Anti-SARS-CoV-2 Activity of Ampelozizyphus amazonicus (Saracura-Mirá): Focus on the Modulation of the Spike-ACE2 Interaction by Chemically Characterized Bark Extracts by LC-DAD-APCI-MS/MS

Traditional medicine shows several treatment protocols for COVID-19 based on natural products, revealing its potential as a possible source of anti-SARS-CoV-2 agents. Ampelozizyphus amazonicus is popularly used in the Brazilian Amazon as a fortifier and tonic, and recently, it has been reported to relieve COVID-19 symptoms. This work aimed to investigate the antiviral potential of A. amazonicus, focusing on the inhibition of spike and ACE2 receptor interaction, a key step in successful infection. Although saponins are the major compounds of this plant and often reported as its active principles, a polyphenol-rich extract was the best inhibitor of the spike and ACE2 interaction. Chemical characterization of A. amazonicus bark extracts by LC-DAD-APCI-MS/MS before and after clean-up steps for polyphenol removal showed that the latter play an essential role in maintaining this activity. The effects of the extracts on viral replication were also assessed, and all samples (aqueous and ethanol extracts) demonstrated in vitro activity, inhibiting viral titers in the supernatant of Calu-3 cells after 24 hpi. By acting both in the SARS-CoV-2 cell entry process and its replication, A. amazonicus bark extracts stand out as a multitarget agent, highlighting the species as a promising candidate in the development of anti-SARS-CoV-2 drugs.


Introduction
Three years have passed since COVID-19 was declared a pandemic. Currently, confirmed cases have reached more than 600 million and led to six million deaths worldwide [1]. SARS-CoV-2, the etiologic agent of COVID-19, is a single-stranded positive-sense RNA virus with a 27-32 kb-long genome [2,3], which is divided into two open reading frames (ORFs) that encode non-structural proteins and subgenomic RNAs that encode the structural proteins [4,5]. Like other enveloped viruses, coronaviruses have unique proteins responsible for the recognition of and binding to host cell receptors [6,7]. The spike protein (SP) has 180 to 200 kDa and is a structural glycoprotein present in all coronaviruses and which is formed by a signal peptide and two subunits: S1, which contains the receptor binding domain (RBD), and S2, in which the fusion peptide is present [8,9]. They are found in the homotrimeric form, as a crown-like ornamentation around the viral particle, which gave the name to this group of viruses [9].
In the viral particle, SP exists in its pre-fusional conformation, and its specific cleavage is mandatory to trigger the conformational change that leads to viral and cellular membrane fusion [9,10]. Two possible entry pathways into host cells have already been described, and both start with the recognition and binding of SP RBD to its major ligand, the angiotensinconverting enzyme 2 (ACE2) [9,10].
Although the vaccine arsenal is widely available to the population and complete vaccination clearly reduces severe cases, it is still imperative to consider mild and moderate infections, whose needs are currently unmet. Controlling non-severe cases will help to reduce the viral spread and reduce the social and economic impact of COVID-19. Therefore, investments in new therapeutic approaches are welcome.
Since the pandemic's beginning, natural products' relevance as a potential source of active compounds in the fight against SARS-CoV-2 has exponentially grown, and records in the literature show how these substances represent a promising approach. Plants' secondary metabolites, such as saponins, terpenes, alkaloids, and phenolic acids, have been shown to display antiviral activity against several specimens of the Coronaviridae family [11][12][13]. To date, a few works have described several representatives from different classes of natural products, such as flavonoids and alkaloids, that clearly exhibit a promising activity against SARS-CoV-2 [14][15][16][17], reinforcing the necessity of continued efforts to find potential alternatives to be used in the clinical management of mild and moderate COVID-19 cases.
In 2020, a great number of in silico studies emerged as an alternative to evaluate natural products' ability to act as protease, helicase, and replicase inhibitors, especially because of the limited infrastructure to manipulate SARS-CoV-2. In silico analyses provided valuable and reliable information on a range of substances with potential anti-SARS-CoV-2 activity in a relatively short time; these are currently being tested in in vitro and in vivo assays [18][19][20][21][22]. Despite viral enzymatic inhibition being the main mechanism of action for most of the currently available and in development antivirals, blockage of SP interaction with ACE2 is becoming an attractive approach for a new generation of antivirals since it has the potential to hamper viral recognition and infection of host cells, preventing viral propagation inside the organism as well as contributing to a reduced mutation rate and thus the appearance of new variants [23].
Lately, in vitro methodologies using isolated viral components have gained space in an attempt to increase the rate of the discovery of potential inhibitors, whether they come from natural or synthetic sources. Some of the technologies used for this purpose include pseudoviruses to investigate the cell entry process, FRET (fluorescence resonance energy transfer) assays to evaluate enzymatic activity, and colorimetric or luminescent assays to evaluate protein interactions [24].
As part of a broader ongoing screening project on natural inhibitors of the interaction between SP and ACE2, a library of plant extracts belonging to Brazilian biome biodiversity (in the Amazon, Atlantic Forest, and Cerrado region) was evaluated, from which the extracts from A. amazonicus Ducke stood out as the most promising. A. amazonicus is one of 984 species of the Rhamnaceae Juss. family [25], which is endemic to the Amazonian states of South America; its distribution includes the regions of Peru, Colombia, Venezuela, Guyana, Suriname, French Guiana, and Brazil [26][27][28]. The bark and roots of the plant, popularly known as "cerveja de indio" or "saracura-mirá", have been used in Brazil for over 30 years by the Amazonian traditional communities to prepare a drink with stimulant, tonic and fortifying properties to combat and prevent malaria, among other uses [29]. Studies with a spray-dried aqueous bark extract of the plant have demonstrated its properties as an immune enhancer and adaptogen [30,31]. More recently, the species has been recorded as being used for the prevention and treatment of COVID-19 by the Sateré Mawé indigenous people, as well as by riverine communities in the Brazilian Amazon during the pandemic [32]. Reports on the chemical composition of the plant focus basically on its triterpenes and saponins, which are believed to be responsible for the wide range of biological activities reported for the species [29][30][31][33][34][35]. In fact, aqueous extracts prepared from the plant produce abundant foaming, due to the presence of up to 48% of saponins [31,33], which explains one of the popular names "cerveja de indio" which means "Indian beer". However, although never studied, phenolic compounds are also present in these plant extracts, especially ethanol compounds (personal observations).
Since the literature shows the potential of A. amazonicus as a source of useful compounds for the attenuation of COVID-19 symptoms, we aimed in the present study to evaluate the potential of its chemically characterized bark extracts to inhibit the receptorbinding domain (RBD) of SARS-CoV-2 spike protein and ACE2 interaction before and after clean-up steps with lead acetate for the removal of phenolics, and to inhibit SARS-CoV-2 replication in vitro.

