Introduction

Alzheimer’s disease (AD) is the most common dementia type, and it is characterized by progressive cognitive decline and psychiatric symptoms [1, 2]. AD affects approximately 10 to 30% of people aged > 65 years, posing a significant challenge for both individuals and society [3]. The pathological changes in AD include the accumulation of amyloid-beta (Aβ) aggregates, neurofibrillary tangles, neuroinflammation, and loss of neurons [4]. Nevertheless, the key mechanisms underlying its pathogenesis and progression remain poorly understood. Therefore, medications that delay AD progression or improve its symptoms are still limited.

Triterpenoids, including oleanolic acid, are a large family of compounds synthesized in some plants; they comprise pentacyclic motifs and potent immune-modulating activities [5, 6]. To increase the utility of natural triterpenoids, researchers have developed several synthetic derivatives [5]. Omaveloxolone (Omav) is a second-generation semi-synthetic oleanane triterpenoid with anti-cancer and anti-inflammation activities [6, 7]. To note, Omav can achieve meaningful concentrations in the brain and exert therapeutic effects on neurological diseases, such as epilepsy and amyotrophic lateral sclerosis [8,9,10]. Moreover, in 2023, the FDA has approved Omav as the first treatment for Friedreich’s ataxia [11]. Although the possibility of using Omav for neurological diseases has been raised, studies in AD are scarce.

Network pharmacology is an approach that helps elucidate the mechanisms of drugs. It aims to refine the treatment strategies by studying the intricate biological network of disease characteristics and drug targets connected to each other [12]. Molecular docking, on the other hand, is a theoretical technique that explores the interaction and recognition between ligands and proteins; it holds significant importance in understanding the molecular mechanisms underlying pharmacological activities and predicting the structure of protein–ligand complexes [13]. The integrated use of these two methods has gained attention in providing a comprehensive framework for drug discovery and optimization.

In this study, we aimed to investigate whether Omav ameliorated cognitive impairment in APP/PS1 mice. Utilizing network pharmacology, we predicted the potential targets and biological pathways of Omav for AD treatment. Finally, we prioritized two of the predicted targets for further validation using molecular docking and in vivo experiments, to understand the underlying mechanisms.

Materials and Methods

Animals and Drug Treatment

A total of 16 APP/PS1 double transgenic mice and 8 wild-type (WT) mice with the same genetic background of 30-week-old were purchased from Beijing Zhishan Co., Ltd. (Beijing, China). These mice were maintained in cages with three to four per cage, ensuring free access to food and water. They were housed under monitored conditions, maintaining a constant temperature range of 20–24 °C and a regular 12-h light/dark cycle. After a week of adaptation to the environment, APP/PS1 mice were assigned randomly into two treatment groups: APP/PS1-vehicle (APP/PS1) and APP/PS1-Omaveloxolone (APP/PS1-Omav). Omav (20 mg/kg) or vehicle (corn oil) was administered daily via oral gavage for 4 weeks. The Omav dose was selected based on previous literature and dosage conversion from human to mice using a calculator [14, 15]. After the 4-week period, all mice were subjected to behavioral testing, and the administrations were continued throughout the test.

Morris Water Maze Test

The Morris water maze tests, as described by Vorhees and William [16], were conducted in a circular tank containing a submerged platform (diameter, 10 cm). The water temperature was controlled at 21 ± 1 °C. The experiments were conducted in two sections: (i) spatial acquisition section (days 1–5) and (ii) probe trial section (day 6). In the spatial acquisition section, each mouse was released into the water facing the wall and allowed to find the platform within 60 s. The escape latency was recorded as the time taken to find the platform. Mice that successfully found the platform within the time limit were permitted to remain for 5 s; otherwise, they were gently placed on the platform and allowed to stay for 20 s. Each mouse underwent four trials daily, and the escape latency on each training day was averaged over four trials. For the probe trial, the platform was removed, and mice were released into the water from the opposite quadrant and recorded for 60 s to evaluate the spatial memory ability. The time spent and the number of crossings in the target zone were recorded. Trajectories were video-tracked and analyzed using SMART3 software (Panlab Harvard, USA).

