The Usnic Acid Analogue 4-FPBUA Enhances the Blood–Brain Barrier Function and Induces Autophagy in Alzheimer’s Disease Mouse Models

Abstract

Preclinical and clinical studies have indicated that compromised blood–brain barrier (BBB) function contributes to Alzheimer’s disease (AD) pathology. BBB breakdown ranged from mild disruption of tight junctions (TJs) with increased BBB permeability to chronic integrity loss, affecting transport across the BBB, reducing brain perfusion, and triggering inflammatory responses. We recently developed a high-throughput screening (HTS) assay to identify hit compounds that enhance the function of a cell-based BBB model. The HTS screen identified (S,E)-2-acetyl-6-[3-(4′-fluorobiphenyl-4-yl)acryloyl]-3,7,9-trihydroxy-8,9b-dimethyldibenzo-[b,d]-furan-1(9bH)-one (4-FPBUA), a semisynthetic analogue of naturally occurring usnic acid, which protected the in vitro model against Aβ toxicity. Usnic acid is a lichen-derived secondary metabolite with a unique dibenzofuran skeleton that is commonly found in lichenized fungi of the genera Usnea. In this study, we aimed to evaluate the effect of 4-FPBUA in vitro on the cell-based BBB model function and its in vivo ability to rectify BBB function and reduce brain Aβ in two AD mouse models, namely, 5xFAD and TgSwDI. Our findings demonstrated that 4-FPBUA enhanced cell-based BBB function, increased Aβ transport across the monolayer, and reversed BBB breakdown in vivo by enhancing autophagy as an mTOR inhibitor. Induced autophagy was associated with a significant reduction in Aβ accumulation and related pathologies and improved memory function. These results underscore the potential of 4-FPBUA as a candidate for further preclinical exploration to better understand its mechanisms of action and to optimize dosing strategies. Continued research may also elucidate additional pathways through which 4-FPBUA contributed to the amelioration of BBB dysfunction in AD. Collectively, our findings supported the development of 4-FPBUA as a therapeutic agent against AD.

Graphical Abstract

The blood–brain barrier (BBB) is a highly selective permeable barrier formed by a monolayer of endothelial cells that lines the brain microvasculature and plays a dynamic role in regulating the exchange of molecules between the peripheral circulation and the brain.² A major function of the BBB is to prevent the entry of blood components such as plasma proteins, and neurotoxic and proinflammatory factors into the CNS, as well as the elimination of waste products and toxic molecules across the BBB.^3,4 Tight junction proteins, including, occludin, claudins, and ZO-1, enhance this barrier by regulating paracellular transport and maintaining selective permeability.⁵ Recent clinical and experimental studies have supported BBB breakdown as one of the earliest events in the pathogenesis of Alzheimer’s disease (AD),³ associated with a reduction in cerebral blood flow (CBF), and hemodynamic dysfunction.⁶ AD is a progressive neurodegenerative disorder that primarily affects older adults, accounting for 60–70% of cases of dementia worldwide.⁷ As the population ages, the prevalence of AD is expected to increase, with projections suggesting that the number of people affected could rise to over 152 million by 2050,⁸ underscoring the urgent need for effective therapies targeting early disease mechanisms like BBB dysfunction. Further, recent imaging and biomarker studies have indicated that early BBB disruption and vascular breakdown in AD are detectable before cognitive decline and/or other brain pathologies.⁹ Emerging research has highlighted several biomarkers indicative of early BBB dysfunction, such as the presence of soluble platelet-derived growth factor receptor β (s-PDGFRβ),¹⁰ matrix metal-loproteinases (MMPs),¹¹ and S100 calcium-binding protein B (S100B), which may serve as early indicators of AD before the onset of cognitive decline.¹² Several AD reports have indicated that a correlation between disrupted BBB integrity and functional activity that triggers an imbalance in the levels of the two major AD hallmark proteins, amyloid-β (Aβ) and tau protein, which subsequently leads to Aβ accumulation and tau hyperphosphorylation observed in AD.^13,14 Both hallmarks are associated with further disruption of cerebrovascular function. According to the “two-hit vascular hypothesis of AD”, initial cerebrovascular damage, characterized by reduced blood flow and BBB breakdown, precedes and may even trigger Aβ pathology. The first hit is sufficient to initiate neuronal injury and cognitive impairment by altering the brain microenvironment and exacerbating vascular reactivity. The second hit involves alterations in Aβ production and impaired Aβ clearance, driven by the ongoing vascular dysfunction.¹⁵ The concept of the “two-hit” hypothesis of AD offers potential targets for intervention. Despite current advances, clinically relevant approaches for the prevention or treatment of BBB breakdown in AD are not available.

Recently, we have developed a high-throughput screening (HTS) assay to identify hit compounds that have the potential to restore the function of a cell-based BBB with cerebral amyloid angiopathy (CAA) characteristics.¹⁶ Results from this screen identified multiple hit compounds, among which is an (S,E)-2-acetyl-6-[3-(4′-fluorobiphenyl-4-yl)acryloyl]-3,7,9-trihydroxy-8,9b-dimethyldibenzo-[b,d]furan-1(9bH)-one (4-FPBUA) analogue of the naturally occurring usnic acid as a potential compound to protect the BBB model against Aβ toxicity. Usnic acid is a lichen-derived secondary metabolite with a unique dibenzofuran scaffold and is commonly found in lichenized fungi of the genera Usnea, Ramalina, and Cladonia.¹ We have reported the discovery of several analogues of usnic acid using the Claisen–Schmidt condensation reaction, including the 4-FPBUA analogue (Figure 1A), which demonstrated a potent inhibitory effect on the mammalian/mechanistic target of the rapamycin (mTOR) pathway.¹ The mTOR pathway has recently gained much attention as a target for the prevention of BBB breakdown through autophagy restoration in several models of age-associated neurodegeneration, suggesting a significant role of mTOR in BBB dysregulation in AD.^17,18 Recent studies have demonstrated that pharmacological inhibition of the mTOR pathway with rapamycin, an FDA-approved drug that inhibits mTOR, exerts a significant protective effect in mouse models of AD.^19,20 In AD, enhancing autophagy is crucial for the removal of Aβ plaques. Key proteins like mTOR and AMPK regulate this process; mTOR inhibits autophagy, while AMPK promotes autophagy. Proteins such as ATG5 and LC3B facilitate autophagosome formation, which is crucial for degrading toxic proteins, thus potentially managing AD progression by improving cellular cleanup.²¹

Figure 1.

Chemistry of 4-FPBUA and its effects on BBB function in an in vitro bEnd3 cell-based model (A) The chemical structure of 4-FPBUA and its correlation with its natural parent usnic acid. (B) 4-FPBUA reduced LY permeation through a bEnd3 cell-based BBB model at concentrations of 5 and 10 μM. (C) 4-FPBUA increased B to A transport of 125I-Aβ40 and 125I-Aβ40 TQ and decreased the transport of 14C-inulin. (D) 4-FPBUA treatment increased the expression of the tight junction protein ZO-1, and the transport proteins P-gp and LRP1. (E) Densitometric analysis of (D) demonstrated that 4-FPBUA caused a significant increase in ZO-1 expression and the transport proteins P-gp and LRP1. Data are presented as the mean ± SD. Statistical analysis was performed using the Student’s t-test for data obtained from 3 independent experiments, **P < 0.01 and ***P < 0.001 versus control (Ctrl).