A. amazonicus Extracts Inhibit at Least Half of RBD:ACE2 Complex Formation In Vitro
The ability of four different extracts of A. amazonicus to interfere in the formation of RDB:ACE2 complex was assessed with the aid of a Lumit™ kit (Promega), a no-wash, add-and-read bioluminescence-based kit suitable for the trial of potential inhibitory samples [36].The tested samples included two aqueous extracts prepared according to traditional medicine, which were previously shown to have immunomodulatory and antiinflammatory properties [30,31], and two ethanol extracts obtained according to the literature [34,35]. At 250 µg·mL −1 , all the available extracts were able to inhibit at least 50% of the interaction between RBD and ACE2 ( Figure 1). roots of the plant, popularly known as "cerveja de indio" or "saracura-mirá", have been used in Brazil for over 30 years by the Amazonian traditional communities to prepare a drink with stimulant, tonic and fortifying properties to combat and prevent malaria, among other uses [29]. Studies with a spray-dried aqueous bark extract of the plant have demonstrated its properties as an immune enhancer and adaptogen [30,31]. More recently, the species has been recorded as being used for the prevention and treatment of COVID-19 by the Sateré Mawé indigenous people, as well as by riverine communities in the Brazilian Amazon during the pandemic [32]. Reports on the chemical composition of the plant focus basically on its triterpenes and saponins, which are believed to be responsible for the wide range of biological activities reported for the species [29][30][31][33][34][35].
In fact, aqueous extracts prepared from the plant produce abundant foaming, due to the presence of up to 48% of saponins [31,33], which explains one of the popular names "cerveja de indio" which means "Indian beer". However, although never studied, phenolic compounds are also present in these plant extracts, especially ethanol compounds (personal observations).
Since the literature shows the potential of A. amazonicus as a source of useful compounds for the attenuation of COVID-19 symptoms, we aimed in the present study to evaluate the potential of its chemically characterized bark extracts to inhibit the receptor-binding domain (RBD) of SARS-CoV-2 spike protein and ACE2 interaction before and after clean-up steps with lead acetate for the removal of phenolics, and to inhibit SARS-CoV-2 replication in vitro.

