Natural Products and Derivatives as Potential Zika virus Inhibitors: A Comprehensive Review

Zika virus (ZIKV) is an arbovirus whose infection in humans can lead to severe outcomes. This article reviews studies reporting the anti-ZIKV activity of natural products (NPs) and derivatives published from 1997 to 2022, which were carried out with NPs obtained from plants (82.4%) or semisynthetic/synthetic derivatives, fungi (3.1%), bacteria (7.6%), animals (1.2%) and marine organisms (1.9%) along with miscellaneous compounds (3.8%). Classes of NPs reported to present anti-ZIKV activity include polyphenols, triterpenes, alkaloids, and steroids, among others. The highest values of the selectivity index, the ratio between cytotoxicity and antiviral activity (SI = CC50/EC50), were reported for epigallocatechin gallate (SI ≥ 25,000) and anisomycin (SI ≥ 11,900) obtained from Streptomyces bacteria, dolastane (SI = 1246) isolated from the marine seaweed Canistrocarpus cervicorni, and the flavonol myricetin (SI ≥ 862). NPs mostly act at the stages of viral adsorption and internalization in addition to presenting virucidal effect. The data demonstrate the potential of NPs for developing new anti-ZIKV agents and highlight the lack of studies addressing their molecular mechanisms of action and pre-clinical studies of efficacy and safety in animal models. To the best of our knowledge, none of the active compounds has been submitted to clinical studies.


Introduction
Zika virus (ZIKV) and other arboviruses belonging to the Flavivirus genus (Flaviviridae family) are transmitted by mosquitoes, causing serious diseases in humans [1]. Some of them such as Zika, Dengue and Chikungunya viruses have spread globally and became major public health problems in tropical and subtropical regions of the globe [2].
ZIKV is genetically and serologically related to other flaviviruses such as Dengue, Yellow Fever, and West Nile viruses [3]. It is a positive-sense single-stranded RNA virus, presenting approximately 10.8 kb, consisting of two non-coding regions 3 and 5 , three structural proteins (capsid, pre-membrane/membrane, and envelope) responsible for the structure of the viral particle, and seven non-structural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) that participate in the replication and packaging of the viral genome [4,5].
The classic transmission of ZIKV occurs by the bite of mosquitoes of the genus Aedes, whereas transmission by blood transfusion, perinatal and through urine have been also described. ZIKV is the only member of the Flavivirus genus that is also transmitted by sexual and transplacental routes [4,6,7]. To date, there is no evidence of transmission by breastfeeding [8].
The first confirmation of ZIKV infection in Brazil dated from May 2015 [9]. Afterwards, the relationship between intrauterine infection by ZIKV and early microcephaly was established by physicians in the Brazilian northeast. Subsequently, ZIKV infection was associated with other severe neurological complications in babies born to ZIKV-infected mothers [10] and in ZIKV-infected adults [11]. The contact of pregnant women with ZIKV may result in fetuses and babies with brain calcifications, microcephaly and other congenital Zika syndromes (CZSs). Many authors suggest that even in babies considered normocephalic at birth, CZSs may cause complications such as delayed neurological development, loss of visual and auditory acuity, and difficulties in language proficiency [12][13][14]. In adults, one of the most important neurological manifestations is the Guillain-Barré syndrome, which is characterized by the peripheral neuropathy associated with progressive paralysis [15].
ZIKV continues to circulate in Brazil and in other countries with high potential for dissemination. Predictive models suggest that the geographic distribution of the Aedes aegypti mosquito, the main vector of ZIKV, continues to expand because of population growth and movement, uncontrolled urbanization, low prevention level, inadequate sanitary conditions, and climate changes [16]. Current actions to prevent infection by ZIKV encompass the use of barrier measures against the vector bite (use of repellents, mosquito nets, screens, long clothing), the elimination of possible breeding sites for larvae of Aedes aegypti, and the use of condoms to prevent sexual transmission of the virus [17,18].
To date, there is neither effective therapy against ZIKV infection and resulting complications such as microcephaly and Guillain-Barré syndrome nor vaccines for its prevention; therefore, the World Health Organization (WHO) considers urgent the research and development of new drugs against this virus [19,20]. More accurate diagnostic tests are also demanded, since ZIKV infection is difficult to diagnose and to differentiate clinically from Dengue, Chikungunya, and other viral infections [21].
The research of natural products (NPs) has been proved to be a valid strategy for developing important drugs currently used in therapy, and they also play an important role in the discovery of new antiviral agents [22][23][24][25][26][27]. In this context, the search for compounds active against ZIKV is relevant and may result in the development of new antivirus drugs to be adopted for the prevention and treatment of future generations. Several classes of NPs have been described in the literature as active against ZIKV in different experimental models, including polyphenols, triterpenes, alkaloids, and steroids, among others. Therefore, the goal of the present work was to undertake a comprehensive revision on NPs that have been reported to exhibit potential effect against ZIKV by using predictive computational tools, as well as by in vitro and in vivo models of viral infection, along with their mechanisms of antiviral activity.

