Keap1 expression in ED+NP mice compared with ED+CP mice (P < 0.0001). In contrast, Keap1 expression remained largely unchanged in Nrf2–/– samples (Figure 2E).

Nrf2 peptide activates the potential antioxidant genes in alcohol ingestion in vivo. Since Nrf2 has a regulatory role in the activation of antioxidant genes (43), in this study, we analyzed how NP treatments activate the antioxidant gene expression in alcohol-ingested mice. Immunofluorescence analysis of HO-1 expression in brain microvessels revealed a significant main effect of treatment (F(2,27) = 58.82, P < 0.0001). Post hoc analysis showed a reduced expression of HO-1 in ED animals treated with CP, whereas NP activated the expression of HO-1 in ED animals in the brain microvessels (P < 0.0001; Figure 2, F and G). Next, we analyzed the changes in the antioxidant proteins HO-1, GPx1, and GSTm1 by Western blotting to validate the regulatory role of Nrf2 in ED-fed mice. ANOVA revealed significant main effects of treatment for all 3 proteins (HO-1: F(7,88) = 65.56, P < 0.0001; GPx1: F(7,88) = 28.18, P < 0.0001; GSTm1: F(7,88) = 32.20, P < 0.0001) (Figure 2, H–K). Post hoc analysis showed a significant decrease in HO-1, GPx1, and GSTm1 expression in ED+CP WT animals (reduced 0.35-, 0.4-, and 0.28-fold, respectively) compared with CD+CP control WT animals (P < 0.0001; Figure 2, H–K). However, NP treatment in ED animals activated the expression of these 3 antioxidant proteins in comparison with CP-treated ED animals (P < 0.0001; Figure 2, H–K), whereas, in Nrf2−/− mice, the expression of HO-1, GPx1, and GSTm1 was not improved in NP-treated ED animals (Figure 2, H–K).

Next, using RT-qPCR in WT and Nrf2−/− mice, we investigated the effect of alcohol on the expression of the antioxidant genes HO-1, GPx1, and GSTm1. ANOVA revealed significant main effects of treatment for all 3 genes (HO-1: F(3,28) = 13.01, P < 0.0001; GPx1: F(3,28) = 13.53, P < 0.0001; GSTm1: F(3,28) = 31.65, P < 0.0001) (Figure 3, A–C). Post hoc analysis showed a significant reduction in the mRNA expression of HO-1, GPx1, and GSTm1 in ED+CP WT animals compared with CD+CP WT controls, and NP treatment attenuated the downregulation of these 3 antioxidant genes (P < 0.0001). A further decrease in the expression of HO-1, GPx1, and GSTm1 was found in ED+CP Nrf2−/− mice compared with the ED+CP WT group (at least P < 0.05). Moreover, NP had no effect on the expression of these 3 antioxidant genes in ED+NP Nrf2−/− mice (P < 0.0001; Figure 3, A–C).

Nrf2 peptide improves the expression of antioxidant genes in alcohol ingestion. (A–C) The mRNA expression level of HO-1 (A), GPx1 (B), and GSTm1 (C) in the CD- and ED-fed mice of WT and Nrf2–/– groups with CP or NP treatment (n = 8 per group). (D) Agarose gel electrophoresis of purified cross-linked chromatin samples extracted from WT mouse frontal cortex. (E–G) ChIP-qPCR analysis of the DNA binding activity of HO-1 (E), GPx1 (F), and GSTm1 (G) in the CD- and ED-fed animals of WT and Nrf2–/– groups with CP or NP treatment (n = 8 per group). Anti-IgG antibody was used as a negative control. All values are expressed as mean ± SD. Statistical analysis was performed by 3-way ANOVA followed by Bonferroni’s post hoc test. P < 0.05 was considered statistically significant. @P < 0.05, @@P < 0.01, @@P < 0.001 vs. representative groups in WT animals. Exact P values are shown between the compared groups.

