Naturally Occurring Acetylcholinesterase Inhibitors and Their Potential Use for Alzheimer's Disease Therapy

Alzheimer's disease (AD) is a main cause of dementia, accounting for up to 75% of all dementia cases. Pathophysiological processes described for AD progression involve neurons and synapses degeneration, mainly characterized by cholinergic impairment. This feature makes acetylcholinesterase inhibitors (AChEi) the main class of drugs currently used for the treatment of AD dementia phase, among which galantamine is the only naturally occurring substance. However, several plant species producing diverse classes of alkaloids, coumarins, terpenes, and polyphenols have been assessed for their anti-AChE activity, becoming potential candidates for new anti-AD drugs. Therefore, this mini-review aimed to recapitulate last decade studies on the anti-AChE activity of plant species, their respective extracts, as well as isolated compounds. The anti-AChE activity of extracts prepared from 54 plant species pertaining 29 families, as well as 36 isolated compounds were classified and discussed according to their anti-AChE pharmacological potency to highlight the most prominent ones. Besides, relevant limitations, such as proper antioxidant assessment, and scarcity of toxicological and clinical studies were also discussed in order to help researchers out with the bioprospection of potentially new AChEi.


INTRODUCTION
Alzheimer's disease (AD) is a main cause of dementia, accounting for up to 75% of all dementia cases and has become a population aging-related concern for policymakers and public health systems around the world by its both direct and indirect costs (Takizawa et al., 2015;Fiest et al., 2016;Scheltens et al., 2016). Nowadays, AD prevalence among people over 60 years old is estimated in 40.2 per 1000, while its incidence proportion is 34.1 per 1,000 (Prince, 2015;Fiest et al., 2016). Those values mean that over 45 million people is suffering from AD symptoms worldwide, whereas this scenario is expected to double every 20 years until at least 2050 (Scheltens et al., 2016). AD is mainly characterized by progressive neurodegenerative disorder, clinically demonstrated by cognitive and memory decline, progressive impairment of daily activities, and a variety of neuropsychiatric symptoms and behavioral disturbances (Tarawneh and Holtzman, 2012).
Pathophysiological processes described for AD progression involve neurons and synapses degeneration resulting from beta-amyloid (Aβ) protein aggregation and neurofibrillary tangles, as well as, neuroinflammation, mitochondrial damage, oxidative stress and excitotoxicity, which interfere with several neurotransmitters signaling pathways (Madeo and Elsayad, 2013;Godyn et al., 2016;Henstridge et al., 2016). Among the latter, cholinergic dysfunction is the most studied and has been closely associated with the early cognitive decline found in AD patients (Craig et al., 2011). In fact, early in the 70's, it was observed that cholinergic neurons were prematurely lost in AD process, arising the Alzheimer's Cholinergic Hypothesis (Bartus et al., 1982). This hypothesis was further corroborated by observations that cholinergic neurons in basal forebrain are severely damaged during AD progression (Bartus, 2000).
Despite the huge research on AD, supportive care from family and other caregivers is still the mainstay treatment, though pharmacotherapy has importantly evolved during the last decade. Four drugs are currently used for the treatment of the dementia phase: the acetylcholinesterase (AChE) inhibitors (AChEi)-donepezil, rivastigmine, and galantamine-and the glutamate antagonist memantine. AChEi increase synaptic acetylcholine (ACh) levels and improve cholinergic function in the brain (Anand and Singh, 2013;Andrieu et al., 2015). Amongst those clinically relevant AChEi, galantamine is the only naturally occurring substance, consisting of an alkaloid extracted from Amaryllidaceae family (Heinrich, 2010;Murray et al., 2013). Galantamine reversibly and competitively inhibits AChE (Thomsen and Kewitz, 1990) and allosterically modulates nicotinic ACh receptors (Schrattenholz et al., 1996). Notwithstanding, besides its anti-AChE activity, most of natural AChEi molecules generally present additional pharmacological properties, particularly antioxidant, which enable them to be applied as multi-target strategies against AD onset and progression (Orhan et al., 2011;Ayaz et al., 2017;Sahoo et al., 2018).
Several studies have been carried out toward identification and isolation of natural molecules applicable for design and development of new anti-AD drugs, particularly those pertaining to the classes of alkaloids, terpenes, coumarins and polyphenols (Huang et al., 2013). Therefore, this mini-review recapitulates last decade studies on the anti-AchE activity of plant species, their respective extracts, as well as isolated compounds, in order to settle down the state-of-the-art in the field and to help researchers out with the bioprospection of potentially new AChEi candidates applicable for anti-AD drug design and pharmacotherapy.