Sample Collection

After behavioral testing, the mice were anesthetized and transcardially perfused with cold PBS. On an ice plate, brains were carefully removed and dissected in half. One half was fixed in 4% polyformaldehyde for morphological studies, whereas the remaining half was frozen at − 80 °C for biochemical analyses.

Network Pharmacology and Molecular Docking

The chemical structure and SMILES information of Omav were obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). Potential targets of Omav were predicted using two databases: Super-PRED [17] and Pharmmaper [18]. Targets with a probability ≥ 60% or a norm fit > 0.5, respectively, were selected for further analysis. The potential targets of AD were obtained from the GeneCards database (https://www.genecards.org/). AD targets were selected with a relevance score > 10. A Venn diagram was drawn to obtain the intersection genes. The STRING database was used to construct a protein–protein interaction (PPI) network [19]. Then the PPI network was imported into Cytoscape 3.10.1 for further analyses and visualization. Key targets were obtained using the Centiscape 2.2 plugin according to three topological features (degree, betweenness, and closeness). The graph was arranged and optimized based on degree values. The DAVID database was then used for GO and KEGG enrichment analysis to elucidate the core mechanisms and pathways of Omav against AD [20]. Finally, the AutoDock Vina software was used to perform molecular docking between the candidate proteins with Omav. A binding energy of less than − 5.0 kcal/mol − 7.0 kcal/mol indicates a good or strong binding activity, respectively [21]. The docking results were visualized using the PyMOL software.

Immunohistochemistry and Immunofluorescence

Brains were fixed in 4% polyformaldehyde for 24 h and underwent a gradient dehydration process before being paraffin-embedded. Serial 5-µm coronal sections were prepared, deparaffinized, hydrated, and treated with 0.01 M citric acid for antigen retrieval. After three washes with PBS, sections were submerged in 3% H2O2 for 30 min. For blocking, the sections were incubated with 5% goat serum for 1 h. Next, they were incubated with primary antibody (anti-Aβ 1–42, Abcam, ab201060) at 4 °C overnight. After rinsing three times with PBS, horseradish peroxidase-labeled secondary antibody was used for incubation for 1 h, followed by colorization using a DAB kit. Images were captured using an optical microscope (Olympus, Tokyo, Japan). Three non‐overlapping visual fields were randomly selected, and the number of senile plaques in each field was measured using the ImageJ software (Image Pro Plus software, USA). For immunofluorescence, the sections were incubated with the primary antibody (anti-Iba-1, Invitrogen, MA527726) overnight at 4 °C, followed by incubation with CoraLite488-conjugated secondary antibody for 2 h. DAPI was used for nuclear staining, and images were captured using an Olympus fluorescence microscope (Olympus, Japan).

Western Blot

For protein collection, mouse brain tissues were lysed in a RIPA buffer comprising a cocktail of protease and phosphatase inhibitors. After a brief sonication, the lysates were centrifuged at 12,000 g for 15 min (4 °C). The samples were then prepared by mixing with 5 × loading buffer and boiled for 5 min. Equal amounts of proteins were added into each well of the SDS-PAGE gels for electrophoresis and transferred onto 0.45-µm PVDF membranes. After incubation with blocking buffer, membranes were incubated with the following primary antibodies: cleaved caspase 3 (Wanleibio, WL02117), Bax (Proteintech, 50599-2-Ig), Bcl-2 (Proteintech, 26593-1-AP), β-actin (Proteintech, 66009-1-Ig), p-mTOR (Abcam, ab109268), t-mTOR (Abcam, ab134903), LC3 (Abcam, ab192890), and P62 (MedChemExpress, HY-P80518) at 4 °C overnight. Membranes were then incubated with secondary antibody for 1 h. Protein bands were visualized using ECL reagent, and the abundance of proteins was analyzed using the ImageJ software.

