Potential Role of Phenolic Extracts of Mentha in Managing Oxidative Stress and Alzheimer’s Disease

With an increase in the longevity and thus the proportion of the elderly, especially in developed nations, there is a rise in pathological conditions that accompany ageing, such as neurodegenerative disorders. Alzheimer’s disease (AD) is a neurodegenerative disease characterized by progressive cognitive and memory decline. The pathophysiology of the disease is poorly understood, with several factors contributing to its development, such as oxidative stress, neuroinflammation, cholinergic neuronal apoptotic death, and the accumulation of abnormal proteins in the brain. Current medications are only palliative and cannot stop or reverse the progression of the disease. Recent clinical trials of synthetic compounds for the treatment of AD have failed because of their adverse effects or lack of efficacy. Thus, there is impetus behind the search for drugs from natural origins, in addition to the discovery of novel, conventional therapeutics. Mints have been used traditionally for conditions relevant to the central nervous system. Recent studies showed that mint extracts and/or their phenolic constituents have a neuroprotective potential and can target multiple events of AD. In this review, we provide evidence of the potential role of mint extracts and their derivatives as possible sources of treatments in managing AD. Some of the molecular pathways implicated in the development of AD are reviewed, with focus on apoptosis and some redox pathways, pointing to mechanisms that may be modulated for the treatment of AD, and the need for future research invoking knowledge of these pathways is highlighted.

pathways was observed in the brains of individuals with AD (neuroinflammation) [21]. Furthermore, alteration in acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) and the abnormal acetylation of histone were involved in the pathophysiology of AD [1,22]. Histone deacetylases (HDAC) comprise a family of enzymes that catalyze the hydrolysis of acetyl groups from an N-acetyl lysine amino acid on a histone, which allows the histones to wrap the DNA more tightly [23]. Histones are the building blocks of chromatin at which transcriptional regulation occurs. The post-translational modification of histones, such as methylation, phosphorylation, and acetylation, affects transcriptional regulation [23]. Insulin resistance might also be associated with AD pathology, as impaired neurotrophic effects of insulin were reported in aged brains, which reduced the ability of the brain to repair neuronal damage, resulting in synaptic, metabolic, and immune response dysfunctions [15]. Mitochondrial dysfunction also plays an important role in AD pathology and progression [24]. The reduced expression of mitochondrial enzymes and oxidative phosphorylation genes with altered mitochondrial structure, potential, and permeability were observed in AD brains [17]. Increased mitochondrial membrane permeability leads to the excessive production of ROS and leakage of proteins, which initiates a cascade of reactions leading to cell death [17,24]. The accumulation of Aβ plaques and tau tangles was linked to mitochondrial impairment in AD; however, there is still a debate whether they are the cause or a consequence of mitochondrial dysfunction [24].
Although there are a number of controversial hypotheses and factors used to explain the initiation and progression of AD, neuronal death is the common factor. As discussed above, oxidative stress plays a pivotal role in the disease development and can exacerbate key events in AD. Thus, this review highlights the signaling cascade of apoptosis and some cellular antioxidant defense mechanisms and how mint extracts can alter these systems towards a potential AD treatment.

Apoptosis and AD
Neuronal death from apoptosis has been found in the brains of patients suffering from neurodegenerative diseases. Apoptotic cell death is characteristic in neurons and glial cells in AD [25]. Apoptosis is a genetically programmed cell death (PCD). There are two pathways of PCD: the stimulation of death receptors by external ligands (extrinsic pathway) and an internal mitochondrial pathway (intrinsic pathway) ( Figure 1). Both pathways lead to the recruitment and activation of initiator caspases (that transmit the initial apoptotic signals), then effector caspases (responsible for the final phase of cell death) [26]. The morphological characteristics of an apoptotic cell include chromatin condensation, DNA fragmentation, nuclear fragmentation, and membrane blebbing [27].
Caspases (cysteine-aspartic proteases) are a family of protease enzymes that play the major role in apoptosis. Caspases have a cysteine protease activity, cleaving a target protein after an aspartic acid residue, leading to changes in the structure (degradation) and function of the protein. A cascade involving the activation of about nine different caspases is involved in the stimulation of the pro-apoptotic pathway [26]. The caspase cleavage of APP and tau may promote the formation of Aβ and neurofibrillary tangles [28]. Therefore, preventing the activation and execution of apoptosis may provide an effective means of treating AD [28].
The extrinsic pathway is activated when death receptors, such as the tumour necrosis factor receptor (TNFR), are stimulated, then activating the death-inducing signaling complex (DISC). The DISC, in turn, activates initiator caspase-e.g., caspase 8-leading to the activation of effector caspases-e.g., caspase 3 [26].
Caspases (cysteine-aspartic proteases) are a family of protease enzymes that play the major role in apoptosis. Caspases have a cysteine protease activity, cleaving a target protein after an aspartic acid residue, leading to changes in the structure (degradation) and function of the protein. A cascade involving the activation of about nine different caspases is involved in the stimulation of the proapoptotic pathway [26]. The caspase cleavage of APP and tau may promote the formation of Aβ and neurofibrillary tangles [28]. Therefore, preventing the activation and execution of apoptosis may provide an effective means of treating AD [28].
The extrinsic pathway is activated when death receptors, such as the tumour necrosis factor receptor (TNFR), are stimulated, then activating the death-inducing signaling complex (DISC). The DISC, in turn, activates initiator caspase-e.g., caspase 8-leading to the activation of effector caspases-e.g., caspase 3 [26].
The intrinsic pathway can be initiated by apoptotic stimuli, such as DNA damage, chemotherapeutic agents, ROS, the removal of survival factors (such as cytokines and hormones), and ageing. The cascade starts when apoptotic stimuli activate the p53 protein, which activates p21 and the pro-apoptotic members of the Bcl-2 protein family (Bax (Bcl-2-associated X protein) and Bak (Bcl-2 homologous antagonist killer)). Pro-apoptotic proteins form pores in the mitochondria, affecting the outer membrane permeability and inducing the release of cytochrome C from the mitochondria. Cytochrome C binds with apoptotic protease activating factor 1 (Apaf-1) to form an apoptosome. The apoptosomes then recruit and activate the inactive procaspase 9. Once activated, this initiator caspase can then activate effector caspases and trigger a cascade of events that lead to apoptosis [26,27]. This process is still incompletely understood, although it has been studied extensively [27] (Figure 1).