Next, binding of Nrf2 to the ARE gene and activation of antioxidant genes were further confirmed by ChIP-qPCR analyses. Figure 3D shows purified cross-linked chromatin samples for ChIP-qPCR extracted from WT mouse frontal cortex. The ChIP-qPCR results indicated that both in CD+CP and in ED+CP WT mice, Nrf2 binds proximal to HO-1, GPx1, and GSTm1. ANOVA revealed significant main effects of treatment on HO-1, GPx1, and GSTm1 DNA enrichment (HO-1: F(4,35) = 197.9, P < 0.0001; GPx1: F(4,35) = 105.5, P < 0.0001; GSTm1: F(4,35) = 33.07, P < 0.0001; Figure 3, E–G). Notably, ED+CP WT mice showed significantly reduced expression levels of HO-1, GPx1, and GSTm1 DNA compared with CD+CP WT mice (P < 0.0001; Figure 3, E–G). These results demonstrate that Nrf2 binds to the regulatory regions of these antioxidant genes, and alcohol reduces the expression of these antioxidant genes. As expected, Nrf2–/– mice showed a negligible expression of HO-1, GPx1, and GSTm1 in CP-treated CD and ED groups (Figure 3, E–G). Interestingly, NP-treated ED WT animals showed a significantly higher level of HO-1, GPx1, and GSTm1 genes compared with CP-treated ED WT animals, and the effect of NP was not observed in Nrf2–/– mice (P < 0.0001; Figure 3, E–G). This confirms that NP activates the antioxidant pathway in alcohol-ingested animals.

Nrf2 peptide reduces the expression of oxidative stress markers. Since Nrf2 regulates the expression of antioxidant genes, we aimed to investigate the effect of NP on alcohol-induced oxidative stress. In immunofluorescence staining, the expression of the lipid peroxidation marker 4-hydroxynonenal (4-HNE) was significantly higher in the brain microvessels of ED animals treated with CP (F(2,15) = 52.65, P < 0.0001; Figure 4, A and B). As expected, the absence of Nrf2 in Nrf2–/– mice led to a further increase in the expression of 4-HNE in CP-treated ED groups (Figure 4, A and B). In Western blotting, statistical analysis indicated a significant effect of treatment on NADPH oxidase 1 (NOX1) and its lipid peroxidation product 4-HNE (NOX1: F(3,28) = 14.42, P < 0.0001; 4-HNE: F(3,28) = 12.51, P < 0.0001). Subsequent comparisons showed a significantly increased expression of NOX1 and its product 4-HNE in ED+CP WT animals compared with CD+CP WT controls (P < 0.0001). However, treatment with NP decreased the expression of NOX1 and 4-HNE in the ED+NP WT group, and this effect was not observed in ED+NP Nrf2–/– mice (P < 0.0001; Figure 4, C–E). Next, we tested the effect of alcohol and NP on the formation of the lipid peroxidation product MDA in the brain frontal cortex tissue lysate and blood plasma of CD or ED WT and Nrf2–/– mice (F(3,20) = 5.51, P < 0.0001). In brain frontal cortex tissue lysates, the level of MDA was significantly increased in ED+CP WT mice compared with CD+CP WT controls (P < 0.0001; Figure 4F). As predicted, the level of MDA was significantly reduced in NP-treated ED WT mice. However, no significant effect of NP was observed in ED Nrf2–/– mice. A similar trend was observed in blood plasma samples (F(3,20) = 9.36, P < 0.0001; Figure 4G).

Nrf2 peptide attenuates oxidative stress. (A and B) Immunofluorescence staining of lipid peroxidation marker 4-HNE in the intact brain microvessels of mice fed with CD or ED with CP or NP treatments. Scale bar: 10 μm. (B) Quantification of 4-HNE staining analyzed using ImageJ software (n = 10 per group). (C–E) Western blot analysis of NOX1, 4-HNE, and β-actin from the mouse cortex tissue lysate of CD- and ED-fed animals in WT and Nrf2-knockout animals (Nrf2–/–) with CP or NP treatments. Bar graphs with dot plot show the ratio densitometry of NOX1 (D) and 4-HNE (E) to β-actin (n = 8 per group). All values are expressed as mean ± SD. (F and G) Level of MDA generation from the mouse frontal cortex tissue lysate (F) and blood plasma (G) by ELISA in CD- and ED-fed animals of WT and Nrf2 knockout groups with CP or NP treatments (n = 8 per group). All values are expressed as mean ± SD. Statistical analysis was performed by 3-way ANOVA, followed by Bonferroni’s post hoc test. P < 0.05 was considered statistically significant. @P < 0.05, @@P < 0.01, @@P < 0.001 vs. representative groups in WT animals. Exact P values are shown between the compared groups.