Hematoxylin–Eosin (H&E) Staining

H&E staining was performed using a commercial kit (Solarbio, Beijing, China). Briefly, after deparaffinization and hydration, the brain sections were dyed with a hematoxylin solution for 10 min, washed, and differentiated for 30 s. Then, they were re-dyed with eosin solution for 1 min. After another wash with tap water, the sections underwent dehydration, transparentization, and sealing procedures for observation.

Enzyme-linked Immunosorbent Assay (ELISA)

Briefly, brain tissues were homogenized in cold PBS consisting of a cocktail of protease and phosphatase inhibitors. After a brief sonication, the homogenates were centrifuged at 5000 g for 10 min (4 °C). The levels of tumor necrosis factor-alpha (TNF-α), interleukin (IL)-1β, and IL-6 in the supernatants were determined using commercial mouse ELISA kits (Elabscience, China) according to the manufacturer’s instruction.

Statistical Analysis

All values were presented as mean ± SEM. The escape latencies in the Morris water maze were calculated using the area under the curve, and group differences were analyzed using one-way analysis of variance (ANOVA). All other data were analyzed using a t-test or one-way ANOVA, followed by the LSD post hoc test. P < 0.05 indicated statistical significance. All statistical analyses were performed using SPSS 26.0 (SPSS Corp., Armonk, NY, USA).

Results

Omav Treatment Ameliorates Cognitive Dysfunction in APP/PS1 Mice

The spatial learning and memory abilities of the mice were evaluated using the Morris water maze test. In the spatial acquisition section, APP/PS1 mice exhibited a longer time searching for the platform compared with the WT littermates. The administration of Omav significantly improved the impaired learning ability in APP/PS1 mice (Fig. 1A). Moreover, the probe trial performed on the last day suggested that APP/PS1 mice spent less time in the target zone and had fewer crossovers than the WT littermates. However, Omav-treated mice demonstrated an increase in the time spent in the target zone and had more crossovers compared with the APP/PS1 controls (Fig. 1B, C). Figure 1D illustrates the representative trajectories in the probe trial. These results suggested that Omav treatment rescued cognitive dysfunction in AD mice.

Fig. 1
figure 1

Omav improved cognitive deficits in APP/PS1 mice. The Morris Water Maze test was used to evaluate the effect of Omav on APP/PS1 mice. A Escape latency during the spatial acquisition section. The number sign (#) indicates P < 0.05 between the APP/PS1 and WT group. The asterisk (*) indicates P < 0.05 between the APP/PS1 and APP/PS1 + Omav group. B and C Time and crossings in the target zone. *P < 0.05; **P < 0.01. Data presented as mean ± SEM. D Representative tracing graphs during the probe trials

Full size image

Omav Treatment Reduces Aβ Deposition in APP/PS1 Mice

To investigate the effects of Omav treatment on Aβ deposition in APP/PS1 mice, immunohistochemical staining was conducted using a specific antibody against Aβ 1–42. The results revealed a significant elevation in the number of plaques in the hippocampus and cortex of APP/PS1 mice compared with WT littermates. Notably, Omav treatment dramatically reduced Aβ deposition compared with APP/PS1 mice (Fig. 2A, B). These results suggested that Omav treatment decreased Aβ burden in APP/PS1 mice.

Fig. 2
figure 2

Omav reduced Aβ burden in APP/PS1 mice. A Representative images of Aβ42 deposits in the hippocampus and cortex using an antibody against Aβ42. Scale bar: 200 µm. B Quantification of plaque numbers in the hippocampus and cortex. *P < 0.05; **P < 0.01. Data presented as mean ± SEM

Full size image

Network Pharmacological Analysis of Omav Against AD

We obtained 94 and 120 genes from Super-PRED and Pharmmapper, respectively. After eliminating the duplicates, 208 genes were retained for further analysis. Additionally, we selected 2022 genes with a relevance score > 10 by searching the GeneCards database. Targets of both Omav and AD were subjected to a Venn diagram, and 112 common genes were screened (Fig. 3A). The PPI network indicated 110 nodes and 1229 edges. The key targets were screened using Centiscape 2.2 plugin. Finally, the key target subnetwork with 26 nodes and 268 edges was obtained based on a closeness ≥ 0.005 (average), betweenness ≥ 100.364 (average), and degree ≥ 22.345 (average) (Fig. 3B). The size and darkness of a node directly correlate with its degree value, with larger and darker nodes indicating higher degree values.

