The role of natural flavonoids on neuroinflammation as a therapeutic target for Alzheimer’s disease: a narrative review

Abstract

Alzheimer’s disease is a neurodegenerative disease that affects a large proportion of older adult people and is characterized by memory loss, progressive cognitive impairment, and various behavioral disturbances. Although the pathological mechanisms underlying Alzheimer’s disease are complex and remain unclear, previous research has identified two widely accepted pathological characteristics: extracellular neuritic plaques containing amyloid beta peptide, and intracellular neurofibrillary tangles containing tau. Furthermore, research has revealed the significant role played by neuroinflammation over recent years. The inflammatory microenvironment mainly consists of microglia, astrocytes, the complement system, chemokines, cytokines, and reactive oxygen intermediates; collectively, these factors can promote the pathological process and aggravate the severity of Alzheimer’s disease. Therefore, the development of new drugs that can target neuroinflammation will be a significant step forward for the treatment of Alzheimer’s disease. Flavonoids are plant-derived secondary metabolites that possess various bioactivities. Previous research found that multiple natural flavonoids could exert satisfactory treatment effects on the neuroinflammation associated with Alzheimer’s disease. In this review, we describe the pathogenesis and neuroinflammatory processes of Alzheimer’s disease, and summarize the effects and mechanisms of 13 natural flavonoids (apigenin, luteolin, naringenin, quercetin, morin, kaempferol, fisetin, isoquercitrin, astragalin, rutin, icariin, mangiferin, and anthocyanin) derived from plants or medicinal herbs on neuroinflammation in Alzheimer’s disease. As an important resource for the development of novel compounds for the treatment of critical diseases, it is essential that we focus on the exploitation of natural products. In particular, it is vital that we investigate the effects of flavonoids on the neuroinflammation associated with Alzheimer’s disease in greater detail.

Introduction

Alzheimer’s disease (AD) is a neurodegenerative disease that is characterized by memory loss, progressive cognitive impairment, and various behavioral disturbances. From a clinical perspective, the characteristics of AD include memory loss, progressive cognitive impairment, and various behavioral and neuropsychiatric disturbances that eventually result in disability in daily functional activities (Goedert and Spillantini, 2006). Although the pathological mechanisms underlying AD are complex and remain unclear, previous research has identified two widely accepted pathological characteristics: extracellular neuritic plaques containing amyloid beta peptide, and intracellular neurofibrillary tangles containing tau. The clinical and pathological effects of AD, along with its overall complexity, make it one of the most expensive, lethal, and burdensome diseases of the 21^st century (Scheltens et al., 2021).

Flavonoids are plant-derived secondary metabolites that contain polyphenolic heterocyclic compounds and are found in abundance in fruits, flowers, seeds, and vegetables. Thus far, more than 5000 types of natural flavonoids have been identified in nature. These metabolites possess a common parent nucleus C6-C3-C6 structure that is made up of two aromatic rings (rings A and B) connected by an oxygen-containing pyran ring (ring C). Based on the degree of oxidation in the central C ring, substitution at three positions in the C ring, and the hydroxylation pattern of the A/B rings, the flavonoids include several main categories, including: flavones, flavonol, flavanones, flavanonols, isoflavanones, dihydroisoflavanones, chalcones, dihydrochalcones, flavan-3-ols, flavan-3,4-diols, anthocyanidins, xanthones, aurones, and homoisoflavones. The chemical structures of the flavonoid subclasses are shown in Figure 1.

Figure 1.

Chemical structure of the C6-C3-C6 parent nucleus (shown in the rectangle) and subclasses of flavonoids.

The scaffolds and structural diversity of flavonoids result in a wide range of pharmacological effects, including antioxidant (Boots et al., 2008), antibacterial (Cushnie and Lamb, 2005), antiviral (Badshah et al., 2021), antinociceptive (Alghamdi, 2020), anticarcinogenic (Lall et al., 2016), antiulcerogenic (Jung et al., 2007), and anti-inflammatory (Maleki et al., 2019) activities. Furthermore, numerous researchers have reported the neuroprotective effects and possible biochemical mechanisms of flavonoids in age-associated neurodegenerative disorders, including AD. In this review, we summarize the latest scientific evidence relating to the use of natural flavonoids as treatments for AD.

Retrieval Strategy

We performed a computer-based online search to retrieve articles published prior to the December 31, 2022. A combination of “AD,” “neuroinflammation,” and “flavonoids” was used as keywords to search the PubMed database. The results were further screened by title and abstract, and only studies that investigated the relationship between neuroinflammation and flavonoids in AD were included. Articles involving only other mechanism in AD without also examining neuroinflammation were excluded. In total, 13 natural flavonoids targeted neuroinflammation in AD treatment were screened out.

Pathophysiology and Inflammatory Processes in Alzheimer’s Disease

The pathophysiology of AD

The Aβ theory was first proposed in the 1980s and is now considered a landmark event in understanding the pathophysiology of AD. The production of Aβ is the result of the amyloidogenic processing of amyloid precursor protein (APP), a single-pass transmembrane protein that is primarily found in the synapses of neurons. APP is mainly metabolized through two pathways: the non-amyloidogenic processing pathway and the amyloidogenic processing pathway. The non-amyloidogenic processing pathway is usually utilized in healthy scenarios. APP is cleaved by α-secretase to generate a soluble extracellular N-terminal fragment (soluble APPα [sAPPα]) and a C-terminal fragment (CTF-α (C83)) (O’Brien and Wong, 2011; Zhang et al., 2011). Previous research has demonstrated that sAPPα can exert neuroprotective effects (Habib et al., 2017). The amyloidogenic processing pathway plays a leading role in the abnormal state of AD. APP is cleaved by β-secretase to generate a soluble extracellular N-terminal fragment (sAPPβ) and an intracellular C-terminal fragment (CTF-β (C99)). Then, CTF-β produces an abundance of Aβ peptides containing 36–43 amino acid residues via γ-secretase-induced cleavage (Zhang et al., 2011). In humans, approximately 90% of Aβ is the Aβ_1–40 form containing 40 amino acids, followed by Aβ_1–42. Research has shown that Aβ_1–40 and Aβ_1–42 increasingly accumulate to sequentially form oligomers and fibrils; the soluble Aβ oligomer is the most neurotoxic (Zhang et al., 2019). These factors all contribute to the formation of extracellular plaques which are viewed as one of the hallmarks of AD pathology.

Compared with the “amyloid cascade hypothesis,” intracellular neurofibrillary tangles (NFT) in AD pathophysiology has attracted significant levels of attention over recent years. Histopathological analysis revealed that the presence of tau inclusions is more correlated with cognitive impairment than amyloid plaques (Bakota and Brandt, 2016). Tau is highly soluble and a natively unfolded microtubule-associated protein that is primarily distributed in neuronal axons. Tau binds to microtubules, thereby stabilizing structure under a healthy state (Pedersen and Sigurdsson, 2015). In abnormal situations, the hyperphosphorylation of tau results in microtubule destabilization and the formation of NFT; these processes usually occur in the neuronal body (Chong et al., 2018).

There is some debate on whether tau pathology acts parallel to Aβ deposition, thus enhancing their toxicity, or whether tau pathology refers to a downstream effect of Aβ deposition (Giacobini and Gold, 2013). Either way, Aβ deposition and tau hyperphosphorylation are well acknowledged as seminal events in AD and therefore represent significant processes for the development of related drugs.

Inflammation in AD

The relationship between AD and neuroinflammation was discovered more than 30 years ago. Neuroinflammation refers to a chronic and self-sustaining form of inflammation in a microenvironment characterized by Aβ plaques, activated microglia and astrocytes, stressed neurons, as well as various inflammatory cytokines and chemokines (Rubio-Perez and Morillas-Ruiz, 2012). It is still not clear whether neuroinflammation is a driving factor for AD or a response to pathogenic events. However, the current evidence clearly emphasizes the important role of neuroinflammation in the progression of AD.

Pattern recognition receptors

During the onset stage, the presence of Aβ induces microglia and astrocytes by several sensors. The first sensor is toll-like receptors (TLRs). TRL4 and coreceptors (CD14 and MD2) that are expressed in microglia (Jin et al., 2008; Gambuzza et al., 2014), along with TLR2 and TLR9 (Gambuzza et al., 2014), could be activated by fibrillar Aβ. The second sensor for Aβ is receptor for advanced glycation end-products (RAGE), a cell surface receptor belonging to the immunoglobulin superfamily. In AD, Aβ peptides and oligomers bind to RAGE and activate glia cells, especially microglia (Paudel et al., 2020). Some other coreceptors are involved, including CD14, CD36, CD47, α6β1 integrin, and class A scavenger receptor (Bamberger et al., 2003; Sheedy et al., 2013).

