Neuropharmacology of Secondary Metabolites from Plants with Anxiolytic and Antidepressant Properties

Depression and anxiety currently rank as the second and fifth most common causes worldwide of years lived with disability-a reality that has intensified the search for new treatments. There are many studies of herbal extracts and secondary metabolites from plants used in traditional medicine due to their antidepressant and anxiolytic properties. Clinical and preclinical studies about some of the mechanisms of action of metabolites like alkaloids, terpenes, flavonoids, and sterols, among others, have documented effects similar to those produced by clinically effective drugs. These metabolites have shown anxiolytic and antidepressant effects in various experimental models of anxiety by interacting with γ -aminobutyric acid subtype A receptors (GABA A -receptors) and by stimulating the serotonergic, noradrenergic, and dopaminergic neurotransmitter systems. These pharmacological effects can be attributed to plant metabolites that share structural similarities with monoamines, which allow them to bind to receptors. The objective of this chapter is to summarize the various mechanisms of action that have been identified in secondary metabolites with anxiolytic and antidepressant properties. Terpenes, alkaloids, flavonoids, and sterols can interact at different levels of the neurotransmission systems involved in the neurobiology of anxiety and depression, suggesting their potential for treating these mental illnesses. on MAO-B in the three


Introduction
According to the Global Burden of Disease, depression and anxiety are currently the second and fifth most common causes worldwide of years lived with disability in both sexes in the age range of 15-49 years [1]. In 2015, 4.4% (322 million people) of the world's population suffered from depressive disorders, while 3.6% (264 million) were affected by anxiety [2]. In that year, the World Health Organization (WHO) estimated that by 2020 depression would be the second leading cause of extracts, fractions, isolated compounds, or derivatives using ex vivo, in vitro, and in vivo assays (e.g., preliminary pharmacological screening, classic animal models of anxiety like the light dark box (LDB) or elevated plus-maze (EPM) tests, etc., or the forced swim test (FST) and tail suspension tests (TST), among others). These tools have allowed researchers to analyze the possible metabolites responsible for the anxiolytic or antidepressant properties of plants used by different populations, and identify how their mechanisms of action affect the functioning of the central nervous system (CNS). Their studies contribute to advancing scientific understanding of the neurobiology of depression and anxiety, and to developing new pharmacological treatments that may favorably impact public health.
Plants used in traditional medicine contain compounds in their secondary metabolism [13] such as alkaloids, phenols, sterols, carbohydrates, tannins, terpenes, and phytoalexins, all of which have important biological activities [14]. The most widely studied metabolites are terpenes, flavonoids, alkaloids, and sterols, whose mechanisms of action stimulate the serotonergic, noradrenergic, dopaminergic, or GABAergic neurotransmission systems, acting on receptors or the synthetic pathways of neurotransmitters and their transporters. However, they may also stimulate other neurotransmission systems. For example, terpenes can stimulate at the same time serotonergic, dopaminergic, and noradrenergic neurotransmission systems [12] that can produce a similar effect on mood regulation, perhaps leading to an overstimulation that may generate undesirable collateral effects.
This chapter reviews and discusses the findings from research on several metabolites of medicinal plants that have shown potential anxiolytic and antidepressant activities once screened for their biological mechanisms at various levels: receptor, transporter, synthesis, gene, protein, or metabolic. The studies analyzed were identified by a preliminary search in PubMed, Scopus and Ovid for articles on (i) the dose effects and possible mechanisms of action of metabolite(s) isolated from parts of plants with previously identified anxiolytic or antidepressant effects; and (ii) standard chemical tests performed with specific metabolites. NE = no explorated. Table 2.

Metabolite
Secondary metabolites with anxiolytic properties in preclinical study.
Carnosol generated an antidepressant-like effect at doses of 0.01 and 0.1 mg/kg [16] on TST, and 10, 30, and 100 mg/kg, i.p., on TST and FST, as did the oleanolic and ursolic acids. Meanwhile, 100 mg/kg of ursolic acid showed an effect similar to that of imipramine at 60 mg/kg [17]. Betulinic acid at 10 mg/kg was evaluated in male mice on TST [16]. None of those metabolites had effects on locomotor activity. Ursolic acid at doses of 0.01 and 0.1 mg/kg produced an effect similar to that of fluoxetine (10 mg/kg), imipramine (1 mg/kg), and bupropion (10 mg/kg) on TST [18]. Only the 10-mg/kg dose had an antidepressant effect on FST, which was similar to that of bupropion at 10 mg/kg. Studies exploring the mechanism of action of ursolic acid found the involvement of D1 and D2 receptors and a pharmacological synergism with bupropion at 1 mg/kg, p.o. (dual dopamine/noradrenaline reuptake inhibitor, DDNRI) [18], 5 mg/kg of fluoxetine, or 2 mg/kg of reboxetine (SRNI) [19]. They also observed a related increase in noradrenaline and dopamine synthesis on TST [19]. These findings are noteworthy because they suggest the possibility that various mechanisms of action may be stimulated by the same metabolite.