Peroxiredoxins (Prx) are a family of thioredoxin peroxidase enzymes, and their activity is linked to Trx [36]. There are six Prxs in mammalian cells that play important roles in the response to H2O2 [37]. Prxs are upregulated by stress agents such as H2O2. Prxs are able to reduce H2O2 to H2O; meanwhile, the peroxidatic cysteine becomes oxidized to a sulphenic acid. The sulphenic acid then reacts with the cysteine of the adjacent Prx molecule, forming an intermolecular disulphide bond. The resulting Prx disulphides are reduced by Trx using electrons from NADPH [36]. Prx can also catalyze the oxidation of ASK1 in response to peroxide, forming the Prx-ASK1 disulphide complex, resulting in p38 phosphorylation [43]. Altering the expression of Prx1 or Prx2 had different effects; a decrease in the Prx1 expression with an increase in Prx2 inhibits the activation of ASK1/p38 induced by H2O2 [37]. The signaling pathways that regulate the expression of Prx are not clear [44]. The overexpression of Prx1 in PC12 cells and primary neurons attenuated Aβ-induced toxicity and protected dopaminergic neurons from apoptosis [37]. Additionally, the overexpression of Prx2 protected against Aβ-induced toxicity in a transgenic mouse model for AD [37]. The expression levels of Prx1, Prx2, and Prx6 were high in the brains of AD patients, while that of Prx3 was low [45][46][47]. The redox states of Prx2 and Prx6 in the serum and brains of AD patients were more oxidized [37].

Heme Oxygenase-1 (HO-1)
Heme oxygenase (HO) is an enzyme responsible for the catalytic degradation of heme into biliverdin, ferrous ions, and carbon monoxide. There are two forms of HO: HO-1 (an inducible form) and HO-2 (a constitutively expressed form) [48]. HO-1 can be expressed by a variety of stressors, including H2O2. HO-1 has antioxidant, anti-inflammatory, anti-apoptotic, and antiproliferative (inhibition). Compounds in yellow boxes are inhibitors.
The process of apoptosis is regulated through the anti-apoptotic members Bcl-2 and Bcl-xL, which inhibit the mitochondrial pathway and prevent apoptosis. Anti-apoptotic proteins antagonize the effects of the pro-apoptotic proteins and prevent the release of cytochrome C. It was found that the level of both the pro-apoptotic and anti-apoptotic proteins is high in AD brains [29]. The rise in the level of the anti-apoptotic proteins could be a compensatory response in AD to protect the remaining neurons from apoptosis [29]. Inhibiting apoptosis might prevent or treat a wide range of common degenerative disorders, but no successful treatment has been developed to date [26].
The Bcl-2-associated death promoter (Bad) protein is another protein that can indirectly promote apoptosis through the intrinsic mitochondrial pathway [30]. Active Bad heterodimerizes with anti-apoptotic proteins, preventing their action; hence, Bad is considered a pro-apoptotic protein.
The phosphorylation of Bad by protein kinase B (Akt) causes its inactivation. The phosphorylation of Bad induces a conformational change in Bad, which allows binding with the 14-3-3 protein, which sequesters Bad in the cytoplasm and prevents its interaction with Bcl-2 or Bcl-xL. The dephosphorylation of Bad by phosphatases such as Ca 2+ -stimulated phosphatase calcineurin and protein phosphatase 2A (PP2A) drive its activation through its release from sequestration [30].

The Redox System and AD
ROS are generated normally through cell metabolism. At physiological concentrations, ROS perform normal functions, such as the induction of mitogenic response, the regulation of signal transduction, and involvement in protection against infectious agents [31]. Under normal conditions, the production of ROS is balanced by the antioxidant system. Imbalance by the over-production of ROS and/or insufficient antioxidants leads to oxidative stress. Excessive ROS can cause cell membrane damage through lipid peroxidation; the modification of signal and structural proteins, leading to misfolding and aggregation; and the oxidation of RNA/DNA, which interrupts transcription and results in gene mutation [31]. Oxidative stress can change the expression and activities of enzymes interacting with different signaling pathways [16]. As the brain is the organ in which the consumption of oxygen occurs at the highest rate, it is highly susceptible to oxidative stress, leading to neurodegenerative diseases [31]. Impaired natural antioxidant defense systems were identified in AD patients [20]. Glutathione is one of the most prevalent cellular antioxidants and exists in reduced (GSH) and oxidized (GSSG) forms [24]. Glutathione peroxidase catalyzes the oxidation of GSH into GSSG, while glutathione reductase catalyzes the reverse reaction [24]. The reduction in H 2 O 2 is coupled to the oxidation of GSH [20]. Under conditions of oxidative stress and in the brains of AD patients, a decreased GSH/GSSG ratio was detected, which indicated the altered antioxidant ability of the GSH system [20,24]. It was also found that the severity of the disease is proportional to the magnitude of the changes in the GSH redox cycle [24]. Superoxide dismutase (SOD) and catalase (CAT) are other cellular antioxidant enzymes that catalyze the conversion of superoxide into H 2 O 2 and then into water, respectively [24]. The decreased activity of these enzymes was observed in cognitively impaired patients [24]. Altered expression levels of the antioxidant proteins, thioredoxin, peroxiredoxin, and heme oxygenase, were also reported in AD brains [24] and are covered in more detail in this review.
At normal physiological concentrations, ROS are key components of cellular signaling cascades for controlling cell proliferation, survival, and migration. The phosphorylation of Bad via Akt inhibits the pro-apoptotic cascade by making the anti-apoptotic proteins available to exert their action ( Figure 1). PP2A dephosphorylates Akt at threonine 308 and serine 473, thus blocking the phosphatidylinositol 3-kinase (PI3K)/Akt pathway. H 2 O 2 can inactivate PP2A at the cysteine site and therefore activates the Akt pathway, which facilitates cell survival [31].