Microglia and astrocytes

Microglia and astrocytes are the major source of cytokines in AD. Microglial activation, in particular, has been extensively evaluated as an index of neuroinflammation (Zimmer et al., 2014). Following the activation of TLR4, RAGE, and coreceptors on the glial cells, signal-dependent transcription factors are subsequently activated. These transcription factors then drive the expression of downstream inflammatory response genes, including interleukin-1α (IL-1α), IL-1β, and tumor necrosis factor-α (TNF-α) (Swanson et al., 2018). The increased expression levels of cytokines in AD induce the further production of APP and higher levels of Aβ, thus creating a vicious circle (Cacquevel et al., 2004). Activated microglia also produce reactive oxygen and nitrogen species, neurotoxic substances, and proteolytic enzymes (such as cathepsin B), which induce synaptic dysfunction, extracellular matrix damage, and neuronal death (Chaney et al., 2019).

Recent studies reported that astrocytes constitute an important aspect of the inflammatory microenvironment. However, the contribution of astrocytes to the progression of AD is largely overlooked when compared with the contribution made by microglia. Reactive astrocytes release several types of molecules, including cytokines (ILs, TNF-α, transforming growth factor-β [TGF-β], and others), chemokines (CXCL and CCL family), growth factors [brain-derived neurotrophic factor (BDNF), nerve growth factor, and others], gliotransmitters (glutamate, D-serine, and ATP), and small molecules (NO and prostaglandins) (Sofroniew, 2014; Harada et al., 2016). Astrocytes are also important for the clearance and degradation of Aβ by providing trophic support to neurons and forming a protective barrier against Aβ (Yamazaki and Kanekiyo, 2017; Bandyopadhyay, 2021). Moreover, reactive astrocytes increase the levels of sAPPα and β-site APP-cleaving enzyme 1 (BACE1), thus leading to a further increase in Aβ production (Lesné et al., 2003; Chacón-Quintero et al., 2021). Therefore, further studies are now needed to investigate the specific role of astrocytes in the progression of AD. Nevertheless, astrocytes represent a potential target for AD.

Complement system

In addition to glial cells, the role of the complement system in AD pathology is indisputable. As an essential component of the innate immune system, the complement system acts as a bridge between innate and adaptive immunity (Torvell et al., 2021). Previous studies showed that activation of the complement system in the brain leads to neuroinflammation, neuronal and synaptic loss, subsequently resulting in neurodegeneration in AD patients (Jiang and Bhaskar, 2017; Torvell et al., 2021; Carpanini et al., 2022). The co-localization of complement proteins with Aβ plaques and NFT have been reported in many studies, as well as the elevated levels of C1q, C3, C4 proteins, and mRNA, in the brain tissues of patients with AD (Rogers et al., 1992). Complement activation products (C3b) and terminal products (C5b–C9) of membrane attack complex (MAC) have been specifically detected in brain tissues from AD patients, thus indicating that MAC potentially induces neuronal loss and neurodegeneration (Walker and McGeer, 1992).

There is a notable vicious cycle between Aβ plaques, tau hyperphosphorylation, and neuroinflammation, with regard to the progression of AD (Rubio-Perez and Morillas-Ruiz, 2012). The neuroinflammatory microenvironment consists of activated microglia, astrocytes, complement cascade, various chemokines, and cytokines, reactive oxygen intermediates, proteolytic enzymes, and excitatory amino acids, which in turn leads to increased Aβ production and aberrant tau hyperphosphorylation (Heppner et al., 2015; Figure 2). In other words, neuroinflammation has an additive effect and acts as a risk factor with regard to increased AD severity.

Figure 2.

The neuroinflammatory process in Alzheimer’s disease.

In the pre-symptomatic stage, the amyloidogenic processing of APP and hyperphosphorylated tau lead to the accumulation of Aβ, the formation of amyloid plaques, and the development of intracellular neurofibrillary tangles. Activated microglia, astrocytes, the complement cascade, chemokines, cytokines, and reactive oxygen intermediates, constitute the inflammatory microenvironment in brain, which promotes pathological processes and aggravates the severity of Alzheimer’s disease. APP: Amyloid precursor protein; Aβ: amyloid beta; MAC: membrane attack complex; RAGE: receptor for advanced glycoxidation endproducts; TLR: toll-like receptor.

Flavonoids in Alzheimer’s Disease Treatment

As one of the most important diseases worldwide, there are still very few drugs available to treat AD. The current drugs that are mainly used in the clinic are tacrine, donepezil, rivastigmine, galantamine, memantine, and namzaric (Rafii and Aisen, 2015). As a complex multifactorial disease, there is an urgent need to develop more efficient drugs for the treatment of AD. Numerous small molecules that can be isolated from nature represent an exciting resource for the discovery of new drugs.

Epidemiological evidence and clinical trials have confirmed that a higher consumption of dietary flavonoids is associated with lower prevalence rate of dementia in 23 developed countries, even after adjusting for genetic background (Beking and Vieira, 2010). The Mediterranean diet is rich in flavonoids from fruits, vegetables, and wine, and has been proven to reduce the aggravation of mild cognitive impairment to AD (Singh et al., 2014; Scarmeas et al., 2018; Nagpal et al., 2019; Shishtar et al., 2020). Thus, there is a clear correlation between a higher intake of flavonoids and a lower incidence rate of AD. However, the mechanisms underlying this correlation have yet to be elucidated. Considering the good performance of flavonoids for the treatment of other inflammation-related diseases, an increasing number of studies have been conducted to investigate the anti-neuroinflammatory effects of AD; these are summarized in the following pages.

Apigenin

Apigenin (4′,5,7-trihydroxyflavone, Figure 3) is widely available in plants, including vegetables (parsley, celery, onions), oranges, herbs (chamomile, thyme, oregano, basil), and plant-based beverages (tea, beer, and wine). Numerous studies have reported that apigenin exhibits various pharmacological functions and has the potential to be a therapeutic agent for inflammation and neurodegenerative-related diseases, as well as autoimmune function, and even several types of cancers (Salehi et al., 2019).

Figure 3.

Structures of 13 natural flavonoids, including apigenin, luteolin, naringenin, quercetin, morin, kaempferol, fisetin, isoquercitrin, astragalin, rutin, icariin, mangiferin, and anthocyanins.

In a previous study, an isoflurane/pentylenetetrazole-induced rat/mouse model with cognitive dysfunction and behavioral impairments, was treated with different doses of apigenin (50 and 100 mg/kg per day, intraperitoneal injection (i.p.), for 1 week; 10 and 20 mg/kg per day, i.p., every fifth day in 20 days). Analysis revealed significant enhancements in partial learning and memory, cognitive deficit, and behavioral impairments, as determined by the Morris water maze (MWM), T-maze, elevated plus-maze (EPM), forced swim test, and tail suspension test (Chen et al., 2017; Sharma et al., 2020). Similar effects of treatment with apigenin (40 mg/kg, intragastrically (i.g.), for 5 days/week for 12 weeks) on learning and memory improvement were observed in APP/PS1 double transgenic mice; the mechanisms of action underlying these effects were directly related to the intervention of APP processing and the prevention of Aβ burden by downregulating APP and BACE1 (the major β-secretase), and β-CTF levels, relieving Aβ deposition, and decreasing the levels of insoluble Aβ (Zhao et al., 2013). Apigenin has also been found to inhibit the expression of glycogen synthase kinase-3 (GSK-3β), thereby reducing tau hyperphosphorylation (Alsadat et al., 2021).

In addition, the use of apigenin (50 and 100 mg/kg/day, i.p. for 1 week; 1 μM, 24 hours) during in vivo and in vitro experiments on an isoflurane-induced rat model, and co-cultures of neurons and glial cells from the rat cortex, respectively, led to the satisfactory suppression of neuroinflammation, as characterized by a reduction in microglial activation and proliferation, the modulation of microglial morphology, a reduction in the expression levels of M1 inflammatory marker/cytokines (CD68, OX42, IL-6, IL-1β, CCL5, and gp130), and an increase in the expression of M2 inflammatory cytokines (IL-10). Apigenin has also been shown to exhibit a superoxide anion scavenging effect and improved the antioxidative enzyme activities of superoxide dismutase (SOD) and glutathione peroxidase (GSH-PX) (Chen et al., 2017; Dourado et al., 2020). The mechanism of action involved restoration of the extracellular signal regulated kinase/cAMP response element-binding protein/BDNF (ERK/CREB/BDNF) pathway in the cerebral cortex (Zhao et al., 2013; Sharma et al., 2020).