A. amazonicus Extracts Inhibit at Least Half of RBD:ACE2 Complex Formation In Vitro
The ability of four different extracts of A. amazonicus to interfere in the formation of RDB:ACE2 complex was assessed with the aid of a Lumit™ kit (Promega), a no-wash, add-and-read bioluminescence-based kit suitable for the trial of potential inhibitory samples [36].The tested samples included two aqueous extracts prepared according to traditional medicine, which were previously shown to have immunomodulatory and anti-inflammatory properties [30,31], and two ethanol extracts obtained according to the literature [34,35]. At 250 µg.mL −1 , all the available extracts were able to inhibit at least 50% of the interaction between RBD and ACE2 ( Figure 1). were assessed with the Lumit™ immunoassay (Promega). #74-Aqueous bark extract, prepared on a laboratory scale to mimic the traditional beverage; #75-Ethanol bark extract, obtained from the remaining "cake" after the aqueous extraction of the plant material which generated extract #74. #76-Ethanol bark extract, prepared by percolation; #156-Aqueous bark extract, prepared in a pilot-scale industrial plant, according to the traditional method. These data are the result of three independent technical replicates. were assessed with the Lumit™ immunoassay (Promega). #74-Aqueous bark extract, prepared on a laboratory scale to mimic the traditional beverage; #75-Ethanol bark extract, obtained from the remaining "cake" after the aqueous extraction of the plant material which generated extract #74. #76-Ethanol bark extract, prepared by percolation; #156-Aqueous bark extract, prepared in a pilot-scale industrial plant, according to the traditional method. These data are the result of three independent technical replicates.
The most active extract (#76) was able to inhibit 92 ± 0.03% of this interaction at this concentration. Notably, a very similar inhibition pattern can be observed for both aqueous extracts (#74 and #156) and ethanol extract #75, which was prepared from the resulting cake after the aqueous extraction of the barks, while ethanol extract #76 was obtained from direct extraction of the plant material with ethanol [34].

Chemical Characterization of A. amazonicus Bark Extracts Reveal the Presence of Saponins, Triterpenes, and Phenolic Compounds
The aqueous (#74 and #156) and ethanol extracts (#75 and #76) of A. amazonicus were analyzed by ultra-high-performance liquid chromatography with PDA and coupled to mass spectrometry (UHPLC-PDA-APCI-MS/MS). All extracts showed a very similar saponin profile (Figure 2), which is the main class of compounds described for the species, as reported before [31,34,35,37,38], but a different profile of UV absorbing compounds (as recorded by the PDA detector) not yet reported in the literature for this species.  Table 1 shows the main saponins annotated in the studied extracts, all of which bear dammarane-type triterpene aglycones (Figure 3), as previously reported for A. amazonicus  The UHPLC-APCI(-)MS/MS analyses show a complex chemical profile and two characteristic regions in the chromatogram, one formed by UV-absorbing compounds (0-5 min), and the other one represented by triterpenes and saponins (10-25 min). Table 1 shows the main saponins annotated in the studied extracts, all of which bear dammaranetype triterpene aglycones (Figure 3), as previously reported for A. amazonicus [31,34,35].
Saponin structures were proposed based on the MS fragmentation profile. Compound 1 shows a characteristic MS/MS spectrum of the aglycone 16-keto-tetrahydroxydammar-23-ene, with the loss of 142Da corresponding to its side chain (C 8 H 14 O 2 ) ( Figure S1). Compound 2 contains an acetyl group attached to the konarigenin aglycone and exhibits characteristic losses of this residue (C 2 H 4 O 2 ) ( Figure S2). Saponins 3 and 4 bear a C-31 aglycone of the 16-keto-tetrahydroxydammar-24-methylene type, whose fragmentation profiles show losses of C 9 H 16 O 2 (156Da) corresponding to its side chain ( Figures S3 and S4).  [35]. This process occurs through common fragmentation reactions in API sources such as APCI. In all MS/MS spectra of the annotated saponins, characteristic neutral losses of sugar residues were observed (162Da for hexoses, 146Da for deoxyhexoses, and 132Da for pentoses).  Saponin structures were proposed based on the MS fragmentation profile. Compound 1 shows a characteristic MS/MS spectrum of the aglycone 16-keto-tetrahydroxydammar-23-ene, with the loss of 142Da corresponding to its side chain (C8H14O2) ( Figure S1). Compound 2 contains an acetyl group attached to the konarigenin aglycone and exhibits characteristic losses of this residue (C2H4O2) ( Figure  S2). Saponins 3 and 4 bear a C-31 aglycone of the 16-keto-tetrahydroxydammar-24-methylene type, whose fragmentation profiles show losses of C9H16O2 (156Da) corresponding to its side chain ( Figures S3 and S4).  [35]. This process occurs through common fragmentation reactions in API sources such as APCI. In all MS/MS spectra of the annotated saponins, characteristic neutral losses of sugar residues were observed (162Da for hexoses, 146Da for deoxyhexoses, and 132Da for pentoses).
Contrary to the saponins, the MS profile of the extracts in the region of the UV absorption compounds (0-5 min) is not similar ( Figure S7). Preliminary analysis of the λmax of the UV-absorbing peaks suggested that compounds eluted in this zone of the chromatogram could be attributable to a phenolic class of compounds such as flavonoids. In Contrary to the saponins, the MS profile of the extracts in the region of the UV absorption compounds (0-5 min) is not similar ( Figure S7). Preliminary analysis of the λ max of the UV-absorbing peaks suggested that compounds eluted in this zone of the chromatogram could be attributable to a phenolic class of compounds such as flavonoids. In fact, the total phenolic content (TPC) of the most active extract #76 was 46.3 mgEAG/g, confirming the presence of phenols in the extract.