A study of the essential oil of Origanum majorana (OMEO) identified 14 component terpenes that represented 92.32% of yield. The five components that were more abundant in OMEO were terpinene-4-ol (32.69%), γ-terpinene (12.88%), transsabinene hydrate (8.47%), α-terpinene (7.98%), and sabinene (6.211%). Administered in a single acute dose of 40 or 80 mg/kg, OMEO increased swimming and climbing times in male mice on FST. These findings were interpreted as showing an antidepressant-like effect. In that study [20], pretreatment with antagonist drugs demonstrated that the terpenes act through several mechanisms: first, by activating the dopaminergic receptors D 1 and D 2 , the noradrenergic α1and α2-adrenoceptor, and the 5-HT 1A and 5-HT 2A receptors, and, second, by activating or increasing 5-HT synthesis and monoamine vesicular storage involved in reducing the immobility time produced by 80 mg/kg of OMEO [20]. A pharmacological synergism at the combined subthreshold OMEO doses of 10 mg/kg plus fluoxetine or imipramine (5 mg/kg, i.p.) was also seen. It reduced immobility time but increased swimming and climbing times, similar effects to those produced by OMEO at 80 mg/kg [20]. (A more detail view about the OMEO components can be reviewed at Tables 1 and 2 as summary). (R)-(À)-linalool produced antidepressant effects in male mice on TST at a single dose of 100 or 200 mg/kg, i.p., and on FST when administered 3 times at 100 mg/kg, i.p. [21]. That effect was produced by activation of the 5-HT 1A receptor and α2-adrenoceptors [22], while (1S)-(À)-β-pinene produced an antidepressant effect after three applications at a dose of 100 mg/kg, i.p., by activating βadrenoceptors, 5-HT 1A and D1-receptors, and noradrenergic neurotransmission in the cerebral cortex [22].
However, due to the terpene's capacity to stimulate several neurotransmission systems-especially high doses of monoterpenes-possible collateral or undesirable effects as serotonergic syndrome need to be explored. The monoterpenic oxide, 1,4cineole, for example, produced a prodespair effect at doses of 200 mg/kg, i.p., on FST and 400 mg/kg, i.p., and FST and TST, but did not induce any significant deficit in motor coordination on the rota-rod test (RRT). It did, however, have an anxiolytic effect at a dose of 400 mg/kg in male mice evaluated in EPM that was not associated with any sedative effect [23]. These findings require additional study in light of potential depressor effects on the CNS, and to elucidate the mechanisms of action involved.

Terpenes with anxiolytic effects: preclinical research
Anxiolytic properties have also been attributed to terpenes. A combination of two monoterpenoids, p-cymene + thymol, both at doses of 3mg/kg i.p., produced anxiolytic effects on the hole-board test (HBT) and EPM [24]. Studies have also demonstrated that (À)-myrtenol, a monoterpenoid alcohol, produced an anxiolytic effect on EPM at doses of 25, 50, and 75 mg/kg, i.p., though only the 25-and 50-mg/kg doses did so on LDB. On both tests, the anxiolytic effect of 25 mg/kg of (À)-myrtenol was mediated by GABA A receptors [25]. In another study, rosmanol produced an anxiolytic effect at doses of 10, 30, and 100 mg/kg in EPM and LDB, and the 10-and 30-mg/kg doses showed an effect similar to those of diazepam at 1 mg/kg [15]. This mechanism of action acts on GABA A receptors at a distinct site from the high-affinity benzodiazepine-binding region [15]. Triterpenes as ursolic acid, oleanolic acid, and carnosol produced anxiolytic effects at doses of 10, 30, and 100 mg/kg in EPM and LDB. The effect of the 30-and 100-mg/kg doses of these three metabolites was similar to that of diazepam at 1 mg/kg. No significant effect was seen on locomotor activity. We know that this effect was produced through GABA A receptors of the α1β2γ2L conformation because 2.5 mg/kg of flumazenil blocked the anxiolytic effects of 10 mg/kg of all three in EPM [17].
Although most terpenes have a GABAergic mechanism, their action may also occur through the serotonergic system, as evidenced in the study by Costa et al. [26]. In an experiment with male rats, those authors identified that acute administration of 5 mg/kg, p.o., or 14-day repeat doses (1 mg/kg/day), of the essential oil of ripe fruits of C. aurantium (Rutaceae) (whose chemical composition includes 98.66% limonene, 0.53% β-myrcene, 0.41% β-pinene, and 0.41% unidentified compounds) produced an anxiolytic-like effect in LDB that was mediated by the serotonergic system through 5-HT1A receptors (WAY 100635 0.5 mg/kg, i.p.), not the GABAergic system (flumazenil 2 mg/kg, i.p.). That study also analyzed the antidepressant effect on FST after oral and inhaled treatment, but found that it did not modify immobility. Their results suggest that distinct mechanisms of action exist for the anxiolytic and antidepressant effects [26]. In this regard, some of the terpenes in certain essential oils produce anxiolytic effects and are often used in aromatherapy to reduce anxiety in animals and humans [27]. Inhaling emulsions of linalool oxide (a monocyclic alcohol) at 0.65, 1.25, 2.5, and 5.0% w/w during 7 min of exposure in the inhalation chamber, for example, produced anxiolytic-like effects in EPM and LDB, but no significant motor deficit on RRT [28].
Finally, songorine, a C20 diterpenoid alkaloid, produced an anxiolytic-like effect at 0.25 mg/kg v.o. for 5 days with greater efficacy than phenazepam on Vogel's conflict test (VCT) [29]. For the terpenes described so far, we know that both the serotonergic and GABAergic systems are involved in their mechanisms of action, and these are the same systems that are activated by other groups of metabolites, such as flavonoids and sterols (see Tables 1 and 2 for summaries).