Luteolin

Luteolin is widely available in vegetables (broccoli, onion leaves, carrots, peppers, cabbages), fruits (apple), and medicinal herbs (Chrysanthemum indicum var. albescens, Codariocalyx motorius (Houtt.) H. Ohashi, and Artemisia dubia var. asiatica Pamp.) (Aziza et al., 2018). A hydroxyl group at the 3′-position distinguishes luteolin from apigenin (Figure 3). According to modern pharmacological research, luteolin exhibits antioxidant, antimicrobial, anti-inflammatory, chemotherapeutic, cardioprotective, antidiabetic, neuroprotective, and antiallergic effects. Plants with a high luteolin content have been used for a long time in Iran, Brazil, and China, to treat inflammation-related diseases (Nabavi et al., 2015).

Luteolin has attracted significant attention in the field of AD treatment due to its significant effects against cognitive impairment. For example, luteolin treatment (at doses of 100 and 200 mg/kg per day, i.g., for 17 days) reduced cerebral Aβ accumulation in the Tg2576 mouse model (Rezai-Zadeh et al., 2009), increased the levels of choline acetyl transferase, and decreased the activity of acetylcholinesterase (AChE), in Aβ-induced rats (Yu et al., 2015). Luteolin was also found to alleviate amyloidogenesis by reducing BACE1 expression and Aβ_1–42 accumulation. Synaptic dysfunction was also improved after treatment with luteolin, as characterized by an increase in the levels of two synaptic markers: postsynaptic density protein 95 (PSD95) and synaptosomal-associated protein (SNAP)-25 (Ahmad et al., 2021).

With regard to the inhibition of neuroinflammation, treatment with luteolin (80 mg/kg per day, i.g., for 2 weeks) resulted in a significant reduction in the expression levels of TNF-α, IL-1β, and p-nuclear factor-κB p65 [p-NF-κB p65 (Ser536)] in the frontal cortex and the hippocampus of a Aβ_1–42-induced mouse model of Alzheimer’s disease. In vitro experiments further showed that neuroinflammation was alleviated after treatment with luteolin (5, 10, 20, 25, and 50 μM, for 1 hour), as characterized by a reduction in the levels of IL-6, IL-1β, TNF-α, inducible nitric oxide synthase (iNOS), cyclooxygenase-2 (COX-2), and prostaglandin E2 (PGE2), the alleviation of microglial activation and proliferation, and a reduction in neuronal apoptosis (Jang et al., 2008; Zhu et al., 2011). The mechanisms underlying these actions were associated with suppression of the NF-κB signaling pathway and c-Jun N-terminal kinase (JNK) signaling pathway, and the activation of AP-1 (Jang et al., 2008; Zhu et al., 2011; Ahmad et al., 2021).

Naringenin

Naringenin (5,7,4′-trihydroxyflavanone, Figure 3) belongs to the family of flavanones and is found in abundance in citrus fruits, such as grapefruits (Citrus paradisi) and oranges (Citrus sinensis). According to previous studies, naringenin has effective antimicrobial, antioxidant, and anti-inflammatory properties, and has been shown to regulate lipoprotein metabolism, diabetes, atherosclerosis, and insulin resistance (Patel et al., 2018).

Pre-treatment with naringenin (50 mg/kg per day, i.g., for 2 or 8 weeks; 100 mg/kg, i.g., for 1 hour before surgery) was shown to improve the acquisition and retention of memory in various model of AD, including streptozotocin-induced memory impairment in rats, D-galactose-induced aging in mice, and AlCl3/D-galactose-induced neurotoxicity in rats. Several behavioral experiments were employed to evaluate the effects of naringenin, such as the MWM, EPM, forced swim, Y-maze, open-field, passive avoidance, and nesting behavior tests (Khan et al., 2012; Ghofrani et al., 2015; Zhang et al., 2017; Haider et al., 2020). Biochemical analysis proved that naringenin significantly reduced SOD and malondialdehyde (MDA) activities, as well as increased catalase (CAT), glutathione peroxidase (GPx), and glutathione (GSH) activities in the hippocampus. These results showed that naringenin prevented the brain from lipid peroxidation and oxidative stress (Khan et al., 2012). Furthermore, 24 hours of treatment with naringenin at a dose of 25 μM, remarkably alleviated amyloidogenesis and the direct neurotoxic effects of Aβ on SH-SY5Y human neuroblastoma cells by downregulating the expression levels of APP and BACE1. Moreover, naringenin reduced the levels of phosphorylated tau in SH-SY5Y cells (Md et al., 2018).

Quercetin

Quercetin is categorized as a flavonol (2-(3,4-dihydroxyphenyl)-3,5,7-trihydroxy-4H-chromen-4-one, Figure 3) and is commonly found in fruits (apples, berries, and grapes), vegetables (onions, shallots, and tomatoes), and medicinal plants (Ginkgo biloba, Hypericum perforatum, and Sambucus canadensis) (Li et al., 2016). As an important bioflavonoid, quercetin is known for its anti inflammatory, antihypertensive, vasodilatory, anti-obesity, anti-hypercholesterolemic, and antiatherosclerotic effects (Batiha et al., 2020).

Over recent years, the neuroprotective effects of quercetin have been extensively studied in multiple neurodegenerative disorders, including AD. Quercetin significantly improved learning, memory, and cognitive functions when used to treat a mouse model of AD at a dose of 25 mg/kg per day, i.p., every 48 hours for 3 months or 40 mg/kg per day, per os (p.o.), for 16 weeks (Wang et al., 2014; Sabogal-Guáqueta et al., 2015). Quercetin also prevented the degradation of acetylcholine and reduced Aβ production by inhibiting the levels of AChE and secretase enzymes (Khan et al., 2009). In the age-triple transgenic (3×Tg) mouse model of AD, the administration of quercetin (25 mg/kg, i.p., every 48 hours for 3 months) reduced the levels of paired helical filament, Aβ_1–40, and Aβ_1–42, and the BACE1-mediated cleavage of APP to CTFβ (Sabogal-Guáqueta et al., 2015). In a scopolamine-induced model of memory impairment, and an LPS-induced model of neuroinflammation, the administration of quercetin (0.25 mg/kg per day, i.p., for 1 week) attenuated the activation of microglia and astrocytes, as characterized by specific markers (Iba-1 and GFAP) (Khan et al., 2018). Another treatment using quercetin (0.25 mg/kg per day, i.p., for 1 week; 25 mg/kg per day p.o. + 3 mg/kg/day, i.p., for 1 week) also attenuated the increase of COX-2, nitric oxide synthase-2, IL-1β, TNF-α, and IL-6 in hippocampal subregions and the prefrontal cortex of the mouse brain (Khan et al., 2018; Olayinka et al., 2022). Therefore, the effects of quercetin on the reduction in extracellular β-amyloidosis, tauopathies, astrogliosis, and microgliosis in AD are due to its inhibitory effect on neuroinflammation. The underlying mechanism of action is associated with suppression of the TLR4/NF-κB pathway, JNK pathway, nuclear factor (erythroid-derived 2)-like 2 (Nrf2) pathway, and the PI3K/Akt pathway (Zaplatic et al., 2019).

Morin

Morin (3,5,7,2′,4′-pentahydroxyflavone, Figure 3), also known as morin hydrate, is a flavonol that is isolated as a yellow pigment from a variety of plants, particularly the Moraceae family. Morin possesses anti-inflammatory, antioxidant, and free radical scavenging properties, and is used to treat various diseases, such as cancer, acute liver injury, neuroinflammatory disorders, diabetes, gastritis, mastitis, and myocardial infarction (Komirishetty et al., 2016; Bachewal et al., 2018; Verma et al., 2019; Rajput et al., 2021).

Following the administration of morin at a dose of 20 mg/kg per day, i.p., for 12 weeks, the MWM test revealed that spatial learning and memory deficits in APPswe/PS1dE9 mice had significantly recovered. Moreover, morin inhibited Aβ production and alleviated the burden created by Aβ plaques in the cortex and hippocampus through several targets, including promotion of the non-amyloidogenic APP processing pathway by increasing the expression of ADAM10 (the major form of α-secretase). Morin also inhibited the amyloidogenic APP processing pathway by reducing the expression levels of BACE1 and PS1 (the major form of γ-secretase) and facilitated the degradation of Aβ by enhancing the expression of Aβ-degrading enzyme (Du et al., 2016). Dose-independent inhibitory effects of morin (12.5, 25, 50, and 100 μM, for 24 hours) against β-secretases and 50 μM for 24 hours against γ-secretases activities, were observed in MC65 cells and the membranes of APPswe cells (Carmona et al., 2020). In addition, the administration of morin (10 mg/kg per day, i.p., for 7 days; 1 and 10 μM, for 6 hours) had a significant inhibitory effect on GSK-3β activity and attenuated GSK3β-induced tau phosphorylation in a dose-dependent manner in the hippocampi of 3×Tg-AD mice and in SH-SY5Y cells (Gong et al., 2011). The effects of morin on both Aβ deposition and tau pathology is speculated to be associated with its effect on neuroinflammation. Previous studies reported that the administration of morin (100 and 200 mg/kg per day, p.o., for 2 days; 20 mg/kg per day, i.g., for 14 days) significantly reduced the expression levels of MDA, TNF-α, IL-1β, IL-6, NF-κB, and N-methyl-D-aspartate receptor (NMDAR) subunits 2A and 2B, and increased the expression levels of CAT, SOD, GSH, GPx, BDNF, and the α7 nicotinic acetylcholine receptor (nAChR), in the hippocampi of Aβ_1–42 rats (Çelik et al., 2020; Mohammadi et al., 2021).