Clean-Up Procedure with Lead Acetate Removes Phenolic Compounds and Promotes Differences in Biological Activity
Taking into account the results of the inhibition rates of the formation of the RBD:ACE2 complex by A. amazonicus extracts (Figure 1), and their chemical characterization by mass spectrometry showing a similar saponin profile, a clean-up step on the most active ethanol extract (#76) was performed to separate the UV-absorbing compounds (probably phenolic) from the saponins, aiming to identify whether they could be responsible for the enhanced activity of this extract.
Treatment with lead acetate is a classic method used to precipitate phenolic compounds [39], and therefore it was used to clean up the resulting extract from these substances. In sequence, extracts #76 and #76Pb (after clean-up with lead acetate) were evaluated in the RBD:ACE2 complex formation at different concentrations ( Figure 4). most active ethanol extract (#76) was performed to separate the UV-absorbing compounds (probably phenolic) from the saponins, aiming to identify whether they could be responsible for the enhanced activity of this extract.
Treatment with lead acetate is a classic method used to precipitate phenolic compounds [39], and therefore it was used to clean up the resulting extract from these substances. In sequence, extracts #76 and #76Pb (after clean-up with lead acetate) were evaluated in the RBD:ACE2 complex formation at different concentrations ( Figure 4). The data show that extract #76 interferes in the RBD:ACE2 complex formation in all concentrations tested (25 to 125 µg.mL −1 ), reducing the interaction by at least 40% in a dose-dependent pattern ( Figure 4). The highest concentration (125 µg.mL −1 ) reduced the interaction by 92.45 ± 3.65%, which is similar to what was observed at 250 µg.mL −1 ( Figure  1). Interestingly, the removal of phenolic compounds by lead acetate clean-up abolished the inhibition potential of #76 extract by 70% at the highest concentration (125 µg.mL −1 ), as shown in Figure 4, suggesting that phenolic compounds and not saponins are responsible for hampering RDB and ACE2 interaction.
Aiming to compare the chemical profiles of ethanol extracts before and after clean-up, the samples were subjected to ultra-high performance liquid chromatography coupled to tandem MS in the negative ion mode (UHPLC-APCI(-)MS/MS), followed by alignment in MZmine 2.53. Figure 5 clearly shows that the UV-absorbing compounds' region was simplified upon the clean-up step, indicating the removal of phenolic compounds, the main ones of which were tentatively identified. The data show that extract #76 interferes in the RBD:ACE2 complex formation in all concentrations tested (25 to 125 µg·mL −1 ), reducing the interaction by at least 40% in a dose-dependent pattern ( Figure 4). The highest concentration (125 µg·mL −1 ) reduced the interaction by 92.45 ± 3.65%, which is similar to what was observed at 250 µg·mL −1 (Figure 1). Interestingly, the removal of phenolic compounds by lead acetate clean-up abolished the inhibition potential of #76 extract by 70% at the highest concentration (125 µg·mL −1 ), as shown in Figure 4, suggesting that phenolic compounds and not saponins are responsible for hampering RDB and ACE2 interaction.
Aiming to compare the chemical profiles of ethanol extracts before and after clean-up, the samples were subjected to ultra-high performance liquid chromatography coupled to tandem MS in the negative ion mode (UHPLC-APCI(-)MS/MS), followed by alignment in MZmine 2.53. Figure 5 clearly shows that the UV-absorbing compounds' region was simplified upon the clean-up step, indicating the removal of phenolic compounds, the main ones of which were tentatively identified.
The tandem mass spectrum produced by compounds 7 and 8, showed a pattern typical of a C-glycosyl derivative. The fragmentation profiles of these C-  Figure S9). Compound 8, which exhibits a λ max at 285 nm, as is typical of the chalcones, was annotated as 3 ,5 di-C-glucosyl phloretin ( Figure 6). The similarity of the mass fragmentation patterns of 7 and 8 suggests the hypothesis that compound 7 is probably a C-glucosyl flavonoid derivative.  (Table 2) [40,41]. Compound (7) remained unidentified, but it was possible to hypothesize its structure as a C-glycosyl flavonoid based on the similarity of its mass fragmentation pattern when compared with 3′,5′di-C-glucosyl phloretin, (8).  exhibits a λmax at 285 nm, as is typical of the chalcones, was annotated as 3′,5′di-C-glucosyl phloretin ( Figure 6). The similarity of the mass fragmentation patterns of 7 and 8 suggests the hypothesis that compound 7 is probably a C-glucosyl flavonoid derivative. While compound 8 has been previously described in Ziziphus spina-christi, a member of the Rhamnaceae family [40], no data were found in the literature for the mass and fragmentation patterns observed for compound 7, suggesting that this may be an as yet undescribed C-glucosyl flavonoid.  While compound 8 has been previously described in Ziziphus spina-christi, a member of the Rhamnaceae family [40], no data were found in the literature for the mass and fragmentation patterns observed for compound 7, suggesting that this may be an as yet undescribed C-glucosyl flavonoid. This is the first report on the phenolic composition of bark extracts of A. amazonicus, in which only saponins and triterpenes have been previously identified [31,34,35,37,38]. Despite saponins being the major compounds in the plant extracts, the phenolic compounds described in this study seem to be responsible for the inhibition of SP and ACE2 interaction.