Fig. 3
figure 3

Network pharmacology of Omav against AD. A The Venn diagram showed 112 common targets between Omav and AD. B PPI network analyses revealed 26 key targets. C Bubble graph of GO enrichment analysis. D Bar chart of KEGG enrichment analysis

Full size image

GO enrichment analysis included biological progress (BP), cellular components (CC), and molecular functions (MF). A total of 319 BP, 57 CC, and 91 MF satisfied the required p-value < 0.05. The leading 10 BP, CC, and MF catalogs were visualized (Fig. 3C). Results of BP analysis suggested that the function of Omav against AD was primarily focused on positive regulation of cell migration, inflammatory response, positive regulation of gene expression, negative regulation of apoptotic process, and protein phosphorylation. The inflammatory response and negative regulation of apoptotic process were the leading BP items; thus, Omav might exert therapeutic effects by regulating neuroinflammation and apoptosis. In addition, we identified 142 signaling pathways enriched by KEGG analysis (p-value < 0.05) and selected the leading 20 for visualization (Fig. 3D). To note, Alzheimer’s disease was also one of the top pathways enriched by KEGG, supporting the potential role of Omav in the treatment of AD. Next, we tried to select key targets for further validation. We found that CASP3 and MTOR were the top two key targets obtained by PPI network, and were also involved in the inflammatory response and negative regulation of apoptotic process in the GO analysis as well as the Alzheimer’s disease pathway in the KEGG analysis. Therefore, we selected the leading two targets, CASP3 and MTOR, for further verification.

Molecular Docking Analysis

Molecular docking was performed to identify the binding ability of the candidate proteins to Omav for verification. The binding energies of Omav for CASP3 and MTOR were − 10.2 and − 9.8 kcal/mol, respectively. Notably, both binding energies were less than − 7 kcal/mol, indicating strong binding affinities. Figure 4 presents the three-dimensional docking conformations of the ligands with their respective targets, providing a visual representation of the molecular interactions.

Fig. 4
figure 4

Docking results of Omav with CASP3 (A) and MTOR (B)

Full size image

Omav Ameliorates Neuronal Apoptosis and Damage in APP/PS1 Mice

H&E staining was performed to evaluate the effects of Omav on neuronal damage. Neurons in the WT group displayed a normal structure; they were neatly arranged, and uniformly stained, with round nuclei and obvious nucleoli (Fig. 5A). In contrast, the APP/PS1 group exhibited smaller numbers of neurons, irregular neuronal arrangements, and abnormal neuronal morphology, including a dark appearance, the lack of clear nuclei, and triangulated and shrunken neuronal bodies. These pathological changes were mitigated upon the administration of Omav. Caspase 3, the key target explored by network analyses, is one of the executing enzymes in apoptosis, and is regarded as a proximate mediator of apoptosis [22, 23]. Therefore, we examined the levels of cleaved caspase 3 and other apoptosis-related proteins, including Bax and Bcl-2, in the brains. The results showed that the cleaved caspase 3 and Bax (a pro-apoptotic protein) were highly expressed in the brains of APP/PS1 mice than in WT littermates, whereas Bcl-2 (an anti-apoptotic protein) was downregulated. Omav treatment downregulated the expression of cleaved caspase 3 and Bax and upregulated that of Bcl-2 (Fig. 5B). Taken together, Omav regulated apoptosis and alleviated neuronal damage in APP/PS1 mice.