Kaempferol

Kaempferol, chemically known as 3,5,7-trihydroxy-2-(4-hydroxyphenyl)-4H-1-benzopyran-4-one (Figure 3), is commonly found in numerous vegetables (such as cabbage, kale, broccoli, fruits (strawberries, gooseberries, citrus fruits, grapefruit, apples, dry raspberry, grapes, tomatoes), beans, and teas. Kaempferol can also be isolated from Astragali Complanati Semen (Astragalus complanatus R. Br.), Ginkgo Folium (Ginkgo biloba L.), and Propolis (Apis mellifera L.) in Traditional Chinese Medicine (Montaño et al., 2011). Kaempferol is known to possess anti-inflammatory, antioxidant, antimicrobial, anticancer, and antiviral properties; its neuroprotective effects have been increasingly investigated over recent years (Santos et al., 2021).

In an ovariectomized rat model of sporadic dementia, kaempferol treatment (10 mg/kg per day, i.p., for 21 days) significantly improved spatial memory, as evidenced by a shorter total time spent and a longer-latency to the platform in the MWM test. Kaempferol treatment also increased the levels of SOD and GSH, and reduced the levels of TNF-α and MDA in the hippocampi of rats (Kouhestani et al., 2018; Babaei et al., 2021). In LPS-induced mice, the levels of IL-6, IL-1β, TNF-α, monocyte chemotactic protein-1 (MCP-1), intercellular cell adhesion molecule-1, and COX-2, in the striatum tissues were reduced following the administration of kaempferol (50 mg/kg per day, i.p., for 7 days) (Yang et al., 2019). These results supported the beneficial effects of kaempferol on neuroinflammation in AD. The reduction of neuroinflammation alleviated neuronal injury, maintained brain-blood barrier integrity, and increased the levels of tyrosine hydroxylase (TH) and PSD95 in the striatum of mice, by downregulating the HMGB1/TLR4 pathway (Yang et al., 2019). These findings were confirmed both in vitro and in vivo. Phagocytosis and the production of reactive oxygen species (ROS), along with the levels of inflammatory mediators (NO, PGE2, COX-2, iNOS, TNF-α, and IL-1β) in LPS-stimulated microglial BV-2 cells were significantly reduced after kaempferol treatment at different doses (30, 50, and 100 μM, for 1 hour). Analysis demonstrated that TLR4, mitogen activated protein kinases (MAPK) cascades, PKB (AKT), and the NF-κB signaling pathway, were all involved in the mechanism of action associated with these findings (Park et al., 2011).

Fisetin

Fisetin (3,3,4,7-tetrahydroxyflavone, Figure 3) is widely available in several common fruits (strawberries, persimmons, apples, grapes, kiwis, peach) and vegetables (onions, cucumber, lotus root). Thus, fisetin is commonplace in daily diets with an estimated mean daily intake by humans of around 0.4 mg (Elsallabi et al., 2022). Fisetin exhibits a large plethora of bioactivities, including anti-inflammatory, hypolipidemic, hypoglycemic, antioxidant, neuroprotective, antiangiogenic, and antitumor properties. Furthermore, fistein is known to exert satisfactory effects for the treatment of neurological complications and neurodegenerative diseases (Ravula et al., 2021).

With regard to studies evaluating the behavioral effects of fisetin in AD, treatment with fisetin at a dose of 25 mg/kg per day, p.o., for 7 or 9 months, and 20 mg/kg per day, i.p., for 2 weeks, was previously shown to prevent cognitive and locomotive deficits in senescence-accelerated mouse prone 8 (SAMP8) mice, and in APPswe/PS1dE9 mice, as determined by the MWM test, EPM test, Barnes’ maze test, probe test, open field test, and object recognition test (Currais et al., 2014, 2018; Ahmad et al., 2017). The administration of fisetin at a dose of 25 mg/kg per day, p.o., for 9 months, 20 mg/kg per day, i.p., for 2 weeks, and 10 μM, for 24 hours, was previously shown to significantly reduce the accumulation of Aβ and BACE1, the level of phosphorylated tau (Ser 413) in the hippocampi of APPswe/PS1dE9 mice, and Aβ-induced cognitive dysfunction in mice (Currais et al., 2014; Kim et al., 2016; Ahmad et al., 2017). Similarly, in vitro experiments with varying doses of fisetin (10 μM, for 24 hours; 10 and 20 μM, for 36 hours) verified its inhibitory effect on tau aggregation and the disaggregation of tau filaments in mouse/rat cortical cells (Kim et al., 2016; Xiao et al., 2021).

In terms of neuroinflammation, the administration of fisetin (25 mg/kg per day, p.o., for 9 months, and 5 μM, for 24 hours) notably suppressed various activated neuroinflammatory mediators and gliosis, including p-IKKβ, p-NF-κB, TNF-α, IL-1β, Iba-1, and GFAP. In addition, the levels of oxidative stress in APPswe/PS1dE9 mice and BV-2 microglial cells were both alleviated by the administration of fisetin; this observation was marked by protein carbonylation and chelating activity on iron and copper (Currais et al., 2014; Maher, 2020). The mechanisms of action involved in this process included a reduction in the level of cyclin-dependent kinase 5 (Cdk5) activator p35 cleavage product (p25), and activation of the autophagy pathway by transcription factor EB and Nrf2 (Currais et al., 2014; Kim et al., 2016).

Isoquercitrin

Isoquercitrin (quercetin-3-O-β-D-glucopyranoside, Figure 3) is another major glycosidic form of quercetin. Compared with rutin (section 4.10.), very few studies have investigated the actions of isoquercitrin. Research has shown that Isoquercitrin can be detected in various plants, including Rosa soulieana Crépin, Eucommia ulmoides Oliv., Crataegus pinnatifida Bge. (Chinese hawberry), Crataegus azarolus L. (azarole), Caragana arborescens Lam. (Siberian peashrub), Arbutus unedo L., and various Allium species (Valentová et al., 2014). Several studies have reported the beneficial health effects of isoquercitrin, including its antioxidant, anti-inflammatory, anticancer, antidiabetic, antiallergic, and neuroprotective properties (Bombardi Duarte et al., 2018). The neuroprotective effects of isoquertin in neurodegenerative diseases have attracted particular attention from many researchers.

In Aβ_25–35-induced mice, isoquercitrin treatment (10 mg/kg per day, i.g., for 4 weeks) significantly improved spatial cognitive and object recognition ability, as well as learning and memory function, as determined by the MWM, T-maze, and novel object recognition (NOR) tests (Kim et al., 2019). Isoquercitin was also shown to reduce Aβ aggregation and the expression of amyloidosis-related proteins in the brains of mice injected with Aβ, including β-secretase, presenilin (PS)-1, and PS-2 (Lim et al., 2019; Kim et al., 2020).

Isoquercitrin treatment has also been shown to significantly alleviate neuroinflammation in AD. In LPS-induced microglia, RAW 264.7 macrophages cells, and mice hippocampus, the administration of isoquercitrin (2.5 and 5 μg/mL, for 12 hours; 6.25, 12.5, and 25 μM, for 24 hours; 1 and 5 mg/kg per day, i.p., for 4 days) effectively reduced NO production; lipid peroxidation; the levels of iNOS, ROS, COX-2, prostaglandin E synthase 2 (PTGES2), and the expression of multiple pro-inflammatory cytokines (TNF-α, IL-1β, IL-6, and MCP-1). The mechanism involved was associated with the inhibition of MAPK kinase phosphorylation by extracellular signal-regulated kinase, c-Jun N-terminal kinase, and p38 signaling proteins (Kim et al., 2022; Lee et al., 2022).

Astragalin

Astragalin (kaempferol-3-O-glucoside, Figure 3) is found in certain medicinal plants (Astragalus mongholicus, Astragalus smicus, Cuscuta chinensis, Cassia alata, Rosa chinensis Jacq.), green tea (Camellia sinensis), beans (Phaseolus vulgaris L.), persimmons (Diospyros kaki), and cherry (Prunus serotina). It has been reported that astragalin can exert antioxidant and anti-inflammatory activity against a variety of diseases (Riaz et al., 2018).