Polyphenolic compounds are among the most promising chemical class of SARS-CoV-2 inhibitors [43]. A recent review compiled the structure-activity relationships of plant phenolics with antiviral activities against human coronaviruses [44], showing that data on viral entry inhibitors are scarce. The flavonoid luteolin has been shown to inhibit the entry of HIV-luc/SARS pseudotyped virus and wild-typed SARS-CoV into Vero E6 cells without substantial cytotoxicity. Emodin, an anthraquinone, inhibited the interaction between SARS-CoV S protein and ACE2 receptor in a dose-dependent manner.
However, most of the studies with SARS-CoV-2 SP and ACE2 interactions are in silico studies. Computational studies showed that catechin and curcumin bind the interface of the 'RBD:ACE2-complex', suggesting the efficacy of these two polyphenols in hindering the formation of SP and ACE2 complex [45]. Structure-based virtual screening of the Taiwan Database of Extracts and Compounds identified 39 natural products targeting the viral receptor-binding domain (RBD) of the SARS-CoV-2 spike protein; among them, amentoflavone was selected as the best one [46]. Recently, surface plasmon resonance of candidate molecular binders has been performed to determine the binding affinities and kinetics for RBD of a group of five compounds previously selected by a virtual screening of a natural product database, one of which was 5,4 -di-hydroxy-6,7-di-glucosyl-flavanone [47].
Concerning chalcones, a recent in silico study screened a library of natural and synthetic chalcones against homology-modeled S protein. Analysis of protein-ligand docking revealed that phloretin can interact with the spike proteins' key RBD, pointing to this class of natural compounds as promising [48].
Interestingly, in another recent in silico study, phloretin showed a strong inhibitory effect on the crucial residues ASN417, SER496, TYR501, and HIS505 of the omicron receptorbinding domain (RBD), which are supported for the inviolable omicron and angiotensinconverting enzyme II (ACE2) receptor interaction [49]. Additionally, the authors state that phloretin is suggested as the antiviral drug against omicron B.1.1.529, which strongly supports our findings that 3 ,5 di-C-glucosyl phloretin is in part responsible for blocking the SP and ACE2 interaction in extract #76.
On the other side, flavones such as quercetin and their derivatives are currently considered promising compounds for COVID-19 treatment [50]. Quercetin possesses a 15-carbon skeleton with a chromone core comprising bicyclic 1,4-benzopyrone (A-and Crings) substituted on carbon 2 with a catechol moiety (B-ring), with hydroxyl groups at positions 3 , and 4 with electron-donating activity [51]. The presence of hydroxyl groups in the B-ring reflects positively on its binding affinities to the viral target [52]. In silico and in vitro studies have been focused on quercetin, testing it experimentally against some targets such the spike protein of SARS-CoV-2, with promising results showing its potential as an antiviral agent [53,54].
In an in silico study [55], the authors suggest that beyond the interaction with active sites of the spike protein of SARS-CoV-2, polyphenols from Geranii herba could interact in the cell surface receptor glucose-regulated protein 78 (GRP78) to regulate the cell signaling to release ER stress and other processes. They could also interact either directly or indirectly with the Cys/His dyad of the main protease to inhibit viral replication. In this way, polyphenols could interfere with viral pathogenesis in various stages [56].
Moreover, flavonoids are known for their potential ability to regulate inflammatory mediators and inhibit endothelial activation, among other factors, which might be beneficial in regulating the cytokine storm during SARS-CoV-2 infection, and may positively influence inflammatory changes associated with COVID-19 [57]. Additionally, phenol-rich extracts from plants have also demonstrated antiinflamatory activity [58,59].
The results presented here show evidence that supports that compounds belonging to chalcones and flavones in the A. amazonicus extracts were able to inhibit the binding of the viral spike protein to its ACE2 receptor, which is one vital mechanism for the entry and replication of SARS-CoV-2 viruses.