Alcohol augments the activation of the ICAM-1 signaling pathway, and Nrf2 peptide modulates it. Since we observed in one of our previous studies that ICAM-1, along with its receptors LFA-1 and Mac-1, is crucial for transmigration (44), we next analyzed the effect of alcohol on this ligand and receptors and how Nrf2 regulates it. ANOVA of Western blot data showed a significant main effect of treatment on the expression of ICAM-1, LFA-1, and Mac-1 (ICAM-1: F(3,28) = 47.88, P < 0.0001; LFA-1: F(3,28) = 59.09, P < 0.0001; Mac-1: F(3,28) = 36.07, P < 0.0001), and pairwise analysis showed significantly elevated expression of ICAM-1, LFA-1, and Mac-1 in the ED+CP group in both WT and Nrf2–/– mice compared with the CD groups of WT and Nrf2–/– mice, respectively (P < 0.0001; Figure 5, A–D). However, treatment with NP significantly decreased the expression of ICAM-1, LFA-1, and Mac-1 in the ED+NP WT group (P < 0.0001; Figure 5, A–D). NP treatment did not cause any significant change in the expression of ICAM-1, LFA-1, and Mac-1 in the ED group of Nrf2–/– mice (P < 0.0001; Figure 5, A–D). Immunofluorescence staining of ICAM-1 and LFA-1 showed a significant increase in their expression in ED+CP WT animals compared with the CD+CP WT group (F(3,20) = 14.61, P < 0.0001; Figure 5, E and F). A further increase in the expression of ICAM-1 and LFA-1 was observed in Nrf2–/– animals (Figure 5, E and F). Next, we analyzed the level of ICAM-1 protein in blood plasma using ELISA. These results revealed a significantly increased level of ICAM-1 in the ED+CP WT group compared with the CD group of both WT and Nrf2–/– mice (F(3,20) = 11.86, P < 0.0001; Figure 5G). However, NP treatment reduced ICAM-1 expression in ED+NP WT group mice. Notably, NP treatment did not alter ICAM-1 expression in the ED+NP group of Nrf2–/– mice (Figure 5G).

Alcohol augments the activation of ICAM-1 and its receptors LFA-1 and Mac-1. (A–D) Western blot analysis of ICAM-1, LFA-1, Mac-1, and β-actin from the mouse cortex tissue lysate of CD- and ED-fed animals in WT and Nrf2–/– with CP or NP treatments. Bar graphs with dot plot show the ratio densitometry of ICAM-1 (B), LFA-1 (C), and Mac-1 (D) to β-actin (n = 8 per group). (E) Immunofluorescent staining of ICAM-1 (green) merged with its receptor LFA-1 (red) and DAPI (blue) in the brain cortical tissue sections of CD- and ED-fed animals in WT and Nrf2–/– with CP or NP treatments. Scale bar: 20 μm. (F) Quantification of ICAM-1 and LFA-1 staining analyzed using ImageJ software (n = 6). (G) ELISA quantification of ICAM-1 in the tissue lysate of CD- and ED-fed animals in WT and Nrf2–/– with CP or NP treatments (n = 8 per group). All values are expressed as mean ± SD. Statistical analysis was performed by 3-way ANOVA followed by Bonferroni’s post hoc test. P < 0.05 was considered statistically significant. @P < 0.05, @@P < 0.01, @@P < 0.001 vs. representative groups in WT animals. Exact P values are shown between the compared groups.

Nrf2 peptide protects the brain from pericyte loss and maintains BBB integrity in alcohol ingestion. Since brain pericytes play a key role in maintaining neurovascular stability and integrity (45), we next aimed to assess the expression levels of PDGF-B (expressed in brain endothelial cells) and PDGFR-β (expressed in pericytes) to investigate how alcohol compromises pericyte function in maintaining BBB integrity, and to determine how NP protects this integrity. In double immunofluorescence, we studied the expression of PDGF-B and PDGFR-β (46, 47) in intact brain microvessels. We found that the levels of PDGF-B and PDGFR-β1 were significantly reduced in ED+CP mice compared with the CD+CP group (P < 0.0001; Figure 6, A–C). NP treatment protected the brain microvessels by preserving PDGF-B and PDGFR-β1 protein levels in ED+NP animals compared with ED+CP (Figure 6, A–C).