Apoptosis signal-regulating kinase 1 (ASK1) is a member of the MAPKKK family that is activated by oxidative stress induced by ROS, tumour necrosis factor (TNF) α, lipopolysaccharide, endoplasmic reticulum stress, and calcium influx; it selectively activates the JNK and p38 MAPK pathways [32]. Oxidative stress activates ASK1, which activates JNK and p38 MAPK and decreases the activity of PP2A, which leads to tau hyperphosphorylation, one of the major hallmarks of AD brains [31]. Additionally, the activation of JNK and p38 stimulates β-secretase, causing Aβ accumulation and neuronal cell death [31,35]. The thioredoxin (Trx) family proteins are oxidoreductases that act as cofactors in the activity of other enzymes and facilitate the reduction of several transcription factors and signaling molecules [36]. Mammalian genomes encode two Trx systems: Trx1 and TrxR1 (thioredoxin reductase 1) present in the cytosol and Trx2 and TrxR2 in the mitochondria [37]. Trx1 contains two redox-active cysteine residues in their conserved active centres that can be oxidized to form intramolecular disulphide bonds (Trx-S2) ( Figure 2). The reduction of Trx-S2 is catalyzed by TrxR, with NADPH (nicotinamide adenine dinucleotide phosphate) as the electron donor. The thiol oxidoreductase activity of Trx makes it suitable to reduce protein disulphides and H 2 O 2 , thus functioning as an electron carrier for the catalytic actions of peroxidases and as a protector protein against oxidant-mediated disulphide bond formation [38]. Reduced Trx1 can bind to the N-terminal domain of ASK1, which inhibits its kinase activity, preventing cell apoptosis [32]. Thus, Trx1 can protect against oxidative stress and prevent cell apoptosis by removing H 2 O 2 . ROS, TNFα, and Aβ oxidize Trx, resulting in the dissociation of the Trx-ASK1 complex and the activation of ASK1, which induces apoptosis via the JNK and p38 MAPK pathways [37,39]. ASK1 also induces a mitochondria-dependent apoptotic pathway, but this pathway is not well understood [39]. The treatment of SH-SY5Y cells with Aβ resulted in the oxidation of Trx1 [37]. The over-expression of Trx1 in SH-SY5Y cells and in rat primary hippocampal neurons resulted in a protective effect against Aβ-induced cell death [37]. It was found that Trx1 protein levels were decreased in patients during amnestic mild cognitive impairment and in several regions of AD-affected brains [40][41][42]; however, in another study, there was no significant difference in the Trx levels between the brains of control and AD patients, although the TrxR1 activity was enhanced in the AD patients [37].
Apoptosis signal-regulating kinase 1 (ASK1) is a member of the MAPKKK family that is activated by oxidative stress induced by ROS, tumour necrosis factor (TNF) , lipopolysaccharide, endoplasmic reticulum stress, and calcium influx; it selectively activates the JNK and p38 MAPK pathways [32]. Oxidative stress activates ASK1, which activates JNK and p38 MAPK and decreases the activity of PP2A, which leads to tau hyperphosphorylation, one of the major hallmarks of AD brains [31]. Additionally, the activation of JNK and p38 stimulates -secretase, causing Aβ accumulation and neuronal cell death [31,35].

Thioredoxin and Peroxiredoxin
The thioredoxin (Trx) family proteins are oxidoreductases that act as cofactors in the activity of other enzymes and facilitate the reduction of several transcription factors and signaling molecules [36]. Mammalian genomes encode two Trx systems: Trx1 and TrxR1 (thioredoxin reductase 1) present in the cytosol and Trx2 and TrxR2 in the mitochondria [37]. Trx1 contains two redox-active cysteine residues in their conserved active centres that can be oxidized to form intramolecular disulphide bonds (Trx-S2) ( Figure 2). The reduction of Trx-S2 is catalyzed by TrxR, with NADPH (nicotinamide adenine dinucleotide phosphate) as the electron donor. The thiol oxidoreductase activity of Trx makes it suitable to reduce protein disulphides and H2O2, thus functioning as an electron carrier for the catalytic actions of peroxidases and as a protector protein against oxidantmediated disulphide bond formation [38]. Reduced Trx1 can bind to the N-terminal domain of ASK1, which inhibits its kinase activity, preventing cell apoptosis [32]. Thus, Trx1 can protect against oxidative stress and prevent cell apoptosis by removing H2O2. ROS, TNFα, and Aβ oxidize Trx, resulting in the dissociation of the Trx-ASK1 complex and the activation of ASK1, which induces apoptosis via the JNK and p38 MAPK pathways [37,39]. ASK1 also induces a mitochondria-dependent apoptotic pathway, but this pathway is not well understood [39]. The treatment of SH-SY5Y cells with Aβ resulted in the oxidation of Trx1 [37]. The over-expression of Trx1 in SH-SY5Y cells and in rat primary hippocampal neurons resulted in a protective effect against Aβ-induced cell death [37]. It was found that Trx1 protein levels were decreased in patients during amnestic mild cognitive impairment and in several regions of AD Peroxiredoxins (Prx) are a family of thioredoxin peroxidase enzymes, and their activity is linked to Trx [36]. There are six Prxs in mammalian cells that play important roles in the response to H2O2 [37]. Prxs are upregulated by stress agents such as H2O2. Prxs are able to reduce H2O2 to H2O; meanwhile, the peroxidatic cysteine becomes oxidized to a sulphenic acid. The sulphenic acid then reacts with the cysteine of the adjacent Prx molecule, forming an intermolecular disulphide bond. The resulting Prx disulphides are reduced by Trx using electrons from NADPH [36]. Prx can also catalyze the oxidation of ASK1 in response to peroxide, forming the Prx-ASK1 disulphide complex, resulting in p38 phosphorylation [43]. Altering the expression of Prx1 or Prx2 had different effects; a decrease in the Prx1 expression with an increase in Prx2 inhibits the activation of ASK1/p38 induced Peroxiredoxins (Prx) are a family of thioredoxin peroxidase enzymes, and their activity is linked to Trx [36]. There are six Prxs in mammalian cells that play important roles in the response to H 2 O 2 [37]. Prxs are upregulated by stress agents such as H 2 O 2 . Prxs are able to reduce H 2 O 2 to H 2 O; meanwhile, the peroxidatic cysteine becomes oxidized to a sulphenic acid. The sulphenic acid then reacts with the cysteine of the adjacent Prx molecule, forming an intermolecular disulphide bond. The resulting Prx disulphides are reduced by Trx using electrons from NADPH [36]. Prx can also catalyze the oxidation of ASK1 in response to peroxide, forming the Prx-ASK1 disulphide complex, resulting in p38 phosphorylation [43]. Altering the expression of Prx1 or Prx2 had different effects; a decrease in the Prx1 expression with an increase in Prx2 inhibits the activation of ASK1/p38 induced by H 2 O 2 [37]. The signaling pathways that regulate the expression of Prx are not clear [44]. The overexpression of Prx1 in PC12 cells and primary neurons attenuated Aβ-induced toxicity and protected dopaminergic neurons from apoptosis [37]. Additionally, the overexpression of Prx2 protected against Aβ-induced toxicity in a transgenic mouse model for AD [37]. The expression levels of Prx1, Prx2, and Prx6 were high in the brains of AD patients, while that of Prx3 was low [45][46][47]. The redox states of Prx2 and Prx6 in the serum and brains of AD patients were more oxidized [37].