The Characterized Extracts Can Inhibit SARS-CoV-2 In Vitro Replication
Since plant extracts represent complex matrices composed of a wide range of substances, it is possible that the same extract targets different steps in the virus life-cycle, which results in distinguishable antiviral activity. Considering the RBD:ACE2 inhibition promoted by A. amazonicus extracts, we assessed whether they would be able to act at a different stage of infection. For this purpose, the Calu-3 cells, a well-established model for SARS-CoV-2 in vitro experimental infection, were used to evaluate the cytotoxicity and antiviral activity of all plant extracts [60,61].
Although saponins are known to be potential cytotoxic agents [62], the cell toxicity determined by MTT assay showed that none of the uncleaned extracts were toxic to Calu-3 cells at the highest concentration tested (200 µg·mL −1 ). The only exception was #76_Pb, which reduced cell viability by 40% at the highest concentration ( Figure 7A). Despite this, all the extracts exhibited a concentration required to reduce cell viability by 50%, with values (CC 50 ) higher than 200 µg·mL −1 .
The antiviral activity of the extracts was determined by the reduction in SARS-CoV-2 replication in Calu-3 cells. All the extracts were able to inhibit SARS-CoV-2 replication at 24 h ( Figure 7B). The concentration required to inhibit viral infection-induced cytopathogenicity by half was most promising for the extracts #75 and #156, which presented EC 50 values < 25 µg·mL −1 (Table 3). antiviral activity of all plant extracts [60,61].
Although saponins are known to be potential cytotoxic agents [62], the cell toxicity determined by MTT assay showed that none of the uncleaned extracts were toxic to Calu-3 cells at the highest concentration tested (200 µg.mL -1 ). The only exception was #76_Pb, which reduced cell viability by 40% at the highest concentration ( Figure 7A). Despite this, all the extracts exhibited a concentration required to reduce cell viability by 50%, with values (CC50) higher than 200 µg.mL -1 .   According to the chromatographic profile shown in Figure 2, the samples seem to be very similar regarding the saponins profile, which may explain the fact that they are all active in inhibiting SARS-CoV-2 replication. Although the results obtained previously demonstrate that A. amazonicus extracts can act on the ligand-receptor interaction of a structural protein and that this action is linked to the phenolic compounds in the extract, antiviral activity is a complex process and can exhibit different possibilities of interference.
In this section, the results shown reflect that A. amazonicus extracts (especially #75 and #156) also act on a distinct target, quite possibly non-structural proteins, since the model represents a scenario in which the infection is already established. RBD:ACE2 binding assays demonstrate that saponins are not important for the interaction's inhibition, unlike the polyphenols in the extract. However, it is known that saponins have anti-inflammatory and immunomodulatory effects, properties already demonstrated for these same extracts and that are important in the clinical approach to COVID-19 treatment [30,31,63,64]. Additionally, this action would also support the traditional use of this species by Amazonian communities [32]. Interestingly, there is one record for an anti-SARS-CoV-2 compound in the Rhamnnaceae family: jubanine B, a cyclopeptide alkaloid from Ziziphus spina-christii (L) Desf. was demonstrated to be a promising ligand of 3CL pro protein, the main protease of the SARS-CoV-2 virus [65].
A. amazonicus is an Amazonian adaptogen [29], and plants with this characteristic could consist of interesting models to be explored in drug discovery against SARS-CoV-2. Several traditional Chinese medicine protocols have been evaluated since 2020 in the treatment of COVID-19, using in silico, in vivo, and in vitro approaches [66]. As the basis of these practices are contained in medicinal plants and natural products, the good results presented here show that alternatives from plant secondary metabolites are very promising.
Panax ginseng Meyer, popularly known as Korean ginseng, is a plant with a high content of saponins structurally similar to the ones present in A. amazonicus, with dammarane triterpene structure [67]. It has been shown, in vivo that transgenic mice that were fed with ginseng extracts had much lower virus titers than the ones that were not. Additionally, interferon-gamma (IFN-γ) production was significantly higher in the lung tissue of the mice fed with ginseng extracts [68], demonstrating the potential of this species in SARS-CoV-2 infection, not only in virus replication but also in host response. Similarly to A. amazonicus, P. ginseng is also used in traditional medicine as an immune system fortifier, and has been pointed out as a potential treatment for COVID-19 [69]. Therefore, A. amazonicus is an interesting species to be considered in new drug discoveries to counter SARS-CoV-2, since it presents promising substances with proven action in its replication process.
Four different A. amazonicus bark extracts from our biobank library (registered at the Sistema Nacional de Gestão do Patrimônio Genético e do Conhecimento Tradicional Associado, SISGEN, under the registration number CB8C853), coded with numbers #74, #75, #76 and #156, were used for in vitro assays.
In brief, extract #74 was prepared with 250 g of dried and ground bark submitted to extraction with boiling water (5% w/v) for 15 min and filtered. A second extraction was performed with boiling water (2.5%, w/v, 30 min). The extracts were mixed and infused into a spray-dryer nozzle unit of a Buchi Mini Spray Dryer B-290 (Buchi Laboratorius-Technik AG, Flawil 1, Switzerland) [30]. The extractive yield was 4.5%. Extract #75 was prepared with commercial ethanol (96%) from the cake resulting from the aqueous extraction that generated extract #74, until exhaustion of plant material. It was then filtered and concentrated under reduced pressure in a rotary evaporator, at a temperature not exceeding 40 • C. The extractive yield was 7.3%. For extract #76, ground and dried bark (346.544 g) were extracted with commercial ethanol (96%) by percolation until exhaustion. The extract was filtered, and ethanol was removed by rotary evaporation at 40 •C under reduced pressure [34]. The extractive yield was 8.61%. Finally, to extract #156, dried and ground bark was used for the preparation of extract at the Centroflora Group facility, Botucatu, SP, Brazil. The bark (14.9 kg) was submitted to a two-step extraction process as follows: extraction with boiling water (5% w/v) for 15 min, followed by a second extraction with boiling water (2.5%, w/v) for 30 min. The extracts were mixed and filtered (using a rotary drum filter with a mesh of 30 µm) and then concentrated until the total solid level reached 24% (w/v) (Bernauer evaporator). Final drying was achieved by spray drying as reported in [31]. The extractive yield was 4.52%.
A brief description of the methodologies is presented in Table 4. The work was authorized by Conselho de Gestão do Patrimônio Genético (CGEN) to access the traditional knowledge with bioprospecting purposes, by the Resolution CGEN no. 213 (6 December 2007), renewed in the Resolution CGEN no. 87/2012, and by the authorizations AEAC3E9 in 2018, and A5A49CA in 2022.