Fig. 5
figure 5

Omav ameliorated neuronal apoptosis and damage in APP/PS1 mice. A Representative images of H&E staining; scale bar: 100 µm. B Representative western blots for cleaved caspase 3, Bcl-2, and Bax. The relative density was normalized to β-actin. *P < 0.05; **P < 0.01. Data presented as mean ± SEM

Full size image

Omav Alleviates Inflammatory Response and Induces Autophagy in APP/PS1 Mice

We assessed whether Omav administration regulated the pro-inflammatory cytokine levels in APP/PS1 mice. The levels of IL-1β, IL-6, and TNF-α were markedly increased in APP/PS1; whereas the administration of Omav significantly reduced the elevation (Fig. 6A). Activated glia was reported to upregulate the expression of pro-inflammatory cytokines [24]. Thus, immunofluorescence staining was performed to assess microglial reactivity using an anti-Iba1 (a microglial cell marker) antibody. Results showed a marked increase in the immunoactivity of Iba1 in APP/PS1 mice compared with WT littermates, whereas the staining was less intense in Omav-treated mice (Fig. 6B). Then, we examined the effects of Omav on mTOR activation, since mTOR was the key gene in the PPI network and was also enriched in the GO analysis of the inflammatory response. Western blot analysis suggested that mTOR phosphorylation was significantly increased in the APP/PS1 group and was downregulated by Omav administration (Fig. 6C), indicating the inhibitory effect of Omav on the mTOR signaling pathway. mTOR is a pivotal negative regulator of autophagy, and previous study reported that impaired autophagy could result in inflammation [25, 26]. Therefore, we examined the expressions of autophagy-related proteins, including LC3 (a hallmark of autophagy activation) and P62 (a classic autophagy substrate). We observed a significant decrease in the LC3-II/LC3-I ratio and an increase in P62 expression in APP/PS1 mice, compared with their WT littermates. Notably, Omav treatment enhanced the ratio of LC3-II/LC3-I and decreased P62 expression, indicating an increase in autophagy flux (Fig. 6C). These results showed that Omav could alleviate neuroinflammation and induce autophagy through the inhibition of mTOR phosphorylation.

Fig. 6
figure 6

Omav alleviated neuroinflammation and induced autophagy in APP/PS1 mice. A ELISA analysis of pro-inflammatory cytokines including IL-1β, IL-6, and TNF-α; *P < 0.05; **P < 0.01. B Representative immunostaining of Iba-1 in the different groups. C Representative western blots for p-mTOR, LC3, and P62. *P < 0.05; **P < 0.01. Data presented as mean ± SEM

Full size image

Discussion

AD is a progressive neurological disease with an increasing prevalence, and it affects the most sensitive areas of the brain [27]. However, the development of AD drugs is challenging, with only two innovative drugs approved by the FDA since 2003 [28]. Safe and effective therapeutic strategies are urgently needed to treat AD. In the present study, Omav significantly alleviated cognitive dysfunction and reduced Aβ deposition in APP/PS1 mice. Additionally, based on the results of network pharmacology, Omav ameliorated neuronal apoptosis and inflammatory response by reducing the activation of caspase 3 and mTOR, respectively. This study presents compelling evidence supporting the role of Omav as a future treatment option for patients with AD.

Omav, also termed RTA-408 or CDDO-DFPA, is a novel semi-synthetic oleanane triterpenoid [29]. Partly because of its brain-penetrating properties, increasing evidence has indicated its pharmacological role in treating neurodegenerative diseases [10]. For example, Omav treatment significantly improved neurological function in patients with Friedreich ataxia and, thus, became the first-ever drug approved by the FDA for its treatment [11, 30]. Additionally, Yang et al. reported that Omav alleviated neuronal degeneration in mice with amyotrophic lateral sclerosis harboring the hSOD1G93A mutation [9]. In addition to inducing neuroprotection in neurodegenerative disorders, Omav exerts protective effects against other neurological diseases, including epilepsy and secondary brain injury after intracerebral or subarachnoid hemorrhage [8, 29, 31]. Despite reports on the neuroprotective role, its effects on AD and related critical mechanisms require clarification. The current study demonstrates that Omav significantly enhanced cognitive performance of APP/PS1 mice and also reduced Aβ deposition in both the hippocampus and cortex. These findings highlight the unique potential of triterpenoid derivatives, such as Omav, for the treatment of AD.