Treatment with rapamycin restored microvascular endothelial function,²² improved cognitive function, and increased CBF in mouse models of AD.¹⁸ In these AD models, mTOR activation promoted the accumulation of Aβ in the brain by inhibiting autophagy, which plays a housekeeping role in the removal of misfolded or aggregated proteins, clearance of damaged organelles, and elimination of intracellular pathogens.²³ By inhibition of mTOR, autophagic activity is enhanced, leading to increased degradation and clearance of Aβ, thereby reducing its accumulation in the brain. This upregulation of autophagy due to mTOR inhibition allows more autophagosomes to encapsulate Aβ peptides, which are then directed to the lysosomes for efficient degradation. Additionally, enhancement of autophagy can improve endothelial cell function by removing oxidatively damaged proteins and organelles that contribute to cellular dysfunction. Restoration of endothelial health is crucial for maintaining BBB integrity. Thus, targeting the mTOR pathway represents a dual therapeutic strategy, not only diminishing Aβ plaque formation through the activation of autophagy but also potentially restoring BBB function.²¹

The aims of this study were, first, to in vitro confirm the potency of the usnic acid analogue 4-FPBUA in enhancing the cell-based BBB model function and, second, to investigate its effect in vivo on BBB integrity, and autophagy with a focus on the mTOR pathway, overall brain Aβ load, and memory function in two AD mouse models, namely, 5xFAD and TgSwDI mice. Both mouse models carry APP transgenes to overproduce Aβ and are characterized by BBB leakage; however, unlike 5xFAD mouse, the TgSwDI mouse is characterized by vascular deposition of Aβ.^24,25

RESULTS

4-FPBUA Enhanced Tightness of the Cell-Based BBB Model and Aβ Transport across the bEnd3 Cell Monolayer.

A bEnd3 cell-based BBB model was utilized to evaluate the effect of 4-FPBUA on monolayer integrity. The bEnd3 cell line is an immortalized mouse brain endothelial cell line. As shown in Figure 1B–E, 4-FPBUA added to the apical compartment (A) significantly enhanced the tightness of the in vitro model at two tested concentrations, 5 and 10 μM, as demonstrated by the reduced permeation of Lucifer Yellow (LY), a paracellular permeability marker, across the bEnd3 monolayer (Figure 1B). Consistent with the reduced permeation of LY, 4-FPBUA, tested at 10 μM, significantly reduced the permeation of ¹⁴C-inulin, another permeability marker (Figure 1C). This effect of 4-FPBUA was associated with increased expression of the tight junction protein ZO-1 as determined by western blotting (Figures 1D,E). In addition, 4-FPBUA (10 μM) significantly increased the transport of ¹²⁵I-Aβ₄₀ and transport quotient TQ_B→A across the bEnd3 monolayer by 2- and 4-fold, respectively (Figure 1C). This effect was associated with a significant increase in the expression of P-gp and LRP1 (Figure 1D,E), which are major Aβ transport proteins expressed in endothelial cells of the BBB, as determined by western blotting. Collectively, these results suggest that endothelial cell treatment with 4-FPBUA enhanced the tightness and functionality of the cell-based BBB model, as monitored by increased tightness and Aβ clearance.

Selection of Mouse Models.

In this study, we evaluated the effect of 4-FPBUA on Alzheimer’s disease-related pathology using two distinct human APP transgenic mouse models: the TgSwDI model, which exhibits cerebral microvascular and parenchymal Aβ pathology, and the 5xFAD model, which exhibits parenchymal Aβ pathology. We selected these models because they share the following key features: both mouse models develop Aβ pathology in the same neuroanatomical regions, including the hippocampus, cortex, thalamus, and subiculum, and produce APP and Aβ peptides in the same sets of neurons.^26–28 In addition, TgSwDI and 5xFAD mice have a comparable onset time of about 2–4 months of age in the development of Aβ and related pathologies. Therefore, these features of TgSwDI and 5xFAD mice provide an excellent opportunity to evaluate the ability of 4-FPBUA treatment to enhance BBB integrity and reduce the cerebrovascular and parenchymal amyloid pathology (in TgSwDI mouse), parenchymal amyloid pathology (in 5xFAD mouse) and associated cognitive impairment.²⁶

4-FPBUA Enhanced BBB Integrity In Vivo in TgSwDI and 5xFAD Mouse Brains.

The effect of 4-FPBUA (5 mg/kg; intraperitoneally (IP)) on BBB tightness was evaluated in both the TgSwDI and 5XFAD mouse models (Figure 2). Twenty-eight days of treatment with 4-FPBUA significantly increased the expression of tight junction proteins ZO-1 and occludin in both models (Figure 2A,B). Western blot analysis revealed an increase of 180% and 80% in ZO-1 and occludin levels, respectively, in the TgSwDI. Similarly, 5xFAD mice displayed a significant increase in ZO-1 (150%) and occludin (55%) expression after treatment. Besides, 4-FPBUA treatment increased the expression of claudin-5 by 22% in 5xFAD mice, but not in TgSwDI mice, as determined by western blotting (Figure 2A,B).

Figure 2.

4-FPBUA (5 mg/kg, IP, 28 days) enhanced BBB function in mouse brains. Mice were treated with 4-FPBUA (5 mg/kg/day) for 28 days. (A) Representative western blots and (B) densitometric analysis of the BBB tight junction proteins ZO-1, occludin, and claudin-5, (C) transport proteins P-gp and LRP1, and Aβ degrading enzyme neprilysin in the brains of TgSwDI and 5xFAD mice. Data are presented as the mean ± SEM of n = 4 5xFAD and n = 5 TgSwDI mice in each group. Statistical analysis was performed using the Student’s t-test. *P < 0.05, **P < 0.01, and ***P < 0.001 versus vehicle.

4-FPBUA Reduced Total Aβ Load in Both TgSwDI and 5xFAD Mouse Brains.

Treatment with 4-FPBUA significantly reduced brain Aβ levels compared to the vehicle groups in both TgSwDI and 5xFAD mouse models. As shown in Figure 3, immunostaining with the 6E10 antibody revealed a reduction in total Aβ in both the hippocampus and cortex of both mouse models. Notably, 4-FPBUA treatment reduced cerebrovascular Aβ deposits in the brains of TgSwDI mice, as indicated by arrows in Figure 3. Similarly, thioflavin-S (Thio-S) staining of Aβ plaques showed a pattern similar to that of 6E10 immunostaining, with a significant overall reduction in Aβ plaques in TgSwDI and 5xFAD mouse models (Figure 4A).

Figure 3.