In an AlCl3/D-galactose-induced mouse model, the administration of astragalin (5, 10, and 20 mg/kg per day, subcutaneous injection (s.c.), for 3 weeks; 5 and 10 mg/kg per day, i.g., for 3 weeks) effectively ameliorated cognitive impairment and the aging-like phenotype, as well as improved learning and memory performance, as determined by the MWM, NOR, open-field, sucrose preference, tail suspension, and forced swimming tests. Astragalin also restored oxidative stress-regulating enzymes or markers (T-SOD, T-AOC, CAT, GSH-PX, and MDA) in the brain (Hu et al., 2022; Yao et al., 2022). Similar results were observed in a murine model with ovariectomy-induced perimenopausal depression and LPS-induced neuroinflammation. Astragalin treatment (5 and 10 mg/kg per day, i.g., for 3 weeks; 5 and 20 mg/kg per day, i.p., for 4 days) switched microglia from the M1 phenotype to the M2 phenotype, and reduced the levels of NO, iNOS, ROS, COX-2, PTGES2, nuclear NF-κB p65, and the expression of multiple pro-inflammatory cytokines (TNF-α, IL-1β, IL-6, and MCP-1) in the brain (Kim et al., 2022; Yao et al., 2022).

Similarly, the administration of astragalin (10, 20, and 40 μM, for 2 hours) to LPS/ATP induced BV-2 cells switched microglia from the M1 phenotype to the M2 phenotype, as characterized by the increased mRNA expression of M2 markers (Arg1, Ym1, Fizz1, and Klf4), reduced levels of pro-inflammatory cytokines (TNF-α, IL-1β, IL-6, and IL-12), and increased levels of anti-inflammatory cytokines (IL-10 and IL-4) (Yao et al., 2022). The molecular mechanism of action for astragalin is known to be related to activation of the IL-4 receptor/janus kinase 1/signal transducer and activator of transcription 6 (IL-4R/JAK1/STAT6) signaling pathway, and attenuation of the Notch/HES-1 and NF-κB signaling axis (Hu et al., 2022; Yao et al., 2022).

Rutin

Rutin (3,3′,4′,5,7-pentahydroxyflavone-3-rhamnoglucoside, Figure 3) is found extensively in a variety of plants, including passionflower, buckwheat, tea, and apple. In addition to being a vital nutritive component of food, rutin is also known as rutoside, quercetin-3-rutinoside, or sophorin. As a glycoside consisting of flavonolic aglycone quercetin, along with disaccharide rutinose, rutin also exhibits antioxidant, cytoprotective, vasoprotective, anticarcinogenic, neuroprotective, and cardioprotective properties (Ganeshpurkar et al., 2017).

Previous studies on the treatment of AD, reported that the administration of rutin (4 μM, for 48 hours; 0.8 and 8 μM, for 12 hours) notably inhibited tau aggregation, reduced tau oligomer-induced cytotoxicity, regulated tau hyperphosphorylation by increasing the levels of PP2A, and promoted the microglial uptake of extracellular tau oligomers in primary hippocampal neurons in a mouse model and in SH-SY5Y cells. Moreover, treatment with rutin (4 and 8 μM, for 24 hours; 0.8 and 8 μM for 12 hours) inhibited inflammation in primary microglia or BV-2 microglial cells by reducing the production of pro-inflammatory cytokines (TNF-α, IL-1β, and NO) and attenuating oxidative stress (ROS, MDA, glutathione disulfide [GSSG]; Yu et al., 2015; Sun et al., 2021).

In APPswe/PS1dE9 transgenic mice, Tau-P301S mice, and intracerebroventricular-streptozotocin infused rats, the administration of rutin (100 mg/kg, p.o., for 6 weeks or 30 days; 25 mg/kg, p.o., for 3 weeks) significantly improved cognitive function, as determined by MWM, Y-maze, and NOR tests. Both oligomeric Aβ and pathological tau levels in the brain were reduced after rutin treatment. These effects were characterized by a reduction of microgliosis and astrocytosis, as well as the prevention of microglial synapse engulfment. Rutin treatment also increased SOD activity, and reduced the levels of IL-1, IL-6, IL-8, COX-2, iNOS, MDA, GSSG, and GSH. These results proved the satisfactory effects of rutin with regard to the inhibition of neuroinflammation, thus alleviating synaptic loss in the mouse brains via mechanisms related to the downregulation of NF-κB pathway (Javed et al., 2012; Xu et al., 2014; Sun et al., 2021).

Icariin

Icariin (4′-O-methyl-8-γ,γ-dimethylallylkaempferol-3-rhamnoside-7-glucoside, Figure 3) is the most abundant constituent of Epimedii herba (Epimedium brevicornu Maxim., Epimedium sagittatum (Sieb. et Zucc.) Maxim., Epimedium pubescens Maxim., Epimedium koreanum Nakai). Icariin exhibits multiple biological activities, including antioxidative, antineuroinflammatory, and antiapoptotic effects. Icariin has significant potential for the treatment of various diseases, ranging from neoplasm to cardiovascular disease (Jin et al., 2019).

Previous research has indicated that icariin exerts preventive and therapeutic effects on the nervous system, including AD. In SAMP8 mice, APP/PS1 mice, and transgenic AD model mice (5×FAD), icariin treatment (50 μmol/kg/day, p.o., for 8 days; 60 mg/kg per day, i.g., for 22 days) significantly reduced spatial memory impairment, as determined by MWM, EPM, and open-field tests (Urano and Tohda, 2010; Wang et al., 2019; Wu et al., 2020). In addition, the expression levels of APP and BACE1 in the hippocampus were downregulated after treatment with icariin (60 mg/kg per day, i.g., for 22 days; 30 μmol/kg per day, i.g., for 6 months; 100 mg/kg per day, i.g., for 10 days), thereby reducing the cytotoxicity of Aβ_1–42 in SAMP8 mice and APPV717I transgenic (Tg) mice (Zhang et al., 2014; Wu et al., 2020). Icariin also reduced the levels of pro-inflammatory cytokines (IL-1β, IL-6, and TNF-α), increased the levels of anti-inflammatory cytokines (IL-4, IL-10, and TGF-β1), reduced Aβ plaque accumulation, and changed the microglial M1 phenotype (Iba-1 and iNOS) in the hippocampus and prefrontal cortex. The underlying mechanisms were associated with the suppression of peroxisome proliferator-activated receptor γ (PPARγ) and an increase in sirtuin-1 (Zhang et al., 2014; Wang et al., 2019; Chuang et al., 2021). Long-term treatment with icariin (60 mg/kg per day, p.o., for 8 months) modulated the differentiation of CD4⁺ T cells and the release of inflammatory cytokines in the plasma and brain tissue of APP/PS1 mice, thus indicating regulatory activity over systemic immune function (Zhu et al., 2019). Furthermore, in vitro experiments indicated that icariin treatment (2.5 μM, for 48 hours) attenuated the levels of p-tau and GSK-3β in okadaic acid-induced SH-SY5Y cells (Li et al., 2022).

Mangiferin

Mangiferin (2-β-D-glucopyranosyl-1,3,6,7-tetrahydroxy-9H-xanthen-9-one, Figure 3) is a natural C-glucoside xanthone present in mango, papaya, Coffea arabica, and Anemarrhena rhizome. Previous research indicated that mangiferin exhibits various actions, including antioxidant, anti-inflammatory, anti-lipid peroxidation, antidiabetic, anticancer, and immunomodulatory properties (Sun et al., 2021).

Researchers have also investigated the effects of mangiferin on AD. Mangiferin treatment (40 mg/kg per day, i.p., for 14 days; 100 and 200 mg/kg per day, i.g., for 60 days) significantly improved the learning ability and memory retention in amnesic mice induced by scopolamine, and aging dementia SAMP8 mice, as determined by MWM, EPM, and passive shock avoidance paradigm tests (Biradar et al., 2012; Du et al., 2019). In other studies, the administration of mangiferin (40 mg/kg per day, i.p., for 14 days; 50 mg/kg per day, i.g., for 22 weeks) reduced the expression levels of Aβ_1–40 and Aβ_1–42 in the hippocampi of SAMP8 mice, and also reduced the hyperphosphorylation of tau in the cortex and hippocampi of APP/PS1 mice, thus leading to the amelioration of damage in hippocampal neurons and the mitochondria (Biradar et al., 2012; Infante-Garcia et al., 2017). The increased levels of dopamine, noradrenaline, and AChE, in the brain were significantly reversed following the administration of mangiferin (40 mg/kg per day, i.p., for 14 days).