Heme Oxygenase-1 (HO-1)
Heme oxygenase (HO) is an enzyme responsible for the catalytic degradation of heme into biliverdin, ferrous ions, and carbon monoxide. There are two forms of HO: HO-1 (an inducible form) and HO-2 (a constitutively expressed form) [48]. HO-1 can be expressed by a variety of stressors, including H 2 O 2 . HO-1 has antioxidant, anti-inflammatory, anti-apoptotic, and antiproliferative properties. Non-toxic phytochemicals (e.g., curcumin), polyphenols, and drugs stimulate the expression of HO-1 through the three major MAPK pathways: ERK, JNK, and p38 MAPK [48] (Figure 3). HO-1 can also be upregulated through the PI3K/Akt pathway [49]. HO-1 is highly inducible under oxidative stress through nuclear factor (erythroid-derived 2)-like 2 (Nrf2). Nrf2 is found in the cytoplasm bound to its inhibitor kelch-like ECH-associated protein-1 (Keap-1), which leads to its ubiquitination and degradation under normal conditions [48]. Oxidative stress stimulates the dissociation of Nrf2 from Keap-1 and its translocation into the nucleus, where it mediates HO-1 gene transcription [50]. Preconditioning SH-SY5Y cells with a low concentration of H 2 O 2 before inducing cell injury with a higher concentration protected the cells against this injury through the up-regulation of HO-1 via the PI3K/Akt pathway [49]. The overexpression of HO-1 with an increased serine phosphorylation on the HO-1 protein was found in AD, suggesting that it may be subjected to oxidative damage and undergo post-translational modifications in brain tissue [50].  As AD pathology occurs years before its diagnosis, the process of apoptosis might be different from the classical one [29]. Although the apoptotic and redox pathways have been identified to be involved in the events of AD development and progression, little is known about the predominant apoptosis-regulating factors and oxidative stress, along with the antioxidant defence mechanism that underlies the process in AD brains. The upstream and downstream kinases seem to play a pivotal role in the signaling cascades. The elucidation of the mechanistic basis for the protection of the cells Keap-1 (kelch-like ECH-associated protein-1); Nrf2 (nuclear factor (erythroid-derived 2)-like 2); PI3K (phosphatidylinositol 3-kinase); ROS (reactive oxygen species); → (activation); Antioxidants 2020, 9, x FOR PEER REVIEW Figure 2. A simplified diagram of the regulation of ASK1-mediated apoptosis by thior and peroxiredoxin (Prx). A (amyloid ); ASK1 (apoptosis signal-regulating kinase 1); adenine dinucleotide); Fas (apoptotic receptor); JNK (c-Jun NH2-terminal kinases); M phosphatases); NADP (nicotinamide adenine dinucleotide phosphate); P (protein); T necrosis factor ); ROS (reactive oxygen species); TrxR (thioredoxin reductase); ? (me pathway is still unexplored); (activation); ⟜ (inhibition). Compounds in yellow inhibitors.
Peroxiredoxins (Prx) are a family of thioredoxin peroxidase enzymes, and their to Trx [36]. There are six Prxs in mammalian cells that play important roles in the r [37]. Prxs are upregulated by stress agents such as H2O2. Prxs are able to reduc meanwhile, the peroxidatic cysteine becomes oxidized to a sulphenic acid. The sulp reacts with the cysteine of the adjacent Prx molecule, forming an intermolecular d The resulting Prx disulphides are reduced by Trx using electrons from NADPH [3 (inhibition). Compounds in yellow boxes are inhibitors. As AD pathology occurs years before its diagnosis, the process of apoptosis might be different from the classical one [29]. Although the apoptotic and redox pathways have been identified to be involved in the events of AD development and progression, little is known about the predominant apoptosis-regulating factors and oxidative stress, along with the antioxidant defence mechanism that underlies the process in AD brains. The upstream and downstream kinases seem to play a pivotal role in the signaling cascades. The elucidation of the mechanistic basis for the protection of the cells from oxidative damage and suppressing cell death or enhancing cell survival might be useful aspects in the treatment of the disease, and, therefore, are worth further investigation.

Treatment of AD
No pharmacotherapeutics have yet been able to cure patients with AD [2]. Memory deficits in AD patients results from a defect in the cholinergic system [51]; hence, an important approach to treat AD is to enhance the levels of the neurotransmitter acetylcholine (ACh) in the brain. This can be achieved by the inhibition of cholinesterase (ChE), the enzyme responsible for the hydrolysis of ACh [52]. Approved medications are three AChE inhibitors-rivastigmine, galantamine, and donepezil-and one NMDA (glutamate) receptor antagonist-memantine [2]. These approved medications are only symptomatic and palliative treatments [1]. Many compounds have entered clinical trials in the last few years, with none being efficacious, and in some cases treatments have resulted in the further deterioration of cognitive function [15]. The failure of these trials could be due to the heterogeneous nature, together with the incomplete understanding of the pathophysiology of AD [53]. It was recently suggested to combine a number of therapeutic agents that target different pathogenic mechanisms, but this approach faces some challenges (e.g., the proper selection of doses and drug interactions) [53].