Flavonoids with antidepressant effects: preclinical research
Flavonoids are the most widely studied active metabolites (for a broad review of research, see German-Ponciano et al. [30]). Genistein is an isoflavone that can cross the blood-brain barrier in mice [31] and rats [32]. Acute oral administration of 5, 15, and 45 mg/kg genistein in male mice did not reduce immobility time on FST or TST, but chronic, dose-dependent administration for 21 days produced antidepressantlike effects on both tests, without affecting locomotor activity [33]. This effect was associated with increased NA and 5-HT concentrations in the hippocampus and frontal cortex, and of 5-HT in the hypothalamus, though it decreased the 5-HTIAA/ 5-HT ratio in the hippocampus and frontal cortex. These results suggest an inhibition effect of genistein on MAO-A in the hippocampus, frontal cortex, and hypothalamus and on MAO-B in the hippocampus, three brain structures involved in the neurobiology of depression and anxiety. Their results [33] also demonstrate that central depletion of 5-HT reversed the antidepressant effect of genistein, suggesting a critical role of the serotonergic system, specifically through 5-HT 1A receptors. It is important to note that these results on the serotonergic metabolic ratio (5-HIAA/5-HT) may be dependent on gonadal hormones. Ovariectomized rats (OVX, surgical removal of both ovaries) showed reduced immobility times on FST after administration of genistein (10 mg/kg, p.o. [34], or by i.m.) [35] for 14 days, but the downward tendency of the serotonergic metabolic ratio caused by FST was only evident in the hippocampus [34]. Sapronov and Kasakova [35] found antidepressant-like effects on FST at the same dose of genistein (10 mg/kg) but in non-ovariectomized rats. That effect was more marked in the metestrus and diestrus phases of the estral cycle, which are characterized by low plasmatic concentrations of ovarian hormones, than during the proestrus and estrus stages with their characteristically high hormone concentrations. This suggests that these hormones play a significant role in the antidepressant effect of genistein.
A similar effect on the serotonergic system and MAO-A activity was found with quercetin 4'-O-glucoside or quercetin administered at doses of 10 and 20 mg/kg, p. o., for 7 days in Swiss albino mice of both sexes. These substances produced antidepressant-like effects in mice on FST as well as those subjected to unpredictable, chronic mild stress (CUMS, a mouse model designed to induce depression) and subsequently evaluated on FST. In that study, a 20-mg/kg dose of quercetin 4'-O-glucoside showed a similar effect to that of fluoxetine at 20 mg/kg, p.o., on FST, with or without prior exposure to CUMS [36]. A consequence of CUMS on the sucrose preference test (SPT, a test used to study anhedonia rodents, the main symptom of depression in humans) is a decrease in the consumption of a sweetened solution. In these sense, both doses of quercetin 4'-O-glucoside reverted this effect, and interpreted as being antidepressant. Other consequences of CUMS are metabolic, for example, excessive production of reactive oxygen species that was evidenced by higher brain thiobarbituric acid reactive species (TBARS). Compromising endogenous anti-oxidants, like reduced glutathione (GSH), enhances MAO-A activity in the brain and, consequently, depletes monoamine levels there, especially serotonin 5-HT. The effects observed in that study were blocked by 21 days of treatment with 10 and 20 mg/kg of quercetin 4'-O-glucoside [36], suggesting a possible mechanism of action with an antioxidant effect that impedes ROS production. Another study [37] found that 10 mg/kg of quercetin administered for 14 days reduced immobility time on TST, but not FST, while doses of 25 and 50 mg/kg produced this effect in female mice on both tests. The mechanisms of action were explored on TST, where i.c.v. administration of N-methyl-Daspartate (NMDA at 0.1 pmol/site) and L-arginine (at 750 mg/kg, i.p., a nitric oxide inhibitor) blocked the antidepressant effect of quercetin. Hence, the antidepressant-like effect of quercetin may involve inhibiting NMDA receptors to decrease intracellular calcium that, in turn, inhibits the protein calmodulin, which then inhibits neuronal nitric oxide synthase to decrease nitric oxide levels (NO). This hypothesis is supported by the finding that administering methylene blue (a NO synthase inhibitor) at 20 mg/kg, i.p., and soluble guanylate cyclase or 7-nitroindazole (another NO synthase inhibitor) at 50 mg/kg, i.p., improved quercetin's antidepressant-like effect on TST. This indicates that the antidepressant effect may be dependent on limiting NO synthesis, either by inhibiting the enzyme or by reducing NO production, perhaps via decreased cyclic guanosine monophosphate (cGMP), since sildenafil (a phosphodiesterase 5 selective inhibitor that increases cGMP levels) also canceled this effect [37].
A model of depression induced by olfactory bulbectomy (OB, surgical removal of the olfactory bulbs) reduced the latency to immobility and increased immobility time on FST and TST. This was accompanied by an increase in the levels of the markers of oxidative stress, for example, 116% in the case of lipid hydroperoxide content (LOOH) in the hippocampus. This effect was reverted by 52.25% by administering 25 mg/kg of genistein in the content of LOOH, as observed on the immobility on FST and TST. In sham rats only (i.e., animals subjected to the same surgical procedure but without resection of the olfactory bulbs), genistein reduced glutathione (GSH) levels, in that study by 65.94%. The authors [37] explained that "the reduction of GSH levels caused by OB and, surprisingly, quercetin, can be explained by the fact that glutathione peroxidase, in addition to reducing H 2 O 2 , decreases lipid and nonlipid hydroperoxides at the expense of GSH, causing it to become oxidized and giving rise to glutathione disulfide. Therefore, it is suggested that LOOH activated glutathione peroxidase which, in turn, oxidized GSH to normalize LOOH levels". In this area, increased levels of the markers of oxidative stress in major depression have been associated with poor response to antidepressant treatment [38]. Therefore, a therapy that reduces the levels of markers of oxidative stress and produces antidepressant effects could be a promising form of treatment.