Alcohol impairs BBB integrity and Nrf2 peptide protects it. (A) Immunofluorescence staining of PDGF-B (red) colocalized with its receptor PDGFR-β1 (green) and merged with DAPI (blue) in the intact brain microvessels of mice fed with CD or ED with CP or NP treatments. Scale bar: 25 μm. (B and C) Quantification of PDGF-B (B) and PDGFR-β1 (C) staining analyzed using ImageJ software (n = 6 per group). (D) Western blot analysis of PDGF-B, PDGFR-β1, and β-actin from the mouse cortex tissue lysate of CD- and ED-fed animals in WT and Nrf2–/– groups with CP or NP treatments. Bar graph with dot plot shows the ratio densitometry of PDGF-B and PDGFR-β1 to β-actin (n = 6 per group). (E) Western blot analysis of integrin α6, integrin β1, and β-actin from the mouse cortex tissue lysate of CD- and ED-fed animals in WT and Nrf2–/– groups with CP or NP treatments. Bar graph with dot plot shows the ratio densitometry of integrin α6 and integrin β1 to β-actin (n = 6 per group). All values are expressed as mean ± SD. Statistical analysis was performed by 2-way ANOVA (B) and 3-way ANOVA (C and D) followed by Bonferroni’s post hoc test. P < 0.05 was considered statistically significant. ***P < 0.001 between the bars, as shown in the figure. @P < 0.05, @@P < 0.01, @@@P < 0.001 vs. representative groups in WT animals. Exact P values are shown between the compared groups.

In Western blotting, using Nrf2−/− mice, we validated the role of Nrf2 in activating the PDGF-B/PDGFR-β signaling pathway. The expression levels of PDGF-B and PDGFR-β1 in ED+CP WT animals were significantly decreased (reduced to approximately one-quarter) compared with those in CD+CP WT mice (F(7,48) = 8.48, P < 0.0001) (Figure 6D). As predicted, in NP-treated animals, the expression levels of PDGF-B and PDGFR-β1 were significantly increased in ED+NP WT samples compared with ED+CP WT samples (P < 0.0001). However, in ED+CP Nrf2−/− mice, the expression of PDGF-B and PDGFR-β1 was further reduced in comparison with ED+CP WT animals (P < 0.0001), and NP treatment did not cause any change in the expression of these two proteins in ED+NP Nrf2−/− mice (Figure 6D).

Both pericytes and endothelial cells are attached to the extracellular matrix (ECM) proteins of the basement membrane by different integrins (8, 48). To test the hypothesis that alcohol compromises the expression level of integrins and affects the integrity of the BBB and NP restores the functionality of the BBB, we analyzed the expression level of two different integrins — integrin α6 and integrin β1 — by Western blotting. The expression level of these two high–molecular weight integrins was significantly reduced in ED+CP animals compared with CD+CP control animals (reduced to one-third for integrin α6, and one-half for integrin β1; P < 0.0001; Figure 6E); however, no reduction was observed in the expression of these two integrins in NP-treated ED WT animals. Consistent with the genotype, in ED+NP Nrf2−/− samples, the expression of integrin α6 and integrin β1 was not improved in comparison with ED+NP WT samples (Figure 6E).