Medicinal plants contain a variety of secondary metabolites that provide a promising source for improving cognitive performance and the treatment of neurodegenerative and age-related diseases [54,55]. Traditional uses of medicinal plants for improving memory have led to drug development. Galanthus nivalis L. (Amaryllidaceae) has been used traditionally in Bulgaria and Turkey for the treatment of neurological disorders [56]. This traditional use was the key for performing further studies on this plant, resulting in the isolation of the alkaloid, galantamine [57]. Researchers found that galantamine is an acetylcholinesterase inhibitor, and it was approved in 2004 by the Food and Drug Administration (FDA), USA, as a medication for AD [58], which emphasizes the importance of ethnopharmacology in drug discovery.
Moreover, the adverse effects of synthetic drugs can lead to their discontinuation. For example, tacrine (brand name "Cognex") was the first centrally acting AChE inhibitor (approved for the treatment of AD in 1993), but was discontinued in 2013 [59] because of its hepatotoxic effect [60].

History and Uses of Mentha
Mentha (commonly known as mint) is a genus in the Lamiaceae, with 18 species and 11 hybrids [61] distributed in diverse environments mainly in temperate and sub-temperate regions across the world [62]. Mentha has a long history of cultivation and is widely known for its culinary, medicinal, and aromatherapeutic properties. Since ancient Egyptian times, peppermint has been used to treat indigestion. The ancient Greeks and Romans also used it for soothing the stomach [63]. In the Middle Ages, peppermint was used as a tooth polisher and to keep rats and mice out of storerooms because of its strong smell. Peppermint then became popular in the eighteenth century in Western Europe as a folk remedy for nausea, vomiting, morning sickness, menstrual disorders, and respiratory infections [64]. In 1721, it was listed in the London Pharmacopoeia as a remedy for treating a wide range of ailments including sores, venereal disease, colds, and headaches [64].
Mint was used in Asia and the Mediterranean region for the treatment of gastric and intestinal colic and spasms [65]. Different Mentha taxa have several ethnomedicinal uses (Table S1). For example, M. pulegium was used in Greece and the Mediterranean region for the treatment of dizziness, nausea, Antioxidants 2020, 9, 631 9 of 24 sea sickness, headache, migraine, intestinal pains or inflammation, and toothache. It was also used as an anti-convulsive, anesthetic, sedative, and spasmolytic [66].
Mints are edible plants known for their refreshing taste and aroma. M. pulegium was commonly used in Portugal as a spice in traditional food preparations [67]. M. spicata and M. × piperita are used as teas in the Mediterranean region, while their dry or fresh leaves are used as a flavor enhancer [68,69].
Mints have a history of uses, implicating activity within the central nervous system (CNS) in traditional medicine. In South Africa, the dried leaves of M. aquatica are burned and the smoke inhaled for the treatment of mental illnesses and for protection against and removal of curses and evil spirits, which resemble symptoms of depression [70]. In Mediterranean countries, M. arvensis and M. × piperita are used to treat neuralgia [66,71], M. pulegium and M. suaveolens are used as anticonvulsives, and M. longifolia, M. spicata and M. × villosa-nervata are used as sedatives [66].
The essential oils of mints are commercially important for the pharmaceutical, food, and cosmetic industries. They are often used as flavoring agents in toothpaste, candy canes, chewing gums and beverages [73]. They are commonly used for their cooling, invigorating, and anesthetic qualities. They have been described as being effective in treating a wide range of gastrointestinal tract complaints [73]. Mentha EO have been reported to have anti-inflammatory, hepatoprotective, hypotensive, vasorelaxant, antioxidant, analgesic, antibacterial, antifungal, antiviral, cytotoxic, and insecticidal properties [73].
The phenolic composition of mint extracts has recently gained the attention of researchers, especially after finding that their bioactivities are highly associated with their phenolic content [69,74]. It is believed that the phenolic compounds are the main active ingredients of, for example, mint teas, because they are prepared uncovered, so the volatile substances are lost in the process [75]. Phenolic acids and flavonoids are the main phenolic classes identified in mints [73]. Rosmarinic acid is a characteristic compound of Mentha, being present in almost all species [73]. The main phenolic compounds reported in mints are luteolin, naringenin, apigenin, hesperitin, eriodictyol, and their glycosides, as well as rosmarinic, caffeic, chlorogenic, and salvianolic acids and their derivatives [73] (Table S2, Figure S2).
It has been suggested that 75% of the phenolic compounds present in a mint infusion may be responsible for its biological activities [76]. A strong correlation between the total phenolic content (TPC) of mints and their antioxidant capacity has been reported [69,74,77], indicating that the biophenols may be responsible for the antioxidant activity. Conforti et al. [78] and [79] found that a hydroalcoholic extract of M. aquatica which showed antiproliferative activity in breast cancer cells in vitro and free radical scavenging and antioxidant activities contained high levels of phenolics. The flavonoids (apigenin and luteolin) and their glucosides (rutinoside and glucuronide) isolated from M. longifolia subsp. longifolia showed antimutagenic activity in vitro [80,81]. Peppermint teas that showed antichlamydial activity had a high content of luteolin and apigenin glycosides [82]. Water

Bioactivities of Mentha Extracts Related to CNS, Oxidative Stress and AD
EO of M. suaveolens and M. pulegium may have an effect on the CNS showing AChE inhibitory activity [67,84]. M. aquatica EO have an affinity for GABA-benzodiazepine receptors, indicating sedative or anticonvulsive activities [85]. Moreover, the EO of M. × piperita were reported to enhance memory in healthy participants [86].
Dietary flavonoids and phenolic acids have been reported to protect neurons from injury and promote memory, learning, and cognitive function, indicating a neuroprotective effect [87]. For example, the flavonoid quercetin, which is generally found in fruits and vegetables, has a neuroprotective activity against agent-induced toxicity and enhances the resistance of neurons to oxidative stress and excitotoxicity through a modulation in the mechanisms of cell death in vitro [88]. The flavonoids gallocatechin gallate and theaflavin inhibit Aβ aggregation, and the stilbenes, resveratrol and its glucoside, could suppress the aggregation [89]. Quercetin is also reported to ameliorate AD pathology and protect cognitive and emotional function in aged triple-transgenic AD-model mice [88].
