Advances in Tricyclic Compounds for the Treatment of Alzheimer's Disease

Abstract

Alzheimer's disease (AD) is a complex neurodegenerative disorder involving interconnected pathological pathways, including cholinergic dysfunction, Aβ deposition, tau hyperphosphorylation, neuroinflammation, oxidative stress, and dysregulation of metal ion homeostasis. Recently, the “one drug, one target” paradigm has faced certain clinical limitations, driving a paradigm shift toward multi‐target therapeutic strategies. In this context, tricyclic scaffolds, characterized by rigid fused‐ring skeletons, excellent structural tunability, and favorable neuropharmacological properties, have emerged as promising scaffolds for developing multi‐target AD therapeutics. These molecules can interact with multiple AD‐related targets through stable π–π stacking and hydrophobic interactions. Some tricyclic scaffolds, such as tetrahydrocarbolines, share structural similarity with endogenous neurotransmitters, suggesting potential involvement in modulating cognitive and mood‐related pathways, along with favorable BBB permeability and neuro‐compatibility. Accordingly, this review begins by outlining the distinct pathological mechanisms underlying AD, followed by a summary of recent progress on tricyclic compounds, encompassing both single‐target and multi‐target‐directed molecules, with an emphasis on structure–activity relationships and mechanisms of action, aiming to offer new insights and strategies for combating this devastating disease.

This review summarizes the research progress on tricyclic compounds, including both single‐target and multi‐target‐directed compounds in AD therapy, with an emphasis on structure–activity relationships and mechanisms of action, aiming to provide new insights and strategies for addressing this devastating disease.

1. Introduction

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by progressive cognitive decline and behavioral impairment, representing the most common type of dementia [1, 2]. Based on genetic background and clinical features, AD is broadly classified into three forms: familial AD (FAD), early‐onset AD (EOAD), and late‐onset AD (LOAD). FAD accounts for 1% of cases, driven by pathogenic mutations in the genes encoding amyloid precursor protein (APP), presenilin‐1 (PSEN1) and presenilin‐2 (PSEN2) [3]. EOAD accounts for 5%–10% of cases, while the predominant LOAD (>90%) is multifactorial, involving genetic, aging, metabolic, cerebrovascular pathology, neuroinflammation, oxidative stress and environmental exposures [4, 5, 6, 7, 8]. Emerging evidence also implicates epigenetic dysregulation and alterations in the gut microbiota in the onset and progression of LOAD [9, 10, 11, 12].

With global population aging, dementia has emerged as a critical public health challenge. Over 55 million people live with dementia, of which 60%–70% are attributable to AD. Epidemiological projections indicate 78 million by 2030 and 139 million by 2050 [13]. Meanwhile, AD‐related mortality reached 1.5 million in 2019, and the economic burden exceeded 1.3 trillion USD in 2021, expected to rise to 2.8 trillion USD by 2030 [14, 15]. Despite the alarming situation, the clinical management of AD remains severely limited [16]. This stark contrast underscores an urgent imperative: the development of novel, potent, and safe strategies for AD prevention and treatment has become a top priority in neuroscience and drug discovery.

2. Pathology and Mechanistic Hypotheses of AD

AD is characterized by distinct neuroanatomical and histopathological changes. Microscopically, three cardinal features define the disease: (i) β‐amyloid (Aβ) deposition forming senile plaques (SPs); (ii) aggregation of hyperphosphorylated tau forming neurofibrillary tangles (NFTs); and (iii) synaptic dysfunction with neuronal loss [17, 18, 19]. The pathogenesis of AD is complex, involving multi‐system cascades and intricate molecular interactions. Although a unified theory remains elusive, several key mechanisms have been identified, including cholinergic dysfunction, abnormal Aβ deposition, tau hyperphosphorylation with NFT formation, microglia‐driven neuroinflammation, oxidative stress, ferroptosis, and dyshomeostasis of metal ions (Figure 1), based on which multiple hypotheses have been proposed [20, 21, 22].

FIGURE 1.

Pathology and mechanistic hypotheses of AD.

2.1. Cholinergic Hypothesis

The cholinergic hypothesis has evolved from a focus on neurotransmitter deficiency to a systems‐level understanding of regulatory mechanisms. Early studies revealed extensive loss (40%–90%) of cholinergic neurons in the nucleus basalis of Meynert (NBM), leading to reduced acetylcholine (ACh) levels in the cerebral cortex and hippocampus, and contributing to cognitive impairment [23]. It is characterized by markedly reduced choline acetyltransferase (ChAT) activity and biphasic alterations in acetylcholinesterase (AChE) activity [24]. These findings established the classical framework of the cholinergic deficit hypothesis, representing a linear pathogenic cascade of “neuronal loss‐neurotransmitter deficiency‐cognitive decline.”

Recently, the complex interplay between the cholinergic system and the multifaceted pathological network of AD has been increasingly elucidated. Inestrosa [25] demonstrated the peripheral anionic site (PAS) of AChE can serve as a seed for Aβ fibril formation, and the AChE‐Aβ complex exhibits tenfold higher neurotoxicity than Aβ alone, partly explaining the abnormal accumulation of AChE in SPs. Furthermore, studies have uncovered links between Aβ and nicotinic acetylcholine receptors (nAChRs). In early AD, pathogenic Aβ inhibits presynaptic neurotransmitter release and postsynaptic currents via α7nAChR, impairing long‐term potentiation (LTP) [26]. Concurrently, α7nAChR facilitates neuronal internalization of Aβ, exacerbating Aβ accumulation and Ca²⁺ overload, thereby inducing excitotoxicity and neuronal death [27]. Moreover, Aβ binding to α7nAChR activates the extracellular signal‐regulated kinase (ERK) signaling pathway and upregulates glycogen synthase kinase‐3β (GSK‐3β) activity, accelerating NFT formation [28]. At the neural circuit level, research has moved beyond the neuron‐centric framework to identify anti‐inflammatory signaling pathways mediated by AChRs. ACh acts on α7nAChR expressed on astrocytes and macrophages, activating the JAK2/STAT3 pathway, upregulating anti‐inflammatory gene expression, and suppressing pro‐inflammatory cytokines such as tumor necrosis factor‐alpha (TNF‐α), interleukin‐6 (IL‐6), and interleukin‐1β (IL‐1β). Additionally, ACh inhibits nuclear factor kappa‐B (NF‐κB) signaling by blocking its nuclear translocation, thereby reducing pro‐inflammatory cytokine transcription [29, 30]. Collectively, these mechanisms have shifted the therapeutic paradigm for AD from single‐target intervention toward multi‐system coordinated regulation, driving the development of multifunctional acetylcholinesterase inhibitors (AChEIs) [31, 32, 33].

2.2. Aβ Hypothesis

The Aβ hypothesis posits that aberrant Aβ metabolism initiates the pathological cascade. Under genetic (APP, PSEN1, PSEN2) or environmental influences, abnormal cleavage of APP by β‐ and γ‐secretases generates Aβ peptides (Aβ _1–40 and Aβ _1–42), which aggregate into neurotoxic oligomers (AβO) and form insoluble fibrils that deposit as SPs. The process triggers downstream events, including impaired synaptic plasticity, LTP inhibition, activation of GSK‐3β and CDK5, tau hyperphosphorylation, and cognitive decline [34, 35, 36]. Advances in multi‐omics have expanded the hypothesis from the linear “Aβ‐tau‐neurodegeneration” model into a multidimensional framework. Aβ pathology interconnects with various mechanisms through complex networks: (i) Aβ deposition activates glia, promoting pro‐inflammatory cytokine release and chronic neuroinflammation [37]; (ii) AβO induces mitochondrial dysfunction and ROS accumulation, causing oxidative damage [38]; (iii) AβO disrupts proteostasis by impairing the ubiquitin‐proteasome and autophagy‐lysosome pathways, thereby hindering Aβ clearance [39, 40, 41, 42].

Recently, Gilbert [43] used cryo‐EM to determine the in situ structures of Aβ and tau in AD brains, revealing heterogeneity in Aβ fibril and location‐specific variability in tau filament architecture. These findings challenge the view of plaques as static deposits and inform diagnostic and therapeutic development. Current therapeutic strategies targeting the Aβ cascade fall into three categories: inhibiting Aβ production (e.g., verubecestat, semagacestat), antagonizing Aβ receptors, and enhancing Aβ clearance (e.g., anti‐Aβ antibodies, vaccines) [44]. The approval of aducanumab, lecanemab, and donanemab has boosted confidence in this approach [45]. However, the hypothesis remains debated. First, the temporal disconnection between Aβ burden and cognitive decline, along with weak correlations between Aβ reduction and cognitive improvement, suggest that targeting Aβ alone may be insufficient [46]. Second, 10%–30% of AD patients lack typical Aβ pathology, and tau transgenic models can develop neurodegeneration without Aβ, indicating an Aβ‐independent pathway may exist [47]. Despite these controversies, the Aβ hypothesis has provided a critical foundation for AD research.

2.3. Tau Hypothesis

Tau is a microtubule‐associated protein encoded by MAPT and plays critical roles in microtubule assembly and stabilization, intracellular transport, and synaptic regulation [48]. The tau hypothesis posits that abnormal tau modifications and formation of NFTs constitute the core pathological mechanism underlying cognitive impairment in AD. During AD progression, tau undergoes various post‐translational modifications, including hyperphosphorylation (Thr181, Ser202, Thr205), acetylation, and truncation, leading to dissociation from microtubules and loss of microtubule‐stabilizing function. Dissociated tau progressively aggregates within the neuronal soma, forming oligomers and paired helical filaments (PHFs), which ultimately mature into NFTs [49].

Pathological tau exerts neurotoxicity through multiple mechanisms: (i) disrupting microtubule network integrity and impairing axonal transport; (ii) propagating in a prion‐like manner between neurons, driving pathological spread; (iii) compromising synaptic stability and inhibiting synaptogenesis; and (iv) activating microglia to induce neuroinflammation and oxidative stress [50, 51, 52, 53]. Moreover, the pathological convergence between Aβ and tau has led to the hypothesis of synergistic effects in AD, manifested as cross‐activation between Aβ and tau, mutual enhancement of their pathological spread, coordinated induction of mitochondrial and autophagy‐lysosomal dysfunction, and synergistic synaptic toxicity [54, 55].

Recent advances in tau‐targeted therapies for AD have been notable. Leuco‐methylthioninium (TRx0237), a second‐generation tau inhibitor developed by TauRx, exhibits improved efficacy over its predecessor, methylene blue (Figure 2). However, it failed to meet primary endpoints in two Phase III trials, and a new Phase III trial (LUCIDITY) is ongoing [56]. In immunotherapy, active vaccines AADvac1 and ACI‐35.030 (JNJ‐64042056) target distinct tau epitopes to elicit specific antibody responses, demonstrating favorable immunogenicity and safety profiles [57]. Among passive immunotherapies, although gosuranemab did not meet primary endpoints in a Phase II trial, it effectively reduced free tau levels in cerebrospinal fluid (by 60%) with acceptable safety [58]; semorinemab showed cognitive benefits in a Phase II trial, with a 43.6% improvement on the ADAS‐Cog11 scale [59]. Gene therapy has also achieved significant progress. The antisense oligonucleotide IONIS‐MAPTRx reduced cerebrospinal fluid tau levels by over 50% in a Phase I trial [60], while an AAV‐based BDNF gene therapy (AAV2‐BDNF) has entered Phase I clinical evaluation [61]. BDNF gene therapy holds the potential to reconstruct neural circuits, slow cell loss and promote functional recovery. Additionally, emerging strategies such as the microtubule stabilizer TPI‐287, nanoconjugates and the tau‐targeting PROTAC degraders offer diverse directions for therapeutic development [62, 63, 64, 65].

FIGURE 2.

Structure of Leuco‐methylthioninium (TRx0237).

2.4. Oxidative Stress Hypothesis

Oxidative stress refers to a state of oxidative damage resulting from an imbalance between the generation and elimination of reactive species [66]. It is characterized by elevated levels of superoxide anion (O₂ ⁻), hydroxyl radical (·OH), and reactive nitrogen species (RNS), along with decreased activity of antioxidant enzymes like superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GPx) [67, 68].