Nrf2 peptide ameliorates the BBB tight junction damage in alcohol ingestion. In our previous study, we demonstrated the induction of oxidative stress and BBB damage following alcohol ingestion (11), but here we demonstrate that alcohol-induced BBB damage can be repaired by treatment with NP. We studied the neuroprotective effect of NP by analyzing the BBB components as assessed by the expression level of tight junction proteins such as claudin-5 and occludin and a junctional adhesion molecule, JAM-A, in the brain microvessels of WT animals by immunostaining and Western blotting. In double immunostaining, we colocalized claudin-5 and occludin with vWF (a specific microvessel marker) to highlight the expression in brain microvessels. Variance analysis (2-way ANOVA) revealed significant main effects of alcohol treatment on claudin-5 expression in brain microvessels (F(3,28) = 11.66, P < 0.0001). Follow-up pairwise comparisons showed that the expression level of claudin-5 in brain microvessel cross sections was significantly reduced (to one-half; P < 0.0001) in ED+CP animals compared with CD+CP mice, whereas the decrease in claudin-5 in the ED group was mitigated by NP in ED+NP mice (Figure 7, A and B). Next, when we analyzed the expression level of occludin in intact brain microvessels by immunofluorescence staining, variance analysis indicated a significant main effect of alcohol on occludin expression (F(3,28) = 37.91; P < 0.0001), and the protein expression of occludin was significantly reduced in ED+CP animals compared with CD+CP controls (to one-half; P < 0.0001) (Figure 7, C and D). As expected, the expression of occludin protein was significantly increased (nearly restored to the level of CD+CP) in ED animals treated with NP (P < 0.0001; Figure 7D). Consistent with the highly branched architecture of cerebral microvessels, alcohol-induced injury resulted in discontinuous and fragmented tight junction protein staining (claudin-5 and occludin) (Figure 7, A and C), reflecting pathological disruption of microvascular integrity.

Nrf2 peptide protects the BBB in alcohol ingestion. (A) Immunofluorescence staining of claudin-5 (red) colocalized with vWF (green) and merged with DAPI (blue) in the cross section of intact brain microvessels of mice fed with CD or ED with CP or NP treatments. Scale bar: 25 μm. (B) Quantification of claudin-5 staining analyzed using ImageJ software (n = 8 per group). (C) Immunofluorescence staining of occludin colocalized with vWF (green) and merged with DAPI (blue) in the intact brain microvessels of animals fed with CD or ED with CP or NP treatments. Scale bar: 10 μm. (E–H) Western blot analysis of claudin-5, occludin, JAM-A, and β-actin from the mouse cortex tissue lysate of CD- and ED-fed animals in WT and Nrf2–/– groups with CP or NP treatments. Bar graphs with dot plot show the ratio densitometry of claudin-5 (F), occludin (G), and JAM-A (H) to β-actin (n = 8 per group). All values are expressed as mean ± SD. Statistical analysis was performed by 2-way ANOVA (B and D) and 3-way ANOVA (F and G) followed by Bonferroni’s post hoc test. Statistically significant, *P < 0.05, **P < 0.01, ***P < 0.001 vs. their representative control groups (CD) in B and D. @P < 0.05, @@P < 0.01, @@@P < 0.001 vs. representative groups in WT animals. P < 0.05 was considered statistically significant. Exact P values are shown between the compared groups in F–H.

We validated the expression of these BBB proteins and the regulatory role of Nrf2 in alcohol ingestion by Western blotting. When we examined the expression level of claudin-5 and occludin in ED+CP animals, we observed a significant decrease in the expression of claudin-5 and occludin in WT mice compared with CD+CP WT animals (claudin-5: reduced from 0.88 to 0.255; F(7,40) = 25.34, P < 0.0001; occludin: reduced from 0.8383 to 0.315; F(7,40) = 47.08, P < 0.0001) (Figure 7, E–G). Activation of Nrf2 by NP treatment protected the expression level of claudin-5 and occludin in ED+NP WT samples compared with CP-treated ED samples (P < 0.0001). However, in ED+CP Nrf2−/− mice, the expression of claudin-5 and occludin was further reduced in comparison with ED+CP WT animals (P < 0.0001), and NP treatment did not improve the expression of these two tight junction proteins (Figure 7, E–G). A similar trend was observed in JAM-A in Western blotting, where it was reduced to one-sixth in ED+CP WT samples compared with CD+CP WT controls (F(7,40) = 132.2, P < 0.0001); however, in ED+NP Nrf2−/− samples, the expression of JAM-A was not protected in comparison with ED+NP WT samples (Figure 7, E and H).