Additional mechanisms of the antidepressant action of flavonoids have been explored. Administering chrysin for 14 days at a dose of 20 mg/kg, for example, increased grooming time in male OB C57B/6J mice evaluated on the splash test (ST), where 200 ml of a 10% sucrose solution is squirted on the mouse's snout to initiate grooming behavior. Here, greater grooming time is considered an antidepressant effect. Doses of 5 and 20 mg/kg impeded an increase in immobility time by these OB mice on FST, but increased 5-HT and brain-derived neurotrophic factor concentrations in the hippocampus [39]. In another study, fisetin administered at 5 mg/kg v.o. increased activation of the tropomyosin kinase B receptor (TrkB) by signaling and increasing its phosphorylation in the hippocampus. This suggests that fisetin produced pro-neurogenesis [40] related to its antidepressant effect on FST and TST after 1 or 2 weeks of treatment with a relatively short therapeutic latency compared to clinically-effective antidepressants. Fisetin also reversed depressionlike behaviors induced by spatial restraint stress in mice evaluated on FST and TST [40]. Other studies have found that the chemical standard dihydromyricetin activated the ERK1/2-CREB pathway and increased glycogen synthase kinase-3 beta (GSK-3β) phosphorylation at ser-9 with upregulation of BDNF expression in the hippocampus, while inhibiting neuroinflammation. These findings may be related to the antidepressant effect seen on TST and FST after once-daily administration of 10 and 20 mg/kg, v.o. for 3 days, but not after a single acute dose [41]. Interestingly, dihydromyricetin reverted the depressogenic effect caused by CUMS in mice subjected to SPT and FST, or TST, only after administration of once daily for 7 days, but not 3 days [41]. Another flavonoid analyzed is hesperidin, which increased BDNF levels in the hippocampus after administration of once daily for 21 days (0.3 and 1 mg/kg, i.p.). These doses produced an antidepressant effect on TST similar to fluoxetine (32 mg/kg i.p.) and imipramine (15 mg/kg, i.p.). Another research has also verified that when applied acutely (1 mg/kg after 30 min) or chronically (0.1, 0.3, and 1 mg/kg for 21 days) hesperidin significantly decreased nitrate/nitrite (NOX) levels in the hippocampus of mice, suggesting a possible inhibition of the L-arginine-NO-cGMP pathway [42].
Another flavonoid that has shown effects on the CNS is baicalin, which may promote neuronal differentiation through neuronal maduration and ensure their survival via the associated Akt/FOXG1 pathway, which stimulates dendrite elongation. This is related to findings that indicated that, after 6 weeks of treatment, a 60-mg/kg dose of baicalin had an effect similar to that of fluoxetine (15 mg/kg, v.o.), because it reverted the decrease of sucrose intake on SPT and the increase in immobility on TST produced by CUMS [43]. Another flavonoid that associates antidepressant and antioxidant effects is naringin, which reduced immobility on FST at doses of 2.5, 5, and 10 mg/kg given for 7 days. The antidepressant effect of these doses correlated with enhanced cholinergic transmission due to a decrease in the activity of the enzyme acetylcholinesterase and of the antioxidant defense systems caused by higher GSH levels, as well as an increase in the activity of superoxide dismutase (SOD) and catalase (CAT) in mice brains.
Studies have demonstrated that naringin inhibits lipid peroxidation and nitrosative processes by reducing levels of ROS and nitrogen species [44]. Finally, the extract of Cirsium japonicum at doses of 200 and 400 mg/kg has shown the ability to reduce immobility time on FST in a similar manner to that of 5 mg/kg of the antidepressant imipramine. A major component of this plant is the flavonoid luteolin, which at doses of 5 and 10 mg/kg produced a similar effect to that of the complete extract, likely through a positive modulating effect on the GABA A receptor complex. This was proven in an in vitro study where extracts of both Cirsium japonicum and luteolin increased Cl À influx in an effect impeded by pretreatment with bicuculline, a competitive GABA A receptor antagonist [45].
The varied mechanisms seen in flavonoids make them an important object of study, especially in the search for side effect-free treatments that can compromise their effectiveness or produce toxicity by interacting with other medications or food. This is another area of research that remains to be explored.

Flavonoids with anxiolytic effects
A particularly important fact concerning the potency of the biological activity of plants is that it depends on several factors, for instance, the part of the plant used, the region where it is gathered, the season, and harvesting time, among others [46]. For example, in male mice evaluated by HBT and EPM, a single dose of 100 mg/kg i.p. of the methanolic extract of inflorescences of Tilia americana var. mexicana collected in Morelia, Mexico, produced a more effective anxiolytic effect than those gathered in Honey, Puebla, though the leaves collected in Honey were more effective than those from Morelia or Santa María Ahuacatitlan, Mexico. These three Mexican states are located at different elevations with distinct humidity and soil types. That study quantified quercetin, rutin, and isoquercitrin in the inflorescences and leaves, determining that the concentrations of these substances differed with the part of the plant used and the collection area [46]. It also tested several standard commercial flavonoids: kaempferol (10 mg/kg), quercetin (20 mg/kg), astragalin (10 mg/kg), isoquercitrin (2 mg/kg), quercetin (10 mg/kg), and rutin (15.7 mg/kg), and a mixture of flavonoids (MIX) composed of quercetin 20 mg/kg, rutin 15.7 mg/kg, and isoquercitrin 2 mg/kg and quercetin (20 mg/kg). Results showed that a mixture of quercitin (20 mg/kg), rutin (15.70 mg/kg), and isoquercitrin (2 mg/kg) produced an anxiolytic effect in male mice tested in HBT and EPM [46] by reducing the number of head-dippings but increasing the time spent in the open arms, respectively. Finally, upon testing the anxiolytic effect of the methanolic extract of Tilia americana var. mexicana, those authors found that this produced an effect in EPM through the participation of GABA/BDZ (flumazenil 5 mg/kg) and 5HT 1A serotonergic receptors (WAY 100635 0.5 mg/kg), though they were not involved in the anxiolytic effect on HBT [46].