4-FPBUA (5 mg/kg, IP, 28 days) treatment reduced the Aβ burden in the hippocampus and cortex of TgSwDI and 5xFAD mice. (A) Representative hippocampus sections stained with 6E10 (green) antibody against Aβ to detect total Aβ load, anticollagen IV (red) to stain microvessels, and DAPI as a nuclear counterstain to highlight the dentate gyrus area. White arrows indicate cerebrovascular deposits of Aβ. (B) Representative cortex sections stained with 6E10 (green) antibody against Aβ to detect the total Aβ load and anticollagen IV (red) to stain microvessels. Scale bar: 50 μm.

Figure 4.

Effect of 4-FPBUA (5 mg/kg, IP, 28 days) on Aβ plaques and soluble Aβ monomers in TgSwDI and 5xFAD mouse brains. (A) Representative sections were stained with Thio-S (green) antibody to detect the Aβ plaque load and DAPI as a nuclear counterstain. Scale bar, 100 μm. (B) Brain levels of soluble human Aβ40 and Aβ42, as determined by ELISA. Statistical analysis was performed using the Student’s t-test (n = 4–5 mice per group) *P < 0.05, **P < 0.01, and ***P < 0.001 versus vehicle. Scale bar, 200 μm.

The ELISA analysis quantification of soluble Aβ₄₀ and Aβ₄₂ in brain homogenates from both TgSwDI and 5XFAD mice treated with 4-FPBUA confirmed a significant reduction in brain levels of both peptides. Aβ₄₂ differs from Aβ₄₀ by having two additional C-terminal amino acids, making it more prone to form Aβ plaques in AD brains. 4-FPBUA treatment significantly reduced the brain levels of both Aβ₄₀ (by 72%) and Aβ₄₂ (by 43%) in TgSwDI mice. However, in 5xFAD mice, 4-FPBUA treatment reduced Aβ₄₀ levels by 53% without altering Aβ₄₂ levels (Figure 4B). Collectively, these findings demonstrate that treatment with 4- FPBUA reduced soluble and insoluble Aβ load in both TgSwDI and 5xFAD mouse brains.

4-FPBUA Enhanced the Expression of Aβ Clearance and Protein Production in TgSwDI and 5xFAD Mouse Brains.

To explain the Aβ reduction in mouse brains, the effect of 4-FPBUA on Aβ clearance and protein production was evaluated by western blotting. As shown in Figure 2A,C, 4-FPBUA treatment significantly increased the expression of Aβ major transport proteins, P-gp and LRP1, by 58 and 140%, respectively, in TgSwDI mice and by 35 and 24%, respectively, in 5xFAD mice. In addition, 4-FPBUA significantly increased the levels of the Aβ degrading enzyme neprilysin by 71 and 114% in TgSwDI and 5xFAD mice, respectively.

For Aβ production proteins, 4-FPBUA treatment significantly reduced the expression of BACE1, an enzyme that plays an important role in Aβ production, by 45 and 35% in both TgSwDI and 5xFAD mouse brains, respectively. 4-FPBUA treatment also increased the levels of soluble APPα (sAPPα) by 69 and 43% in TgSwDI and 5xFAD mice, respectively, without altering soluble APPβ (sAPPβ), as shown in Figure 5.

Figure 5.

4-FPBUA (5 mg/kg, IP, 28 days) reduced the Aβ brain load by reducing Aβ production in TgSwDI and 5xFAD mouse brains. Representative blots and densitometry analysis showed that 4-FPBUA treatment significantly increased sAPPα and reduced BACE1 enzyme levels, with no significant effect on the sAPPβ levels in the brain homogenates of TgSwDI and 5xFAD mice. Data are presented as the mean ± SEM of n = 4 5xFAD and n = 5 TgSwDI mice in each group. Statistical analysis was performed using the Student’s t-test. ns = not significant, *P < 0.05, **P < 0.01, and ***P < 0.001 versus the vehicle (Veh) group.

4-FPBUA Treatment-Induced Autophagy Markers in TgSwDI and 5xFAD Mouse Brains.

Autophagy is a cellular process that eliminates waste products and dysfunctional proteins. Multiple studies have shown its important role in regulating Aβ levels, BACE1 turnover, and APP processing,²⁹ which is downregulated with aging and AD.³⁰ As usinic acid has been reported to induce cell arrest and autophagy in cancer cells, both in vivo and in vitro,³¹ we investigated the role of its analogue in modulating autophagy as a mechanism for amyloid reduction. We analyzed several autophagy markers involved in the initiation and maturation of autophagosomes and autophagic flux. For this purpose, the mTOR pathway and its downstream proteins, P70S6K and 4EBP1, were investigated by western blotting. The findings demonstrated that 4-FPBUA significantly reduced the phosphorylated mTOR (p-mTOR) by 47 and 50% in TgSwDI and 5xFAD mice, respectively, without altering the total mTOR levels (Figure 6). This effect was associated with the increased phosphorylation of p-Raptor and reduced phosphorylation of P70S6K to p-P70S6K, and 4EBP1 to p-4EBP1 in both mouse models (Figure 6A–C).

Figure 6.

4-FPBUA (5 mg/kg, IP, 28 days) treatment induced autophagy markers through mTOR pathway downregulation in TgSwDI and 5xFAD mouse brains. (A) Representative western blots and densitometry analysis of mTOR, phosphorylated mTOR (p-mTOR), Raptor, phosphorylated-Raptor (p-Raptor), P70S6K, phosphorylated-P70S6K (p-P70S6K), 4EBP1, and phosphorylated-4EBP1 (p-4EBP1) in the brain homogenates of TgSwDI (B) and 5xFAD (C) mice. Data are presented as the mean ± SEM of n = 4 5xFAD and n = 5 TgSwDI mice in each group. Statistical analysis was performed using the Student’s t-test. ns = not significant, *P < 0.05, **P < 0.01, and ***P < 0.001 versus vehicle.

In addition, 4-FPBUA significantly increased the expression and phosphorylation of the major autophagosome initiation proteins AMPK and ULK1 (Figure 7). This effect was associated with a significant increase in beclin-1, a protein essential for autophagosome maturation, in both TgSwDI (by 23%) and 5xFAD (by 30%) mice. The expression of LKB1 and its phosphorylated form p-LKB1 was also evaluated, and the results showed that 4-FPBUA increased p-LKB1 by 35% in TgSwDI mice, but not in 5xFAD mice, without altering the total levels (Figure 7).

Figure 7.

4-FPBUA (5 mg/kg, IP, 28 days) treatment induced autophagy markers through AMPK pathway activation in TgSwDI and 5xFAD mouse brains. (A) Representative western blots and densitometry analysis of AMPK, phosphorylated AMPK (pAMPK), ULK1, phosphorylated-ULK1 (pULK1), beclin, LKB1, and phosphorylated-LKB1 (p-LKB1) in the brain homogenates of TgSwDI (B) and 5xFAD (C) mice. Data are presented as the mean ± SEM of n = 4 5xFAD and n = 5 TgSwDI mice in each group. Statistical analysis was performed using the Student’s t-test. ns = not significant, *P <. 05, **P <. 01, and ***P <. 001 versus vehicle.