Methodology
A literature review was conducted on NPs and derivatives tested against ZIKV. The following databases were accessed: PubMed/MEDLINE, Scientific Electronic Library Online (SciELO), Scopus, Google Scholar, Web of Science, Science Direct, SciFinder and Literatura Latino-Americana e do Caribe em Ciências da Saúde (LILACS). Articles written in English and published up to December 2022 were reviewed. Indexed keywords from Medical Subject Headings (MeSH) and Descritores em Ciências da Saúde (DeCS) were used to build up search strategies, and the following ones were adopted: "Zika virus" combined with the descriptors "medicinal plants", "plants", "bioactive compound", "antiviral bioactive compounds", "phenolic compounds", "flavonoids", "alkaloids", "terpenoids", "antiviral natural products", "antiviral activity", "biological product", "natural products", "inhibitors", "activity", "antiviral effect", "viral life cycle", "compounds". The "AND" connector was used between the terms as, for example, "Zika virus" and "medicinal plants". The search results were saved and compiled using Microsoft Excel version 2016 software to select potential eligible studies.
The revision was focused on scientific articles reporting NPs from plants, animals, fungi, and bacteria with antiviral action against ZIKV assayed by in silico, in vitro, and in vivo models. Reports on synthetic and semisynthetic derivatives of NPs were also selected for revision, whereas review articles, notes, editorials, abstracts presented at scientific meetings, experience reports, dissertations and theses were excluded. The abovementioned inclusion and exclusion criteria were strictly followed for selecting eligible articles for revision. The title, abstract, and keywords were firstly read to pre-select the articles of interest, which were carefully browsed subsequently for inclusion in this paper. The number of papers reporting NPs tested against ZIKV published between January 1997 and December 2022 is presented in Figure S1 as Supplementary Material.

NPs with Potential Anti-ZIKV Activity Identified by In Silico Tools
The ability of NPs to interact with molecular targets of ZIKV has been assessed by using in silico tools, mainly molecular docking, and the main targets so far identified and its possible effects on the viral life cycle are depicted in Figure 1. In general, literature data point out polyphenols and alkaloids as the best molecular docking ligands for ZIKV target proteins with high affinity for protein receptors. The literature data are presented in the following subsections based on the assayed target protein. Table S1 of the Supplementary Material summarizes NPs that interact with the relevant molecular targets of ZIKV, which is disclosed by in silico assays.

ZIKV NS2B-NS3 Protease and NS3 Proteins
Several alkaloids have been reported to act as inhibitors of the ZIKV NS2B-NS3 protease, including bromocriptine (a derivative of the alkaloid ergoline) [28], hydroxychloroquine (a semisynthetic derivative of quinine) [29], and the bis-indole alkaloids flinderol A and flinderol B [30,31]. Seven novel berberine derivatives were tested as possible inhibitors of the ZIKV NS2B-NS3 protein, and one presented a high value of binding energy [32]. Additionally, the isoquinoline alkaloid cassiarin D, along with the flavonoids 3 -O-methyldiplacone and exiguaflavanone A, in addition to the sesquiterpene lactucopicrin, showed strong anchoring properties with the NS3 helicase of ZIKV [33].
Roy and coworkers [34] reported the flavonoids myricetin, quercetin [35], isorhamnetin, luteolin, and apigenin, along with curcumin, compounds commonly found in different vegetables and fruits, as non-competitive allosteric inhibitors of the ZIKV NS2B-NS3 protease [36]. Interestingly, curcumin also showed the strongest binding with tumor necrosis factor (TNF) in a molecular docking study [36]. Thus, TNF inhibition can reduce the inflammatory process caused by ZIKV infection, decreasing symptoms and the risk of future complications. Another flavanone, naringenin, was able to interact with the NS2B-NS3 protease protein domain of ZIKV, acting as a non-competitive inhibitor [37].
The flavanone hesperidin exhibited significant in silico inhibition of ZIKV NS2B-NS3 protease, which is important in the ZIKV replication process [38], along with the chalcones angusticornin B and curaridin [30,31]. More recently, a set of 150 NPs and semisynthetic derivatives was subjected to the molecular docking against the ZIKV NS2B-NS3 protease. Quercetin, rutin, and pedalitin, isolated from Pterogyne nitens, presented the best affinity to this ZIKV protease [39].
Thirty-five compounds inhibited the ZIKV-NS5 protein in the MTase domain, of which the flavonoids silybin B and silybin C exhibited the highest affinity energy (−10.2 kcal/mol). Additionally, 29 compounds inhibited the ZIKV-NS5 protein in the RNA-dependent RNA polymerase (RdRp) domain, the highest affinity energy being found for the flavonoid isopomiferin (−9.2 kcal/mol). This compound was also identified as a potential inhibitor of both NS5 MTase and NS5 RdRp of ZIKV in addition to four dengue serotypes proteins [47]. Moreover, polydatin, dihydrogenistin, liquiritin, rhapontin, and cichoriin were successfully bound inside the pocket of ZIKV NS5 RdRp [48].
Sinefungin, an adenosine derivative originally isolated from the bacteria Streptomyces griseoleus, was characterized as a competitive inhibitor of ZIKV metiltransferase [49]. In its turn, lycorine was shown to interact specifically with the NS5 RdRp domain of the protein [50].
A set of 72 compounds from Piper nigrum and Salvia rosmarinus was tested against ZIKV. Among them, 26 constituents of P. nigrum and 31 compounds of S. rosmarinus showed effective drug-likeness properties. Piperine and isoscutellarein have shown remarkable inhibitory potential against the ZIKV NS5 RdRp protein [51]. In its turn, the polyphenol theaflavin, isolated from P. nigrum, interacts with ZIKV NS5-MTase by hydrogen bounds [52].