Oxidative stress plays a multifaceted role in AD, acting as an initiator, amplifier, and executor. As an early driver, it precedes Aβ deposition and NFTs, triggering downstream pathological cascades, including upregulation of β‐ and γ‐secretases to promote Aβ production, inhibition of insulin‐degrading enzyme (IDE) to impair Aβ clearance, and activation of GSK‐3β and p38 MAPK signaling to exacerbate tau hyperphosphorylation [69, 70]. Throughout the disease course, oxidative stress contributes to cognitive decline by disrupting neuronal network integrity. Its neurotoxicity is manifested as: (i) neuron‐specific damage, selectively targeting metabolically demanding cholinergic neurons, inducing autophagic abnormalities in hippocampal CA1 neurons, and attacking synaptic vesicle proteins and glutamate receptors to impair synaptic plasticity [71]; (ii) organelle dysfunction, inducing mitochondrial DNA mutations and respiratory chain complex dysfunction, leading to intracellular calcium overload and excitotoxicity [72]; and (iii) interaction with pathological proteins, upregulating β‐ and γ‐secretases via Notch1 signaling inhibition to promote Aβ deposition, while activating calcium‐dependent and calmodulin‐dependent kinases to disrupt the balance between protein phosphatases and kinases, thereby exacerbating tau hyperphosphorylation [73].

To date, therapeutic strategies based on the oxidative stress hypothesis have included antioxidants (e.g., vitamin E, lipoic acid), endogenous antioxidant system activators (e.g., dimethyl fumarate), NADPH oxidase inhibitors, and mitochondria‐targeted antioxidants (e.g., MitoQ, SkQ1) [74, 75, 76]. However, clinical trials targeting this pathway have faced challenges due to limited blood‐brain barrier (BBB) penetration, inadequate metabolic stability, and target specificity. Given the multidimensional pathological network of AD, antioxidant strategies may be better suited as synergistic components rather than standalone therapies. Therefore, the development of multifunctional antioxidants capable of acting on multiple pathological nodes holds promise for delivering more significant clinical benefits.

2.5. Neuroinflammation Hypothesis

Neuroinflammation refers to the CNS immune response to pathological stimuli, involving a complex cascade of cytokines, chemokines, ROS, and second messengers. Recently, the role of the innate immune system in AD pathophysiology has gained increasing attention. Accumulating evidence indicates that neuroimmune activation is not merely a consequence of AD pathology but also a key driver of disease progression [77, 78].

As the primary immune cells, microglia and astrocytes play central roles in AD‐related synaptic damage and neurodegeneration. Under physiological conditions, microglia maintain an M2 anti‐inflammatory phenotype, supporting neural homeostasis by monitoring the microenvironment, clearing cellular debris, and promoting tissue repair. However, in response to chronic stimuli like Aβ deposition and tau pathology, microglia undergo transcriptional shifts toward a pro‐inflammatory M1 phenotype, releasing large amounts of inflammatory mediators, including TNF‐α, nitric oxide (NO), and IL‐1β, thereby inducing neuronal injury and parenchymal atrophy [79, 80]. Moreover, microglia engage in complex bidirectional interactions with core AD pathologies. On one hand, microglial phagocytosis of Aβ via the TREM2 receptor can temporarily alleviate plaque burden, but sustained phagocytic activity activates the NLRP3 inflammasome, leading to caspase‐1‐dependent maturation and release of IL‐1β, ultimately impairing hippocampal neurogenesis and synaptic plasticity [81, 82]. On the other hand, inflammatory cytokines can exacerbate tau phosphorylation by activating GSK‐3β, while pathological tau in turn reactivates microglia, establishing a self‐amplifying inflammatory loop [83, 84]. Scott–Hewitt [85] demonstrated that during aging, the innate immune complement protein C1q secreted by microglia is internalized by neurons and associates with neuronal ribonucleoprotein (RNP) complexes in an RNA‐dependent manner, affecting protein translation homeostasis and synaptic plasticity, ultimately impairing memory formation and consolidation. This mechanism establishes a direct molecular link between aging, immune activation, and synaptic loss, providing important support for the inflammaging hypothesis.

Astrocytes serve as homeostatic and defensive cells in the CNS, maintaining neuronal function, BBB integrity, and neurotransmitter metabolism. In AD pathology, Aβ deposition and microglia‐derived factors such as TNF‐α and C1q drive astrocytic conversion to reactive A1/A2 phenotypes [86]. Reactive astrogliosis impairs glutamate reuptake and potassium buffering while promoting the release of complement proteins (C1q, C3) and inflammatory cytokines via the NF‐κB pathway, thereby exacerbating neuroinflammation, promoting Aβ fibril deposition, and contributing to synaptic damage. This vicious cycle constitutes a major driver of AD progression [87]. Notably, recent studies suggest that modulating astrocytic metabolism or spatial behavior may offer therapeutic opportunities. Minhas [88] found that inhibition of indoleamine 2,3‐dioxygenase 1 (IDO1) restores astrocytic glucose metabolism in AD models, rescues hippocampal synaptic plasticity, and improves hippocampal memory function in AD mice. Huang [89] demonstrated that targeting the Plexin‐B1 signaling pathway optimizes the spatial interaction between glial cells and pathological Aβ plaques. Reducing Plexin‐B1 levels decreases the number of reactive astrocytes and microglia, and enhances the encapsulation of Aβ plaques by the remaining glial cells, forming a denser glial barrier that promotes Aβ sequestration and clearance, thereby alleviating neuroinflammation. Currently, pharmacological interventions targeting neuroinflammation in AD remain at an exploratory stage, with no clinically approved drugs specifically indicated for this purpose. Ongoing research efforts are focused on microglial modulation (e.g., the TREM2 agonist AL002 and the CSF1R inhibitor PLX5622), kinase inhibitors (e.g., baricitinib), and immunomodulatory strategies (e.g., IL‐1β antibody canakinumab and the Tdap vaccine) [90, 91, 92].

2.6. Vascular Hypothesis

The vascular hypothesis posits that cerebrovascular dysfunction is not merely a comorbidity but rather an early and driving event in AD. This notion is supported by growing evidence linking vascular risk factors, such as hypertension, diabetes, atherosclerosis, and cerebral hypoperfusion, to the onset and progression of cognitive decline [93].

Cerebrovascular abnormalities contribute to AD pathology through multiple mechanisms. Chronic cerebral hypoperfusion reduces oxygen and glucose delivery, leading to energy deprivation, mitochondrial dysfunction and oxidative stress, which in turn promote Aβ accumulation and tau hyperphosphorylation [94]. Impaired cerebral blood flow also compromises Aβ clearance across the BBB and via perivascular drainage pathways, thereby exacerbating Aβ deposition [95]. Moreover, BBB disruption facilitates the infiltration of peripheral immune cells and neurotoxic substances into the brain parenchyma, fueling neuroinflammation and neuronal injury. Vascular pathology and classical AD hallmarks interact bidirectionally. Aβ deposits around cerebral vessels, a condition known as cerebral amyloid angiopathy (CAA), further impairs vascular integrity and vasomotor function, perpetuating hypoperfusion and ischemia [96]. Similarly, tau pathology has been linked to vascular smooth muscle cell degeneration and microvascular dysfunction [97]. These interconnections create a vicious cycle in which vascular damage and neurodegenerative pathology reinforce each other, accelerating AD progression.

Despite accumulating evidence, the vascular hypothesis remains complementary to the more established Aβ and tau cascades rather than a competing paradigm. Current therapeutic strategies targeting vascular components include management of vascular risk factors, improvement of cerebral perfusion, restoration of BBB integrity, and enhancement of perivascular clearance pathways [98].

2.7. Metal Ions Dyshomeostasis Hypothesis

Metal ions are essential elements involved in signal transduction, enzymatic reactions, and structural maintenance. Studies have revealed a strong link between metal ion dyshomeostasis and AD pathology [99]. Postmortem analyses show that metal levels in AD brains, particularly in the hippocampus, amygdala, olfactory cortex, and neocortex, are elevated three to fivefold compared with normal controls, with regional accumulation patterns closely matching the anatomical distribution of AD pathologies [100]. Transition metal ions (e.g., Cu²⁺, Fe²⁺/Fe³⁺, Zn²⁺) bind specifically to N‐terminal histidine residues (His6, His13, His14) of Aβ, accelerating its conversion from soluble monomers to β‐sheet structures and promoting SPs formation [101]. Moreover, redox‐active Cu‐Aβ and Fe‐Aβ complexes generate ROS through Fenton‐like reactions, triggering cascading damage, including direct oxidation of neuronal membrane lipids, proteins, and DNA, and activation of the microglial TLR4/NF‐κB pathway, which promotes release of pro‐inflammatory cytokines, ultimately leading to mitochondrial dysfunction and impaired synaptic plasticity [102, 103]. Recent research has found that endogenous lithium deficiency, distinct from excess of other metals, initiates AD pathology via Aβ‐mediated trapping; lithium orotate supplementation could bypass this trap and reverse cognitive decline [104].

Several classes of modulators have been developed to target metal ion dyshomeostasis in AD, including traditional chelators (e.g., deferoxamine), metal transporters (e.g., Cu^IIGTSM), and novel 8‐hydroxyquinoline derivatives (e.g., PBT2) [105, 106]. However, the development of ideal metal‐targeting therapeutics faces substantial challenges, requiring a combination of metal selectivity, moderate affinity, appropriate molecular weight, favorable BBB permeability, and low toxicity. These stringent physicochemical requirements largely account for the failure of many metal‐targeting agents in clinical trials. Although current metal‐targeting agents face challenges, metal ion modulation‐with its potential for multi‐dimensional intervention within the pathological network‐complements existing anti‐Aβ and tau strategies and offers new possibilities for AD treatment.

3. Overview of Clinical Therapeutics for AD

Current clinical treatments for AD fall into two categories: symptomatic therapies and disease‐modifying therapies (DMTs). Approved symptomatic treatments primarily consist of AChEIs and N‐methyl‐D‐aspartate (NMDA) receptor antagonists. AChEIs improve cholinergic neurotransmission by inhibiting AChE and increasing synaptic ACh concentrations, thereby temporarily alleviating cognitive decline. Tacrine, the first AChEI approved for AD (1993) based on the cholinergic hypothesis, was withdrawn in 2013 due to dose‐dependent hepatotoxicity. Subsequent development yielded safer and more effective AChEIs donepezil (1996), rivastigmine (2000), and galantamine (2001), which have become mainstream AD treatments [107] (Figure 3). In 2024, Alpha Cognition's oral extended‐release benzgalantamine received FDA approval for mild‐to‐moderate AD. This prodrug reduces gastrointestinal side effects and exerts dual mechanisms: AChE inhibition and allosteric modulation of α7 and α4β2 nAChRs, providing long‐term cognitive benefits [108]. Memantine, an NMDA receptor antagonist, modulates glutamatergic signaling by attenuating excessive NMDA receptor activation, thereby reducing excitotoxicity and offering symptomatic benefits in moderate‐to‐severe AD [109]. However, these symptomatic agents only temporarily alleviate cognitive decline without halting or reversing disease progression [110, 111].

FIGURE 3.

Overview of AD drug development history.

The past 5 years have witnessed major breakthroughs in DMTs for AD, particularly in anti‐Aβ antibodies. In 2021, aducanumab became the first Aβ antibody to receive accelerated FDA approval [112] (Figure 3). Aducanumab selectively binds and promotes clearance of insoluble fibrils. In the EMERGE trial, it slowed cognitive decline by 22%, but the ENGAGE trial failed to meet its primary and secondary endpoints, resulting in inconsistent efficacy evidence. Its approval was based primarily on Aβ plaque reduction as a surrogate endpoint rather than definitive clinical benefit, limiting its use [113]. In 2023, Biogen discontinued promotion of aducanumab. Nevertheless, the drug laid the groundwork for subsequent Aβ antibodies and prompted regulatory agencies to emphasize the combination of biomarker improvement with clinically meaningful endpoints in AD drug approvals.

Lecanemab, co‐developed by Eisai and Biogen, received full FDA approval in July 2023 as the second Aβ‐targeting DMT. Lecanemab selectively targets neurotoxic AβO and soluble protofibrils. Phase III trials demonstrated clear clinical benefit, with a 27% slowing of cognitive decline [114]. Regarding safety, lecanemab carries a risk of amyloid‐related imaging abnormalities (ARIA), though incidence is relatively low (12.6% for edema, 17.3% for microhemorrhages), with most cases being asymptomatic or mild [115]. The success of lecanemab marks a new phase in AD therapy focused on targeted Aβ clearance, though its long‐term efficacy requires further evaluation.