Notably, LKB1 acts as a key upstream kinase that activates AMPK, which, in turn, can inhibit mTOR signaling, thereby promoting autophagy and cellular energy homeostasis.

To better understand the 4-FPBUA effect on the activity of the enzymes involved in protein phosphorylation, we determined the ratio of the phosphorylated protein to total protein in both mouse models. The results are presented in Figure 8, which demonstrate some variability between the mouse models, suggesting different regulatory mechanisms caused by 4-FPBUA.

Figure 8.

Effect of 4-FPBUA (5 mg/kg, IP, 28 days) treatment on the changes in the ratio of phosphorylated to total proteins related to autophagy in TgSwDI (A) and 5xFAD (B) mouse brains. Data are presented as the mean ± SEM of n = 4 5xFAD and n = 5 TgSwDI mice in each group. Statistical analysis was performed using the Student’s t-test. ns = not significant, *P <. 05, **P <. 01, and ***P <. 001 versus vehicle.

4-FPBUA treatment also significantly increased the levels of ATG3, ATG5, and ATG7, which are proteins involved in autophagosome elongation and completion, by up to 2-fold in both mouse models (Figure 9A–C). In addition, the microtubule-associated protein light chain 3 (LC3I), a protein that undergoes conversion to LC3II during autophagosome formation, was significantly reduced by 30 and 20% in 4-FPBUA-treated TgSwDI and 5xFAD mice, respectively, which was associated with 80 and 120% increase in the level of LC3II, respectively, compared to vehicle-treated mice (Figure 9A–C). This conversion indicates increased autophagosome formation. Collectively, these findings suggest that 4-FPBUA induced autophagy via the suppression of the mTOR pathway and the activation of the AMPK pathway.

Figure 9.

4-FPBUA (5 mg/kg, IP, 28 days) treatment induced autophagy-related markers (ATGs) and modulated LC3–1 and LC3II levels in TgSwDI and 5xFAD mouse brains. (A) Representative western blots and densitometry analysis of ATG3, ATG5, ATG7, LC3–1, and LC3II in the brain homogenates of TgSwDI (C) and 5xFAD (B) mice. Data are presented as the mean ± SEM of n = 4 5xFAD and n = 5 TgSwDI mice in each group. Statistical analysis was performed using the Student’s t-test. ns = not significant, *P <. 05, **P <. 01, and ***P <. 001 versus vehicle.

4-FPBUA Increased the Expression of Synaptic Markers and Morris Water Maze Test Performance.

To evaluate the effects of 4-FPBUA on synaptic markers, the expression levels of two major synaptic markers, postsynaptic density protein 95 (PSD95, a postsynaptic marker) and synaptosomal-associated protein (SNAP-25, a presynaptic marker), were evaluated. 4-FPBUA treatment significantly increased the expression of the postsynaptic marker PSD95 by 280 and 220% in TgSwDI and 5xFAD mice, respectively, compared to vehicle-treated mice (Figure 10A). In contrast, SNAP-25 levels were not significantly affected by treatment. The Morris Water Maze (MWM) test was also performed to assess the effect of 4-FPBUA on learning and memory function in mouse models. 4-FPBUA treatment significantly improved the memory performance of TgSwDI mice, which found the hidden platform much faster (15.37 ± 2.54 s) compared to the vehicle-treated mice (28.47 ± 3.05 s). 4-FPBUA treatment significantly reduced TgSwDI’s mice swimming distance to reach the platform from 10.32 ± 1.21 m in the vehicle-treated group to 6.77 ± 0.96 m in the 4-FPBUA-treated group. These results indicate that 4-FPBUA treatment improved spatial learning and memory in TgSwDI mice. On the other hand, in 5xFAD mice, while there was a trend toward improvement in both time to find the platform and swim the distance, 4-FPBUA did not significantly alter either parameter (Figure 10B); however, 4-FPBUA treatment significantly increased the % time spent in the target quadrant by approximately 60%, while in the remaining groups, the % spent was indistinguishable from chance level (less than 25%) (Figure 10B). It is important to note that swimming speed was not affected by the 4-FPBUA treatment in either model, indicating that the observed behavioral changes were not due to motor deficits.³²

Figure 10.

4-FPBUA (5 mg/kg, IP, 28 days) treatment upregulated neurosynaptic markers in the brains of TgSwDI and 5xFAD mice and significantly improved spatial learning and memory, as determined by the MWM test. (A) Representative western blots and densitometry analysis of the neurosynaptic markers PSD95 and SNAP-25 in the brain homogenates of TgSwDI and 5xFAD mice. (B) Typical swimming patterns of vehicle- and 4-FPBUA-treated mice on the fifth day of the MWM test. Data are presented as the mean ± SEM of n = 4 5xFAD and n = 5 TgSwDI mice in each group. Statistical analysis was performed using the Student’s t-test. ns = not significant, *P < 0.05 and **P < 0.01 versus vehicle groups.

DISCUSSION

Recent preclinical studies in AD mouse models demonstrated the pharmacological enhancement of BBB function and autophagy to reduce brain Aβ levels and improve memory.^33–36 Neuroimaging studies in subjects with mild cognitive impairment (MCI) and early AD have reported a breakdown of BBB in several brain regions, including the hippocampus and several gray and white matter regions before brain atrophy or dementia.^9,37 Furthermore, magnetic resonance imaging studies have revealed loss of BBB integrity, as demonstrated by increased levels of microbleeds in 25–45% of AD subjects, compared to 11.1% in asymptomatic or healthy elderly.^38,39 While the precise mechanism leading to BBB dysfunction in AD is not yet understood, the accumulation of Aβ in the brain parenchyma and vasculature could, directly or indirectly, disrupt the BBB by impairing its ability to clear Aβ across the BBB and damaging tight junction proteins.^40–42 A growing body of evidence strongly supports the idea that therapies that modulate the BBB could offer protection and treatment for AD.^43,44

Using HTS, we identified several hit compounds, including 4-FPBUA, which enhanced the integrity and function of a cell-based BBB model.¹⁶ In this work, we aimed to validate the HTS findings in vitro and further evaluate the effect of 4-FPBUA on the BBB function and Aβ brain levels in vivo using two established AD mouse models: TgSwDI and 5xFAD. The TgSwDI mouse expresses a variant of the APP gene with mutations known to cause familial AD and progressively accumulates insoluble Aβ in the hippocampus, cortex, and microvessels starting around 3 months of age. This mouse model also exhibits learning and memory impairments at this age. The 5xFAD mouse carries mutations in five different genes, including the APP gene, with extracellular Aβ detected in the hippocampus and cortex with positive Aβ plaques appearing earlier in these mice, around 1.5 months of age; impaired memory emerges between 3 and 6 months that become progressively worse with age.^24,25 TgSwDI and 5xFAD mouse models exhibit increased permeability of the BBB at the age of 4 and 2 months, respectively.^24,45,46 Thus, in this study, TgSwDI and 5xFAD mice were used at the ages of 4 and 2 months, respectively, to receive a daily treatment of 5 mg/kg 4-FPBUA IP for 28 days, ending the treatment at 5 and 3 months of age, respectively.

