NADPH Oxidase Inhibition Promotes Brain Resilience by Attenuating Tauopathy and Neuroinflammation in Alzheimer's Disease

Abstract

Alzheimer's disease (AD) associates closely associated with the activation of NADPH oxidase (Nox) isozymes. CRB-2131, a novel oxadiazole derivative, is identified as a potently suppresses Nox isozymes. It inhibits reactive oxygen species production (ROS) by hippocampal neuronal and microglial cells and reduces microglial activation. Prophylactic (starting at 3.5 months of age) and therapeutic (starting at 6 months of age) oral administration with CRB-2131 for 10 weeks in 5XFAD mice reduced hippocampal superoxide levels, lipid peroxidation, Tau phosphorylation, and neuroinflammation. Prophylactic and therapeutic CRB-2131 treatment of 5XFAD mice restored their impaired cognition as shown by the novel-object recognition, Y-maze, and Morris water-maze tests. CRB-2131 treatment increased mature neurons, reduced apoptotic mature neurons, and elevated immature neurons in the hippocampus. Positron-emission tomography/computed-tomography imaging confirmed that CRB-2131 stimulated neuronal regeneration. CRB-2131 suppresses brain oxidation, tauopathy, and neuroinflammation, thereby preventing mature neuron death and promoting neuron regeneration. Ultimately, this fosters a resilient brain and protects cognition.

Keywords: NADPH oxidase inhibitor; alzheimer's disease; brain resilience; neuroinflammation; tauopathy.

Figure 1

Chemical structure of CRB‐2131 and its effect on neuronal and microglial ROS production, Tau phosphorylation, and neuroinflammation. A) Chemical structure of CRB‐2131. B) Concentration‐dependent Nox inhibition curves of CRB‐2131. Drosophila membranes specifically overexpressing hNox1, hNox2, hNox4, hNox5, hDuox1, and hDuox2 were subjected to ROS measurement with Diogene. The IC₅₀ value of CRB‐2131on hNox1, hNox2, hNox4, hNox5, hDuox1, and hDuox2. C and D) Dose‐dependent inhibitory activity of CRB‐2131 on Aβ‐mediated ROS generation (C) and Tau phosphorylation (D) in HT‐22 cells. Three independent experiments (N = 3, mean ± SD, F (5, 12) = 5.604, p = 0.0068). E) Dose‐dependent inhibitory activity of CRB‐2131 on Aβ‐mediated ROS generation in BV2 microglial cells. (N = 3, mean ± SD, F (4, 10) = 6.505, p = 0.0076). F) Transwell migration assay. BV2 microglial cells were seeded into the upper chamber, and stimulating medium containing Aβ was added to the lower chamber. After 24 hrs, the migrated cells were stained with hematoxylin/eosin and counted. Representative images for each group. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001, as determined by one‐way ANOVA followed by Dunnett's multiple comparisons test.

Figure 2

CRB‐2131 inhibits ROS generation and Tau phosphorylation in 5XFAD mice. (A and B) Ability of prophylactic CRB‐2131 (1, 3, or 10 mg kg⁻¹ of CRB‐2131, q.d., P.O., and 10 weeks) to suppress ROS generation in the DG of 5XFAD mice, as determined by DHE staining after sacrifice. The mpk means mg kg⁻¹. A) Representative images. Scale bar = 100 µm. B) Quantification of DHE levels. Data are expressed as mean ± SEM. One‐way ANOVA: F(4, 63) = 6.958, p = 0.0001. WT (n = 16), 5xFAD + vehicle (n = 16), CRB‐2131 1 mpk (n = 7), 3 mpk (n = 8), 10 mpk (n = 21)) (C and D) Ability of prophylactic CRB‐2131 to inhibit 4‐HNE levels in the brain of 5XFAD mice. C) Representative images. Scale bar = 100 µm. D) Quantification of 4‐NHE levels. Data are expressed as mean ± SEM. One‐way ANOVA: cortex; F(4, 26) = 17.26, hippocampus; F(4, 26) = 13.25 p<0.0001. (WT (n = 6), 5xFAD + vehicle (n = 5), CRB‐2131 1 mpk (n = 7), 3 mpk (n = 8), 10 mpk (n = 5)). E) Ability of prophylactic CRB‐2131 to inhibit Tau phosphorylation (pTau) in the brain of 5XFAD mice. (WT (n = 4), 5xFAD + vehicle (n = 5), CRB‐2131 10 mpk (n = 4)). All quantitative data in this figure are shown as mean ± SEM. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001, as determined by one‐way ANOVA followed by Turkey's post‐hoc test.