Clean-Up Procedure with Lead Acetate
Separately, 0.5 g of #76 was mixed with 20 mL of 50% ethanol (EtOH) solution and vortexed for 15 min. Then, 10 mL of 10% lead acetate was added in each tube and vortexed for 25 min at 50 • C. After centrifugation, the supernatant was collected, lyophilized, and coded as #76_Pb.

Sample Preparation for Chemical Investigation
Separately, 10 mg of each sample (#76, and #76_Pb) was transferred to 2 mL microtubes. Then, 1 mL of acetonitrile (ACN) was added to each microtube, sonicated for 15 min and vortexed for 30 s. After that, 200 µL of each solution was transferred to a 1.5 mL microtube and then 300 µL of ACN and 500 µL of purified water were added. The final solution (1:1, v:v) was vortexed for 30 s and centrifuged at 10,000 RPM for 15 min. 750 µL of supernatants were injected and analyzed.

Sample Preparation for RBD:ACE2 Inhibition Assays
Separately, 1 mg of each sample was transferred to a microtube and solubilized in 20 µL of dimethyl sulfoxide (DMSO-Sigma-Aldrich, St. Louis, MO, USA) and 80 µL ultrapure water. The solutions were vortexed for 1 min and then centrifuged (5804 R-Eppendorf) for 5 min at 10,000 RPM. After, the supernatant of each solution was collected, identified, and refrigerated.

LC-DAD-APCI-MS/MS Analysis
Ultra-high-performance liquid chromatography with an ultraviolet detector coupled to tandem mass spectrometry (UHPLC-UV-MS/MS) analyses was performed using a UHPLC Dionex TM UltiMate TM 3000 system, coupled to an LCQ Fleet (ThermoFisher Scientific, Waltham, MA, USA) consisting of an oven, a solvent degasser, an ultra-high-pressure pump, an autosampler, a diode array detector and a column temperature manager. An ACQUITY UPLC ® BEH C18 reversed-phase column (2.1 × 100 mm, 1.7 µm, Waters) was used at a flow rate of 0.35 mL·min −1 . The column temperature was kept at 30 • C, and the mobile phases were 0.1% formic acid in water for A, and acetonitrile for B. The gradient elution mode was as follows: 25% B in 0-2 min, 25-50% in 2-30 min, 50-98% in 30-31 min, 98-98% in 31-33 min, 98-25% in 33-35 min, and 25-25% in 35-38 min. Wavelength channels used in the DAD instrument were 210 nm, 254 nm, 275 nm, and 360 nm with a bandwidth of 2 nm.
The mass spectrometer (MS), equipped with an atmospheric pressure chemical ionization (APCI) source and an ion trap analyser (with 1.000 of resolution), was operated in negative ion mode. High-purity nitrogen (N 2 ) was used as the sheath gas (30 arbitrary units) and auxiliary gas (15 arbitrary units). High-purity helium (He) was used as the collision gas. The MS parameters used were 6.0 kV of source voltage, −2 V of capillary voltage, −116 V of tube lens voltage, 400 • C of capillary temperature and 400 • C of APCI vaporizer temperature. Full scan data acquisition (mass range: m/z 100-1500) and datadependent acquisition (topN = 3) were performed. The normalized collision energy of the collision-induced dissociation (CID) cell was set at 35 eV. Data obtained from LC-MS analyses were converted to mzML format using the Proteowizard software. Then, the data were processed in Mzmine 2.53.