METHODOLOGY
This mini-review revises published studies available in Pubmed between 2007 and 2018 (1st semester), which were retrieved by using the following descriptors combination: "antiacetylcholinesterase and plant extract" and "acetylcholinesterase inhibitors and plant extract and Alzheimer." The only criterion for inclusion was that anti-AChE activity of the plant extract and/or isolated compounds had been assessed by Ellman's methodology (Ellman et al., 1961), which is considered a gold standard for AChEi screening (Holas et al., 2012). On the other hand, two criteria for exclusion were applied: the lack of reliable positive controls, which might include but are not limited to galantamine, huperzine A and B, or physostigmine (Mehta et al., 2012); and the absence of half maximal inhibitory concentration (IC 50 ) assessment, which allow us to compare the anti-AChE potencies among different plant extracts and/or isolated compounds (Colovic et al., 2013).
A total of 207 original studies were retrieved, from which 71 were considered appropriate. All the species Latin names were validated at The Plant List (2013); version 1.1.; http://www. theplantlist.org/ (accessed 15th August, 2018). When the Latin name provided by the study diverged from that accepted at The Plant List, the species was identified by the accepted one followed by the former, which was reported as synonym, between parenthesis. To improve the readability of the text, the identity of the plant taxonomist(s) for each species is informed only in Table 1, excepting those mentioned as the source of isolated compounds, but whose extracts were not assayed.

PLANT SPECIES WITH ANTI-ACETYLCHOLINESTERASE ACTIVITY
Amaryllidaceae is the leading family of genera holding anti-AChE activity, particularly Galanthus spp., which are the primordial source of galantamine (Heinrich, 2010). However, subsequently to galantamine's approval for the treatment of mildto-moderate AD in 2001, a plethora of species have been assessed in a pursuit of new AChEi. In our survey timeframe, a total of 39 studies reporting the anti-AChE activity for 51 species, from 29 different families, were considered. The most prevalent families were Amaryllidaceae, Lycopodiaceae, and Polygonaceae, contributing with 5, 5, and 4 species, respectively. Noteworthy, Huperzia spp. keep drawing ethnopharmacology researchers' attention, despite the consistent basic and clinical evidence already available for Huperzine A on AD treatment (Ha et al., 2011;Sahoo et al., 2018). Table 1 summarizes the contemplated species, which were classified in three categories, in accordance to the IC 50 values determined for their respective extracts/fractions: high potency, IC 50 < 20 µg/mL; moderate potency, 20 < IC 50 < 200 µg/mL; and low potency, 200 < IC 50 < 1,000 µg/mL. Those cutoffs were set according to the average IC 50 value described for galantamine in the literature (∼ 2 µM or 0.575 µg/mL) multiplied by a factor of 10 (Lopez et al., 2002;Ingkaninan et al., 2003;Berkov et al., 2008). Similar criteria have been previously applied by Murray et al. (2013), excepting that they included studies reporting only the maximal anti-AChE inhibitory activity and set the cutoff for low potency at IC 50 > 500 µg/mL.
Huperzia spp. (Lycopodiaceae) have been used for over 1,000 years in China for diverse neuronal-and cognitive-based illnesses (Ma et al., 2007), becoming of major interest for the pharmaceutical industry upon the isolation of the alkaloid Huperzine A from H. serrata (Liu et al., 1986). Thenceforth, huge research has focused on the isolation of Huperzine A and other Lycopodium alkaloids from Huperzia spp. and other Lycopodiaceae species (Ha et al., 2011;Damar et al., 2016;Sahoo et al., 2018). In spite of that, our survey retrieved recent relevant studies on anti-AChE activity of five Huperzia spp.: H. serrata (Ohba et al., 2015), H. squarrosa (Tung et al., 2017), H. brevifolia, H. compacta, and H. tetragona (Armijos et al., 2016; Table 1). In the study by Ohba et al. (2015), alkaloid enriched fraction of H. serrata aerial parts, whose major alkaloidal constituent was Huperzine A (∼0.5%), presented IC 50 = 5.96 µg/mL. On the other hand, in the study by Armijos et al. (2016), alkaloid fraction of H. tetragona aerial parts strongly inhibited AChE (IC 50 = 0.9 µg/mL), meanwhile H. brevifolia and H. compacta presented moderate potency (IC 50 = 39.6 and 62.4 µg/mL, respectively). The authors ascribed the high potency of H. tetragona to other Lycopodium alkaloids, mainly lycopodine, 6-OH-lycopodine and des-Nmethyl-α-obscurine, since Huperzine A was not detected in any of the assessed species. Tung et al. (2017) assessed anti-AChE activity in three different fractions obtained from the ethanol extract of H. squarrosa aerial parts. EtOAc and BuOH fractions presented moderated activity, whose IC 50 values were 23.44 and 50.11 µg/mL, respectively. The n-hexane fraction otherwise presented the lowest AChE inhibitory activity (IC 50 = 257.03 µg/mL).
As showed in the abovementioned studies, anti-AChE activity of plant extracts is significantly variable regardless of the predominant secondary metabolite class or the polarity of the extracting solvent. To cope with these limitations and still screen potentially applicable species, most researchers have also assessed the extracts antioxidant capacity, in order to demonstrate their dual efficacy. Although the present mini-review does not aim to discuss antioxidant aspects, it is noticeable that most studies cited in Table 1 have either quantified total phenolic content or measured antioxidant capacity in their extracts. Such assessments require appropriate methods that address the mechanism of antioxidant activity and focus on the kinetics of the reactions involving the antioxidants (Amorati and Valgimigli, 2018). Contrariwise, phenolic content was predominantly measured by Folin-Ciocalteu method, which also quantifies nonphenolic compounds, such as aromatic amino acids, sugars, ascorbic acid, and organic acids (Pueyo and Calvo, 2009), reason why it is not advisable for total phenol quantification. Similarly, antioxidant capacity was mostly assessed by trapping of the radicals DPPH • and ABTS •+ , which are non-biologically relevant oxidants (Amorati and Valgimigli, 2018). Thus, most plant extracts propelled as dually efficient (Anti-AChE and antioxidant) much probably deserve a biological approach to characterize their preventive instead of scavengering antioxidant capacity.