Donanemab, developed by Eli Lilly, is a monoclonal antibody targeting pyroglutamate‐modified Aβ (pGlu‐Aβ) at the N‐terminus. Phase III trials demonstrated its potential for precision therapy based on tau pathology stratification: cognitive decline was slowed by 35.1% in patients with tau burden, with even greater efficacy (40%) in the low/intermediate tau subgroup [116]. Safety‐wise, donanemab also exhibits mechanism‐related ARIA, with incidence rates of 24% for edema and 31% for microhemorrhages, mostly mild to moderate. Notably, approximately 52% of patients achieved prespecified Aβ clearance criteria after 1 year of treatment and could discontinue dosing [117]. This “finite‐duration” paradigm may improve long‐term safety and reduce economic burden. Approved in July 2024, donanemab joins lecanemab as an Aβ‐targeting therapy with clear clinical benefit, and its “treat‐to‐clear” strategy may reshape the clinical treatment paradigm for AD.

However, current Aβ antibody therapies face several limitations: (i) narrow therapeutic window, restricted to early‐stage patients and requiring stringent Aβ‐positive selection via PET or cerebrospinal fluid testing, limiting the eligible population; (ii) modest cognitive benefits, as lecanemab and donanemab slow decline but neither halt disease progression nor achieve neural repair; (iii) safety concerns, with relatively high ARIA incidence, particularly in APOEε4 carriers, necessitating frequent MRI monitoring and adding to healthcare burden; and (iv) limited impact on tau pathology, with inability to reverse established synaptic damage and neuronal loss, and some patients experiencing continued cognitive decline even after Aβ clearance. These limitations underscore the complexity of AD treatment and highlight the need for more sensitive early diagnostic tools, novel drug delivery systems, optimized therapeutic windows, and individualized monitoring strategies to overcome current therapeutic bottlenecks [118, 119, 120, 121].

4. Tricyclic Scaffold‐Based Compounds for AD

AD is a complex neurodegenerative disorder characterized by a network of interconnected pathological pathways, including cholinergic dysfunction, Aβ deposition, tau hyperphosphorylation, neuroinflammation, and oxidative stress [122]. The traditional “one drug, one target” drug development model has exhibited certain clinical limitations. This therapeutic challenge has prompted researchers to reevaluate the dynamic interplay among various pathological pathways in AD and to explore novel drug development strategies [123]. Advances in systems biology have offered a new perspective for AD drug discovery, laying a theoretical foundation for multi‐target drug design. The emergence of network pharmacology has further underscored the distinctive advantages of multi‐target therapeutics in the treatment of AD, driving a paradigm shift in drug development models and sparking a wave of intensified research into multi‐target drugs [124, 125].

In this context, tricyclic core molecules featuring three fused rings sharing at least two adjacent atoms have emerged as promising scaffolds for AD drug development. The rigid scaffold and extended conjugated systems enable stable insertion into the active pockets of AD‐related proteins via extensive π–π stacking and hydrophobic interactions. Moreover, tricyclic skeletons offer excellent structural tunability; both the ring system and peripheral sites can be selectively modified to fine‐tune affinity and selectivity for distinct targets, facilitating the rational design of multi‐target agents [126, 127]. Some tricyclic scaffolds, such as tetrahydrocarbolines and their analogs, share structural similarity with endogenous neurotransmitters like serotonin and melatonin, suggesting potential involvement in modulating cognitive and mood‐related pathways, along with favorable BBB permeability and neuro‐compatibility [128]. From a physicochemical perspective, tricyclic scaffolds often exhibit desirable drug‐like properties. Many possess favorable lipophilicity and metabolic stability, which translate into acceptable oral bioavailability and brain penetration, as exemplified by chlorpromazine and imipramine. Importantly, the tricyclic core generally lacks structural alerts, thereby reducing the risk of idiosyncratic toxicity and genotoxicity. Thence, leveraging the tricyclic scaffold for rational drug design to discover novel therapeutics that concurrently modulate multiple pathological aspects of AD may represent a promising direction in current drug discovery. This review summarizes the research progress on tricyclic compounds, including both single‐target and multi‐target directed molecules in AD therapy, with an emphasis on structure–activity relationships (SAR) and mechanisms of action, aiming to provide new insights and strategies for addressing this devastating disease.

4.1. Single‐Target Tricyclic Compounds

4.1.1. AChE Inhibitors

The discovery and application of ChE inhibitors represent a significant breakthrough in symptomatic AD therapy. Inhibiting AChE to reduce ACh hydrolysis in the synaptic cleft is one of the most direct strategies for ameliorating cognitive deficits associated with cholinergic dysfunction [129]. In recent years, numerous selective AChE inhibitors based on tricyclic scaffolds have been reported. El‐Sayed [130] reported a class of tricyclic compounds based on the pyrrolizine scaffold and evaluated their inhibitory activities against AChE and butyrylcholinesterase (BChE). Most of the compounds exhibited selectivity for HuAChE over HuBChE (1b K _i  = 6.85 μM for HuAChE; K _i  > 50 μM for BChE), with compound 1a showing the most potent activity (K _i  = 0.4 μM for HuAChE) (Figure 4). Moreover, 1a demonstrated cytotoxicity lower than or comparable to donepezil in human neuroblastoma (SH‐SY5Y) and normal human hepatic (THLE2) cell lines. In vivo studies confirmed that 1a effectively ameliorated cognitive impairment in scopolamine‐induced AD mouse models.

FIGURE 4.

AChE inhibitors based on tricyclic scaffolds.

Ekins [131] employed a combination of high‐throughput screening and machine learning to identify novel AChEIs, the antiviral agent tilorone (2) (Figure 4). Tilorone exhibited potent inhibitory activity against both electric eel AChE (EeAChE) and HuAChE, with IC₅₀ of 14.4 nM and 73.3 nM, while showing no significant inhibition against BuChE. Molecular docking results revealed that tilorone exerts its inhibitory effect by occupying the active‐site gorge of AChE, with selectivity likely attributable to specific π–π interactions with Trp86 in the PAS. Pharmacological safety profiling demonstrated that tilorone (1 μM) inhibited only AChE among 44 toxicology targets and 485 kinases tested, indicating favorable selectivity.

White [132] synthesized a series of N‐benzylpiperidine derivatives bearing novel isoxazole‐containing tricyclic ring systems. These compounds exhibited potent AChE inhibitory activity, with IC₅₀ values ranging from 0.33–3.6 nM, while showing weaker inhibition against BuChE (IC₅₀ = 600–23,000 nM). Among them, pyrrolobenzisoxazole 3 was identified as the most potent AChE inhibitor (IC₅₀ = 0.33 nM) (Figure 4). In vivo studies revealed that 3 significantly increased extracellular ACh levels by 200% at a dose of 0.4 mg/kg. Besides, 3 exhibited a favorable separation between the dose that elevated extracellular ACh levels and induced peripheral side effects (salivation ED₅₀ = 26 mg/kg) or acute lethality (LD₅₀ [1 h] = 42 mg/kg), with selectivity and safety indices exceeding 60‐fold.

Greig [133] utilized physovenine and physostigmine as lead compounds to prepare two novel series of carbamate derivatives (4a‐d, 5a‐d) (Figure 4). Biological evaluation revealed that these compounds acted as potent inhibitors of either AChE or BChE, with the selectivity for the two enzymes being precisely tunable through the N‐substituent groups. The N‐2′‐methylphenyl‐substituted derivatives exhibited exceptionally high selectivity for AChE (4c), showing over 1300‐fold selectivity for AChE over BChE.

4.1.2. BChE Inhibitors

As research has progressed, the role of BChE in AD has become increasingly evident. BChE is not only involved in the late‐stage regulation of the cholinergic system but may also be directly implicated in pathologies like Aβ deposition and neuroinflammation [134]. This expanded understanding has broadened the scope of drug discovery efforts, shifting attention toward the development of selective BChE inhibitors or dual AChE/BChE inhibitors [135]. To date, reported highly selective tricyclic BChE inhibitors primarily include physostigmine derivatives and phenothiazine‐based compounds. Representative phenothiazine derivatives include ethopropazine (6), N‐[(2‐diethylamino)propionyl]phenothiazine (ASTRA1397, 7), and 10‐(9‐anthracenylcarbonyl)phenothiazine (8) [136, 137] (Figure 5). The structural basis for their selective inhibition lies in the larger active‐site gorge of BChE compared to AChE, which better accommodates bulky hydrophobic groups. This enables the bicyclic aromatic system of phenothiazine derivatives to form stable π–π stacking interactions with residues like Phe329 and Tyr332, thereby achieving high selectivity.

FIGURE 5.

BChE inhibitors based on tricyclic scaffolds.

Sparatore [138] designed and evaluated a series of phenothiazine and other tricyclic system derivatives (9A‐D) (Figure 5). The results indicated that these compounds exhibited varying degrees of inhibition against AChE and BChE, with generally more pronounced inhibition against BChE. Among them, compound 9a demonstrated the most potent BChE inhibitory activity (IC₅₀ = 0.15 μM) and favorable selectivity (SI = 47). Notably, some of the compounds have also been shown to possess additional pharmacological properties relevant to AD therapy, such as presynaptic muscarinic receptor antagonism, aminopeptidase N (APN) inhibition, and antipsychotic effects. Thus, this class of scaffolds holds promise as a lead structure for the development of novel multi‐target drugs.

4.1.3. BACE Inhibitors

Sequential cleavage of APP by β‐secretase (BACE1) and γ‐secretase generates Aβ ₄₀ and Aβ ₄₂, which aggregate abnormally and drive AD pathology. This process activates cell death pathways, disrupts axonal transport, and ultimately leads to cognitive decline [139]. Given the critical role of BACE1 in APP processing, BACE1 has emerged as a key target for DMTs in AD [140]. However, BACE1 is primarily localized in acidic intracellular compartments, posing a significant challenge for transition‐state‐based inhibitors [141]. Additionally, early inhibitors based on hydroxyethylamine and hydroxyethylene scaffolds suffered from low brain penetration, poor oral bioavailability, and susceptibility to P‐glycoprotein (P‐gp) efflux, limitations that hindered clinical translation [142].

To address the challenges, Fremeau [143] employed a structure‐based drug design approach to develop a novel class of tricyclic BACE1 inhibitors based on the aminooxazoline xanthene scaffold. Compound 10 exhibited potent BACE1 inhibitory activity (IC₅₀ = 0.008 μM). In a HEK293 cell model, 10 effectively suppressed Aβ ₄₀ production (IC₅₀ = 0.036 μM) (Figure 6). Pharmacokinetic evaluation revealed favorable liver microsomal stability (HLM = 19 μL·min⁻¹·mg⁻¹, RLM = 20 μL·min⁻¹·mg⁻¹), moderate in vivo clearance (CL = 1.08 L·h⁻¹·kg⁻¹), acceptable oral bioavailability (F = 31%), good CNS penetration (AUC_inf = 7.29 μM·h), and low P‐gp efflux liability (P _app = 9.1 × 10⁻⁶ cm/s). In vivo pharmacodynamic studies in Sprague–Dawley (SD) rats further showed that a single oral dose of 10 (30 or 100 mg/kg) significantly reduced central Aβ ₄₀ levels. Compared with controls, Aβ ₄₀ inhibition in cerebrospinal fluid and brain tissue reached 76% and 45%, respectively, at 30 mg/kg, and increased to 81% and 63% at 100 mg/kg.

FIGURE 6.

BACE inhibitors based on tricyclic scaffolds.

4.1.4. γ‐Secretase Inhibitors

γ‐secretase is an intramembrane protease complex composed of four subunits: presenilin (PS), APH‐1, Nicastrin, and PEN‐2. As the key enzyme responsible for Aβ generation, γ‐secretase has emerged as an important target in AD drug development [144]. Preclinical studies have demonstrated that γ‐secretase inhibitors (GSIs) can significantly reduce Aβ levels in plasma, cerebrospinal fluid, and the cerebral cortex in animal models, exhibiting potential disease‐modifying effects [145]. However, while inhibiting APP cleavage, these inhibitors often interfere with signaling pathways such as Notch, leading to severe adverse effects that limit the clinical application [146].