Extracts of different Mentha taxa containing a number of flavonoids and/or phenolic acids showed a wide range of activities relevant to AD (Figure 4). The extracts of 19 Mentha taxa rich in phenolic compounds showed antioxidant and HDAC inhibition activities, which were strongly associated with their biophenol content [74]. The extracts also exhibited AChE and BuChE inhibition activities [74]. The ethanolic extracts from M. longifolia, M. pulegium, and M. × piperita and their phenolic constituents, rosmarinic, caffeic, chlorogenic, and ferulic acids, showed antioxidant and AChE inhibitory activities, with rosmarinic acid highly likely to contribute to the activities of the extracts [54]. The methanolic extracts of the above species and M. aquatica and M. suaveolens showed antioxidant and lipid peroxidation inhibition activities [73], as well as a monoamine oxidase-A (MAO-A) inhibition activity, indicating an antidepressant effect [90]. M. piperita 70% methanol extract containing flavones and phenolic acids inhibited β-secretase (the enzyme involved in the formation of Aβ), glycogen synthase kinase (GSK-3β), and casein kinase 1δ (CK-1δ) (the enzymes involved in the pathological accumulation of tau) [91]. The methanolic extracts of M. pulegium, M. rotundifolia, M. villosa, M. arvensis, M. aquatica, and M. piperita showed antioxidant and metal ion chelating ability [92]. The ethanolic extract and water infusion of M. longifolia var. calliantha showed antioxidant and AChE inhibitory activities and moderate metal chelating activities [93]. The methanolic extracts of M. aquatica, M. longifolia, and M. × piperita also showed an affinity to the GABA-benzodiazepine receptor, indicating an anxiolytic effect [90]. Extracts rich in phenolic compounds of M. australis, M. diemenica, M. spicata var. crispa, M. × piperita, M. × piperita var. officinalis, and M. requienii inhibited β-secretase and Aβ aggregation in vitro [94]. M. aquatica and M. × piperita had a neuroprotective effect against the H 2 O 2 -induced toxicity in pheochromocytoma (PC12 cells) [90]. M. suaveolens aqueous leaf extract reduced the generation of intracellular ROS in H 2 O 2 -induced damage in HaCaT keratinocytes, decreased the levels of H 2 O 2 -induced apoptotic genes (initiator caspase (caspase-9), effector caspase (caspase-3), and poly (ADP-ribose) polymerase) and induced the expression of HO-1 through the translocation of Nrf2 upon H 2 O 2 exposure [95]. M. diemenica and M. requienii extracts lowered the caspase-3/7 activity in H 2 O 2 -induced damage in SH-SY5Y cells, indicating a neuroprotective effect [94]. Pretreating SH-SY5Y cells with M. diemenica extracts resulted in a downregulation of the expression of the pro-apoptotic protein, Bax, and the upregulation of the anti-apoptotic protein, Bcl-xL, and the antioxidant genes, Prx and HO-1, indicating its potential ability to suppress apoptosis and protect against oxidative stress [94].
Freeze-dried extracts of M. piperita were reported to have protective effects in mice against Aβ formation, ageing-induced stress, amnesia, and neurodegeneration [96]. The extract was also reported to improve acquisition and retention in both exteroceptive and interoceptive behavioral models, and to reverse amnesia and inhibit decreases in ACh levels in the brains of ageing and scopolamine-induced amnesic mice [96]. The oral administration of M. piperita and M. arvensis aqueous extracts enhanced the activity of antioxidant defenses (GSH, SOD, or CAT) in the brain tissues of gamma-irradiationand haloperidol-induced oxidative stress in mice, respectively [97,98]. M. spicata extract with 5% rosmarinic acid was reported to have beneficial effects on learning and memory and the brain tissue markers of oxidation [99]. A dried aqueous M. spicata extract containing high levels of rosmarinic acid (≥14.5%) improved working memory and cognitive performance in individuals (50-70 years of age) with age-associated memory impairment in a randomized, double-blind, placebo-controlled study [100].
Antioxidants 2020, 9, x FOR PEER REVIEW 11 of 26 decreased the levels of H2O2-induced apoptotic genes (initiator caspase (caspase-9), effector caspase (caspase-3), and poly (ADP-ribose) polymerase) and induced the expression of HO-1 through the translocation of Nrf2 upon H2O2 exposure [95]. M. diemenica and M. requienii extracts lowered the caspase-3/7 activity in H2O2-induced damage in SH-SY5Y cells, indicating a neuroprotective effect [94]. Pretreating SH-SY5Y cells with M. diemenica extracts resulted in a downregulation of the expression of the pro-apoptotic protein, Bax, and the upregulation of the anti-apoptotic protein, Bcl-xL, and the antioxidant genes, Prx and HO-1, indicating its potential ability to suppress apoptosis and protect against oxidative stress [94]. Freeze-dried extracts of M. piperita were reported to have protective effects in mice against Aβ formation, ageing-induced stress, amnesia, and neurodegeneration [96]. The extract was also reported to improve acquisition and retention in both exteroceptive and interoceptive behavioral models, and to reverse amnesia and inhibit decreases in ACh levels in the brains of ageing and scopolamine-induced amnesic mice [96]. The oral administration of M. piperita and M. arvensis aqueous extracts enhanced the activity of antioxidant defenses (GSH, SOD, or CAT) in the brain tissues of gamma-irradiation-and haloperidol-induced oxidative stress in mice, respectively [97,98]. M. spicata extract with 5% rosmarinic acid was reported to have beneficial effects on learning and memory and the brain tissue markers of oxidation [99]. A dried aqueous M. spicata extract containing high levels of rosmarinic acid (≥14.5%) improved working memory and cognitive performance in individuals (50-70 years of age) with age-associated memory impairment in a randomized, doubleblind, placebo-controlled study [100].