The findings from the in vitro experiments confirmed that 4-FPBUA enhances the intactness of the bEnd3 cell-based BBB model, as demonstrated by a reduction in the transport of permeability markers LY and ¹⁴C-inulin across the monolayer and a corresponding increase in the expression of the tight junction protein ZO-1. 4-FPBUA also enhanced the transport of Aβ across the monolayer, which was associated with the upregulation of the Aβ major transport proteins P-gp and LRP1. These in vitro findings were consistent with those of the in vivo studies in TgSwDI and 5xFAD mice. 4-FPBUA treatment tested in TgSwDI and 5XFAD mice demonstrated comparable results with enhanced BBB integrity as measured by the increased expression of tight junction proteins ZO-1 and occludin in both mouse models. Furthermore, our data suggest that 4-FPBUA reduced total brain Aβ levels likely through multiple mechanisms, including, at least in part, increased Aβ clearance across the BBB mediated by increased P-gp and LRP1, enhanced Aβ degradation by neprilysin, and reduced Aβ production in both mouse models. In addition to its effects on the BBB, further evaluation demonstrated that 4-FPBUA enhanced autophagy, suggesting an additional pathway for Aβ clearance to reduce brain levels.

Recent studies have shown that dysfunctional autophagy contributes to the accumulation of Aβ in the brains of AD mouse models.^47,48 Conversely, other studies have reported that increased Aβ brain levels disrupt autophagic functions. For example, in a comparison study for the expression level of autophagy markers ATG5, LC3B–I, and LC3B–II in the hippocampus tissues of Tg2576 mouse brains, results demonstrated significantly lower levels of the autophagy markers than those in nontransgenic wild-type mice.⁴⁹ Moreover, reduced levels of the autophagy-related protein beclin-1 were associated with increased accumulation of Aβ in the brains of APP transgenic mice, whereas enhanced levels of beclin-1 reduced the accumulation of Aβ,⁵⁰ suggesting that restoring autophagy could be a viable target for drug development and therapeutic intervention to prevent and treat AD.

Autophagy is an important intracellular degradation and recycling process by which cells eliminate damaged organelles and degrade abnormal proteins, such as Aβ, through autophagosome-lysosomal pathways, suggesting that deficits in this process could lead to neurodegeneration. Therefore, the development of molecules that are able to induce autophagy could provide a novel therapeutic strategy to slow the progression and treat Aβ-related pathology and AD. Several compounds have been reported for their ability to activate autophagy and provide neuroprotection against Aβ and related pathology.^51–54 Several studies have reported that rapamycin significantly reduced the accumulation of Aβ in the brains of mouse models of AD and improved cognitive function by inducing autophagy by suppressing mTOR activation.^55–57 Similarly, findings from our study showed that 4-FPBUA induced autophagy by inhibiting mTOR1 activation, as determined by the reduced expression of phospho-mTOR and its downstream pathway. mTOR is a protein complex that regulates cell growth and synthesis through proteins such as P70S6K and 4EBP1. mTOR suppression by 4-FPBUA inactivated P70S6K and 4EBP1 by reducing their phosphorylation and the total levels of 4EBP1, which were collectively suppressed by Raptor phosphorylation. Raptor phosphorylation reduces mTOR-catalyzed P70S6K and 4EBP1 phosphorylation that are necessary for mTOR activity.⁵⁸

In addition, treatment with 4-FPBUA resulted in increased AMPK activation, confirming the potential effect of 4-FPBUA on autophagy during the induction phase. AMPK activation through LKB activation can also reduce mTOR activity by directly phosphorylating Raptor.^35,59 As a downstream target of activated AMPK, ULK1 was similarly upregulated with increased phosphorylation at Ser317.⁶⁰ Besides its effect on mTOR and AMPK, and downstream targets, 4-FPBUA increased beclin-1 levels that is required for autophagy induction.⁶¹ Induced autophagy was also associated with the upregulation of autophagy-related proteins, including ATG3, −5, and −7, essential for phagophore formation and elongation,⁶² and upregulating the conversion of LC3I to LC3II linked to the autophagosomal membrane.⁶³

TgSwDI and 5xFAD mouse model treatment with 4-FPBUA resulted in nuanced effects on the mTOR pathway and autophagy, reflecting the compound’s influence on cellular signaling processes. In both models, mTOR kinase activity appears to be inhibited, as evidenced by consistent reductions in phosphorylated mTOR without changes in the total mTOR levels. This suggests that 4-FPBUA downregulates critical pathways involved in cell growth and autophagy initiation. Moreover, Raptor activation, a key component of the mTOR complex, shows model-specific responses; the TgSwDI model exhibits an increase in both total and phosphorylated forms, although the phosphorylation does not increase proportionally with the total protein level, indicating inefficiencies in activation. Conversely, in the 5xFAD model, while the total Raptor levels remain unchanged, phosphorylation significantly increases, suggesting that more effective enzymatic activation and potentially different regulatory mechanisms are engaged by 4-FPBUA in this setting.

The treatment did not alter the synthesis or degradation of P70S6K in either model. Still, a notable decrease in the phosphorylated form underscores 4-FPBUA’s strong inhibitory effect on downstream mTOR signaling, particularly affecting pathways controlling protein synthesis and cellular growth. Similar effects are observed with 4EBP1, where both total and phosphorylated levels decrease, suggesting a broad inhibitory impact on mTOR signaling across different genetic backgrounds and enhancing the compound’s role in down-regulating processes like protein synthesis and cell proliferation. Furthermore, AMPK shows an increase in phosphorylation in the 5xFAD model without a change in total protein. In contrast, in the TgSwDI model, the total and phosphorylated AMPK levels increased. However, the phosphorylation ratio to total protein decreases, indicating a complex regulatory effect that may not fully maximize the potential activation of AMPK.

The differences in the LKB1 response between the models highlight intrinsic disparities. The TgSwDI model shows increased phosphorylation without changes in total levels, suggesting selective activation of pathways underpinning energy homeostasis and possibly autophagy. In contrast, the 5xFAD model shows no change in phosphorylated or total LKB1 levels, reflecting a potential model-specific resistance or differing underlying regulatory mechanisms affected by 4-FPBUA. Finally, ULK1, which is crucial for autophagy initiation, increases both its expression and phosphorylation in the TgSwDI model, suggesting an effective enhancement of autophagic processes. However, in the 5xFAD model, despite similar increases, the ratio of phosphorylated to total ULK1 decreases, possibly indicating a limitation in activation relative to the upregulated expression.

The observed differences between the two mouse models likely stem from inherent genetic or pathological variations that influence how signaling pathways and molecular mechanisms respond to 4-FPBUA treatment. These variations may reflect the unique pathological landscapes of AD models, such as TgSwDI and 5xFAD, which could affect protein expression patterns, enzymatic activities, and the overall cellular environment, thereby moderating the effectiveness and specificity of drug responses within each model. Indeed, additional studies are necessary to confirm and clarify these results.