Several studies have verified that mangiferin exerts neuroprotective effects via the attenuation of oxidative stress and inflammatory responses. Increased lipid peroxidation and GSH levels in the whole brain of scopolamine-induced mice and SAMP8 mice were reduced after treatment with mangiferin (40 mg/kg per day, i.p., for 14 days; 100 and 200 mg/kg per day, i.g., for 60 days) (Biradar et al., 2012; Du et al., 2019). In vitro experiments verified that the levels of NO, IL-1β, IL-6, and TNF-α, as well as the mRNA and protein expression levels of iNOS and COX-2, were notably reduced after treatment with mangiferin (100 and 150 μg/mL, for 2 hours) in LPS-induced BV-2 microglial cells (Lei et al., 2021). Treatment using mangiferin (1 μM, for 30 minutes) restored the activity of antioxidant enzymes (SOD and CAT), thereby mitigating mitochondrial dysfunction and reducing protein oxidation (total carbonyl content) in Aβ-treated cortical neurons and organotypic slices of the neocortex (Alberdi et al., 2018). The mechanisms of action involved the NF-κB, NLRP3, PI3K/Akt, and ERK1/2 signaling pathways (Feng et al., 2019; Lei et al., 2021).

Anthocyanidin

As a special category of flavonoid, anthocyanin possesses a cationic structure at the B ring, thus differentiating it from other categories of flavonoid. Anthocyanins also exist in acetylated and glycoside forms, such as 3-glycosides, 5-glycosides, and diglucosides, which include but are not limited to glucose, galactose, and rutinose (Figure 3). To date, more than 650 anthocyanins have been identified from fruits and vegetables, including blueberry, grape, eggplant, cranberries, purple cauliflower, plums, and strawberries (Khan et al., 2020). Common anthocyanins are composed of one of six anthocyanidin bases (cyanidin, delphinidin, malvidin, pelargonidin, peonidin, and petunidin) which differ in their molecular structure at the B ring, and a sugar moiety attached at the 3 position of the C ring (Winter and Bickford, 2019; Figure 3). Anthocyanin has many beneficial health effects, including cancer-preventive, immuno-protective, and neuroprotective effects (Afzal et al., 2019).

With regard to AD treatment, anthocyanidins have received significant attention over recent years. The administration of anthocyanins (24 mg/kg per day, i.p., for 14 days; 100 mg/kg per day, i.p., for 7 weeks; 20 mg/kg per day, i.g., for 3 months) resulted in a significant improvement in behavioral performance and prevented the loss of cognitive function in LPS-induced mice, D-galactose-induced rats, and APP/PSEN1 mice, as determined by MWM, Y-maze, and probe tests (Rehman et al., 2017; Khan et al., 2019; Li et al., 2020). In addition, in vivo and in vitro experiments proved that the administration of anthocyanidins (24 mg/kg per day, i.p., for 14 days; 5, 10, and 20 μM, for 1 hour) protected neurons against Aβ-induced and LPS-induced neurotoxicity in the mouse brain, SK-N-SH cells, and mouse hippocampal HT22 cells (Belkacemi and Ramassamy, 2015; Khan et al., 2019). The administration of anthocyanin (24 mg/kg per day, i.p., for 14 days; 20 mg/kg per day, i.g., for 3 months) significantly inhibited the activation of microglia and astrocytes, as characterized by a reduction in the expression of glial cells sensors (TLR2, TLR4, RAGE, and CX3CR1), glial markers (Iba-1, GFAP), and microglia homeostatic factors (TREM2, tyrosine kinase binding protein [TYROBP]) in the hippocampus and cortex of the brain. Thus, neuroinflammation was significantly alleviated by anthocyanins and this occurred by reducing the levels of ROS and lipid peroxidation, as well as suppressing various inflammatory markers, including TNF-α, p-NF-κB, IL-1β, IL-6, COX-2, iNOS, and CD33. Anthocyanin treatment also increased the expression levels of synaptophysin, SNAP-23, SNAP-25, CD68, and IRF7, thereby restoring synaptic and phagocytotic function in a murine model of AD (Rehman et al., 2017; Li et al., 2020). The mechanisms of action associated with these effects involved regulation of the CD33/TREM2/TYROBP signaling pathways in the microglia, as well as upregulation of the phosphorylated-phosphatidylinositol 3-hydroxy kinase/Akt/GSK3β (p-PI3K/Akt/GSK3β) pathway, Nrf2/heme oxygenase-1 (Nrf2/HO-1) pathway, and p-JNK pathway in neurons (Rehman et al., 2017; Ali et al., 2018; Li et al., 2020).

In addition to the 13 natural flavonoids mentioned above, 7,8-dihydroxyflavone (7,8-DHF) has also attracted attention with regard to the treatment of AD and is a naturally-occurring plant-based compound that is found abundantly in fruits and vegetables (Akhtar et al., 2021). Strong preclinical evidence has shown that BDNF is beneficial as a therapeutic agent for various neurological disorders. Researchers identified 7,8-DHF as a small molecule that can mimic BDNF and could specifically bind to the extracellular domain of the TrkB receptor to act as a selective TrkB agonist (Jang et al., 2010). Moreover, 7,8-DHF can repress the expression of BACE1, reduce the levels of Aβ_1–40 and Aβ_1–42, and significantly improve cognitive functions in various animal models of AD (Devi and Ohno, 2012; Chen et al., 2018; Akhtar et al., 2021). However, there is a defined target for 7,8-DHF and the mechanism of the target is not related to neuroinflammation. For this reason, we have not included 7,8-DHF in our list of potential targets for AD.

The names, sources, AD models, doses, and mechanisms of action, for the 13 selected natural flavonoids with regard to neuroinflammation in AD are summarized in Table 1.

Table 1.

Names, sources, AD models, doses, and mechanisms of 13 natural flavonoids on neuroinflammation in Alzheimer’s disease