Network pharmacology was performed to identify the possible targets and signaling pathways regulated by Omav. The results showed that caspase 3 displayed the highest degree value, indicating its pivotal role in the treatment of AD with Omav. Caspase 3 is one of the key proteases that initiates the caspase cascade, and the stimulation of caspase 3 can result in apoptosis, which is considered to underlie the pathological manifestations associated with AD [23, 32]. To date, Omav has been proved to decrease the expression of cleaved caspase 3 in other pathological conditions including subarachnoid hemorrhage, diabetic cardiomyopathy, and H2O2-induced cell injury [31, 33, 34]. Consistent with previous studies, we observed an elevation in cleaved caspase 3 levels in the brains of APP/PS1 mice [22, 35], which was attenuated by Omav treatment. Furthermore, to further investigate the influence of Omav on apoptosis, we evaluated additional apoptosis-related proteins. Our findings revealed that Omav treatment led to an upregulation of Bcl-2 expression and a downregulation of Bax expression, suggesting that Omav effectively mitigated apoptosis in APP/PS1 mice.

Inflammatory response is another enriched BP item in GO pathway enrichment analysis and has been implicated to play a substantial role in AD [36, 37]. Microglia, the immune-derived cells within the brain, are activated in AD and release several pro-inflammatory cytokines, such as IL-1β and TNF-α, triggering neuroinflammation [38]. In the present study, we investigated the effect of Omav on the expression of pro-inflammatory cytokines and microglial activation. The results demonstrated that Omav treatment attenuated the expression of pro-inflammatory cytokines and suppressed microglial activation.

mTOR, a top-ranked gene in the PPI network that was enriched in inflammatory response, emerged as another key target in the treatment of Omav against AD. mTOR is a serine/threonine kinase, which belongs to the phosphoinositide kinase-related family [39]. As a master regulator of numerous cellular behaviors, mTOR plays a significant role in aging-related processes, such as protein synthesis, ribosome biogenesis, and cell proliferation [40]. Additionally, mTOR was also reported to hold a central position in autophagy; it hinders autophagy by inhibiting the autophagy-initiating complex [41]. Previous studies have reported that Omav and other oleanane triterpenoid derivatives could inhibit the phosphorylation of mTOR and promote autophagy [42, 43].

Interestingly, there is a complex relationship between neuroinflammation and autophagy. In the pathological process of AD, autophagy in microglia is inhibited, accompanied by an increase in mTOR signaling pathway activation [44, 45]. The compromised autophagy could lead to sustained inflammation. On the one hand, defect autophagy may increase inflammatory phenotype in AD and disrupt the degradation of various mediators involved in inflammasome activation or pro-inflammatory cytokines [40, 46, 47]. On the other hand, dysregulation of autophagy accelerates amyloidosis and tau pathology, which, in turn, exacerbates microglial activation and neuroinflammation [48, 49]. In the present study, compared with APP/PS1 mice, we found a significant reduction in the phosphorylation of mTOR and an increased autophagic level in those treated with Omav. Improving autophagic dysfunction may contribute to the attenuation of neuroinflammation observed with Omav treatment.

Our results hold crucial translational implications, offering direct evidence of the anti-apoptotic and anti-inflammatory properties of Omav in APP/PS1 mice via the inhibition of caspase 3 and mTOR, respectively. Given its ability to target multifactorial mechanisms underlying neurological injuries, Omav emerges as a promising therapeutic candidate for AD.

Conclusions

The administration of Omav resulted in significant improvements in cognitive dysfunction and reduced Aβ deposition in APP/PS1 mice. Furthermore, it attenuated neuronal apoptosis and suppressed neuroinflammation. Taken together, these results underscore the potential of Omav as an effective treatment for AD.