NATURAL AChEi COMPOUNDS
The active site of AChE contains two main subsites, the "esteratic" and "anionic" subsites, corresponding to the catalytic machinery and the choline-binding pocket, respectively. As illustrated in Figure 1A, the "esteratic" subsite consists in a histidine residue (His 447 ), whereas the "anionic" subsite is an tryptophan residue (Trp 84 ) able to bind quaternary ligands, which may act as competitive inhibitors (Dvir et al., 2010). Most of natural AChEi reported during our delimited survey period belong to the alkaloid group. Anti-AChE activity of alkaloids is ascribed to their complex nitrogen structures, which once positively charged bind to the "anionic" subsite on AChE active site (Hostettmann et al., 2006;Houghton et al., 2006). For instance, galantamine inhibits AChE by stably binding to Trp 84 , as well as phenylalanine residues on the acyl-binding pocket (Greenblatt et al., 1999). On the other hand, non-alkaloidal AChEi, which include terpenes, flavonoids and other phenolic compounds, seem to act as non-competitive inhibitors that bind to peripheral anionic sites (PAS) mainly represented by the residues Tyr 70 , Asp 74 , Try 121 , Trp 279 , and Tyr 334 (Johnson and Moore, 2006). Figure 1B shows the isolated compounds identified as potential natural AChEi, which were classified in three categories, in accordance to their IC 50 values: high potency, IC 50 < 15 µM; moderate potency, 15 < IC 50 < 50 µM; and low potency, 50 < IC 50 < 1,000 µM. As a comparative reasoning, IC 50 values described for galantamine in the surveyed studies where it was used as positive control were averaged, resulting in were classified in accordance with the criteria described at Section Natural AChEi Compounds in: high potency (green background), moderate potency (orange background), and low potency (red background). The diamond symbol represents the IC 50 value for each compound (in the left) as described in the respective study (in the right).
IC 50 = 4.82 ± 1.29 µM. Studies describing discrepant IC 50 values for galantamine were not considered. Sixteen compounds presented anti-AChE potency higher than galantamine, which include 01 terpene, 2 coumarins, and 13 alkaloids. Other 20 compounds, additionally pertaining to flavonoids and phenolic acids, were also selected with IC 50 ranging from 5.33 to >200 µM ( Figure 1B). The dihydropyranocoumarin decursinol isolated from Angelica gigas Nakai (Apiaceae) was the most potent AChEi (IC 50 = 0.28 µM; Anand et al., 2012), such high potency had been previously attributed to characteristics of cyclization of the isoprenyl unit at C-6 and the functional groups attached to the coumarin nucleus, which differ from other coumarins (Kang et al., 2001).
The major alkaloids with recognized anti-AChE activity are the classical galantamine and huperzine A, which have been elegantly reviewed by Gulcan et al. (2015) and Qian and Ke (2014), respectively. However, other recently described alkaloids of various subclasses deserve special emphasis because of their important inhibitory action on AChE. Jung et al. (2009) assessed the anti-AChE activity of five protoberberine alkaloids isolated from the rhizome of Coptis spp. (berberine, palmatine, groenlandicine, jateorrhizine, and coptisine), with IC 50 ranging from 0.44 to 0.80 µM. Interestingly, groenlandicine also strongly inhibited the enzyme responsible for cleaving the β-site of the amyloid precursor protein, adding an important property against AD pathogenesis (Jung et al., 2009). The potentialities of protoberberines alkaloids as natural AChEi were further supported by isolation of 12 isoquinoline alkaloids, including two new nitrotetrahydroprotoberberines (2,9-dihydroxy-3,11-dimethoxy-1,10-dinitrotetrahydroprotoberberine and 4-nitroisoapocavidine), from Corydalis saxicola Bunting (Papaveraceae). All the alkaloids were selectively active against AChE with IC 50 < 10 µM. Structure-activity relationship study indicated that potency differences were related to the presence of phenolic hydroxy groups, which could reduce the anti-AChE activity, whereas nitro substitutions at ring A, especially at C-1, in the tetrahedroprotoberberines could increase it (Huang et al., 2012).
Studies on the molecular mechanisms by which natural AChEi interact with AChE binding subsites are still scant. Nevertheless, some studies have offered important insights on this matter. Serpentine, the main alkaloid found in the roots of Catharanthus roseus (L.) G. Don (Apocynaceae) presented high anti-AChE potency (IC 50 = 0.77 µM), which was attributed to the binding of its quaternary nitrogen to an Asp residue at AChE peripheral anionic site (Pereira et al., 2010). Lai et al. (2013) when evaluating alkaloids from Stemona sessilifolia (Miq.) Miq. roots (Stemonaceae) identified the AChEi stenine B (IC 50 = 2.10 µM) and stenine (IC 50 = 19.8 µM). Authors attributed the stronger activity of stenine B to its ability to build hydrogen bonds with Tyr 130 , similarly to huperzine A. Lastly, bioactivity-guided chromatographic fractionation of Nelumbo nucifera Gaertn. (Nelumbonaceae) leaf extract led to isolation of three aporphine-type alkaloids, an important subclass of natural inhibitors of AChE. Amongst them, N-methylasimilobine displayed a significant anti-AChE activity with IC 50 = 5.33 µM. According to their in silico studies, such potency was due to a hydroxyl group at the alkaloid C-2 position, which makes hydrogen bond with a carbonyl group on Ser 293 in association with another hydrogen bond between its alkaloidal quaternary nitrogen and the hydroxyl group of Tyr 124 .
Salvia spp. (Lamiaceae) have been used for centuries for its beneficial effects on memory disorders (Hamidpour et al., 2014). Wong et al. (2010) demonstrated that the diterpene cryptotanshinone extracted from the root of Salvia miltiorrhiza Bunge is a reversible inhibitor of human AChE (IC 50 = 4.09 µM) and that chronic oral administration can reverse cognitive deficits induced by scopolamine in rats. Flavonoids, a heterogeneous group of polyphenols, are currently considered a prominent source of anti-AD compounds (Khan et al., 2018) because of their potential AChE inhibitory activity allied to the well-known antioxidant activity and low toxicity (Uriarte- Pueyo and Calvo, 2011). However, our survey did not identify any highly potent, and consequently prominent AChEi pertaining to the flavonoid class (Figure 1). For instance, luteolin and 3,5-dicaffeoylquinic acid, phenolic compounds extracted from Phagnalon saxatile Cass. (Compositae) exhibited low activity against AChE with an IC 50 of 88.00 and 105 µM, respectively .