Name	Source	Model	Dose	Mechanism	Reference
Apigenin	Vegetables (parsley, celery, onions), fruits (oranges), herbs (chamomile, thyme, oregano, basil), and plant-based beverages (tea, beer, and wine)	In vivo: Isoflurane/pentylenetetrazole-induced rat/mouse model; APP/PS1 double transgenic AD mice; In vitro: LPS-induced co-cultures of primary neurons and glial cells from the cortex of newborn and embryonic Wistar rats	In vivo: 50 or 100 mg/kg/d, i.p., 1 wk; 35 mg/kg, i.p., every 5th d in 20 d; 40 mg/kg, i.g., 5 d/wk for 12 wk In vitro: 1 μM, 24 h	In vivo: ↓ APP processing, BACE1, β-CTF, Aβ deposition, insoluble Aβ levels; ↓ BACE1 mRNA, GSK-3β; ↑ ERK/CREB/BDNF pathway in the cerebral cortex In vitro: ↓ CD68, OX42, IL-6, IL-1β, CCL5, gp130, ↑ IL-10; ↑ Enzyme activity of SOD and GSH-PX	Zhao et al., 2013; Chen et al., 2017; Dourado et al., 2020; Sharma et al., 2020; Alsadat et al., 2021
Luteolin	Vegetables, fruits, and medicinal herbs, including broccoli, onion leaves, carrots, peppers, cabbages, apple skins, Chrysanthemum indicum var. albescens, Codariocalyx motorius (Houtt.) H. Ohashi, and Artemisia dubia var. asiatica Pamp.	In vivo: Tg2576 mouse model; Aβ peptide induced AD model rats In vitro: LPS-induced rimary murine microglia and BV-2 cells	In vivo: 100 or 200 mg/kg/d, i.g., 17 d; 80 mg/kg/d, i.g., 2 wk In vitro: 5, 10, 20, 25, or 50 μM, 1 h	In vivo: ↓ Cerebral Aβ accumulation, ↑ choline acetyl transferase and AchE; ↓ p-NF-kB p65 (Ser536), TNF-α, IL-1β in the frontal cortex and the hippocampus; ↑ PSD-95 and SNAP-25; ↓ JNK signaling pathway In vitro: ↓ IL-6, IL-1β, TNF-α, NO, iNOS, COX-2, PGE2; ↓ Microglial activation and proliferation; ↓ NF-κB signaling pathway, JNK signaling pathway, activation of AP-1	Jang et al., 2008; Rezai-Zadeh et al., 2009; Zhu et al., 2011; Yu et al., 2015; Ahmad et al., 2021
Naringenin	Citrus fruits, such as grapefruits (Citrus paradisi) and oranges (Citrus sinensis)	In vivo: streptozotocin-induced memory impairment rats; D-galactose-induced aging mice; AlCl3/D-galactose-induced neurotoxicity rats In vitro: Aβ-induced SH-SY5Y cells	In vivo: 50 mg/kg/d, i.g., 2 or 8 wk; 100 mg/kg, i.g., 1 h before surgery In vitro: 25 μM, 24 h	In vivo: ↓ SOD, MDA, ↑CAT, GPx, and GSH activity in the hippocampus; ↓ APP and BACE expressions, phosphorylated tau level	Khan et al., 2012; Ghofrani et al., 2015; Zhang et al., 2017; Md et al., 2018; Haider et al., 2020
Quercetin	Friuts (apples, berries, grapes), vegetables (onions, shallots, tomatoes), and medicinal botanicals (Ginkgo biloba, Hypericum perforatum, Sambucus canadensis)	In vivo: 3×Tg mice; Scopolamine-induced mice; LPS-induced mice	In vivo: 25 mg/kg/d, i.p., every 48 h for 3 mon; 40 mg/kg/d, p.o., 16 wk; 0.25 mg/kg/d, i.p., 1 wk; 25 mg/kg/d p.o. + 3 mg/kg/d, i.p., 1 wk	In vivo: ↓ Acetylcholine degradation, Aβ production, AChE, secretase enzymes; ↓ Paired helical filament, Aβ1–40 and Aβ1–42 levels, BACE1-mediated cleavage of APP; ↓ Activation of microglia and astrocytes; ↓ COX-2, NOS-2, IL-1β, TNF-α, IL-6, nitric oxide synthase-2 in the hippocampal subregions and the prefrontal cortex of mice brain; ↓ TLR4/NF-κB pathway, JNK pathway, Nrf2 pathway, PI3K/Akt pathway	Khan et al., 2009, 2018; Wang et al., 2014; Sabogal-Guáqueta et al., 2015; Zaplatic et al., 2019; Olayinka et al., 2022
Morin	Plants of Moraceae family	In vivo: APPswe/PS1dE9 mice;3×Tg-AD mice; Aβ1–42-induced AD rats; ifosfamide-induced rats In vitro: MC65 cells, APPswe cells, SH-SY5Y cells	In vivo: 20 mg/kg/d, i.p., 12 wk; 100 or 200 mg/kg/d, p.o., 2 d; 20 mg/kg/d, i.g., 14 d; 10 mg/kg/d, i.p., 7 d In vitro: 12.5, 25, 50, or 100 μM, 24 h; 1 or 10 μM, 6 h	In vivo: ↑ ADAM10 expression, Aβ-degrading enzyme expression; ↓ BACE1 and PS1 expressions; ↓ GSK-3β activity in the hippocampus; ↓ expressions of MDA, TNF-α, IL-1β, IL-6, NF-κB, and NMDAR subunits 2A and 2B; ↑ expressions of CAT, SOD, GSH, GPx, BDNF, and nAChR in the hippocampus In vitro: ↓ β-secretases and γ-secretases activities; ↓ GSK-3β activity	Gong et al., 2011; Du et al., 2016; Carmona et al., 2020; Çelik et al., 2020; Mohammadi et al., 2021
Kaempferol	Beans, cabbage, grapes, broccoli, strawberries, kale, gooseberries, citrus fruits, grapefruit, apples, dry raspberry, tomatoes; Astragali Complanati Semen (Astragalus complanatus R. Br.), Ginkgo Folium (Ginkgo biloba L.), Propolis (Apis mellifera L.)	In vivo: Ovariectomized rat model of sporadic dementia; LPS-induced mice In vitro: LPS-stimulated microglial BV2 cells	In vivo: 10 mg/kg/d, i.p., 21 d; 50 mg/kg/d, i.p., 7 d In vitro: 30, 50, or 100 μM, 1 h	In vivo: ↑ Levels of SOD and glutathione, ↓ TNF-α and MDA in the hippocampus; ↓ Levels of IL-6, IL-1β, TNF-α, MCP-1, COX-2, and intercellular cell adhesion molecule-1 in the striatum; ↑ Levels of TH and PSD95; ↓ HMGB1/TLR4 pathway In vitro: ↓ NO, PGE2, COX-2, iNOS, TNF-α, IL-1β, ROS productions; ↓ Phagocytosis; ↓ NF-κB activation, ↓ p38 MAPK, JNK and AKT phosphorylation	Park et al., 2011; Kouhestani et al., 2018; Yang et al., 2019; Babaei et al., 2021
Fisetin	A number of commonly fruits and vegetables, such as strawberries, persimmons, apples, grapes, kiwis, peach, onions, lotus root, and cucumber	In vivo: SAMP8 mice; APPswe/PS1dE9 mice; Aβ-induced mice In vitro: BV-2 microglial cells; mouse/rat cortical neuronal cells	In vivo: 25 mg/kg/d, p.o., 7 or 9 mon; 20 mg/kg/d, i.p., 2 wk In vitro: 5 μM, 24 h; 10 μM, 24 h; 10 or 20 μM, 36 h	In vivo: ↓ Accumulation of Aβ and BACE-1; ↓ Level of phosphorylated tau (serine 413) in the hippocampus; ↓ p-IKKβ, p-NF-κB, TNF-α, IL-1β, Iba-1, GFAP; ↓ Cyclin-dependent kinase 5 activator p35 cleavage product (p25) In vitro: ↓ tau aggregation, disaggregation of tau filaments; ↓ Protein carbonylation, chelating activity with iron and copper; ↑ Autophagy pathway by TFEB and Nrf2	Currais et al., 2014, 2018; Kim et al., 2016; Ahmad et al., 2017; Maher, 2020; Xiao et al., 2021
Isoquercitrin	Rosa soulieana Crépin, Eucommia ulmoides Oliv., Crataegus pinnatifida Bge. (Chinese hawberry), Crataegus azarolus L. (azarole), Caragana arborescens Lam. (Siberian peashrub), Arbutus unedo L., and various Allium species	In vivo: Aβ-injected mice In vitro: LPS-induced microglia; Raw 264.7 macrophages cells	In vivo: 10 mg/kg/d, i.g., 4 wk; 1 or 5 mg/kg/d, i.p., 4 d In vitro: 2.5 or 5 μg/mL, 12 h; 6.25, 12.5, or 25 μM, 24 h;	In vivo: ↓ Aβ aggregation, β-secretase, PS-1, PS-2 in the brain; ↓ NO production, lipid peroxidation, iNOS, ROS, COX-2, TNF-α, IL-1β, IL-6, MCP-1, PTGES2; ↓ MAPK signaling pathway In vitro: ↓ NO production, lipid peroxidation, iNOS, ROS, COX-2, TNF-α, IL-1β, IL-6, MCP-1, PTGES2; ↓ MAPK signaling pathway by extracellular signal-regulated kinase, c-Jun N-terminal kinase, and p38 signaling proteins	Kim et al., 2019, 2020, 2022
Astragalin	Medicinal plants (Astragalus mongholicus, Astragalus smicus, Cuscuta chinensis, Cassia alata, Rosa chinensis Jacq.), green tea (Camellia sinensis), beans (Phaseolus vulgaris L.), persimmons (Diospyros kaki), cherry (Prunus serotina), etc	In vivo: AlCl3/D-galactose-induced mouse; Ovariectomy-induced perimenopausal depression murine; LPS-induced mice In vitro: LPS/ATP-induced BV2 cells	In vivo: 5, 10, or 20 mg/kg/d, s.c., 3 wk; 5 or 10 mg/kg/d, i.g., 3 wk; 5 or 20 mg/kg/d, i.p., 4 d In vitro: 10, 20, or 40 μM, 2 h	In vivo: ↓ T-SOD, T-AOC, CAT, GSH-PX, MDA, TNF-α, IL-1β, IL-6 in the brain; ↓ Levels of NO, iNOS, ROS, COX-2, nuclear NF-κB p65, TNF-α, IL-1β, IL-6, MCP-1, PTGES2 in the brain; ↓ Notch/HES-1 and NF-κB signaling axis; ↑ IL-4R/JAK1/STAT6 signaling pathway in the hippocampus In vitro: ↓ TNF-α, IL-1β, IL-6, IL-12, ↑ IL-10, IL-4; ↑ Arg1, Ym1, Fizz1, Klf4 mRNA; ↑ IL-4R/JAK1/STAT6 signaling pathway	Hu et al., 2022; Kim et al., 2022; Yao et al., 2022
Rutin	Various plants, such as passion flower, buckwheat, tea, and apple	In vivo: APPswe/PS1dE9 mice; Tau-P301S mouse; Tracerebroventricular-streptozotocin infused rats In vitro: Primary mouse neurons in the hippocampus; SH-SY5Y cells; Primary microglial cells in the brain; BV-2 microglial cells	In vivo: 100 mg/kg, p.o., 6 wk or 30 d; 25 mg/kg, p.o., 3 wk; In vitro: 4 μM, 48 h; 4 or 8 μM, 24 h; 0.8 or 8 μM, 12 h	In vivo: ↓ Oligomeric Aβ level, pathological tau level; ↓ Microgliosis, astrocytosis, ↑ microglial synapse engulfment; ↑ SOD activity, ↓ levels of IL-1, IL-6, IL-8, COX-2, iNOS, MDA, GSSG, GSH; ↓ NF-kB pathway In vitro: ↓ tau aggregation, tau oligomer-induced cytotoxicity, PP2A level; ↑ Microglial uptake of extracellular tau oligomers; ↓ Psroduction of TNF-α, IL-1β, NO, ROS, GSSG, MDA	Javed et al., 2012; Xu et al., 2014; Yu et al., 2015; Sun et al., 2021
Icariin	Mainly present in Epimedii herba [Epimedium brevicornu Maxim., Epimedium sagittatum (Sieb. et Zucc.) Maxim., Epimedium pubescens Maxim., Epimedium koreanum Nakai]	In vivo: SAMP8 mice; APP/PS1 mice; 5×FAD transgenic mouse; APPV717I transgenic (Tg) mice In vitro: Okadaic acid induced SH-SY5Y cells	In vivo: 50 μmol/kg/d, p.o., 8 d; 60 mg/kg/d, i.g., 22 d; 30 μmol/kg/d, i.g., 6 mon; 100 mg/kg/d, i.g., 10 d; 60 mg/kg/d, p.o., 8 mon In vitro: 2.5 μM, 48 h	In vivo: ↓ Expressions of APP and BACE-1 in hippocampus; ↓ IL-1β, IL-6, TNF-α, ↑ IL-4, IL-10, TGF-β1; ↓ Aβ plaque accumulation, expressions of Iba-1 and iNOS in hippocampus and prefrontal cortex; ↓ PPARγ expression, ↑ sirtuin1; modulated differentiation of CD4+ T cells and release of inflammatory cytokines in plasma and brain In vitro: ↓ p-tau and GSK-3β levels	Urano et al., 2010; Zhang et al., 2014; Wang et al., 2019; Zhu et al., 2019; Wu et al., 2020; Chuang et al., 2021; Li et al., 2022
Mangiferin	Mango and papaya, Coffea arabica, and Anemarrhena rhizome	In vivo: scopolamine-induced mice; SAMP8 mice; APP/PS1 mice In vitro: LPS-induced BV2 microglial cells; Aβ-induced cortical neurons; Organotypic slices of neocortex	In vivo: 40 mg/kg/d, i.p., 14 d; 100 or 200 mg/kg/d, i.g., 60 d; 50 mg/kg/d, i.g., 22 wk In vitro: 100 or 150 μg/mL, 2 h; 1 μM, 30 min	In vivo: ↓ Aβ1–40 and Aβ1–42 expressions in the hippocampus; ↓ tau hyperphosphorylation in the cortex and the hippocampus; ↓ Dopamine, nor-adrenaline and AChE levels in the brain In vitro: ↓ NO, IL-1β, IL-6, TNF-α productions, mRNA and protein expressions of iNOS and COX-2; ↑ Activities of SOD, CAT; ↓ Mitochondrial dysfunction, protein oxidation in cortical neurons and the neocortex; ↓ NF-κB, NLRP3, PI3K/Akt, ERK1/2 signaling pathways signaling pathways	Biradar et al., 2012; Infante-Garcia et al., 2017; Alberdi et al., 2018; Du et al., 2019; Feng et al., 2019; Lei et al., 2021
Anthocyanin	Blueberry, grape, eggplant, cranberries, purple cauliflower, plums, strawberries, etc.	In vivo: LPS-induced mice; D-galactose-induced rats; APP/PSEN1 mice; In vitro: SK-N-SH cells; mouse hippocampal HT22 cells	In vivo: 24 mg/kg/d, i.p., 14 d; 100 mg/kg/d, i.p., 7 wk; 20 mg/kg/d, i.g., 3 mon In vitro: 5, 10, or 20 μM, 1 h	In vivo: ↓ Neurotoxicity; ↓ Expressions of TLR2, TLR4, RAGE, CX3CR1, Iba-1, GFAP, TREM2, TYROBP; ↓ Levels of ROS and lipid peroxidation; ↓ TNF-α, p-NF-κB, total-NF-κB, IL-1β, IL-6, COX-2, iNOS, CD33; ↑ SNAP-23, SNAP-25, CD68, IRF7, TREM2, TYROBP; ↓CD33, p-PI3K/Akt/GSK3β pathway, Nrf2/HO-1 pathway, p-JNK pathway In vitro: ↓ Neurotoxicity	Belkacemi et al., 2015; Rehman et al., 2017; Ali et al., 2018; Khan et al., 2019; Li et al., 2020