Nrf2 peptide protects the brain from BBB permeability in alcohol ingestion. Alcohol-induced disruption of BBB integrity was analyzed using the permeability of sodium fluorescein (NaFl) and Evans blue (EB) tracers across the BBB (7–9). Variance analysis revealed significant main effects of alcohol treatment on BBB permeability for both NaFl (MW = 376 Da; NaFl: F(7,40) = 225.4, P < 0.0001) and EB (MW = 961 Da; EB: F(7,40) = 84.65, P < 0.0001). Follow-up pairwise comparisons indicated that ED+CP mice showed a marked increase in the permeability of small–molecular weight NaFl (3.36-fold) and large–molecular weight tracer EB (3.5-fold) across the BBB compared with CD+CP controls (P < 0.0001; Figure 8, A and B). However, NP-treated ED mice exhibited a significantly reduced BBB permeability to both tracers compared with CP-treated ED animals (P < 0.0001), whereas all Nrf2–/– animal groups showed a significantly increased permeability of NaFl and EB compared with their respective WT animals (P < 0.05) (Figure 8, A and B). Moreover, as expected, NP treatment in the ED group of Nrf2–/– animals did not reduce the permeability to NaFl and EB compared with NP treatment in the ED WT group (Figure 8, A and B).

Nrf2 peptide attenuates BBB permeability and the transmigration of leukocytes to the brain in alcohol ingestion. (A and B) Graphical representation of BBB permeability showing the leakage of Evans blue (EB; 5 μM) (A) and sodium fluorescein (NaFl; 5 μM) (B) in mice fed with CD or ED with CP or NP treatments (n = 6 per group). (C) Transmigration of leukocytes was analyzed by infusion of cultured GFP monocytes (green) through the jugular vein of mice; immunostaining images show colocalization with Mac-1 (red). The second panel in each row shows selected, enlarged, and colocalized staining of GFP and Mac-1. The third and fourth panels of each row show GFP and Mac-1 staining, respectively. Scale bars: 400 μm in first panel of each row; 80 μm in second, third, and fourth panels. In second panels, the colocalized yellow cells are infused GFP cells stained with Mac-1 (yellow arrows); and red-stained cells (Mac-1 alone) are endogenous blood cells (white arrows). (D and E) Quantitative analysis of the number of GFP+ (D) and Mac-1+ (E) cells. Data are shown as mean ± SD. Statistically significant, *P < 0.05 and ***P < 0.001 vs. their representative control groups (CD) in WT and Nrf2–/–. @P < 0.05, @@P < 0.001, and @@@P < 0.001 vs. representative groups in WT animals. Statistical analysis was performed by 3-way ANOVA with Bonferroni’s post hoc test.

Nrf2 peptide protects the brain from the transmigration of immune cells to the brain in alcohol ingestion. Next, we focused on investigating the mechanisms underlying the transmigration of leukocytes into the brain after alcohol ingestion. We validated the role of Nrf2 in leukocyte transmigration using Nrf2–/– animals, and the therapeutic effect of NP was confirmed. For this experiment, we used GFP+ bone marrow–derived macrophages that were differentiated from monocytes isolated and cultured from GFP-transgenic mice. We infused these GFP+ macrophages through the jugular vein in WT or Nrf2–/– ED or CD mice with CP or NP treatments. The GFP+ macrophages were detected under a fluorescent microscope. We validated the transmigration of blood cells by detecting GFP+ cells through immunofluorescence using an anti–Mac-1 antibody (a monocyte/macrophage marker).

A 2-way ANOVA revealed significant main effects of treatment and genotype on the number of infiltrated GFP+ and Mac-1+ macrophages (GFP+ cells: F(7,40) = 86.84, P < 0.0001; Mac-1+ cells: F(7,40) = 113.1, P < 0.0001). Post hoc analyses showed a markedly higher number of the GFP+ and Mac-1+ macrophages in the brain tissue sections of ED+CP WT mice compared with CD+CP control mice (P < 0.0001; Figure 8, C–E). Interestingly, NP treatment reduced the alcohol-induced transmigration of leukocytes into the brain, and very few GFP+Mac-1+ cells were observed in the ED+NP group compared with ED+CP animals (P < 0.0001; Figure 8, C–E). Most of the GFP+ cells were colocalized with Mac-1 (yellow arrows); however, the Mac-1+ cells that were not colocalized with GFP (white arrows) indicated the transmigration of endogenous blood cells into the brain (Figure 8, C–E). In CP- or NP-treated ED Nrf2–/– animals, the number of GFP+ and Mac-1+ cells was significantly higher in comparison with their respective WT groups, and NP treatment had no effect in these Nrf2–/– animals (P < 0.0001; Figure 8, C–E). These data demonstrate that Nrf2 plays a significant role in regulating immune cell transmigration during alcohol exposure.