Many phenolic compounds identified in Mentha taxa (Tables 1 and Figure 5) have been reported to have AD-related bioactivities. For example, linarin (acacetin-7-O-β-D-rutinoside) isolated from a flower extract of M. arvensis showed AChE inhibitory effects [101]. It also had neuroprotective effects  [101]. It also had neuroprotective effects against Aβ25-35-induced neurotoxicity in cultured rat PC12 cells through the activation of the PI3K/Akt pathway [102]. The ethanolic extract of the aerial parts of M. aquatica and the isolated flavanone, naringenin, showed affinity for the GABA-benzodiazepine receptor, suggesting an anti-convulsive activity or a sedative effect [85]. Naringenin isolated from a 70% ethanol extract of M. aquatica leaves inhibited MAO-A and MAO-B, indicating its antidepressant and anti-Parkinson's disease effects, respectively [70]. Naringenin can pass the blood-brain barrier, so it can have an effect on the CNS [85]. Caffeic acid was reported to protect mice from the neuronal loss and memory impairment induced by focal cerebral ischaemia, which includes a complex cascade of events such as inflammation and oxidative stress [103]. Salvianolic acid A was reported to have a neuroprotective effect against acute ischemic stroke in mice, as it increased their survival rate, enhanced their activity levels, and improved the severity of brain infarction and apoptosis [104]. Rosmarinic acid showed a neuroprotective effect against H 2 O 2 -induced neurotoxicity in SH-SY5Y cells [105] and against Aβ-induced damage in PC12 cells [106]. Hamaguchi et al. [107] found that rosmarinic acid prevented the development of AD pathology by decreasing the Aβ deposition in the brains of AD model transgenic mice (Tg2576). Danshensu (salvianic acid A) was reported to improve cognitive performance in an animal model of AD [108] and protect PC12 cells against neurotoxicity [109]. Danshensu also increased the expression of HO-1 and suppressed 6-hydroxydopamine-induced oxidative damage via the PI3K/Akt/Nrf2 signaling pathway [110]. Rutin (quercetin-3-O-rutinoside) was reported to have antioxidant capacity, and could upregulate the expression of antioxidant enzymes and enhance memory in transgenic mice as well as inhibit Aβ aggregation in vitro [111]. Several studies showed a neuroprotective effect of rutin in vitro and in vivo [111]. The administration of hesperidin improved learning and memory function by enhancing the recognition index in a transgenic mouse model [112]. It reduced the levels of Aβ 1-40 and βand γ-secretases in the hippocampus and the cortex of the brain of rats and modulated apoptosis via upregulating Bcl2 and downregulating Bax [112]. Chlorogenic acid can pass the blood-brain barrier and was reported to exhibit a protective effect on neuronal cells against oxidative damage and inhibit apoptotic regulators such as caspase-3 [113]. It inhibited AChE and BuChE in rat brain homogenates in vitro [113]. Chlorogenic acid was also found to prevent the overexpression of Aβ in human neuroblastoma cells and improve spatial learning and memory in an aging-mice model [113]. In addition, neoponcirin (isosakuranetin-7-O-rutinoside) was reported to possess a free radical scavenging activity and a neuroprotective effect against H 2 O 2 -induced toxicity through the activation of antioxidant defense enzymes and the inhibition of apoptosis [114]. Apigenin can cross the blood-brain barrier and was found to protect neurons against Aβ and inflammatory stimuli [115]. The treatment of a rat model of spinal cord injury with apigenin recovered neuronal function and showed anti-apoptotic effects, representing in a decrease in the expression levels of pro-apoptotic markers (Bax and caspase 3) and an elevation in the expression level of the anti-apoptotic protein Bcl-2 [116]. Luteolin was also reported to have a neuroprotective effect against apoptosis induced by 6-hydroxy-dopamine in rat PC12 cells [117].   Over the last 20 years, the search for pharmacologically relevant bioactivity in plants has often incorporated antioxidant testing as a way to preliminarily screen plant extracts and fractions. This approach has been based on the fact that oxidative stress is associated with many diseases. However, when dealing with a complex disease such as AD, which has multiple causes, relying on screening with antioxidant tests may lead to the rejection of extracts with other activities which may be relevant to AD. This idea was supported when a principal component analysis (PCA) was carried out to represent the broad spectrum of bioactivities exhibited by mint extracts [74]. Extracts of M. australis and M. diemenica showed a moderate antioxidant capacity [74] but a strong β-secretase inhibition activity [94]. The M. requienii extract had a high antioxidant capacity and Aβ-aggregation and caspase inhibition activities, but low AChE and moderate β-secretase inhibition activities, while M. × piperita var. officinalis had almost opposite responses [74,94]. The M. spicata var. crispa extract showed a high antioxidant capacity and histone deacetylase inhibition; however, it could not inhibit caspase to protect the cells from apoptotic death [74,94].

Variation in Phenolic Composition
When investigating the potential of plant extracts for treatment of diseases, it is important to also understand the factors that may affect the levels of bioactive compounds in the plant. The TPC of different extracts of several Mentha species and their hybrids showed a wide range of variability from 0.124 to 246.7 mg gallic acid equivalent (GAE)/g (Table S3), with an average of approximately 100 mg GAE/g [118,119]. For instance, the TPC of different extracts of M. spicata from India was 2.9-3.4 mg GAE/g [120] and 2.0-11.8 mg GAE/g [121]. Yi and Wetzstein [122] reported higher contents for M. × piperita and M. spicata cultivated in a greenhouse (76 and 63 mg GAE/g DW, respectively) than those grown in the field (39 and 55 mg GAE/g DW, respectively).
Several factors could influence the variation in TPC. While variation between species is expected, variation due to sample preparation was also observed. Different extraction techniques and conditions will yield different phenolic contents qualitatively and quantitatively for the same samples (Table 2) [123,124]. Bimakr et al. [125] found that different bioactive flavonoid compounds were obtained from M. spicata leaves by using conventional Soxhlet extraction (CSE) and supercritical carbon dioxide (SC-CO 2 ) extraction at different extraction parameters such as temperature, pressure, and dynamic extraction time. CSE showed a higher crude extract yield than SC-CO 2 extraction; however, the SC-CO 2 extraction contained more main flavonoid compounds at a higher concentration than the CSE. Extraction using an accelerated solvent extractor, by hydrodistillation, or by heating at 60 • C gave a high TPC (115-247 mg GAE/g) [77,83,119]. Additionally, the polarity of extracting solvents can influence the solubility of chemical constituents and alter the analytical estimation of TPC from the same sample [126]. A 70% acetone extract of M. arvensis contained approximately 16 times the TPC of other solvents (70% ethanol, 70% methanol or water) [126]. Moreover, the drying technique may influence the estimation of TPC. Arslan et al. [127] found that drying M. × piperita in an oven at 50 • C or under direct sunlight at 20-30 • C or in a microwave oven resulted in approximately 53% lower, 16% higher, and 45% higher estimates, respectively, in the TPC as compared to a fresh sample. The increase in the estimated TPC could be due to the release of phenolic compounds from the matrix during the process, and drying might have facilitated the release of the bound compounds from the breakdown of cellular components, but the high temperature of oven drying might cause the thermal degradation of the phenolic compounds, which lowers the TPC of the oven-dried samples [127]. On the other hand, Riachi and De Maria [103] reported that drying peppermint at 60 • C in an oven led to a preserved total phenolic composition as compared to drying in shade, sunlight, or at 75 • C. The microwave oven drying at 900 W of M. piperita leaves yielded a higher TPC (406.7 mg GAE/100 g fresh weight (FW)) than drying using a hot-air oven at 80 • C (183.5 mg GAE/100 g FW), while the TPC of the fresh sample was 229 mg GAE/100 g FW [128]. The methanol extracts of M. longifolia produced higher TPCs by ultrasonic extraction than the 70% aqueous ethanol extracts, while the later gave better TPCs by Soxhlet [124].