AChE: Acetylcholinesterase; AD: Alzheimer’s disease; APP: amyloid precursor protein; Aβ: amyloid-beta; BACE1: β-site APP-cleaving enzyme 1; BDNF: brain-derived neurotrophic factor; CAT: catalase; COX-2: cyclooxygenase-2; CREB: cAMP response element-binding protein; CTF: C-terminal fragment; ERK: extracellular signal regulated kinase; GSH: glutathione; GSH-PX: glutathione peroxidase; GSK-3: glycogen synthase kinase-3; GSSG: glutathione disulfide; HO-1: heme oxygenase-1; i.g.: intragastrically; IL: interleukin; iNOS: inducible nitric oxide synthase; i.p.: intraperitoneal injection; JNK: c-Jun N-terminal kinase; MAPK: mitogen activated protein kinases; MCP-1: monocyte chemotactic protein-1; MDA: malondialdehyde; NF-κB: nuclear factor-κB; nAChR: nicotinic acetylcholine receptor; NMDA: N-methyl-D-aspartate receptor; NO: nitric oxide; Nrf2: nuclear factor (erythroid-derived 2)-like 2; RAGE: receptor for advanced glycoxidation endproducts; PGE2: prostaglandin E2; PI3K: phosphorylated-phosphatidylinositol 3-hydroxy kinase; PPARγ: peroxisome proliferator-activated receptor; p.o.: per os; PS: presenilin; PSD95: postsynaptic density protein 95; PTGES2: prostaglandin E synthase 2; s.c.: subcutaneous injection; SAMP8: senescence-accelerated mouse prone 8; SNAP: synaptosomal-associated protein; SOD: superoxide dismutase; TH: tyrosine hydroxylase; TLR: toll-like receptor; TNF-α: tumor necrosis factor-α; TYROBP: tyrosine kinase binding protein.

Limitations

This paper has some limitations that need to be considered. Considering the diversity and structural complexity of flavonoids, very few studies have focused on anti-neuroinflammation in AD. Only 13 natural flavonoids are discussed in the current review. The specific role of neuroinflammation in the pathogenesis of AD, and the specific mechanisms involved, need to be investigated in greater depth; this research should be highly beneficial for the future development of new drugs.

Conclusion

In this review, we provide an overview of the pathological processes of neuroinflammation in AD. In total, we identified 13 natural flavonoids and summarized their effects and mechanisms against neuroinflammation in AD. As one of the most prevalent chronic diseases in the contemporary world, over 2000 AD-related clinical trials are underway. However, there are very few commercially available drugs available for the treatment of AD at present. Apart from monoclonal antibodies (such as aducanumab and lecanemab) and dietary supplements (neptune krill oil, fish oil, and caprylidene), typical small molecule compounds include donepezil (CAS: 120014-06-4), memantine (CAS: 19982-08-2), and rivastigmine (CAS: 123441-03-2) (https://clinicaltrials.gov). Furthermore, EGb 761 and GV-971 are derived from natural plants and worth mentioning. EGb 761 is a standardized extract of ginkgo leaf in which the main active constituents are flavonoids and ginkgolides (Tomino et al., 2021). Thus, as an abundant resource for medicines, numerous plant-based bioactive compounds, especially flavonoids, have significant potential for drug discovery and need to be investigated further with regard to their potency and bioactivities. We need to gain a better understanding of the structure-activity relationships of flavonoids in order to identify drugs to develop for the treatment of AD. In addition, we need to solve the problem of brain-targeted drug delivery across the BBB if we are to achieve satisfactory therapeutic effects; several studies have focused on this issue as solving this problem would be highly beneficial for the treatment of AD (Ho et al., 2013; Ouyang et al., 2022). It is essential that we gain a better understanding of the pathological mechanisms underlying AD, and investigate the potential of naturally derived flavonoids as potential treatments for the neuroinflammation in AD.