The inhibition of the mTOR pathway by rapamycin has shown a significant protective effect in preclinical models of AD. Rapamycin treatment enhanced microvascular endothelial function and cognitive function in mouse models of AD. mTOR inhibition with rapamycin abolished BBB breakdown in vivo and in vitro and upregulated tight junction proteins.^64,65 Furthermore, Yang et al. reported that the induction of autophagy by serum starvation protected and enhanced BBB integrity by maintaining functional claudin-5 and removing cytoplasmic aggregated claudin-5 and ROS, while autophagy downregulation impaired the endothelial barrier function, as determined by reduced expression of claudin-1 and ZO-1, an effect that was associated with increased paracellular permeability and reduced transendothelial electrical resistance (TEER).⁶⁶ Similarly, while further mechanistic studies are required, our findings suggest that 4-FPBUA directly or indirectly enhances the BBB function by inducing autophagy via mTOR inhibition and AMPK activation. Our findings from the in vitro experiments support that 4-FPBUA can directly improve the function of the cell monolayer by inducing autophagy mediated by LRP1 induction. Besides its role in Aβ transport across the BBB, the endothelium LRP1 has been shown to play a role in autophagy.^67–69 The cellular knockdown of LRP1 caused a significant reduction in AMPK phosphorylation and LC3II, while the upregulation of LRP1 enhanced autophagy through AMPK activation and the generation of LC3II from LC3I.^67,69 In addition, induced autophagy could indirectly enhance the BBB function by reducing brain Aβ levels, which has been shown to negatively disrupt the BBB.

The enhanced BBB function and reduced Aβ levels observed with 4-FPBUA treatment provided a neuroprotective effect against Aβ, as evidenced by the increased expression of the synaptic marker, PSD95, in both mouse models. This effect was associated with improved memory function, as monitored by the MWM test. Interestingly, however, the effect of 4-FPBUA on memory function was better in TgSwDI mice than in 5xFAD mice. TgSwDI mice required less time and swam a shorter distance to reach the hidden platform, while in 5xFAD mice, 4-FPBUA did not affect these parameters, but increased the time spent in the platform quadrant. In TgSwDI mice, memory impairment was observed at 3 months of age,^24,26–28 while that was observed at the age of 4 to 6 months in 5xFAD mice.²⁵ In our study, 5xFAD mice were treated with 4-FPBUA at 2 months and ending the treatment at 3 months of age. It is possible that, at this younger age (investigated age), 5xFAD mice may not have developed significant cognitive deficits. The inclusion of wild-type mice without AD pathology (control) would have helped to clarify this assumption, which was not a part of the study, thus representing a limitation. Other limitations of this study include the use of only female mice and a small number of animals per group. Indeed, to develop 4-FPBUA as a potential AD drug, additional confirmatory studies are needed using both female and male mice with a larger number of animals per group. Pharmacokinetic studies are also necessary to understand the 4-FPBUA bioavailability and brain penetration. Future studies will also include dose-dependent studies to fully understand the 4-FPBUA dose–response relationship and its safety across a broader range. Additional studies are needed to compare the effect of 4-FPBUA with rapamycin on AD pathology. Furthermore, as this study proposes 4-FPBUA to modulate autophagy by inhibiting mTOR, future studies will be designed to clarify the relationship between mTOR inhibition by 4-FPBUA, autophagy, and Aβ levels, and the effect of changes in Aβ levels on modulating autophagy.

In conclusion, the findings of this study demonstrate the neuroprotective effect of 4-FPBUA in two mouse models of AD. 4-FPBUA appears to enhance BBB integrity, reduce Aβ brain levels by increasing its clearance across the BBB, and induce autophagy via mTOR inhibition and AMPK activation. These findings suggest that 4-FPBUA is a promising candidate for further development as a therapeutic agent for AD and related disorders. Figure 11 summarizes the proposed autophagy mechanisms of action of 4-FPBUA.

Figure 11.

4-FPBUA induces autophagy by suppressing mTOR and activating the AMPK pathway. 4-FPBUA suppresses mTOR signaling, as demonstrated by the reduced phosphorylation of P70S6K1 and 4EBP1. This leads to the activation of the ULK1 complex along with beclin-1, initiating autophagosome formation. Additionally, 4-FPBUA enhances AMPK activation by LKB1, further suppressing mTOR. The figure outlines the resulting increase in the levels of ATG3 and ATG5-ATG7 complex, which is essential for autophagosome elongation. It also shows a decrease in LC3I and a concurrent increase in LC3II, indicative of autophagosome membrane expansion and maturation. This elevation in autophagy markers reflects 4-FPBUA’s potential to boost autophagic flux, a process that may counteract the accumulation of misfolded proteins in AD. Increased autophagy decreases the Aβ load, thus decreasing the insult to the BBB and increasing the functionality and integrity by increasing the expression of tight junction proteins and Aβ transport proteins.

MATERIALS AND METHODS

Materials.

4-FPBUA was synthesized, as reported previously.¹ A purity of >95% was established based on 1H NMR and HPLC analyses.¹ Fibronectin from bovine plasma and LY were purchased from Sigma-Aldrich (St. Louis, MO). HTS 3 μm polycarbonate membrane Transwell 96-well plates and 0.45 μm polyester membrane Transwell 24-well plates were purchased from Corning (Corning, NY). Rat-tail collagen type-I, Dulbecco’s modified Eagle’s medium (DMEM), sterile PBS, and penicillin/streptomycin antibiotics were obtained from Gibco (Grand Island, NY). Fetal bovine serum (FBS) was purchased from Atlanta Biologicals (Flowery Branch, GA). ¹⁴C-inulin ([carboxyl-¹⁴C]-inulin, M_W = 5000 Da) was purchased from American Radiolabeled Chemicals (St. Louis, MO).

Synthetic monoiodinated and nonoxidized Aβ₄₀ (¹²⁵I-Aβ₄₀; human, 2200 Ci/mmol) was purchased from PerkinElmer (Boston, MA). Bovine serum albumin (BSA) and Thioflavin-S (Thio-S) were purchased from Sigma-Aldrich. Total protein measurement reagents using the bicinchoninic acid (BCA) method were obtained from Pierce (Rockford, IL). All other chemicals and reagents were of analytical grade and readily available from commercial sources. All antibodies used for western blotting and immunostaining are listed in Table 1.

Table 1.