Another factor that could explain the variations observed in the phytochemical content of plants is the different edaphoclimatic conditions to which they were subjected, which can influence the plant secondary metabolism [103]. Plants grown in temperate areas with extended hours of daylight contained a high content of phenolic acids [103]. As an example, M. spicata exposed to 16 h of daylight contained a higher content of rosmarinic acid than that exposed to 12 h [130]. Additionally, the main active constituent differed under different amounts of daylight in M. × piperita [103]. Peppermint grown under shorter photoperiods (8 h) had gardenin B as the main constituent, while that grown under longer photoperiods (16 h) had pebrellin [131].
Thus, the variables outlined above are important when considering the bioactivity of specific and batch to batch variations of plant-sourced compounds. The data reviewed here exemplifies one of the major challenges when considering multi-component plant extracts as therapeutics. Abundant evidence exists that mixed bioactives have synergistic and in some cases antagonistic effects, among the various components. Rigor in the characterization of plant extracts is essential, not only in the discovery phase, but importantly also (where the efficacy and absence of off-target effects are demonstrated) in the production of putative therapeutics.

Toxicity of Mentha Extracts
Despite the historical use of mints, toxicological studies on mint extracts are scarce. Treating neuroblastoma cells for 24 h with up to 1280 µg/mL 50% aqueous methanol extracts of M. australis, M. diemenica, M. spicata, M. × piperita, M. × piperita var. officinalis, or M. requienii had no toxic effect on the cell viability [94]. Exposure to 0.54, 5.4, and 53.8 mg/mL aqueous extract of M. × piperita for 24 h reduced the viability of Tetrahymena pyriformi cells in a dose-dependent manner [132]. The acute oral administration of 1000 mg/kg aqueous extract of M. spicata did not cause mortality, neurological alterations in rats, or changes in their normal behavior [133]. A histological analysis revealed no toxic effects at the hepatic or renal levels [133]. On the other hand, the chronic administration of 20 g/L M. piperita tea to rats (drinking at all times) during 30 days did not show signs of nephrotoxicity but induced hepatic damage and hormonal changes; however, 20 or 40 g/L M. spicata tea changed both the kidney and liver functions as well as the hormonal levels [134][135][136]. Considering the proportion of a human body weight to that of a rat, Akdogan, Kilinç, Oncu, Karaoz and Delibas [134] precluded a risk of nephrotoxicity of M. spicata tea. Neves [132] recommended a moderate human intake of M. × piperita, and Caro, Rivera, Ocampo, Franco and Salas [133] indicated the safety of M. spicata. Although the extensive use of Mentha extracts as herbal medicines implies their safety, the further evaluation of their potential adverse effects is essential.

Conclusions and Perspectives
The lack of a proper treatment or a prophylaxis for AD are drivers for further discovery and innovation, especially with the increasing number of patients suffering with AD. The exact etiology of AD is complicated and still poorly understood, with multiple factors contributing to neuronal cell death, making finding the way to a therapeutic approach challenging. With this in mind, it is worthwhile addressing various factors while searching for a potential treatment. The failure of recent clinical trials for synthetic compounds due to toxicity or worsening cognition suggests the need for an alternative approach. The ethnomedicinal value of Mentha points to its potential role in neurological conditions. The phenolic extracts and phenolic compounds present in Mentha species showed beneficial effects on the CNS. Many Mentha taxa had bioactivities linked to the different pathophysiological events of AD. They have a strong antioxidant capacity. They were able to inhibit the enzymes AChE, BuChE, HDAC, and β-secretase, meaning that they can modulate acetylcholine degradation, histone acetylation, and the formation of abnormal proteins. They also inhibited Aβ-accumulation and interfered with the pathological accumulation of tau through inhibiting GSK-3β and CK-1δ. Extracts from some taxa reduced ROS generation, exerted a neuroprotective effect against H 2 O 2 -induced oxidative stress, and enhanced the antioxidant enzymes in various cell lines. They could protect cells from apoptotic damage through modulating apoptotic regulators: suppressing caspases and pro-apoptotic proteins and enhancing anti-apoptotic proteins. Moreover, extracts from some taxa improved learning and cognition in mice and AD patients.
However, minimal research has been conducted that may indicate promising pathways towards developing either extracts from Mentha or in the discovery of specific compounds that may be refined and improved as bases for therapeutics for AD. The further explanation of the underlying molecular mechanism is still required. The determination of the pharmacokinetics, pharmacodynamics, bioavailability, and route of administration of mint extracts might provide further insight into the use of mint in food for brain health. The toxicity and ability of mint extracts to cross the blood-brain barrier to reach the neurons in the brain is an area that needs to be explored. The safety and efficacy still need to be assessed. Drinking mint tea or adding mint in food preparations is useful for the stomach but might be useful for brain health as well. How much is needed to exert its action is still to be discovered. Mentha can provide a promising source for AD medications as crude extracts or through the identification of novel bioactive compounds. The discovery and development of such components as a basis of pharmaceutical preparations is to be addressed. Instead of combining a number of agents to treat such a multifactorial disorder, which might not be practicable, Mentha extracts may have the potential to target multiple mechanisms of AD.