Early studies identified the arylsulfonyl GSI MRK‐560 (11, IC₅₀ = 14.3 nM) as a compound with favorable therapeutic efficacy in mouse models and a lack of typical Notch‐related toxicity upon long‐term administration, a property attributed to its selective inhibition of the PS1 isoform [147, 148] (Figure 7). Given the favorable safety profile, MRK‐560 became an ideal lead for structural optimization. In 2010, Zhao [149] introduced a bridging structure between the cyclohexane and difluorobenzene rings, leading to the discovery of a novel tricyclic sulfonyl GSI (compound 12, IC₅₀ = 27.1 nM). In an acute CRND8 mouse model, a single oral dose of 30 mg/kg achieved 90% inhibition of plasma Aβ ₄₀ and 42% inhibition of cortical Aβ ₄₀. In a chronic dosing regimen, 12 consistently reduced plasma and cortical Aβ ₄₀ levels without inducing Notch‐related toxicity. That same year, Hyde [150] designed compound 13 (IC₅₀ = 41.7 nM) by relocating the trifluoromethanesulfonamide group to the C‐7 position. A single oral dose of 10 mg/kg resulted in 101% and 71% inhibition of plasma and brain Aβ ₄₀, respectively. Additionally, 13 (30 μM) showed no significant inhibition of P450 enzymes or the hERG channel. In 2015, Parker [151] retained the key pharmacophore of sulfonyl GSIs while introducing a spirocyclic structure to attach a polar group at the C‐7 position, leading to the design of novel tricyclic GSIs (14). Representative compound 14a (IC₅₀ > 20 μM) not only circumvented the CYP2C9 inhibition risk associated with MRK‐560 but also exhibited favorable PS‐1/PS‐2 isoform selectivity (SI = 33). Furthermore, incorporation of a cyclic phosphonate group in 14a simultaneously optimized its potency (Membr. Aβ ₄₂: IC₅₀ = 9.5 nM) and pharmacokinetic properties (AUC_6 h = 2473 nM·h, B/P = 0.2). In vivo studies confirmed that a single oral dose of 10 mg/kg significantly reduced brain Aβ ₄₂ levels by 39% in wild‐type rats, demonstrating promising disease‐modifying potential.

FIGURE 7.

γ‐secretase inhibitors based on tricyclic scaffolds.

4.1.5. γ‐Secretase Modulators

The clinical failures of semagacestat and avagacestat prompted a strategic shift from “complete inhibition “to” precise modulation,” thereby advancing research into γ‐secretase modulators (GSMs) [152]. GSMs do not interfere with the Notch signaling pathway. Instead, they modulate APP‐CTFβ cleavage to promote the generation of shorter, more soluble Aβ peptides, reducing the production of neurotoxic Aβ ₄₂. This mechanism offers a new paradigm for the development of small‐molecule anti‐Aβ therapeutics [153].

Thompson [154] designed a series of novel tricyclic GSMs based on lead compound (15, IC₅₀ = 29 nM; Figure 8). They employed a conformational restriction strategy to covalently link the aniline and triazine rings, followed by replacement of the triazine core with a thiazole bioisostere, yielding a novel series (16a‐i). Among them, compound 16a exhibited potent Aβ ₄₂ inhibitory activity in H4 cell models (IC₅₀ = 35 nM). Preclinical evaluation revealed favorable metabolic stability (human liver microsomal metabolic rate: 0.035 nmol·min⁻¹·mg⁻¹) and no significant hERG channel inhibition. However, 16a showed strong inhibition of CYP3A4 (IC₅₀ = 0.7 µM) and exhibited high plasma protein binding (>99.8%) in both human and mouse. In 3xTg‐AD mouse models, oral administration of 16a (1–100 mg/kg) demonstrated good BBB penetration (B/P = 1.1). At doses ≥ 10 mg/kg, 16a significantly reduced brain Aβ ₄₂ levels, with a half‐maximal effective dose (ED₅₀) of 100 mg/kg, without affecting total Aβ production. These findings further validate the therapeutic advantage of precise modulation offered by GSMs.

FIGURE 8.

γ‐secretase modulators based on tricyclic scaffolds.

4.1.6. ACAT Inhibitors

Genetic studies suggest that intracellular cholesterol metabolism plays a significant role in regulating Aβ aggregation. Acyl‐coenzyme A: cholesterol acyltransferase (ACAT), a key enzyme in maintaining cholesterol homeostasis, catalyzes the esterification and storage of free cholesterol. This process indirectly affects the cholesterol pool available for ABCA1 transport and may alter the membrane lipid microenvironment, thereby promoting APP processing and Aβ generation [155, 156]. Studies have demonstrated that ACAT inhibitors restore cholesterol homeostasis, inhibiting atherosclerosis progression while significantly reducing cerebral amyloid pathology [157].

Hua [158] designed a series of derivatives based on the tricyclic pyranone (TP) scaffold. Compound 17 exhibited significant neuroprotective effects in MC65 cells expressing a partial APP fusion protein (EC₅₀ = 0.15 μM), but showed no inhibitory activity against ACAT (IC₅₀ > 500 μM) (Figure 9). Using 17 as a lead, the researchers introduced a benzylamine moiety into the C‐7 isopropyl side chain, yielding tricyclic pyranone derivatives (18a‐u), and evaluated the cytoprotective and lipid metabolism‐modulating activities [159]. SAR analysis revealed that para‐hydroxy or para‐hydroxymethyl substitutions on the benzene ring, as well as meta‐fluoro modification, were critical for maintaining high potency. Compounds 18b, 18e, 18m, and 18s exhibited potent protection against Aβ‐induced toxicity at nanomolar concentrations (EC₅₀ = 67–145 nM), while also effectively inhibiting ACAT activity (IC₅₀ = 0.3–1.8 μM) and upregulating ABCA1 cholesterol transporter expression (EC₅₀ = 0.6–2.5 μM). By inhibiting ACAT and activating the LXR‐ABCA1 signaling axis, this class of compounds operates through a unique dual mechanism distinct from conventional single‐target agents.

FIGURE 9.

ACAT modulators based on tricyclic scaffolds.

4.1.7. RXRα Agonists

Retinoid X receptor alpha (RXRα) is a key regulator within the nuclear receptor superfamily, and its agonists have demonstrated significant therapeutic potential in promoting Aβ clearance [160]. Studies have shown that the RXRα agonist bexarotene specifically activates the RXRα/LXR and RXRα/PPARγ pathways, leading to upregulation of APOE and ATP‐binding cassette transporter A1 (ABCA1), thereby enhancing microglia‐mediated phagocytosis and clearance of soluble Aβ [161]. AD animal models have further confirmed that bexarotene treatment effectively reduces cerebral Aβ plaque burden and improves cognitive function, establishing RXRα as a promising therapeutic target for promoting Aβ clearance [162].

Tian [163] isolated 12 pairs of structurally novel terpenoids, (±)‐elodeoidileons A‐L, from Hypericum elodeoides. These compounds feature a unique linear or angular 6/6/6 tricyclic core skeleton (Figure 10). Using spectroscopic techniques, the researchers elucidated their planar structures and absolute configurations. Bioactivity assays revealed that compound 19a‐1 significantly activated RXRα transcriptional activity and upregulated ABCA1 expression in a dose‐dependent manner. Fluorescence titration experiments confirmed a direct interaction between 19a‐1 and the ligand‐binding domain (LBD) of RXRα, with a dissociation constant (K _d) of 5.85 μM. Molecular docking results indicated that 19a‐1 forms a key hydrogen bond with Asn306 in the RXRα‐LBD via its C‐13 hydroxyl group, revealing a binding mode distinct from that of conventional RXRα ligands and suggesting a novel mechanism of action. Cellular thermal shift assay (CETSA) validated that 19a‐1 significantly enhanced the thermal stability of RXRα protein at the cellular level, providing additional evidence for their direct interaction under physiological conditions.

FIGURE 10.

RXRα agonists based on tricyclic scaffolds.

4.1.8. p53 Inhibitors

Nonamyloid‐β component (NAC), the second major component of Aβ plaques, colocalizes with Aβ and promotes its aggregation, playing an important role in the early stages of AD. Studies have shown that NAC significantly elevates p53 levels in rat pheochromocytoma cells. As a key regulator of cell survival, p53 induces cell cycle arrest and apoptosis by activating genes such as p21, Bax, and PUMA, as well as by interacting with mitochondrial proteins. Notably, the p53 functional inhibitor pifithrin‐β (20) has demonstrated significant anti‐apoptotic activity in models of neurodegenerative diseases [164]. Importantly, in AD‐associated microglia and Aβ‐expressing neurons, apoptosis has been shown to be p53‐mediated, and elevated p53 transcript levels correlate closely with Aβ deposition in neurons [165].

Based on this mechanistic understanding, Greco [166] employed a pharmacophore‐based virtual screening strategy to identify 20 drug‐like small molecules structurally related to pifithrin‐β (20) from a compound library (Figure 11). Using an NAC‐induced SH‐SY5Y cell injury model, they systematically evaluated the regulatory effects of compounds 21a‐c on the p53 signaling pathway. The results showed that 21b exhibited superior neuroprotective effects compared to pifithrin‐β at both 1 and 10 μM, with average cell survival rates of 89.1% and 96.7%, respectively. Dose‐response studies revealed that 21b inhibited NAC‐induced cell death with a half‐maximal effective concentration (EC₅₀) of 23.4 nM, demonstrating approximately fourfold greater neuroprotective potency than pifithrin‐β (EC₅₀ = 89.1 nM). Quantitative real‐time PCR analysis further confirmed that pretreatment with 21b (10 μM) effectively suppressed NAC‐induced upregulation of p53 target gene mRNA levels (MDM2, p21, BAX), thereby blocking cell cycle arrest and apoptosis. These findings indicate that the neuroprotective effects of 21b are primarily mediated through inhibition of the p53 pathway.

FIGURE 11.

p53 inhibitors based on tricyclic scaffolds.

4.1.9. MAPRE3 Modulator

As vital components of traditional Chinese medicine, animal‐derived medicines have gained widespread recognition for their clinical value. Recent advances have drawn increasing attention to animal‐derived small‐molecule constituents such as alkaloids and steroids, owing to their structural diversity and promising druggability [167]. Cheng [168] identified a structurally novel 5/6/5 tricyclic zwitterionic compound, aspongopyrimidine A (22), from the medicinal insect Aspongopus chinensis (Figure 12). Its absolute configuration was elucidated through a combination of spectroscopic analysis, computational chemistry, and X‐ray diffraction.

FIGURE 12.

MAPRE3 modulator based on tricyclic scaffolds.

Chemical proteomics analysis revealed that compound 22 specifically binds to the microtubule‐associated protein MAPRE3 (also known as EB3) in HEK‐293T cells, suggesting potential anti‐AD activity. Biological evaluation confirmed that 22 effectively inhibits Aβ ₄₂ deposition in N2aSW cells and enhances glutathione peroxidase 4 (GPX4)‐mediated cellular antioxidant activity. To investigate the functional role of MAPRE3 in AD pathogenesis, the researchers employed siRNA to knock down MAPRE3 expression in N2aSW cells. MAPRE3 silencing not only significantly reduced Aβ plaque formation but also attenuated the anti‐Aβ effect of 22, indicating that its pharmacological activity is MAPRE3‐dependent. Further investigation revealed that 22 does not affect MAPRE3 protein expression levels but instead exerts its anti‐Aβ effect through functional modulation of the protein. These findings position 22 as an anti‐AD lead compound that acts via targeting and modulating MAPRE3, opening new avenues for the study of insect‐derived natural products.

4.1.10. Autophagy Activators

Ganoderma, a traditional Chinese medicinal fungus, has attracted considerable attention in pharmacological and phytochemical research due to its abundance of structurally unique triterpenoids with broad biological activities. Ganoderma triterpenoids (GTs), the primary bioactive constituents, exhibit significant potential in antitumor, antiaging, hepatoprotective, and neuroprotective effects [169]. Studies have shown that GTs alleviate lipopolysaccharide (LPS)‐induced neuroinflammation and apoptosis, promote Aβ clearance, and inhibit tau pathology through activation of the autophagy pathway, suggesting promising applications in AD therapy [170].