CLINICAL STUDIES
Besides galantamine, huperzine A is the most clinically studied alkaloidal AChEi (Qian and Ke, 2014). The efficacy of huperzine A was demonstrated in the treatment of 447 patients with agerelated memory impairment or dementia (Shu, 1998;Ma et al., 2007). However, in another phase II study, the results were not conclusive on its beneficial cognitive effects for patients with moderate AD, requiring further investigation (Rafii et al., 2011). A clinical trial with Salvia officinalis L. administered to patients with mild to moderate AD for a 16-weeks period led to improved cognitive performance (Perry et al., 2003). Of importance, S. officinalis also attenuated cognitive impairment in patients suffering from moderate to severe AD when used for up to 1 year. However, authors recognized that long-term efficacy, safety and administration strategy still require further investigation (Tune, 2001). Salvia spp. are particularly rich in terpenes, whose anti-AChE capacity has been assessed through enough pre-clinical tests, but are awaiting clinical trials (Rollinger et al., 2004;Kennedy and Scholey, 2006). On the other hand, a 22-weeks randomized, double-blind, multicenter trial, including 54 individuals suffering from mild-to-moderate AD, showed that daily intake of Crocus sativus L. (Iridaceae) dried extract (30 mg/day) significantly improved cognitive capacity comparable to that observed in donepezil-treated patients (Akhondzadeh et al., 2010).