List of Antibodies Used in the Study

antibody	application	source
BACE1 (cat# ab108394)	WB	Abcam (Cambridge, MA)
Alexa-fluor 488-labeled 6E10 (cat# 803013)	IHC	BioLegend (San Diego, CA)
mTOR (7C10) (cat# 2983)	WB	cell signaling (Danvers, MA)
Phospho-Mtor (Ser2448) (D9C2) (cat# 5536)	WB	cell signaling
p70 S6 kinase (49D7) (cat# 2708)	WB	cell signaling
Phospho-p70 S6 kinase (Ser371) (cat# 9205)	WB	cell signaling
4EBP1 (53H11) (cat# 9644)	WB	cell signaling
Phospho-4EBP1(Ser65) (D9G1Q) (cat# 2855)	WB	cell signaling
AMPKα (D63G4) (cat# 5832)	WB	cell signaling
Phospho-AMPKα (Thr172) (cat# 2535)	WB	cell signaling
ULK1 (cat# 8054)	WB	cell signaling
Phospho-ULK1 (Ser555) (cat# 5869)	WB	cell signaling
Beclin-1 (cat# 3495)	WB	cell signaling
ATG7 (cat# 8558)	WB	cell signaling
ATG5/12 (cat# 2011)	WB	cell signaling
ATG3 (cat# 3415)	WB	cell signaling
LC3I (cat# 3868)	WB	cell signaling
LC3II (cat# 12741)	WB	cell signaling
LKB1 (cat# 3050)	WB	cell signaling
Phospho-LKB1 (cat# 3055)	WB	cell signaling
Raptor (cat# 2280)	WB	cell signaling
Phospho-Raptor (cat# 2083)	WB	cell signaling
Anticollagen-IV (cat# AB756P)	IHC	EDM-Millipore (Burlington, MA)
APP-TOTAL (cat# MAB348)	WB	EDM-Millipore
SNAP-25 (cat# GTX113839)	WB	GeneTex (Irvin, CA)
PSD95 (cat# GTX42033)	WB	GeneTex
sAPPα (cat# 11088)	WB	Immuno-Biological Laboratories (Minneapolis, MN)
sAPPβ (cat# 18957)	WB	Immuno-Biological Laboratories
HRP-conjugated antigoat IgG, (cat# HAF109)	WB	R&D Systems (Minneapolis, MN)
Actin-β (C4) (cat# sc-47778)	WB	Santa Cruz (Santa Cruz, CA)
Tubulin-β (5F131) (cat# sc-55529)	WB	Santa Cruz
CFL 594-conjugated antirabbit IgG (cat# sc-516250)	IHC	Santa Cruz
HRP-conjugated antirabbit IgG (H +L) secondary antibody (cat# PI31460)	WB	ThermoFisher
HRP-conjugated antimouse IgG (H +L) secondary antibody (cat# PI31430)	WB	ThermoFisher
P-glycoprotein (P-gp) monoclonal antibody (C219) (cat# SIG-38710)	WB	BioLegend
LRP1 (cat# ab28320)	WB	Abcam
Neprilysin (NEP) (cat# sc-65990)	WB	Santa Cruz

Cell Culture.

The immortalized mouse brain endothelial cell line bEnd3 was obtained from ATCC (Manassas, VA). bEnd3 cells, passage 25–35, were cultured in DMEM supplemented with 10% FBS, penicillin G (100 IU/mL), streptomycin (100 g/mL), 1% w/v nonessential amino acids, and 2 mM glutamine. The cultures were maintained in a humidified atmosphere (5% CO₂/95% air) at 37 °C, and the medium was changed every other day. The bEnd3 cell-based BBB model for HTS cell seeding was performed, as reported previously.¹⁶ In brief, bEnd3 cells were plated onto the HTS Transwell 96-well plate at a seeding density of 50,000 cells/cm² to the apical side (A). The inserts were coated with 50 μL of fibronectin solution (30 μg/mL in PBS) 2 h before seeding. Fresh medium (200 μL) was added to the basolateral side (B). The cells were incubated and cultured at 37 °C and 5%CO₂/95% air for 5 days to achieve the optimal barrier integrity of bEnd3 cells. On the fifth day, cells were treated with 5 and 10 μM 4-FPBUA and added to the apical side; 24 h later, the intactness of bEnd3 cells was evaluated by measuring the permeation of LY, a small, water-soluble molecule that diffuses passively between cells and serves as an indicator of monolayer tightness.

The LY permeation measurement involved loading 50 μL of transport buffer containing 100 μM LY onto the apical side and 200 μL of the same prewarmed buffer to the basolateral side of each transwell filter. Transwell plates were then incubated for 1 h. Postincubation, the LY concentration was measured in both the apical and basolateral compartments. This was done by recording the fluorescence intensities at 485 nm for excitation and 529 nm for emission wavelengths and comparing them to known LY standards.¹⁶ The data were captured using a Synergy 2 microplate reader and processed with Gene5 software, facilitating the calculation of the apparent permeation coefficient (Pc), providing a quantitative measure of the LY permeation coefficient (Pc) and, consequently, the integrity of the bEnd3 cell monolayer after treatment with 4-FPBUA. The following equation was used to calculate Pc

$$P_{\text{c}}\left(\text{cm}/\text{s}\right)=\frac{V_{\text{b}}\times C_{\text{b}}}{C_{\text{a}}\times A\times T}$$

where V_b represents the volume on the basolateral side (200 μL), C_b denotes the concentration of LY (μM) on the basolateral side, C_a is the concentration of LY on the apical side, A is the membrane area (0.143 cm²), and T is the time duration of the transport process (3600 s).

Aβ Transport across bEnd3 Cell-Based BBB Model.

Aβ transport studies were performed, as previously reported by us.⁷⁰ Briefly, bEnd3 cells were plated onto 0.45 μm polyester membrane Transwell 24-well plates and then treated with 10 μM 4-FPBUA on day 5 for 24 h. The next day, the Aβ40 transport assay was initiated by the addition of a mixture of 0.1 nM ¹²⁵I-Aβ₄₀ and 0.05 mM ¹⁴C-inulin (to monitor the monolayer integrity) in media to the basolateral side for 30 min. Aliquots from both sides were separately collected for radioactivity analysis using a Wallac 1470 Wizard γ and Wallac 1414 WinSpectral Liquid Scintillation β Counter (PerkinElmer; Waltham, MA). The transport quotients (TQ) from the basal (B) side to the apical (A) side of ¹²⁵I-Aβ₄₀ (TQ_B→A) were determined as previously reported,⁷¹ representing the brain and blood compartments, respectively, using the following equation. ¹²⁵I-Aβ_40(A) and ¹⁴Cinulin_(A) are the concentrations of ¹²⁵I-Aβ₄₀ and ¹⁴C-inulin in the apical compartment, compared to their total concentrations added to the basal compartment (¹²⁵I-Aβ_{40 total} and ¹⁴C-inulin _total).

$${}^{125}\text{I}-\text{A}{\beta }_{{\text{TQ}}_{\text{B}\to \text{A}}}=\frac{\left\{\frac{{}^{125}\text{I}-\text{A}{\beta }_{40\left(\text{A}\right)}}{{}^{125}\text{I}-\text{A}{\beta }_{40\text{total}}}\right\}}{\left\{\frac{{}^{14}\text{C}{-\text{inulin}}_{\left(\text{A}\right)}}{{}^{14}\text{C}{-\text{inulin}}_{\text{total}}}\right\}}$$

Animals.