Peng [171] isolated ten structurally novel A‐ring‐seco‐nortriterpenoids from Ganoderma cochlear, encompassing six distinct skeleton types. In SH‐SY5Y cells stably expressing the human MAPT‐p.P301S mutant, intervention with the ATPase inhibitor bafilomycin A1 (BAFA1) reversed the tau P301S reduction induced by compounds 23a‐d (Figure 13). Although BAFA1 alone increased the LC3‐II/LC3‐I ratio, it did not significantly alter tau P301S protein levels, indicating that 23a‐d mediate the clearance of pathological tau through activation of the autophagy pathway. Further mechanistic studies revealed that 23a‐d did not affect the expression or phosphorylation levels of key proteins in the mTOR signaling pathway but significantly upregulated the phosphorylation levels of AMPKα, AMPKβ1, and ULK1 at multiple sites (Ser777, Ser317, Ser555) in SH‐SY5Y MAPT cells, suggesting that they induce autophagy via activation of the AMPK‐ULK1 pathway.

FIGURE 13.

Autophagy activators based on tricyclic scaffolds.

4.1.11. Neuroprotective/Neuroregenerative Agents

The challenges encountered in AD drug development have prompted researchers to reconsider the therapeutic relevance of targets beyond classical neurodegenerative hallmarks. Alongside typical “positive” pathological features such as Aβ plaques and NFTs, widespread neuronal and synaptic loss, regarded as “negative” pathological markers, has gained increasing attention [172]. The marked decline in hippocampal neuronal density due to cell death serves as a key indicator of AD progression, suggesting that reparative strategies aimed at activating endogenous neuroprotective pathways and sustaining neuroregeneration may offer a viable approach to counter neuronal loss [173, 174].

Bolognesi [175] constructed a focused chemical library of 36 compounds based on the known neuroprotective and neuroregenerative properties of tricyclic alkylamine antipsychotics. It aimed to identify molecules capable of crossing the BBB and eliciting neuroprotective and neuroregenerative phenotypes in diseased neurons. Using a primary neuron phenotypic screening platform, they identified the phenothiazine derivative 24a as displaying superior overall activity across multiple assays (Figure 14). Safety assessments confirmed good biocompatibility of 24a in primary rat cerebellar granule neurons (CGNs) and HepG2 hepatocytes at concentrations of 1–50 µM. In a low‐serum/low‐potassium‐induced apoptosis model of CGNs, 24a exhibited a clear dose‐dependent neuroprotective effect at 1–5 µM. Neurosphere assays in mouse neural stem cells further revealed that 24a (1 µM) increased neurosphere size by 40%, while at 5 µM it boosted neurosphere number to 157% of the control value. In CGN toxicity models induced by Aβ and 6‐hydroxydopamine (6‐OHDA), pretreatment with 24a (5 µM) increased cell viability from 80% to 129% and 120%, respectively, demonstrating superior protective effects relative to the positive controls galantamine and apomorphine. Moreover, in an LPS‐induced microglial neuroinflammation model, 24a effectively modulated microglial phenotypic polarization by dose‐dependently downregulating iNOS expression and maintaining TREM2 levels, thereby exerting indirect neuroprotective effects. These findings underscore the potential of tricyclic alkylamine scaffolds in neuroprotection and neuroregeneration, offering new avenues for AD drug discovery.

FIGURE 14.

Neuroprotective/neuroregenerative agents based on tricyclic scaffolds.

4.1.12. PDE Inhibitor

Phosphodiesterases (PDEs) constitute a superfamily of enzymes comprising 11 subfamilies (PDE1−11) responsible for the hydrolysis of cyclic guanosine monophosphate (cGMP) and cyclic adenosine monophosphate (cAMP). This enzyme superfamily has been extensively investigated as therapeutic targets for erectile dysfunction, pulmonary arterial hypertension, cancer, and schizophrenia [176, 177]. Among them, PDE5 inhibitors such as sildenafil and tadalafil, well established as vasodilators, have garnered increasing attention in AD therapy due to their multifaceted neuroprotective potential [178]. Early studies by Landry [179] elucidated the role of the NO/cGMP/PKG/CREB signaling pathway in AD pathology and demonstrated that inhibition of PDE5, which specifically catalyzes cGMP hydrolysis, significantly improves learning and memory in AD mouse models. Building on these findings, they [180] employed a cyclization strategy to structurally modify lead compound 25a, aiming to enhance potency and selectivity for PDE5 through conformational constraint. This led to the design of a series of novel 1,2,3,4‐tetrahydrobenzo[b][1,6]naphthyridine‐based tricyclic inhibitors (25b‐h) (Figure 15). Biological evaluation revealed that these compounds exhibited potent PDE5 inhibitory activity. Compound 25b demonstrated IC₅₀ values of 0.056 and 30.1 nM against PDE5A1 and PDE6C, representing significant improvements in both enzyme inhibitory potency and aqueous solubility (>500 μM) compared with 25a (PDE5A1: IC₅₀ = 0.27 nM; PDE6C: IC₅₀ = 339 nM). Foot‐shock experiments showed that 25b at doses of 3 mg/kg and 10 mg/kg increased hippocampal cGMP levels by 1.5 and 1.7‐fold, respectively. Pharmacokinetic studies revealed that following intraperitoneal injection, the plasma concentration of 25b peaked at 0.25 h, with a half‐life (T_1/2) of 1.07 h. Drug concentrations in brain tissue and plasma were 492.25 ng/g and 1571.44 ng/mL, respectively (logBB = −0.50), confirming favorable BBB penetration. In APP/PS1 AD mouse models, 25b significantly enhanced synaptic plasticity, restoring LTP from 164.11% to 226.22%. In behavioral assessments using fear conditioning and radial arm water maze tests to evaluate associative and spatial memory, respectively, 25b‐treated mice exhibited a recovery of contextual freezing behavior to approximately 30% and a significant reduction in spatial memory errors, indicating marked cognitive improvement that persisted even after drug withdrawal.

FIGURE 15.

PDE inhibitors based on tricyclic scaffolds.

PDE9 is a subfamily that hydrolyzes only cGMP and exhibits the highest binding affinity, with K _m = 70 nM and a single isoform, PDE9A [181]. PDE9A has emerged as a promising therapeutic target for both diabetes and AD. However, most of the reported PDE9A inhibitors share the same pyrazolopyrimidinone scaffold, which limits structural diversity and hinders the development of novel inhibitors [182]. To address this, Luo [183] employed a combinatorial strategy integrating pharmacophore modeling, molecular docking, molecular dynamics simulations, and bioassay to identify novel PDE9A inhibitors (26a) with a tricyclic scaffold from the SPECS database (Figure 15). Binding pattern analyses indicate that it can bind to the same active site pocket of PDE9A as classical PDE9A inhibitors. In addition, structural modification of 26a (IC₅₀ = 8.0 μM) led to 26b, with an improved inhibitory activity of 2.1 μM, as expected. The novel scaffold discovered in the present study can be used for the rational design of PDE9A inhibitors with higher affinities.

4.1.13. β‐Carboline Based Compounds

Carbolines are natural alkaloids characterized by a pyridoindole tricyclic core and exist in four isomeric forms: α, β, γ, and δ. Among these, β‐carbolines have garnered significant attention due to their diverse biological activities, including sedative, anxiolytic, hypnotic, anticonvulsant, antitumor, antiviral, antiparasitic, and antibacterial effects [184, 185]. Owing to their remarkable bioactivity, β‐carbolines have emerged as a privileged scaffold in medicinal chemistry, leading to the development of several marketed drugs, such as reserpine [186]. Recently, with the growing adoption of multi‐target strategies for treating neurodegenerative diseases, β‐carbolines have demonstrated unique advantages in AD lead discovery, attributed to their inherent affinity for multiple neuropharmacological targets, including AChE, monoamine oxidase, dual‐specificity tyrosine phosphorylation‐regulated kinase 1A (DYRK1A), 5‐HT receptor, and NMDA receptor [187, 188].

In the context of AD pathogenesis, hyperphosphorylation of tau protein induces conformational changes that lead to the formation of NFTs. These highly deleterious aggregates accumulate in the brain, ultimately resulting in neuronal dysfunction. To intervene in this process, research efforts have increasingly focused on developing inhibitors of tau‐associated kinases, with key targets including DYRK1A, CDK5, GSK‐3β, CK1/2, and CDC‐like kinase (CLK1) [189]. Regarding DYRK1A inhibition, harmine (27) exhibits potent activity (IC₅₀ = 80 nM), whereas its analogs norharmane (28) and harmane (29) show weaker inhibition (IC₅₀ > 10 μM) (Figure 16). SAR studies indicate that the N−9 position can accommodate substituents of varying sizes provided that a hydroxyl or methoxy group is retained at the 7‐position; compounds 30 and 31a‐c displayed excellent inhibitory activity (IC₅₀ = 35–59 nM) [190]. Crystal structure analysis of the β‐carboline‐DYRK1A complex revealed that the planar tricyclic core of harmine inserts into the hydrophobic cavity of the enzyme active site, forming key hydrogen bonds with the main chain of Leu241 and the side chain of Lys188, with additional flexible space around the N−9 atom capable of accommodating bulky groups.

FIGURE 16.

DYRK1A and CLK1 inhibitors based on tricyclic scaffolds.

In the context of CLK1 inhibition, a recently identified spirotricyclic compound 32 demonstrated excellent activity (K _i  = 0.14 μM) and high kinase selectivity, with IC₅₀ exceeding 10 μM against 50 kinases [191] (Figure 16). Crystal structure analysis revealed that its spiropiperidine moiety and cyano group engage in multiple hydrogen bonds with residues such as Lys191 and Asp325, while the 7‐Cl atom participates in a halogen bond with Glu242, collectively contributing to the unique binding mode of this class of compounds.

Histone deacetylases (HDACs) constitute a family of enzymes that catalyze the removal of acetyl groups from lysine residues on both histone and nonhistone proteins, playing essential roles in epigenetic regulation and diverse cellular processes [192, 193]. The classical zinc‐dependent HDACs consist of 11 isoforms categorized into four subclasses: Class I (HDAC1, 2, 3, 8), Class IIa (HDAC4, 5, 7, 9), Class IIb (HDAC6, 10), and Class IV (HDAC11). HDAC6 is predominantly cytoplasmic and regulates the acetylation status of multiple AD‐related cytoplasmic proteins, including heat shock protein 90, α‐tubulin, tau, GSK‐3β, and β‐catenin [194]. Overexpression of HDAC6 in AD brains has been shown to disrupt microtubule stability and impair axonal transport, synaptogenesis, and neuronal trafficking, leading to synaptic structural and functional degeneration. Inhibition of HDAC6 not only enhances axonal transport and reduces tau phosphorylation but also promotes neurite outgrowth and protects neurons from oxidative stress‐induced damage [195].

Tetrahydro‐β‐carboline (THβC) has been widely utilized in AD drug design to improve brain penetration due to its structural resemblance to endogenous neurotransmitters such as serotonin and melatonin [196, 197]. In previous work, Gao [198] identified a series of 1,3‐disubstituted THβC derivatives (33) as potent and selective HDAC6 inhibitors (Figure 17). To further optimize BBB permeability, they simplified the architecture by removing the polar amide bond, leading to the design of novel THβC‐based compounds [199]. Biological evaluation revealed that most N‐2‐substituted THβC derivatives exhibited potent HDAC6 inhibitory activity (IC₅₀ < 20 nM). Among them, compound 34 exhibited an IC₅₀ of 15.2 nM with optimal selectivity: it showed no inhibition against HDAC4, 5, 7, 9, or 11, and weak inhibition against HDAC1, 2, 3, 8, 10. In PC12 cell models, 34 significantly promoted the accumulation of acetylated α‐tubulin in neuronal processes, enhanced neuronal excitability and synaptic transmission, and upregulated the expression of neurogenesis markers like growth‐associated protein 43 (GAP‐43) and β3‐tubulin. In H₂O₂‐induced oxidative stress models, 34 dose‐dependently increased cell viability from 56.8% and 33.7% to 89.7% and 78.0%, and inhibited H₂O₂‐induced apoptosis by upregulating the anti‐apoptotic factor Bcl‐2 and downregulating the pro‐apoptotic factor PARP. Further mechanistic studies demonstrated that 34 reduced lactate dehydrogenase (LDH) release in PC12 oxidative stress models, restored SOD activity, and activated the Nrf2/catalase antioxidant pathway. In the zebrafish novel tank diving test, 34 (1 mg/L) significantly increased swimming distance and speed, alleviating scopolamine‐induced anxiety‐like behavior. T‐maze test further confirmed that 34 enhanced exploratory activity in the enriched chamber and effectively ameliorated scopolamine‐induced memory deficits in zebrafish. These findings provide a solid experimental foundation for the development of tetrahydro‐β‐carboline‐based therapeutics for AD.