Figure 3

Prophylactic and therapeutic CRB‐2131 suppresses Tau phosphorylation and neuronal inflammation in 5XFAD mice. A) Prophylactic CRB‐2131‐treated 5XFAD mice (1, 3, or 10 mg kg⁻¹ of CRB‐2131, q.d., P.O., and 10 weeks) were subjected to immunohistochemistry of brain Aβ (Scale bar = 500 µm), phosphorylated Tau (pTau) in the cortex and the hippocampus, Iba1⁺ for microglial cells, and GFAP⁺ for astrocytes (Scale bar = 100 µm). Immunohistochemistry of Aβ (WT (n = 10), 5xFAD + vehicle (n = 5), CRB‐2131 1 mpk (n = 7), 3 mpk (n = 8), 10 mpk (n = 7)), pTau in cortex and DG (WT (n = 6), 5xFAD + vehicle (n = 8), CRB‐2131 1 mpk (n = 7), 3 mpk (n = 8), 10 mpk (n = 10)), Iba1 in DG (WT (n = 10), 5xFAD + vehicle (n = 5), CRB‐2131 1 mpk (n = 7), 3 mpk (n = 8), 10 mpk (n = 7)), and GFAP in DG (WT (n = 10), 5xFAD + vehicle (n = 5), CRB‐2131 1 mpk (n = 7), 3 mpk (n = 8), 10 mpk (n = 7)) of 5XFAD mice with or without CRB‐2131 dose‐dependent administration. The mpk means mg kg⁻¹. B) Therapeutic CRB‐2131‐treated 5XFAD mice (6.0‐month‐old mice, 10 mg kg⁻¹ of CRB‐2131, q.d., P.O., and 10 weeks) were subjected to immunohistochemistry of brain phosphorylated Tau (pTau) in the cortex and the hippocampus, Iba1⁺ for microglial cells, and GFAP⁺ for astrocytes. Scale bar = 100 µm. (WT (n = 5‐6), 5xFAD + vehicle (n = 2), CRB‐2131 10 mpk (n = 5), donepezil 10 mpk (n = 5)).

Figure 4

Prophylactic CRB‐2131 improves behavioral‐test outcomes in 5XFAD mice. Untreated WT, vehicle‐treated 5XFAD, and prophylactic CRB‐2131‐treated 5XFAD mice (3.5‐month‐old mice, 1, 3, or 10 mg kg⁻¹ of CRB‐2131, q.d., P.O., and 10 weeks) were subjected to cognition tests. The mpk means mg/kg. A) Novel object recognition (NOR) test. The time spent by the mice exploring the two same‐shaped objects (A and A*) in the familiarization period and the familiar (A) and novel (B) objects in the second stage was quantified. Data are expressed as mean ± SEM. One‐way ANOVA: F(9^, 82) = 4.979, p<0.0001. (WT (n = 9‐10), 5xFAD + vehicle (n = 11), CRB‐2131 1 mpk (n = 7), 3 mpk (n = 7‐8), 10 mpk (n = 12‐13)). (B) Y‐maze test. The frequency of mouse explorations that involved sequentially investigating each of the three arms of the Y‐shaped box (termed spontaneous alteration) was measured. Data are expressed as mean ± SEM. One‐way ANOVA: F(4, 44) = 12.23, p<0.0001. (WT (n = 10), 5xFAD + vehicle (n = 11), CRB‐2131 1 mpk (n = 7), 3 mpk (n = 8), 10 mpk (n = 13)). C–E) Morris water maze (MWM) test. (C) The mice underwent 5 consecutive days of training (4 trials per day) with a fixed platform. Data represent mean ± SEM. Two‐way repeated measures ANOVA: group effect, F(4, 15) = 1.602, p = 0.2251; day effect, F (3.446, 51.68) = 58.60, p < 0.0001; drug effect, F (15, 60) = 1.849, p = 0.0481, WT versus 5xFAD+Vehicle on day 5: p = 0.0446. (WT (n = 10), 5xFAD + vehicle (n = 19), CRB‐2131 1 mpk (n = 7), 3 mpk (n = 8), 10 mpk (n = 23)) (D‐E) The platform was removed on the sixth day, and the following were measured: the average swimming speed of the mice (D) (F(4, 62) = 1.185, p = 0.3262), the time spent in the platform quadrant (E, left) (F(4, 62) = 10.43, p < 0.0001), the time spent at the place the platform used to be (E, middle) (F(4, 62) = 10.45, p < 0.0001), and the number of times the mice swam over the place the platform used to be (E, right) (F(4, 62) = 8.856, p < 0.0001). These quantitative data are shown as mean ± SEM. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001, as determined by one‐way ANOVA followed by Turkey's post‐hoc test.