ZIKV Envelope Protein
Twenty-five compounds comprising flavones, alkaloids, and polyphenols found in species of Asteraceae, Acanthaceae, Combretaceae, Lamiaceae, Phyllanthaceae and Polygonaceae were selected for virtual screening against the ZIKV envelope protein. Tannic acid isolated from Terminalia arjuna (Combretaceae) showed better interaction with the ZIKV envelope E protein [53], and the authors suggested that it is potentially a good inhibitor of virus adsorption and could be further investigated to control ZIKV infection.
Epigallocatechin gallate showed high affinity for the ZIKV E protein (dimers I and III), preventing the conformational change of dimer III, which is required for ZIKV entry into the host cells [54]. Another study reported epicatechin to present good docking scores with the domain III of the envelope protein along with baicalin, isonimolicinolide, madecassic acid, and apigenin-7-O-beta-D-glucopyranoside [41]. In the same direction, pentagalloylglucose, parishin A, and stevioside showed remarkable docking scores with the ZIKV envelope (E) protein that represents the major target for neutralizing antibodies [55].
Compounds found in the aqueous extracts of Momordica charantia (saikosaponin D and benzoyl paeniflorin), Vitex negundo (macrophylloside D), and Blumea balsamifera (cyanidin 3-5-diglucoside) were subjected to docking studies against several ZIKV targets: helicase, protease, methyltransferase, RNA-dependent RNA polymerase, and envelope protein [63]. These compounds exhibit higher binding affinity to the viral replication proteins than to the envelope protein.
Tangeretin is a polymethoxyflavone from the peel oil of Citrus reticulata. This compound was able to interact with the ZIKV NS1 protein by establishing strong hydrogen bonds and with ZIKV NS5 [66].
The prediction results from the in silico studies herein reviewed demonstrate the potential activity of NPs against molecular targets of ZIKV. The judicious validation of compounds exhibiting the best docking scores, through in vitro and in vivo assays, might disclose promising compounds for developing new drugs to fight ZIKV infection and minimize ZIKV-mediated pathologies [67]. Computational approaches generate activity predictions that can contribute to accelerate the development of new active compounds in addition to predictions of pharmacokinetics and the safety of anti-ZIKV compounds [68]. However, data generated by in silico assays must be carefully analyzed, keeping in mind the need to correlate them with in vitro and in vivo test results.