FIGURE 17.

HDAC inhibitors based on tricyclic scaffolds.

4.2. Multi‐Target Tricyclic Compounds

4.2.1. β‐Carboline‐Based GSK3β/AChE Inhibitors

In a previous study, Zhao [200] identified a class of GSK‐3β/DYRK1A dual inhibitors based on the β‐carboline scaffold (35), revealing that the cyclopropylcarboxamide moiety at the 2‐position of the pyridine ring serves as a key pharmacophore (Figure 18). Subsequently, the researchers retained the pharmacophore while incorporating N‐benzylpiperidine units derived from donepezil at the 7‐position via linkers of varying lengths, leading to the design of novel GSK‐3β/AChE dual‐target inhibitors [201]. Biological evaluation showed that these compounds exhibited moderate to good AChE inhibitory activity (IC₅₀ = 0.27–33.48 µM) with high selectivity over BuChE. Among them, compound 36 displayed the most potent AChE inhibition (IC₅₀ = 0.27 µM, SI = 77.11). Additionally, all derivatives demonstrated moderate to good GSK‐3β inhibitory activity at 10 µM, with 36 achieving an IC₅₀ of 6.78 µM. Cytotoxicity assays confirmed that 36 exhibited very low toxicity across 1–20 µM in SH‐SY5Y, HEK‐293T, HL‐7702, and HepG2 cells. Acute toxicity studies further supported its favorable safety profile at 1500 mg/kg. In a HEK‐293T cell model expressing the P301L mutation tau, 36 dose‐dependently reduced tau phosphorylation at Ser396 while concurrently increasing GSK‐3β phosphorylation at Ser9, resulting in a reduced p‐tau/total tau ratio and inhibition of NFT formation.

FIGURE 18.

GSK3β and AChE dual‐target inhibitors based on tricyclic scaffolds.

4.2.2. NLRP3 Inflammasome and ChE Inhibition

ChE and β‐/γ‐secretase have long been central targets in AD drug development. However, single‐pathway intervention strategies against these targets have failed to halt disease progression, prompting a gradual shift in research focus toward upstream pathological mechanisms, particularly neuroinflammation. As a core regulatory component of the innate immune system, the NLRP3 inflammasome has emerged as a highly promising therapeutic target in recent years [202]. Pathological stimuli such as AβO or fibrillar plaques activate the NLRP3 inflammasome, leading to the release of IL‐1β and IL‐18, which directly exacerbate neuronal damage and synaptic dysfunction. Additionally, NLRP3 activation promotes tau hyperphosphorylation, thereby linking the two core pathological events of AD [203, 204].

Against this backdrop, Bharate [205] employed a dual‐target drug design strategy by hybridizing colchicine (37), which exhibits NF‐κB pathway inhibition and NLRP3 inflammatory modulatory activity, with donepezil, leading to the design of colchicine‐aryl/alkylamine tricyclic compounds (Figure 19). In vitro results showed that compound 38 exhibited potent inhibitory activity against all EeAChE (IC₅₀ = 0.372 μM), HuAChE (IC₅₀ = 0.313 μM), and eqBChE (IC₅₀ = 0.297 μM), indicating that 38 acts as a dual‐target inhibitor. Enzyme kinetic studies revealed a noncompetitive inhibition pattern for both ChEs, with inhibition constants (K _i ) of 0.39 μM (HuAChE) and 0.28 μM (eqBChE). In vitro experiments confirmed that 38 exhibited good BBB permeability (P _e = 12.21 × 10⁻⁶ cm/s) and lipophilicity (log P = 4.05), with no significant cytotoxicity observed in neuronal cells. In a J774A.1 cell inflammation model, 38 effectively inhibited NLRP3 inflammasome activation at 0.1–10 μM, significantly reducing IL‐1β release induced by LPS combined with nigericin (EC₅₀ = 1 μM), demonstrating superior efficacy compared to donepezil. These results indicate that 38 holds promising potential as a next‐generation anti‐AD drug candidate.

FIGURE 19.

NLRP3 inflammasome and ChE dual‐target inhibitors based on tricyclic scaffolds.

4.2.3. MAO‐B and AChE Inhibition

Given the complex multifactorial nature of AD pathology, Rashid [206] employed a multi‐target drug design strategy based on an 8‐substituted‐3,4‐dihydropyrimidine‐2‐thione template to construct a tricyclic scaffold. By introducing an arylidene moiety with a rigid double bond at the 2‐position, and incorporating an indole ring at the 4‐position of the pyrimidine ring to enhance interactions with key residues (Trp86, Trp286) of HuAChE, they designed a series of 2‐arylidenethiazolopyrimidine derivatives. Activity studies revealed that compounds 39a‐d and 40a‐d exhibited AChE inhibitory activity at micromolar to submicromolar levels (Figure 20). Notably, 40a‐d remained highly active at nanomolar concentrations, with the most potent 40b‐d showing IC₅₀ of 0.04, 0.08, and 0.07 μM, respectively. All compounds displayed moderate to good BChE inhibitory activity, as well as excellent MAO‐B inhibitory activity with high selectivity. Compounds 39a, 40c, and 40d were the most potent MAO‐B inhibitors, with IC₅₀ of 0.13, 0.10, and 0.14 μM, respectively. Pharmacokinetic evaluation demonstrated that 40b exhibited no significant cytotoxicity against HEK‐293 or SH‐SY5Y cells, showed no acute toxicity at doses up to 2000 mg/kg, and possessed favorable BBB permeability. Collectively, this series of compounds combines multi‐target inhibitory activity, favorable safety profiles, and potential for brain distribution, positioning them as promising lead structures for AD therapy.

FIGURE 20.

MAO‐B and AChE dual‐target inhibitors based on tricyclic scaffolds.

4.2.4. Antioxidant and AChE Inhibition

Among the various pathogenic hypotheses of AD, oxidative stress induced by ROS has been established as a key driver of AD progression. Oxidative stress not only directly damages neuronal structure but also accelerates Aβ deposition and tau hyperphosphorylation, perpetuating a vicious pathological cycle [207, 208]. Consequently, combining antioxidants with AChEIs to enhance cholinergic neurotransmission is now considered a promising therapeutic strategy for AD [209, 210].

Over the past decade, the discovery of anti‐AD lead compounds from natural products has garnered increasing interest [211, 212]. Euphorbia‐derived diterpenoids have attracted considerable attention due to their ability to activate endogenous antioxidant pathways such as Nrf2/ARE, thereby enhancing cellular resistance to oxidative stress [213]. Notably, myrsinane‐type Euphorbia diterpenoids, characterized by a 5/7/6‐fused tricyclic scaffold, have demonstrated clear neuroprotective activity in various neuronal injury models [214]. Gao [215] employed a biomimetic scaffold transformation strategy, using a naturally abundant Euphorbia diterpenoid (41) as a starting material to construct a series of structurally novel myrsinane‐type derivatives (Figure 21). Biological evaluation revealed that most compounds exhibited moderate AChE inhibitory activity (IC₅₀ = 8.3–36.5 μM), with compound 42a showing the most potent inhibition (IC₅₀ = 8.3 μM). Cytotoxicity assays indicated that all compounds were well tolerated at 12.5 μM, with cell viability exceeding 90%. In an H₂O₂‐induced SH‐SY5Y cell injury model, 42a demonstrated remarkable neuroprotective effects, increasing cell viability from 52.1% to 108.4%. Further mechanistic studies revealed that pretreatment with 42a effectively reduced H₂O₂‐induced intracellular ROS levels and inhibited the mitochondrial apoptosis pathway by upregulating Bcl‐2 while downregulating Bax and caspase‐3 expression.

FIGURE 21.

Antioxidant and AChE inhibitors based on tricyclic scaffolds.

Natural and synthetic coumarins have garnered significant attention due to their broad spectrum of biological activities. Recently, researchers have focused on developing coumarin‐based multifunctional molecules that combine AChE inhibition, antioxidant and anti‐Aβ aggregation activities [216, 217]. Amooru [218] designed a series of novel fused tricyclic coumarin amide derivatives as potential multifunctional anti‐AD candidates. Using 7‐hydroxy‐4‐methylcoumarin as the core scaffold, they introduced an angularly fused iminopyran ring at the 7‐ and 8‐positions and incorporated an N‐substituted amide moiety at the 9‐position (Figure 22). ChE inhibitory assays revealed that these compounds exhibited varying degrees of AChE inhibitory activity (IC₅₀ = 0.003–2.6 μM) with high selectivity (SI = 9.54–5793). Compound 43n displayed the most potent inhibitory activity (IC₅₀ = 0.003 μM), outperforming galantamine. In terms of antioxidant activity, most compounds exhibited potent ABTS radical scavenging activity (IC₅₀ = 7.98–34.16 μM), with half of the compounds (IC₅₀ = 7.98–15.99 μM) outperforming Trolox (IC₅₀ = 27.35 μM). In an H₂O₂‐induced oxidative injury model using SK‐N‐SH cells, compounds 43m‐q dose‐dependently alleviated cell damage and reduced apoptosis. In vivo experiments confirmed that 43n (at 16–200 mg/kg) effectively penetrated the BBB and dose‐dependently inhibited AChE activity in the mouse brain.

FIGURE 22.

Antioxidant and AChE inhibitors based on coumarin scaffolds.

4.2.5. H₃ Receptor Antagonism and AChE Inhibition

Histamine H₃ receptor, a member of the G protein‐coupled receptor (GPCR) family, was initially characterized as a presynaptic autoreceptor that negatively regulates histamine release in the brain [219]. Subsequent studies have revealed its additional role as a heteroreceptor, modulating the release of key neurotransmitters such as acetylcholine, glutamate, and serotonin [220]. Consequently, antagonism of central H3 receptors to enhance the release of pro‐cognitive neurotransmitters has emerged as a promising therapeutic strategy for AD. Recently, research efforts have increasingly focused on developing dual‐acting compounds that combine AChE inhibition with H₃ receptor antagonism. Decker [221] designed a series of structurally diverse AChE/H₃ dual‐target inhibitors (44A‐C) based on a tricyclic nitrogen‐bridged scaffold, which was connected via an ether linkage to tertiary amine moieties (Figure 23). Biological evaluation revealed that most compounds exhibited potent HuAChE inhibitory activity (IC₅₀ = 33.9 nM to 9.42 μM) and moderate eqBChE inhibitory activity (IC₅₀ = 1.0–122.3 μM). Compound 44D‐1 displayed the most potent inhibition (HuAChE: IC₅₀ = 33.9 nM; eqBChE: IC₅₀ = 8.2 μM). Moreover, these compounds showed high affinity for the human H₃ receptor, with 44D‐1 demonstrating a K _i of 76.2 nM, reflecting a favorable balance of dual‐target potency.

FIGURE 23.

H₃ receptor antagonism and AChE inhibitors based on tricyclic scaffolds.

SAR analysis indicated that linking the tricyclic quinazolinone core to a cyclic amine via a suitable linker significantly enhanced HuAChE inhibitory activity into the nanomolar range. Complete reduction of the quinazolinone skeleton to a diamine structure substantially increased H₃ receptor affinity while preserving AChE inhibition. In contrast, selective reduction of the carbonyl group yielded compounds that achieved an optimal balance between AChE inhibition and H₃ antagonism, thereby exhibiting ideal dual‐target synergistic properties. Enzyme kinetic studies confirmed that 44D‐1 acts as a reversible competitive inhibitor of AChE, while steady‐state GTPase assays demonstrated that most compounds function as highly selective H₃ receptor ligands with antagonist or inverse agonist activity. Collectively, these findings support the potential of this class of compounds as dual‐functional agents for AD therapy.