Figure 5

Therapeutic CRB‐2131 improves behavioral‐test outcomes in 5XFAD mice. Untreated WT (n = 14), vehicle‐treated 5XFAD (n = 8‐9), therapeutic CRB‐2131‐treated 5XFAD mice (n = 14), and donepezil‐treated 5XFAD mice (n = 5) (6.0‐month‐old mice, 10 mg kg⁻¹ of CRB‐2131 and donepezil, q.d., P.O., and 10 weeks) were subjected to cognition tests. The mpk means mg/kg. A) Novel object recognition (NOR) test. The time spent by the mice exploring the two same‐shaped objects (A and A*) in the familiarization period and the familiar (A) and novel (B) objects in the second stage was quantified. F(7^, 78) = 11.48, p<0.0001 B) Y‐maze test. The frequency of mouse explorations that involved sequentially investigating each of the three arms of the Y‐shaped box (termed spontaneous alteration) was measured. F(3^, 38) = 9.218, p = 0.0001 C–E) Morris water maze (MWM) test. (C) The mice underwent 5 consecutive days of training (4 trials per day) with a fixed platform. Data represent mean ± SEM. Two‐way repeated measures ANOVA: group effect, F (3, 12) = 9.939, p = 0.0014; day effect, F (3.076, 36.91) = 54.53, p < 0.0001; drug effect, F (12, 48) = 1.531, p = 0.1460, WT versus 5xFAD+Vehicle on day 3: p = 0.0429, WT versus 5XFAD+Donepezil on day 3: p = 0.0133, WT versus 5xFAD+Vehicle on day 4: p = 0.0311, 5xFAD+Vehicle versus 5xFAD+CRB‐2131 on day 5: p = 0.0311. (D–E) The platform was removed on the sixth day, and the following were measured: the average swimming speed of the mice (D) (F(3, 38) = 0.5299, p = 0.6691), the time spent in the platform quadrant (E, left) (F (3, 38) = 6.147, p = 0.0016), the time spent at the place the platform used to be (E, middle) (F (3, 38) = 5.617, p = 0.0027), and the number of times the mice swam over the place the platform used to be (E, right) (F (3, 38) = 7.854, p = 0.0003). All quantitative data in this figure are shown as mean ± SEM. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001, as determined by one‐way ANOVA followed by Turkey's post‐hoc test.

Figure 6

Prophylactic and therapeutic CRB‐2131 protects 5XFAD mice from mature neuron death and enhances the regeneration of immature neurons. A) Prophylactic CRB‐2131‐treated 5XFAD mice (3.5‐month‐old mice, 1, 3, or 10 mg kg⁻¹ of CRB‐2131, q.d., P.O., and 10 weeks) were subjected to immunohistochemistry to determine the mature neurons (NeuN⁺) and immature neurons (DCX⁺) in the DG. (WT (n = 8), 5xFAD + vehicle (n = 4), CRB‐2131 1mpk (n = 7), 3mpk (n = 8), 10mpk (n = 5)) Representative images of NeuN (Scale bar = 100 µm) and DCX staining (Scale bar = 250 µm). The mpk means mg/kg. B) The TUNEL staining of the mature neurons was also assessed. Co‐localization of NeuN expression and TUNEL staining in the hippocampus (10 mg kg⁻¹ CRB‐2131). Scale bar = 100 µm. C) Quantification of the mature (NeuN⁺) TUNEL‐stained hippocampal neurons. Data are expressed as mean ± SEM. One‐way ANOVA: F (2^, 21) = 18.44, p<0.0001. (n = WT (n = 8), 5xFAD + vehicle (n = 8), CRB‐2131 10 mpk (n = 8)). D) Therapeutic CRB‐2131‐treated 5XFAD mice (6.0‐month‐old mice, qd, P.O., 10 weeks) were subjected to immunohistochemistry to determine the mature neurons (NeuN⁺) (WT (n = 6), 5xFAD + vehicle (n = 2), CRB‐2131 10 mpk (n = 5), donepezil 10 mpk (n = 5)), (Scale bar = 100 µm) and immature neuron (DCX⁺) in the DG. (WT (n = 8), 5xFAD + vehicle (n = 7), CRB‐2131 10 mpk (n = 9), donepezil 10 mpk (n = 5)), (Scale bar = 250 µm). All quantitative data in this figure are shown as mean ± SEM. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001, as determined by one‐way ANOVA followed by Turkey's post‐hoc test.