Another flavonoid known to have anxiolytic effects is formononetin, an active metabolite of traditional Chinese medicine red clover (Trifolium pratense L.). Wang et al. [47] observed that administering 25 mg/kg of this metabolite to male mice once daily for 8 days blocked the anxiogenic effect on the open field test (OFT) produced by administering Freund's complete adjuvant (CFA), reduced the time spent and distance traveled in the central area, and decreased the time spent in the open arms of EPM [47]. Formononetin did not modify behavior compared to the control group on either test, indicating that it demonstrated an anxiolytic effect. However, it seems that these results should be understood as a neuroprotective effect more than an anxiolytic one, at least under those study conditions. The notion of a neuroprotector effect is supported by the fact that the study found that formononetin attenuated inflammation and neuronal hyperexcitability by inhibiting NMDA receptors and the CREB signaling pathway in the basolateral amygdala (BLA) [47].
Other anxiolytic mechanisms of action seen in flavonoids are dopaminergic in nature. Theaflavins, for example, increased dopamine (DA) turnover to induce activation of the dopaminergic system in the frontal cortex in male mice in EPM and LDB [48], while chrysin at 2 and 4 mg/kg produced an anxiolytic effect in rats at 12 weeks postovariectomy on LDB by increasing the time spent in the light compartment. Those findings resembled the effect of diazepam. At doses of 1, 2, and 4 mg/kg, this flavonoid increased the frequency of entries into, and the time spent in, the open arms of EPM partially through action on GABA A receptors (pretreatment with 1 mg/kg picrotoxin) [49]. On the other hand, neurosteroids and the serotonergic system have also been implicated in the anxiolytic effect of flavonoids, as in the case of puerarin, which increased 5-HT and allopregnanolone levels in the prefrontal cortex and hippocampus in male rats. These results have been associated with the finding that puerarin increased the time spent in the open arms and the percentage of entries into the open arms of EPM, whereas on the VCT test, it produced an increase in the number of shocks received. In both cases, the effect was similar to that of sertraline, which was used as a positive control drug to generate an anxiolytic effect on both tests [50].
In an animal model of surgically-induced menopause, genistein at 0.09 and 0.12 mg/kg, s.c., for seven consecutive days, or the same treatment regimen but with 17β-estradiol, increased the time spent in, and the percentage and frequency of entries into, the open arms of EPM. These effects were caused by activation of the β-estrogenic receptor (ERβ) since pretreatment with tamoxifen (5 mg/kg, an ERβ antagonist) blocked the anxiolytic effect. Also, genistein and 17β-estradiol increased the frequency of rearing and grooming behaviors on the locomotor activity test (LAT), associated with an anxiety-reducing effect manifested in EPM [51]. Genistein tested at doses of 0.25, 0.5, and 1 mg/kg increased the time spent in, and the frequency of exploration of, the light compartment of LDB, while doses of 0.5 and 1 mg/kg increased rearing frequency, and 1 mg/kg increased grooming time. Those studies used rats at 12 weeks postovariectomy [52]. The authors suggest that "genistein is considered a phytoestrogen that acts in a dose-dependent manner with a broader margin of safety at anxiolytic doses. However, more studies are required to take advantage of its potential therapeutic anxiolytic effects" [51].
In a distinct approach, a post-traumatic stress disorder (PTSD) model used a chamber with a grid floor connected to a system that delivered foot shocks, exposing rats to 5 shocks per day. There, an increase in the contextual freezing time on days 7, 14, and 21 indicated the induction of anxiety-like behaviors. The time spent in freezing behavior was calculated with the shock-administering system turned off. In that study, genistein at 4 and 8 mg/kg i.p. administered to male rats from day 7 reduced freezing time at 7, 14, and 21 days. Interestingly, only the 8-mg/kg dose returned freezing times to control levels on day 21 [53]. The stressed rats were also tested in EPM, where they spent less time in the open arms, indicating an anxietylike effect that was reverted by administering 4 and 8 mg/kg of genistein. This reduced anxiety-like behavior in the stressed rats occurred in association with enhanced tryptophan hydroxylase (TPH) and 5-HT levels, but also promoted the 5-HT receptor-related CaMKII/CREB signaling pathway in the amygdala [53], likely reflecting the fact that the amygdala receives serotonergic projections from the raphe, two brain structures to which emotional valence and 5-HT synthesis, respectively, are attributed [54].
A study evaluated the pharmacokinetic profile of 6-methoxyflavanone and calculated the K P value (i.e., the tissue-to-serum partition coefficient). Molecules with K P values >0.30 are thought to be readily distributed in the brain [55]. That study determined that 30 min postadministration of a 30-mg/kg i.p. dose, the 6methoxyflavanone had crossed the blood-brain barrier (BBB) to reach the amygdala with K p = 0.47, and the cerebral cortex with K p = 0.437. These two cerebral structures are known to be involved in the neurobiology of anxiety [56], so these properties were associated with the anxiolytic effect of 6-methoxyflavanone in male and female mice in EPM at doses of 10, 30, and 50 mg/kg, and on the staircase test (ScT). In this model, the lower frequency of rearings, but no reduction in the steps climbed in a 3-min period, was interpreted as indications of an anxiolytic effect [45]. Those authors verified that 6-methoxyflavanone produced its effect by activating GABA A receptors with the α2-subunit, perhaps in the amygdala and brain cortex, since pretreatment with PTZ blocked this anxiolytic effect in EPM and on ScT [56].