TgSwDI and 5xFAD mice were purchased from Jackson Laboratories (Bar Harbor, ME). TgSwDI mice express human APP under the control of the Thy 1.2 neuronal promoter harboring double Swedish mutations and Dutch and Iowa vasculotropic Aβ mutations, leading to early and aggressive Aβ accumulation associated with inflammatory astrocyte activation and cognitive decline. In the brain of TgSwDI mice, Aβ begins to accumulate at the age of 2 to 3 months and deposits extensively at the age of 12 months.²⁴ 5xFAD mutations include APP KM670/671NL (Swedish), APP I716 V (Florida), APP V717I (London), PSEN1M146L, and PSEN1 L286 V, leading to early, starting around the age of 2 months, and aggressive Aβ plaques accumulation associated with inflammatory astrocytes activation and cognitive decline.²⁵ All animal experiments and procedures were approved by the Institutional Animal Care and Use Committee of the University of Louisiana at Monroe (#14APR-AKK-02). All mice were housed in plastic cages under standard conditions, 12-h light/dark cycle, 22 °C, 35% relative humidity, and ad libitum access to water and food.

Animal Treatment.

To evaluate the effect of 4-FPBUA on Aβ and related pathology, TgSwDI (females, 4 months old, n = 5 mice/group) and 5xFAD (females, 2 months, n = 4 mice/group) mice were divided into 2 groups per mouse model: the control group received vehicle (0.1% Tween-20 and 0.05% DMSO in saline) administered intraperitoneally (IP), and treatment groups received 5 mg/kg 4-FPBUA administered IP daily for 28 days. As 4-FPBUA treatment in mice has not been reported previously, we used a low dose. Moreover, as the kinetics of 4-FUBPA were not evaluated, the IP route of administration was selected over the oral dosage form. During the treatment period, the body weights of animals were monitored every week, and the average TgSwDI mouse weights were 24 ± 3 g and 23 ± 2 for the control and 4-FPBUA-treated groups, respectively. The average 5xFAD mouse weights were 22 ± 2 g and 21 ± 3 g for the control and 4-FPBUA-treated groups, respectively. Five days before the end of the treatment period, all mice underwent the MWM test to assess memory function, as described below. At the end of the treatment period, the brains were collected for analysis.

Western Blot Analysis.

For in vitro studies, bEnd3 cells were seeded in 10 mm cell culture dishes (Corning) and allowed to grow in a humidified incubator (5%CO₂/95% air) at 37 °C. At 80–90% confluence, cells were treated with a medium containing 0.1% DMSO as a control or 10 μM 4-FPBUA for 24 h. At the end of the treatment period, cells were washed twice with ice-cold PBS, scraped, collected, lysed in RIPA buffer containing 1% protease inhibitor cocktail on ice, and then centrifuged at 21,000g for 10 min at 4 °C. For western blotting, the total protein content was determined using the BCA protein assay. Protein samples (25 μg) were loaded and resolved on 10% SDS-polyacrylamide gels and then transferred electrophoretically onto a PVDF membrane (Millipore). Membranes were incubated in 2% BSA blocking solution followed by overnight incubation at 4 °C with primary antibodies; the analyzed bEnd3 proteins include claudin-5, ZO-1, LRP1, P-gp, and the housekeeping protein β-actin, which were used for normalization of the samples. The secondary antibodies used were antimouse IgG antibody (1:2000 dilution) for P-gp; antirabbit IgG antibody (1:2000 dilution) for LRP1, claudin-5, and ZO-1; and antigoat IgG antibody (1:1500 dilution) for β-actin, all labeled with horseradish peroxidase (HRP). Three independent western blot experiments were carried out for each treatment group. The antibodies used are listed in Table 1.

For in vivo studies, protein extracts were prepared from the brain homogenates of TgSwDI and 5xFAD mice. Brain homogenates were homogenized in RIPA buffer containing a 1% protease inhibitor cocktail and centrifuged at 21,000g for 10 min at 4 °C. For western blot analysis, 25 μg of protein sample was loaded, resolved, and transferred to PVDF, as described above. Primary and secondary antibodies used to immunoblot proteins associated with Aβ clearance and production, BBB function, and the mTOR pathway in brain homogenate samples are listed in Table 1. All protein blots were developed using a chemiluminescence detection kit (ThermoFisher Scientific). Bands were visualized using the ChemiDoc MP Imaging System (Bio-Rad Hercules, CA), and band intensities were quantified by densitometric analysis using Image Lab Software V.6.0 (Bio-Rad). The intensities of all blotted proteins were normalized to actin-β, which was used as a housekeeping protein.

Immunohistochemical Analysis.

Brain sections of 18 μm-thick were prepared using a Leica CM3050S Research Cryostat (Buffalo Grove, IL). Subsequently, the sections were fixed by incubation in methanol for 10 min at −20 °C. The sections were washed 5 times in PBS and blocked in PBS containing 10% donkey serum for 1 h at room temperature. Immunostaining was performed on the brain hippocampi of TgSwDI and 5xFAD mice. The entire hippocampus region spanning the dentus gyrus (DG) and CA1-CA3 regions was included in the analysis. For total Aβ detection, brain sections were immunostained with Alexa-fluor 488-labeled 6E10 human-specific anti-Aβ at a 1:200 dilution. For detection of Aβ-plaque load, the brain tissue sections were stained with a freshly prepared and filtered 0.02% Thio-S solution in 70% ethanol for 30 min followed by incubation in 70% ethanol, as described previously.⁷² Image acquisition was performed in 4 tissue sections spanning the hippocampus, each separated by 150 μm (total of 40 sections per mouse). Total Aβ load and Thio-S were captured and quantified at a total magnification of 4×, and fluorescence intensity was quantified using ImageJ version 1.6.0 software (Research Services Branch, National Institute of Mental Health/National Institutes of Health, Bethesda, MD) after adjusting for thresholds. All images were visualized using a Nikon Eclipse Ti–S inverted fluorescence microscope (Melville, NY).

Human Aβ₄₀ and Aβ₄₂ Determination by ELISA.

To quantitatively detect soluble Aβ₄₀ and Aβ₄₂ in mouse brains, supernatants from brain homogenate lysates were analyzed for Aβ₄₀ and Aβ₄₂ brain levels using commercially available ELISA kits according to the manufacturer’s instructions (R&D Systems; Minneapolis, MN). All samples were run in duplicate and corrected to the total protein amount in each sample by using the BCA assay.

Behavioral Testing by Morris Water Maze Test.

The Morris water maze (MWM) test was performed for TgSwDI and 5xFAD mice to assess learning and memory performance at the end of treatment using protocols similar to those described previously.^73,74 All mice underwent training 3 times a day for 4 consecutive days. The platform was maintained in the same quadrant during the entire course of the experiment. The mice were required to find a hidden platform using the distal spatial cues available in the room. The conditions were maintained the same during all of the experiments. An overhead camera connected to a computerized tracking system (SMART 3.0 Platform, Panlab Harvard apparatus; Holliston, MA) was used to record the movements of mice. The latency and swimming distance parameters were calculated and analyzed for the treatment effect on the fourth day of training to compare the performance using statistical analysis.

Statistical Analysis.

Data are expressed as the mean ± SEM. The experimental results were statistically analyzed for significant differences using Student’s t-test for two groups and one-way ANOVA with post hoc analysis using Dunnett’s test for more than two groups. Values of P < 0.05 were considered statistically significant. Data analyses were performed using GraphPad Prism, version 6.0.