4.2.6. BACE‐1 and AChE Inhibition

Ma [222] employed an MTDL strategy to fuse the natural alkaloid deoxyvasicine with donepezil and VK‐28, leading to a series of deoxyvasicine‐donepezil hybrids, aiming to develop novel multi‐target anti‐AD candidates capable of concurrently inhibiting AChE, BACE1, and Aβ aggregation. Biological evaluation revealed that most compounds exhibited moderate to potent inhibitory activities against HuAChE, BACE1, and Aβ _1–42 aggregation. Among them, compounds 45i (IC₅₀ = 56.14 nM), 45l (IC₅₀ = 5.91 nM), 45m (IC₅₀ = 3.2 nM), and 45v (IC₅₀ = 8.65 nM) demonstrated outstanding performance against HuAChE. For BACE1, the IC₅₀ values were 0.834, 0.167, 0.129, and 0.085 μM; and against Aβ aggregation, the IC₅₀ values were 13.26, 19.43, 9.26, and 5.41 μM, respectively (Figure 24). Further studies indicated that 45i, 45m, and 45v effectively inhibited AChE‐induced Aβ _1–42 aggregation and exhibited low cytotoxicity toward SH‐SY5Y cells. Notably, 45i and 45m also demonstrated significant neuroprotective activity against Aβ‐induced injury. Collectively, 45i and 45m combine excellent multi‐target inhibitory activity, low toxicity, favorable neuroprotective effects, and promising pharmacokinetic properties, positioning them as highly valuable multi‐target drug candidates for AD therapy.

FIGURE 24.

BACE‐1 and AChE inhibitors based on tricyclic scaffolds.

4.2.7. Tacrine‐Based Multi‐Target Compounds

Tacrine, the first FDA‐approved AChEI for AD, features a tricyclic aminoquinoline scaffold that not only serves as a classic pharmacophore for ChE inhibition, but also represents an ideal lead structure for multi‐target anti‐AD drug design [223, 224]. The tacrine scaffold has been extended beyond conventional AChE inhibition to encompass a broad range of pharmacological activities, including inhibition of Aβ aggregation, modulation of BACE1, MAO‐B inhibition, NMDA receptor antagonism, and antioxidant effects [225, 226, 227, 228]. This expanded utility has significantly enhanced its neuroprotective and disease‐modifying potential in both cellular and animal models. Previously, Torrero [229] developed the potent antioxidant 7‐methoxy‐2,2‐dimethylchroman‐6‐ol (46), which exhibited strong antioxidant activity in vitro and demonstrated efficacy in reducing oxidative damage in a rat cerebral ischemia/reperfusion model. However, its poor BBB permeability resulted in suboptimal brain concentrations, limiting its neuroprotective efficacy. To address this limitation, the researchers hybridized this scaffold with lipophilic 6‐chlorotacrine (Cl‐THA) to generate novel compounds with enhanced brain penetration and multi‐target activity (Figure 25). HuAChE inhibition assays revealed that all compounds 47a‐s exhibited superior inhibitory activity compared to Cl‐THA (IC₅₀ = 14.5 nM). 47b showed an IC₅₀ of 3.69 nM for AChE and 170 nM for BChE. DPPH radical scavenging assays demonstrated that compounds bearing a free hydroxyl group retained the antioxidant activity of Cl‐THA (IC₅₀ = 17.4 μM). Several O‐benzylated derivatives (47l‐o) also exhibited moderate BACE1 inhibitory activity (IC₅₀ = 7–18 μM). In an E. coli cell model, long‐chain amines (47n‐o) and O‐benzylated amides (47q‐r) showed 30%–50% inhibition of Aβ and tau aggregation at 10 μM. PAMPA‐BBB assays further confirmed that these compounds displayed significantly improved BBB permeability (P _e = 7.0–12.5 × 10⁻⁶ cm/s) compared to Cl‐THA, validating the design strategy of enhancing brain penetration through scaffold hybridization. In APP/PS1 AD mouse models, 47b and 47f demonstrated potential for improving cognitive function, modulating amyloid pathology, and alleviating oxidative stress. Although most endpoints did not reach statistical significance, significant modulation was observed in BACE1 gene expression, sAPPβ levels, and key signaling pathways such as GPX1, Nrf2, and Hmox1.

FIGURE 25.

Antioxidant and AChE inhibitors based on tacrine scaffolds.

Korabecny [230] employed a multi‐target directed ligand strategy to hybridize tacrine derivatives with the phenothiazine (PHT) scaffold, which possesses anti‐Aβ aggregation and antioxidant activities, leading to the design of novel tacrine‐PHT hybrids (Figure 26). Biological evaluation revealed that these compounds exhibited potent inhibitory activity against HuAChE and HuBChE. Compounds 48‐1dA, 48‐1dC and 48‐2dC showed the most potent HuAChE inhibitory activity, while 48‐1aA, 48‐1aC, 48‐1dA and 48‐2dA exhibited the strongest HuBChE inhibition. Within this series, 48‐1dC was the most potent HuAChE inhibitor (IC₅₀ = 8 nM, SI = 24), and 48‐1aA was the most potent HuBChE inhibitor (IC₅₀ = 19 nM, SI = 100). Thioflavin T fluorescence assays revealed that 48‐1aC, 48‐1bC, 48‐1cC, 48‐1dC, and 48‐2dC effectively inhibited tau fibril formation at 50 μM (50.5–62.1%), outperforming tacrine and PHT. Moreover, these compounds not only reduced the total amount of fibrils but also delayed the aggregation lag phase by approximately 1.5–2 h. In Aβ aggregation inhibition assays, 48‐1dC and 48‐2dC exhibited over 70% inhibition of Aβ self‐aggregation. Pharmacokinetic studies demonstrated that 48‐1dC was able to cross the BBB and enter the CNS, with its maximum brain concentration approaching its own IC₅₀ for HuAChE inhibition, and exhibited a good pharmacokinetic profile.

FIGURE 26.

Antioxidant, anti‐Aβ and AChE inhibitors based on tacrine scaffolds.

In neuroinflammation‐related metabolic pathways, the arachidonic acid metabolites epoxyeicosatrienoic acids (EETs) exhibit potent anti‐inflammatory and antioxidant activities. However, their physiological functions are limited by rapid hydrolysis into less active diol derivatives via soluble epoxide hydrolase (sEH) in vivo [231]. Preclinical studies have shown that sEH inhibitors effectively improve cognitive function and reduce neuroinflammation, tau hyperphosphorylation, and Aβ burden in AD mouse models. The marked upregulation of sEH expression in AD brains further supports its validity as a therapeutic target for AD intervention [232, 233]. Conversely, AChE inhibitors may positively modulate arachidonic acid metabolism through activation of muscarinic M1 receptors, thereby promoting endogenous EET production. This interplay suggests a potential synergistic effect between AChE and sEH inhibition in the EET‐mediated anti‐neuroinflammatory pathway [234].

Torrero [235] designed novel sEH/AChE dual‐functional agents by hybridizing Cl‐THA/huprine Y with the sEH inhibitor TPPU, aiming to simultaneously achieve neuroinflammatory modulation and cholinergic enhancement (Figure 27). In vitro activity assays revealed that these molecules retained the high potency of TPPU, exhibiting strong inhibitory activity against human sEH (hsEH) with IC₅₀ ranging from 0.4–4.6 nM. Moreover, 49a‐f demonstrated significant HuAChE inhibitory activity and also displayed submicromolar potency against HuBChE. In vivo experiments further confirmed that all compounds exhibited excellent inhibitory activity against mouse sEH (msEH; IC₅₀ = 12–34 nM) and mouse AChE (mAChE; IC₅₀ = 2.61–35.4 nM). Druggability evaluation showed that these compounds possessed excellent BBB permeability, moderate aqueous solubility (5.6–42.6 μM), and favorable biosafety profiles. Stability assays indicated that 49c exhibited low clearance in human, mouse, and rat liver microsomes. In vivo pharmacodynamic evaluation demonstrated that in the novel object recognition test (NORT), 49c (2 mg/kg) significantly ameliorated both short‐term and long‐term memory deficits in SAMP8 mice and downregulated the expression of NF‐κB, IL‐6, IL‐1β, and glial fibrillary acidic protein (GFAP). Furthermore, 49c significantly increased hippocampal levels of synaptophysin (SYN) and soluble APPα (sAPPα) while reducing sAPPβ levels and tau phosphorylation at Ser396, underscoring its potential for synergistic modulation of multiple pathological pathways in AD.

FIGURE 27.

sEH and AChE inhibitors based on tacrine scaffolds.

4.2.8. 5‐HT₆/5‐HT₃ Receptor and MAO‐B Triple‐Target Inhibition

Blockade of serotonin 5‐HT₆ receptors (5‐HT₆R) on GABAergic neurons has been shown to promote ACh and glutamate release in the corticolimbic region, thereby enhancing cognitive function [236]. Clinical studies have demonstrated that the 5‐HT₃ receptor (5‐HT₃R) antagonist ondansetron exerts pro‐cognitive effects and exhibits significant neuroprotective activity in both in vitro and in vivo models [237]. Meanwhile, the reversible MAO‐B inhibitor lazabemide has shown promising efficacy in Phase III clinical trials for AD, with mechanisms potentially involving suppression of astrocytic hyperplasia and enhancement of synaptic transmission, which play a critical role in modulating neurodegenerative and neuroinflammatory processes [238, 239].

Building on these mechanistic insights, Zajdel [240] employed a fused ligand strategy to identify 1‐(3‐chlorobenzyl)‐4‐(piperazin‐1‐yl)‐1H‐pyrrolo[3,2‐c]quinoline (PZ‐1922, 50) (Figure 28). This compound exhibits triple activity as a 5‐HT₆R (K _i  = 17 nM) and 5‐HT₃R (K _i  = 0.45 nM) antagonist and a reversible MAO‐B inhibitor (pIC₅₀ = 8.93). In behavioral studies, 50 dose‐dependently ameliorated scopolamine‐induced short‐term memory deficits in mice in the NORT, demonstrating clear pro‐cognitive effects. In an Aβ‐induced rat memory deficit model, 50 exhibited superior cognitive improvement compared to the selective 5‐HT6R antagonist intepirdine, suggesting that its multi‐targeted action produces synergistic benefits in alleviating Aβ‐induced memory deficits. At the molecular level, 50 reversed the downregulation of postsynaptic density protein 95 (PSD‐95) and synaptophysin (SYN), thereby improving synaptic structural integrity, while concurrently reducing caspase‐3 and pro‐caspase‐3 levels to inhibit neuronal apoptosis. It also effectively suppressed Cdk5 overexpression and reduced p25 protein levels and the p25/p35 ratio. Furthermore, 50 significantly inhibited the expression of Iba1, a marker of microglial activation induced by oligomeric Aβ _25–35, thereby effectively alleviating neuroinflammation. Collectively, 50 demonstrates superior neuroprotective and pro‐cognitive effects compared to single‐target agents across multiple levels, further validating the therapeutic potential of simultaneously targeting 5‐HT6R, 5‐HT3R, and MAO‐B in AD treatment.

FIGURE 28.

5‐HT₆/5‐HT₃ receptor and MAO‐B triple‐target inhibitors based on tricyclic scaffolds.

4.2.9. AChE, BACE1, and GSK3β Triple‐Target Inhibition

In a previous study, Zhao [241] identified notopterol, a natural product isolated from Notopterygium, as a dual BACE1/GSK3β inhibitor (BACE1: IC₅₀ = 20 µM; GSK‐3β: IC₅₀ = 20 µM). Building on these findings, they employed a structure‐based approach to design 48 notopterol derivatives featuring a furacoumarin scaffold, aiming to achieve balanced AChE/BACE1/GSK3β inhibitory activity (Figure 29). Fortunately, compound 51 exhibited effective inhibitory activity against AChE (IC₅₀ = 1.0 μM), BACE1 (IC₅₀ = 20 μM), and GSK3β (IC₅₀ = 15 μM). Furthermore, 51 demonstrated favorable BBB penetrability, suitable PK profile (T_1/2 = 20 h), acceptable bioavailability (F = 9.8%, 100 mg/kg), and oral safety. Notably, 51 ameliorated the impaired learning and memory in Aβ‐induced AD mice.

FIGURE 29.

AChE, BACE1, and GSK3β triple‐target inhibitors based on tricyclic scaffolds.