Another flavonoid with anxiolytic effects is rutin at doses of 300 and 562 mg/kg, i.p., or 16 nmol/site, in the basolateral amygdala of male rats tested in EPM. This involves partial GABAergic neurotransmission that was not associated with BDZ binding in the GABA A receptors [57]. Finally, viscosine administered to male mice assessed in EPM and LDB was seen to exert its action through the α1β2γ2L and α2β2γ2L modulates of the GABA A receptors at a site distinct from the one classically associated with benzodiazepine [58].

Alkaloids with antidepressant activity
Alkaloids purified from crude acid-base extracts have diverse chemical structures. They may contain one or more nitrogen atom(s) (in the heterocyclic ring) in the form of salt [59]. Pseudoalkaloids that possess nitrogen exist. They are not synthesized from amino acids, but by nitrogen transfer in the form of ammonia to a compound of terpenic origin, steroids, polyketides, monosaccharides, or fatty acids [59].
In another work, the alkaloid derivatives of the β-carbolines (harmane 5, 10, and 15 mg/kg, norharmane 2.5, 10, and 15 mg/kg, and harmine 5, 10, and 15 mg/kg, all i.p.) showed antidepressant effects in mice that were dose-dependent, suggesting that the effect occurs through an inverse agonist mechanism of the benzodiazepine receptors due to flumazenil antagonism (5 mg/kg, i.p.) [61]. In addition, anhedonia was reversed in rats subjected to the CUMS model after harmine treatment at 15 mg/kg/day for 7 days. They showed increased adrenal gland weight, ACTH levels, and BNDF protein levels produced by the CUMS [62]. Treatment for 14 days with harmine (5, 10, and 15 mg/kg) and imipramine (10, 20, and 30 mg/kg) in rats subjected to FST produced antidepressant-like effects, while harmine (10 and 15 mg/kg), but not imipramine, increased BDNF protein levels in the hippocampus of rats [62]. These results indicate that the main mechanism involved in harmine's antidepressant effect is an increase in hippocampal BDNF, though this may be dependent on treatment time and the precise region of the hippocampus.
An infusion of harman (1-methyl-beta-carboline) into the hippocampus of rats, or through systemic administration, increased the concentration of 5-HT [63]. In addition, metabolite levels of 5-HT degradation decreased dose-dependently, probably due to inhibition of MAO-A [63]. Another study showed that harman bonds to type 5-HT 2A serotonergic receptors but shows no affinity to dopaminergic or BZ receptors [64]. Injections of 2.5 and 10 mg/kg of harmane in rats under fear conditioning have been shown to increase plasma ACTH, corticosterone, 5-HT, and NA levels in limbic system structures. These results suggest that harman can modulate behavioral alterations, brain neurochemistry, and neuroendocrine functions through a mechanism that inhibits MAO-A [65].
Subchronic, oral administration for 21 days of the lyophilized extract of Rhazya stricta and alkaloid fractions (akuammidine, rhaziminine, and tetrahydrosecamine) to male rats inhibited the activity of the MAO-B enzyme, a mechanism through which the antidepressant-like effect may occur [66]. Another alkaloid in the Mitragyna speciosa plant has shown antidepressant effects. Administered to mice at doses of 10 and 30 mg/kg, i.p., mitragynine decreased immobility time on then TST and FST [67] and the release of corticosterone. Mitragynine's effect appears to be mediated through the neuroendocrine HPA (hypothalamus-adrenal-pituitary) axis [67].
Punarnavine administered at doses of 20 and 40 mg/kg, v.o., for 14 days decreased immobility on FST, MAO-A activity, and corticosterone levels in both stressed and unstressed mice [68], while treatment with evodiamine at 10 and 20 mg/kg in rats exposed to CUMS reversed the decrease in their preference for sugared water and immobility time on FST, but increased 5-HT and NA levels and the protein expression of BNDF in the hippocampus. However, it reduced corticosterone levels, suggesting that it likely modulates monoamines and BDNF-TrkB signaling in the hippocampus [69]. Chronic administration of piperine in rats at 5, 10, and 20 mg/kg has shown antidepressant-like effects on FST, probably due to a serotonergic mechanism [70]. At a dose of 30 mg/kg, protopine produced an antidepressant effect on TST in mice, perhaps by inhibiting the 5-HT and NA transporters, since in vitro studies have reported that it produces an inhibitory effect on these elements [71].
Because alkaloids have powerful antidepressant effects, many are used in clinical practice with effective therapeutic results. Preclinical studies have clearly demonstrated the antidepressant effects of alkaloids, but evidence of their mechanisms of action is still deficient or unclear. Alkaloids isolated from plants are an option for treating depression, but more studies are needed at the preclinical level to evaluate their potency, efficacy, and safety before they can be incorporated into clinical practice.