TOXICOLOGICAL STUDIES
A recent systematic review and meta-analysis of 43 randomized placebo-controlled clinical trials showed that AChEi improved cognitive function, global symptomatology, and functional capacity, as well as decreased patients' mortality (Blanco-Silvente et al., 2017). However, patients taking AChEi presented higher discontinuation due to adverse events, denoting an important issue on anti-AChE therapy. As showed in Table 1, the majority of the plant extract-based studies mentioned in this mini-review has not assessed their toxicity in animals or humans, although species like S. officinalis (Kennedy and Scholey, 2006) and P. hydropiper (Huq et al., 2014) have been considered as non-toxic. Amongst the main natural AChEi compounds herein mentioned, berberine and safranal seem to ally more advantages than disadvantages. Nevertheless, berberine has been shown to cause mild gastrointestinal reactions, including diarrhea and constipation, besides other less frequent side effects (Imenshahidi and Hosseinzadeh, 2016); and safranal has toxic effects on hematological and biochemical indices, as well as induced embryonic malformation in animal's models at high doses (Bostan et al., 2017).

CLOSING REMARKS AND PERSPECTIVES
The present mini-review demonstrated that during last decade several plant species and their potentially active compounds have been screened for anti-AChE activity. Amongst the most active extracts (Table 1), it is noticeable the use of extracting solvents of distinct polarities, which suggests that their active compounds might pertain to a wide range of secondary metabolites classes. However, having a look at the isolated substances summarized in Figure 1B, most high potent compounds assessed during this period pertain to alkaloid class, exception made to the highest potent decursinol, a dihydropyranocoumarin. Alkaloids indisputably are the most studied class of natural AChEi, what seemly has trapped the researcher's attention in this class when in pursuit of new potential AChEi candidates, a vision that urges to be changed. Notwithstanding, the search for secondary AD-relevant pharmacological properties, such as antioxidant, deserves experimental approaches addressing their capacity to prevent oxidants generation and oxidative damage, instead of their mere scavengering capacity.
Finally, despite the undoubted relevance of new AChEi discovery for AD palliative pharmacotherapy, there is scanty knowledge on their structure-activity relationships, as well as toxicological assessments that would enable them to phase II studies. For instance, berberine and related protoberberine alkaloids have been consistently assessed for their anti-AChE activity, but no phase II study has been conducted so far. Such knowledge is capital both to promote higher safety and to guide the design of new (semi-) synthetic AChEi. Thus, given the plethora of plant species and compounds already described, their assessment through clinical trials certainly represent the main barrier to be transposed in order to expand and improve the pharmacological care of AD patients.

AUTHOR CONTRIBUTIONS
TS conceived the proposal, discussed mini-review's structure, surveyed and selected relevant articles, tabulated the data and drafted the manuscript. TG surveyed and selected relevant articles, tabulated the data. BP supervised articles selection, analysis and data tabulation. AC discussed mini-review's structure, supervised articles selection, analysis and data tabulation. AP conceived the proposal, discussed mini-review's structure, oriented the selection of relevant articles, analyzed tabulated data, and drafted the manuscript. All authors read and approved the final format of the manuscript.

ACKNOWLEDGMENTS
Authors are thankful to Foundation for the Support of Research, Scientific, and Technological Development of the State of Maranhão-FAPEMA, which has importantly funded their research on ethnopharmacology of regional plant species for AD therapy through the grant Universal-00651/15.