4.2.10. Antioxidant, Anti‐Aβ, and AChE Inhibition

Tin [242] designed a series of tricyclic derivatives modified at the N−10 and C−2 positions based on phenothiazine and its bioisostere phenoselenazine scaffolds, in the context of the close association between selenium and oxidative stress in AD. They conducted multi‐target activity evaluation and mechanistic studies focusing on cholinesterase inhibition, Aβ aggregation inhibition, and antioxidant activity (Figure 30). SAR revealed that the NH group at the N−10 position is critical for maintaining potent anti‐Aβ aggregation and excellent DPPH radical scavenging activity, with inhibition rates of 45%–62% at 25 μM and a maximum antioxidant scavenging rate of 92.1% at 50 μM, superior to that of ebselen. N−10 methoxyphenyl acyl‐modified derivatives (52j, 53j) exhibited balanced, nonselective dual AChE/BuChE inhibitory activities, with IC₅₀ in the range of 4.6–5.9 μM, and C‐2 chlorine substitution further enhanced BuChE inhibition. In addition, replacement of sulfur with selenium in the core scaffold generally improved both antioxidant and cholinesterase inhibitory activities. Molecular docking results showed that the tricyclic scaffold can embed into the catalytic site and anionic subsite of cholinesterase, forming stable interactions through π–π stacking, van der Waals forces, and sulfur/selenium–π interactions, with the methoxyphenyl side chain engaging the PAS to confer multi‐target effects. SH‐SY5Y neurocytotoxicity assays confirmed that most derivatives had good safety profiles, with only a few chlorine‐substituted compounds showing slightly higher toxicity.

FIGURE 30.

Antioxidant, anti‐Aβ, and AChE inhibition based on tricyclic scaffolds.

4.2.11. AChE, β‐Secretase, COX‐2, and LOX‐5 Multi‐Target Inhibition

Furanocoumarin derivatives (54) have been reported to exhibit diverse biological activities, including antitumor, antioxidant, and anti‐inflammatory effects [243]. The benzopyran‐2‐one lactone unit can engage in noncovalent interactions with residues, thereby enhancing the ChE inhibitory activity [244]. Pterostilbene (55), a natural bioactive molecule, has demonstrated broad modulatory effects in age‐related diseases and inhibits Aβ self‐aggregation [245]. Based on these properties, Mphahlele [246] designed a series of furanocoumarin‐pterostilbene hybrids through pharmacophore fusion and evaluated their multi‐target therapeutic potential for AD. Screening results revealed that these compounds exhibited moderate to good AChE (IC₅₀ = 1.8–21.6 μM) and BChE inhibitory activity (IC₅₀ = 3.5–17.6 μM). Compounds 56c (AChE: IC₅₀ = 1.8 μM; BChE: IC₅₀ = 3.5 μM) and 56e (AChE: IC₅₀ = 1.9 μM; BChE: IC₅₀ = 5.3 μM) displayed the most potent inhibitory effects (Figure 31). In β‐secretase inhibition assays, these compounds exhibited good activity (IC₅₀ = 9.6–30.3 μM), with 56c and 56e showing the strongest inhibition. In COX‐2 inhibition assays, all compounds showed activity (IC₅₀ = 8.6–21.6 μM) inferior to the positive control celecoxib (IC₅₀ = 0.007 μM); 56b and 56h exhibited the highest potency, with IC₅₀ of 9.7 μM and 8.6 μM. In LOX‐5 inhibition assays, the compounds displayed moderate inhibitory activity (IC₅₀ = 13.9–27.1 μM), with 56e and 56h being the most potent (IC₅₀ = 15.6 and 13.9 μM, respectively). DPPH radical scavenging assays indicated that these compounds possessed moderate to good antioxidant activity (IC₅₀ = 6.8–29.7 μM). 56h not only exhibited good inhibitory activity against AChE (IC₅₀ = 2.8 μM) and β‐secretase (IC₅₀ = 15.3 μM) but also displayed the strongest COX‐2 and LOX‐5 inhibitory activities, as well as the highest DPPH radical scavenging activity (IC₅₀ = 6.8 μM). In LPS‐induced oxidative stress models, 56c and 56h showed IC₅₀ of 2.1 μM and 1.4 μM against MCF‐7 cells, respectively. In LPS‐induced ROS inhibition assays, 56c and 56h (1 μM) achieved inhibition rates of 66% and 68% against ROS production in MCF‐7 cells. This class of furanocoumarin‐pterostilbene hybrids provides a valuable structural basis and design reference for the development of natural product‐based AD therapeutics.

FIGURE 31.

AChE, β‐secretase, COX‐2, and LOX‐5 triple‐target inhibitors based on tricyclic scaffolds.

4.2.12. Aβ, Tau, ChE, and MAO‐B Multi‐Target Inhibition

Natural and synthetic anthraquinones, owing to their conjugated tricyclic planar structure, have shown significant potential in targeting the complex pathological network of AD. This structural feature enables effective insertion into Aβ peptides and tau via π–π stacking and hydrophobic interactions. Moreover, the planar aromatic ring system serves as an ideal pharmacophore for key enzymes such as ChE and MAO‐B [247, 248, 249]. Based on this, Tonelli [250] employed a fragment‐based strategy to link the anthraquinone scaffold with aromatic (hetero)cyclic fragments, yielding novel tricyclic derivatives with multi‐target therapeutic potential for AD. Activity screening revealed that these compounds strongly inhibited Aβ ₄₀ self‐aggregation. Compounds 57a and 57b showed the most potent activity, with IC₅₀ values of 2.1 and 1.9 μM (Figure 32). This potency is primarily attributed to the strong hydrophobic interactions and π–π stacking between the anthraquinone scaffold and Aβ peptide chains, which block β‐sheet formation and extension. Further evaluation revealed that 57a possesses a broad pharmacological profile. It acts as a dual inhibitor of EeAChE (IC₅₀ = 7.3 μM) and eqBChE (IC₅₀ = 1.7 μM), while also inhibiting tau aggregation, with IC₅₀ of 1.78 and 0.36 μM against the PHF6 and R3 peptides, respectively. Moreover, it exhibited potent and selective MAO‐B inhibition (IC₅₀ = 0.57 μM). Atomic force microscopy (AFM) studies confirmed that 57a and 57b not only inhibit Aβ ₄₂ fibrillation but also disaggregate preformed fibrils. In an Aβ‐induced CGN injury model, both compounds alleviated neurotoxicity, increasing cell viability from 51% to 70%.

FIGURE 32.

Aβ, Tau, ChE and MAO‐B multi‐target inhibitors based on tricyclic scaffolds.

4.2.13. Aβ, AChE, Antioxidant, and Anti‐Neuroinflammation Multi‐Target Inhibition

Building upon previous studies, Kassab [251] hybridized the thienopyrimidine scaffold and bioisosteric thienotriazine, both known for their potent antioxidant activity, with an N‐benzylpiperidine moiety to develop 12 multi‐functional ChE inhibitors featuring a tricyclic scaffold. In vitro evaluation revealed that this series exhibited significant activity across multiple AD‐relevant targets. In AChE inhibition assays, compounds 58a‐b, 58d, 58f‐g, and 58j‐k demonstrated subnanomolar IC₅₀ values, comparable to donepezil (Figure 33). Among them, 58j displayed the most potent activity against both AChE (IC₅₀ = 0.124 nM) and BChE (IC₅₀ = 0.379 nM). In terms of Aβ aggregation inhibition, 58c, 58d, 58f, 58j, and 58k exhibited IC₅₀ of 7.12, 1.92, 6.07, 4.13, and 3.81 μM, respectively. Moreover, the series demonstrated significant antioxidant capacity, with 58e and 58b reducing ROS production by 59.0% and 57.2%. The ROS scavenging activity of 58e (78.93 pg/mL) was superior to the positive control malvidin (81.87 pg/mL), while 58b (82.35 pg/mL) showed comparable activity. Their excellent activity across multiple key pathological aspects highlights their promising potential as candidate therapeutics for AD.

FIGURE 33.

Aβ, AChE, antioxidant and anti‐neuroinflammation inhibitors based on tricyclic scaffolds.

4.2.14. AChE, Antioxidant, and Anti‐Neuroinflammation Multi‐Target Inhibition

To develop multi‐target therapeutics for AD, Ramsis [252] designed a series of novel tricyclic compounds based on the cyclohepta[b]thiophene scaffold. Activity screening revealed that most compounds exhibited good AChE inhibitory activity (IC₅₀ = 0.51–12.6 μM), with compounds 59a and 59d showing the most potent effects (IC₅₀ = 0.51 μM, 0.55 μM, respectively) (Figure 34). In the Morris water maze test using AlCl₃‐induced AD mice, 59a, 59e, and 59f significantly shortened escape latency by 32%, 22%, and 30%, respectively, comparable to donepezil (28%). 59b, 59c, and 59d demonstrated even greater improvement in learning ability, reducing escape latency by 47%, 43%, and 46%, respectively. In memory function assessment, 59a‐f significantly prolonged the time spent in the target quadrant, outperforming donepezil. Additionally, in the Y‐maze test, the percentage of spontaneous alternation (SAP%) was significantly increased across all treatment groups, with 59f showing efficacy comparable to donepezil (SAP = 85%). Assessment of cerebral inflammatory markers revealed that 59e and 59f reduced brain levels of IL‐1β (decreased to 46–54.5 pg/g) and TNF‐α (reduced by 70.8%–73.9%), and effectively inhibited Aβ deposition. 59e exhibited a total antioxidant capacity (TAC) of 37.13 μmol/g, restored SOD activity, and reduced malondialdehyde (MDA) levels to 18.35 nmol/g. Regarding neurotransmitter regulation, 59e effectively elevated brain levels of dopamine (DA), serotonin (5‐HT), norepinephrine (NE), and brain‐derived neurotrophic factor (BDNF).

FIGURE 34.

AChE, antioxidant and anti‐neuroinflammation multi‐target inhibitors based on tricyclic scaffolds.

5. Conclusion and Perspectives

AD is characterized by complex pathological heterogeneity, involving multiple interconnected mechanisms such as Aβ deposition, hyperphosphorylated tau protein, neuroinflammation, oxidative stress, epigenetic dysregulation, and impaired neuroprotection and regeneration. This multifaceted nature poses certain challenges for single‐target therapeutic strategies. In this context, tricyclic cores have emerged as a highly adaptable class of scaffolds in AD drug discovery. Their unique rigid skeletons, excellent structural plasticity, and diverse pharmacological activities enable them to transition from early single‐target inhibitors to advanced agents capable of precisely modulating multiple pathological pathways. Looking ahead, future drug design focusing on tricyclic scaffolds should prioritize several key directions: multi‐target synergy via pharmacophore fusion or fragment linking to address Aβ, tau, AChE, and neuroinflammation; improved BBB permeability through structural optimization; modulation of protein–protein interactions using the planarity and π–π stacking of tricyclic rings to block tau oligomerization or Aβ fibrillogenesis; PROTAC‐based degradation of pathological tau; and AI‐assisted scaffold hopping to discover novel asymmetric or fused tricyclic systems with better selectivity and pharmacokinetics. By integrating structural biology, computational chemistry, and AI‐driven design, tricyclic scaffolds are well positioned as next‐generation multi‐target anti‐AD agents to tackle the complex pathology of this disease.

Author Contributions

Yichun Shi: writing – original draft, investigation, formal analysis. Xindan Liao: writing – review and editing, supervision. Guangjun Yu: investigation, methodology, project administration, supervision, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Biographies

Yichun Shi is currently a Ph.D. candidate in Medicinal Chemistry at West China School of Pharmacy, Sichuan University. His research focuses on the discovery and development of novel therapeutic agents for both oncology and neurodegenerative diseases. He is experienced in multi‐target drug design, natural product‐inspired hybrid synthesis, and biological evaluation of small molecules.

Xindan Liao, received her master's degree in Pharmaceutical Analysis from West China School of Pharmacy, Sichuan University in 2014. She is currently working at the Institute of Materials, China Academy of Engineering Physics, and her research focuses on radiation environment monitoring.

Guangjun Yu is currently an assistant researcher at the Institute of Materials, China Academy of Engineering Physics. He received his doctoral degree from West China School of Pharmacy, Sichuan University in 2025. Since 2019, his main research has focused on the design, synthesis and evaluation of multi‐target agents against oncology and neurodegenerative diseases.

Data Availability Statement

No data was used for the research described in the article.