Alkaloids with anxiolytic effect
The alkaloids gelsemine, koumine, and gelsevirine exerted anxiolytic effects in single doses of 2 and 10 mg/kg in mice in EPM and LDB [72]. Gelsemine in low doses (10 À6 M and 10 À14 M) administered to male rats for 7 days also showed anxiolytic effects in EPM [73]. Koumine has shown this effect on VCT in mice at doses of 0.167, 0.5, or 1.5 mg/kg [74]. Other reports indicate that the decoction of the African peach root (Nauclea latifolia) injected intraperitoneally in mice produces dose-dependent anxiolytic-like effects (16,40,80, and 160 mg/kg) in EPM. Its effect has been attributed to isoquinoline-type alkaloids [75], but no reports have yet substantiated this claim. One study reported that the isoquinoline alkaloid berberine hydrochloride has both antipsychotic and anxiolytic properties. In this regard, studies have shown that a dose of 100 mg/kg/day produces anxiolytic effects and can modulate the gratifying effects induced by methamphetamine in rats [76].
The aqueous extract of Eschscholzia californica Cham (200 mg/kg, p.o.) has shown anxiolytic-like effects on the LDB test in mice that have been attributed to action on GABA A receptors [77]. Administration of the aqueous extract of the stem of Uncaria rhynchophylla (200 mg/kg), which contains the alkaloid rhynchophylline, in a single dose, or for 7 days, produced an anxiolytic effect in EPM by acting on the 5-HT 1A receptor [78]. Another example has been implicated to alkaloids with anxiolytic effects were the hydroethanolic extract of Davilla rugosa produced anxiolytic-like effects in EPM when administered to rats at 15 mg/kg [79]. Two other plants that contain alkaloids with anxiolytic effects (erythravine and 11a-hydroxy-eritravine) are Erythrina velutina and Erythrina mulungu. A study in mice showed that chronic administration (23-26 days) of the hydroalcoholic extract of the stem of E. velutina at 100 mg/kg produced an anxiolytic effect in EPM [80], while acute treatment with 200 mg/kg of E. mulungu showed an anxiolytic response in LDB comparable to that of diazepam [81]. Nevertheless, in this study did not identify the content or type of alkaloids in these extracts.
Turning to the species Magnolia (Magnolia spp.), we find that at least four anxiolytic components have been identified: honokiol, 4-O-methylhonokiol, magnolol, and obovatol. Administering honokiol (1 mg/kg) for 7 days had an anxiolytic-like effect on mice tested in EPM with results similar to those of diazepam [82]. That treatment increased the activity of the enzyme glutamic acid decarboxylase (GAD-subtype 65) in the hippocampus, but not the cortex of the mice brains. This, in turn, increased the release of GABA and reduced anxiety behavior. GAD65 is located on the terminal nerve and regulates the release of GABA to the synaptic cleft [83]. On this topic, there are reports that GAD65-deficient mice show higher anxiety levels [83]. Administering 4-O-methylhonokiol (0.1, 0.2, and 0.5 mg/kg) to mice in a single dose or during 7 days produced anxiolytic effects in EPM through the benzodiazepine site by binding to the GABA A receptor [84]. This is similar to observations of obovatol at doses of 0.2, 0.5, and 1 mg/kg [85]. In addition, an increase in the expression of the GABA A receptor α1 subunit [84] and of the α1 subunit in the amygdala and Cl (À) currents was observed [85]. The diterpene alkaloid songorine has shown anxiolytic effects when male mice were tested on VCT at a dose of 0.25 mg/kg [29], revealing an effect similar to that of phenazepam.
Numerous reports attribute anxiolytic activity to a broad range of plants. However, isolating and identifying the alkaloids responsible for this activity have not advanced substantially. Preclinical reports point to a common mechanism of action that modulates the GABAergic and serotonergic systems. The data described here justifies the need to conduct preclinical and clinical studies using alkaloids as alternative treatments for some anxiety disorders.

Sterols with anxiolytic and antidepressant effects
Plants synthesize a class of sterols called phytosterols, whose chemical structure is similar to that of cholesterol. Phytosterols are found in nuts, vegetable oils, cereals, fruits, vegetables, and various plants [86]. Some 40 different types have been identified, including β-sitosterol, campesterol, fucosterol, and stigmasterol. Due to their lipidic nature and glycosylated forms, they are able to cross the bloodbrain barrier and impact the CNS [87]. Trevisan et al. [88] suggest that α-spinasterol has the ability to cross the blood-brain barrier and exert an antagonistic effect on the transient potential receptor V1 (TRPV1). When these receptors are expressed in various areas of the brain-prefrontal cortex, amygdala, hypothalamus, and hippocampus-their activation augments the release of glutamate and, consequently, that of GABA, DA, or other catecholamines [89]. This fact involved TRPV1 receptors in the mechanism underlying the etiology of depression and anxiety. This was corroborated by Socała and Wlaź [90] by administering (1 and 2 mg/kg, i.p.) α-spinasterol to male mice and testing them on FST. Their results suggest an antidepressant effect. Also, intracerebroventricular (i.c.v.) coadministration of 50 μg of the TRPV1 receptor antagonist capsazepine/mouse with an ineffective dose of 0.5 mg/kg, i.p., of α-spinasterol, also reduced immobility time on FST, indicating the involvement of TRPV1 in the neurobiology of depression. However, α-spinasterol itself (0.5, 1, and 2 mg/kg) was unable to produce anxiolytic-like effects in EPM or LDB. In this sense, TRPV1-knockout mice manifested less anxiety behavior on the same tests [91]. Socała and Wlaź [90] proposed that α-spinasterol may be able to activate CB1 receptors with greater affinity because those neurons coexpress these receptors in various brain structures whose activation could activate TRPV1 receptors simultaneously to block their possible anxiolytic effects. In another work, administering fucosterol (10, 20, 30, or 40 mg/kg, v.o.) to male mice produced antidepressant effects on FST and TST, with the 20 and 30 mg/kg doses achieving an effect of comparable efficacy to 20 mg/kg of fluoxetine, a standard dose in humans [92]. Those doses also exerted an acute effect that increased BDNF levels in the hippocampus, a limbic structure involved in mood regulation. Fucosterol also blocked the decrease in 5-HT, 5-HTIIA, and NA levels in mice brains generated by the stress of FST. The effect of fucosterol on that test was similar to that of the positive control drug, but it was unable to prevent the reduction of DA, another factor caused by FST. These findings suggest that the antidepressant mechanism is mediated by increasing monoamines and reducing the rate of 5-HT metabolism. Fucosterol did not modify either motor or exploratory activity and showed no neurotoxic effects [92]. Similar results were found when administering β-sitosterol at 10, 20, and 30 mg/kg for 7 days. In that case, 30 mg/kg exerted effects similar to those of 20 mg/kg of fluoxetine on FST and TST. Finally, the effects on monoamine levels in mice brains confirm that sterols modify the serotonergic and noradrenergic systems but do not impact the dopaminergic system [93].