Figure 7

Prophylactic and therapeutic CRB‐2131 promotes neurogenesis, as shown by the colocalization of BrdU with mature and immature neurons. A,B) Colocalization of BrdU with mature neuron (NeuN) and immature neuron (DCX⁺) in DG of 5XFAD mice (3.5‐month‐old mice, q.d., P.O., and 10 weeks) with or without CRB‐2131 administration (10 mg kg⁻¹). The mpk means mg/kg. (WT (n = 6), 5xFAD + vehicle (n = 4), CRB‐2131 10mpk (n = 5)) (A) Quantification of mature neuron (NeuN) and BrdU colocalization of hippocampus in 5XFAD mice with or without CRB‐2131 administration (10 mg kg⁻¹). Data are expressed as mean ± SEM. One‐way ANOVA: F (2^, 12) = 15.44, p = 0.0005. B) Quantification of immature neuron (DCX⁺) and BrdU colocalization of the hippocampus in 5XFAD mice with or without CRB‐2131 administration (10 mg kg⁻¹). Data are expressed as mean ± SEM. One‐way ANOVA: F (2^, 12) = 4.709, p = 0.0309. C,D) Colocalization of BrdU with mature neuron (NeuN) and immature neuron (DCX⁺) in DG of 5XFAD mice (6.0‐month‐old mice, q.d., P.O., and 10 weeks) with or without CRB‐2131 administration (10 mg kg⁻¹) or donepezil (10 mg kg⁻¹). C) Quantification of mature neuron (NeuN) and BrdU colocalization of hippocampus in 5XFAD mice with or without CRB‐2131 administration (10 mg kg⁻¹) or donepezil (10 mg kg⁻¹). Data are expressed as mean ± SEM. One‐way ANOVA: F (3^, 14) = 4.853, p = 0.0161. (WT (n = 6), 5xFAD + vehicle (n = 3), CRB‐2131 10 mpk (n = 5), donepezil 10 mpk (n = 4)). (D) Quantification of immature neuron (DCX⁺) and BrdU colocalization of hippocampus in 5XFAD mice with or without CRB‐2131 administration (10 mg kg⁻¹) or donepezil (10 mg kg⁻¹). Data are expressed as mean ± SEM. One‐way ANOVA: F (3, 15) = 4.817, p = 0.0152. (WT (n = 6), 5xFAD + vehicle (n = 3), CRB‐2131 10 mpk (n = 5), donepezil 10 mpk (n = 5)). All quantitative data in this figure are shown as mean ± SD. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001, as determined by one‐way ANOVA followed by Turkey's post‐hoc test.

Figure 8

PET/CT testing for neuronal regeneration in prophylactic CRB‐2131‐treated 5XFAD Prophylactic CRB‐2131‐treated 5XFAD mice (3.5‐month‐old mice, q.d., P.O., and 10 weeks) underwent PET/CT at 6 months of age. (WT (n = 5), 5xFAD + vehicle (n = 5), CRB‐2131 10 mpk (n = 5)) (A and B) Ratio of SUVglc in (A) control vehicle‐treated 5XFAD mice relative to WT mice and (B) vehicle‐treated 5XFAD mice relative to CRB‐2131‐treated 5XFAD mice. C) Functional connectivity between vehicle‐treated and CRB‐2131‐treated 5XFAD mice. D) Proposed model by which CRB‐2131 improves cognitive ability in the 5XFAD model of AD. Specifically, by suppressing Nox‐induced oxidative stress, CRB‐2131 reduces Tau hyperphosphorylation and inhibits neuroinflammatory pathways, thereby facilitating the regeneration and repair of neural tissues.