Total Phenolic Content (TPC)
The phenolic content of the most active extract (#76) was assessed as previously described [70]. Briefly, 5 mg of the ethanol extract was diluted in 1 mL of commercial ethanol. Then, an aliquot of 200 µL of this solution was transferred to a new microtube with 800 µL of ultrapure water. The 1 mg/mL solution was used for quantification. In a flat-bottom 96-well plate, 0.1 mL aliquots were added to 0.5 mL of Folin-Ciocateu reagent (10% w/v). The calibration curve was constructed using gallic acid ranging from 0.5 mg/mL to 0.003 mg/mL. The sample was tested in 500 and 250 µg·mL −1 . The total phenolic content of the extract was expressed in milligrams of gallic acid equivalent per gram of extract (mg GAE/g).

Lumit™ RBD:ACE2 Interaction Assays
To assess if A. amazonicus extracts would have any potential in blocking SARS-CoV-2 entry into cells, we used the commercial kit Lumit™ SARS-CoV-2 spike RBD:ACE2 immunoassay (Promega), following the instructions of the manufacturer. Lumit™ methodology relies on the use of the RBD (receptor-binding domain) portion of the SARS-CoV-2 spike protein, as the one responsible for recognizing and binding to the ACE2 receptor. Briefly, samples were added to a white, flat bottom, 96-well plate in a concentration of 250 µg·mL −1 (80% ultrapure water, 20% dimethylsulfoxide), followed by the addition of RBD and ACE2 reagents (7.5 nM). Detection is performed by secondary antibodies coupled to light particles and the substrate for this reaction. The binding of the antibodies to proteins promotes the reconstitution of NanoBiT particles, a technology developed by Promega, generating bioluminescence which was recorded in a SpectraMax M5 (Molecular Devices) microplate reader.
For the PFU/mL assay, the monolayers of Vero E6 (African green monkey kidney, ATCC CRL-1586) cells (1.5 × 10 4 cell/well) were exposed to different dilutions of viral supernatants (1:200-1:25,600) for 1h at 37 • C, 5% CO 2 . Then, the same well volume of carboxymethylcellulose 2.4% medium (DMEM-HG 10×, 2.4 carboxymethylcellulose and 2% fetal bovine serum) was added, and the cells were incubated at 37 • C, 5% CO 2 for 72 h. After this, the cells were fixed with formalin 10% and stained with crystal violet 0.04%, and the quantification of PFUs was carried out to determine the virus titer. All experimental procedures were executed at a biosafety level 3 (BSL3) multi-user facility, according to WHO guidelines [71].

Statistical Analysis
Graphs were generated using the GraphPad Prism 8.0 software. Values of EC 50 were determined by non-linear regression of log (inhibitor) vs. normalized response. Values correspond to the best curve generated based on R 2 values ≥ 0.9. All experiments were realized with three technical replicates.

Conclusions
Until now, all studies regarding A. amazonicus have attributed the species' promising pharmacological potential to saponins, its class of major compounds. This work is the first to report the phenolic composition of A. amazonicus and to show the important role that these substances hold in inhibiting the binding between SARS-CoV-2 spike protein and ACE2 receptor. However, when evaluating a different target, such as viral replication post-infection, saracura-mirá extracts also present promising activity reducing viral titers, which means that saponins may also contribute to the antiviral potential of A. amazonicus. In addition, the anti-inflammatory and immunomodulatory activity of these extracts has already been proven and attributed to this class of compounds. These two features represent very important approaches to the treatment of SARS-CoV-2 infection, as they are able to control symptoms and prevent evolution into severe cases, as has been highlighted in the popular use of A. amazonicus. Therefore, it is believed that A. amazonicus can contribute to controlling the infection caused by SARS-CoV-2 through several mechanisms, standing out as a multitarget agent. The data presented here, together with the literature and recent reports on the medicinal use of A. amazonicus, strongly suggest that this species could be a possible candidate and model for the development of new drugs to treat COVID-19.  Acknowledgments: We would like to thank the Instituto de Biofísica Carlos Chagas Filho and the Central Analítica, Faculdade de Farmácia of the Universidade Federal do Rio de Janeiro for allowing the use of their multi-user facilities. Additionally, thanks are due to the Oswaldo Cruz Institute for assessments related to BSL3 facility (pav. Leonidas Deane, IOC, Fiocruz). We also thank Laboratório de Investigação de Substâncias Bioativas (Faculdade de Farmácia, UFRJ) for helping us with phenolics quantification. Finally, the authors are especially thankful to the quilombola communities from Oriximiná, on behalf of the Association of Remnant Communities of Quilombos of Oriximiná (ARQMO).

Conflicts of Interest:
The authors declare no conflict of interest.