Another case involved αand β-amyrin (αβAMY) isolated from the resin of the stem of Protium heptaphyllum plants obtained and identified from hexane-ethyl acetate fractions analyzed by TLC. That process produced 450 mg of αβAMY made up of 67% αand 33% β-amyrin, which were further purified and tested on FST. Administering 2.5 and 5 mg/kg of αβAMY via i.p. or p.o. decreased immobility time, but the most effective treatment was the 2.5-mg/kg dose via the p.o. route. However, the effects produced by this route were similar to those of imipramine at 30 and 10 mg/kg. Imipramine is a tricyclic antidepressant (TCA) that blocks reuptake of both serotonin and norepinephrine. In addition, a pharmacological synergism between 1 or 2.5 mg/kg of αβAMY and 10 mg/kg of imipramine was observed, but not between 2.5 mg/kg of αβAMY and 4 mg/kg of paroxetine (SSRIs). These effects were blocked by pretreatment with 2 mg/kg of reserpine, an inhibitor of the vesicular catecholamine transporter that facilitates vesicular storage. This result suggests a possible mechanism of action through activation of the noradrenergic system [94]. The base structure that cholesterol and sterols share allows the latter to exert actions at the level of the CNS, as in the case of cholesterol. Cholesterol is a vital substance for neurons because it is required for vesicle transport and neurotransmitter release and as a precursor to neurosteroids. It is also implicated in synaptic plasticity in relation to the formation of new synapses. For these reasons, studying sterols and their mechanisms of action on the CNS is extremely important because of the anxiolytic and/or antidepressant effects they exert.

Final comments and conclusion
This broad review constitutes a significant contribution to our understanding of the mechanisms of action that allow plants to produce antidepressant and anxiolytic effects (see Figure 1 for a summary). However, most of the studies reviewed were conducted with mice, due to the low yields achieved when isolating the metabolites of plant extracts [95], which limit the amount of testing that can be done. Another concern is that some infusions or extracts used as household remedies lose their antidepressant or anxiolytic effects when fractioned [96]. These findings indicate that in some cases it may be necessary to keep the metabolites together at the concentrations present in the original infusion or extract that has a proven therapeutic effect. As explained herein, some metabolites share pharmacological targets, which explains why they lose their effect when separated and emphasizes the importance of using standardized extracts with demonstrated therapeutic effects in animal and human studies [97,98]. Unfortunately, very few clinical studies have evaluated the potential antidepressant or anxiolytic effects of isolated metabolites, so a great deal of work remains to be done.
Several observations suggest that active metabolites share the mechanism of action of antidepressant and anxiolytic drugs like SSRIs, SNRIs, MAOIs, DDNRI, and BZDs, but we should emphasize that some metabolites-at least in preclinical studies-produced better effects than conventional drugs, even at lower doses, while others presented a pharmacological synergism between both types at suboptimal doses that improved the effects exerted separately at higher doses. A second shared characteristic is that they contain active stereoisomers and probably, some metabolites, once metabolized, could become more active. This encourages us to consider a significant number of substances with anxiolytic and/or antidepressant pharmacological profiles and invites us to take on the challenge of evaluating their pharmacokinetics, pharmacodynamics, and safety.
In conclusion, terpenes, flavonoids, alkaloids, and sterols share mechanisms of action that include activation of the critical enzyme for catecholamine synthesis (e.g., tyrosine hydroxylase) or the inhibition of their limiting enzymes, MAO-A and MAO-B, and transporters, thus stimulating the vesicular storage monoamine and the release of neurotransmitters toward the synaptic cleft. Finally, they can prevent the production of ROS and inhibit NO synthesis and, further downstream, interact with the 5-HT 1A , 5-HT 2A , D1, D2, GABA A receptors, and α1, α2, β-adrenoceptors that contribute to stimulating PKA. One consequence is that CREB increases BDNF levels, which foster the appearance of dendritic contacts that improve cerebral neurotransmission and modulate the emotions.

Perspective
This chapter discusses the efficacy of some plant metabolites in treating anxiety and depression disorders, as demonstrated in preclinical studies. In the future, this option for treating such disorders will allow us to reduce treatment costs and moderate the side effects produced by drugs currently in use. However, our review also points out that few clinical studies have focused on the pharmacokinetic and pharmacodynamic processes involving metabolites that would permit the safe use of these extracts. Despite this, research has shown that traditional medicine, especially forms that use medicinal plants that have been passed down through several generations, constitutes an important alternative for health care.