Extracts, NP, and NP-Derivatives Active In Vitro against ZIKV
The selectivity index (SI) is the ratio that measures the window between cytotoxicity and antiviral activity. Thus, SI is the ratio of median cytotoxic concentration (CC 50 ) to median effective concentration (EC 50 ), which is related to a specific biological effect: in this case, the activity against ZIKV. Theoretically, a high SI ratio would indicate an effective and safe compound for the in vivo treatment of a certain virus infection. An SI value ≥ 10 is recommended for selecting a compound/extract potentially useful to be further investigated as an antivirus agent [69]. Therefore, we included in this review mostly samples that presented SI values ≥ 10, comprising isolated compounds from natural sources, derivatives, and some plant extracts, as described in this section. On the other hand, data for samples with SI below the adopted cutoff value are reported on Table S2 as Supplementary Material. The SI values of several samples are being described in the published data; for some of them, there are only EC 50 values reported, whereas for others, only the percentage of inhibition is informed. These data are compiled in Table 1, along with the cell lineage used in the assays, and the antivirus mechanism of action, when available. The chemical structures of some selected compounds, which elicited potent anti-ZIKV activity in vitro and/or in vivo, are depicted in Figure 2.

Plant Extracts
Extracts obtained using solvents of different polarity and essential oils from plant species have already been tested in vitro against ZIKV, and a wide range of SI values has been reported. Hence, the aqueous extract of the aerial parts of Aphloia theiformis showed anti-ZIKV effect and was not cytotoxic (CC 50 = 3000 µg/mL) to Vero cells. The concentration-dependent activity of A. theiformis extract was observed through a 2-log reduction in viral load recovered from ZIKV African strain MR766 in comparison with the viral inoculum (EC 50 of 100 µg/mL, SI = 30). In Vero cells, this extract (500 µg/mL) reduced the infectivity of the ZIKV Asian PF13 strain at 2-log viral load; and for the Brazilian ZIKV strain (ZV BR 2015/15261), the extract reduced 3-log viral load in Huh7.5 cells. The A. theiformis extract inhibited ZIKV adsorption in Vero cells and also showed antiviral effect against Dengue virus (DENV) [70].
The aqueous extract of Psiloxylon mauritianum aerial parts, tested at 100 µg/mL, inhibited ZIKV infection in Vero cells, decreasing by 2-log the production of infectious viral particles from African (MR766) and Asian (PF13) strains [72]. This extract maintained 90% viability of Vero and A549 cells and did not cause DNA damage to A549 cells even at cytotoxic concentration (500 µg/mL), which were determined using MTT and COMET methods, respectively. The obtained results in Vero cells (CC 50 = 1044.00 ± 106.20 µg/mL, EC 50 = 19.50 ± 4.80 µg/mL and SI = 53.5) showed that the antiviral activity of P. mauritianum extract significantly affects the attachment of ZIKV to the cell surface and also that its action is mediated by an interaction between the extract and the ZIKV particles. The extract was also active against four dengue serotypes [72]. It is important to highlight that the P. mauritianum extract is rich in phenolic compounds, such as gallic acid, quercetin and kaempferol.
The essential oil from the aerial parts of Ayapana triplinervis was not cytotoxic for A549 cells and showed anti-ZIKV activity (EC 50 = 38.0 µg/mL, SI = 12.5). Dimethylthiohydroquinone ether, isolated from the essential oil of A. triplinervis, also showed antiviral activity against ZIKV (EC 50 = 45.0 µg/mL, SI = 9.1). The inhibition of ZIKV mediated by dimethylthio-hydroquinone ether occurs after the virus binds to the cell membrane and is explained by the inability of the ZIKV-bound particles to be internalized in the host cell [74].
The methanolic extract of Psychotria viridis leaves induced anti-ZIKV activity in Vero cells with SI > 28. 85. The extract showed virucidal activity and inhibited the virus at intracellular stages of the viral cycle [76]. Mono and dimethyltryptamine are the main bioactive compounds found in the methanolic extract of P. viridis leaves.
The aqueous extract of cranberry pomace showed anti-ZIKV activity (CC 50 = 865.1 µg/mL, EC 50 = 26.0 µg/mL, SI= 33.2) in A549 cells. At 100 µg/mL, the extract reduced virus progeny production by 2-log. Moreover, the extract prevented ZIKV entry into the host cell by inhibiting the viral binding step. The extract also showed antiviral effect against different serotypes of DENV [77].
The essential oil of Lippia alba showed virucidal activity against ZIKV assayed in Vero cells (CC 50 = 1789.90 ± 1.74 µg/mL, EC 50 = 32.20 ± 0.54 µg/mL, SI = 55.6). β-caryophyllene is the major constituent of the essential oil, but it has not been tested, and the virucidal activity of the essential oil was ascribed to the mixture of constituents [60].
Some plant extracts have been also reported to be active in vitro against ZIKV, but their selectivity indexes were not determined. Hence, Carica papaya fruit pulp inhibited ZIKV infection in A549 cells without loss of cell viability [80]. The extract promoted a concentration-dependent inhibition of ZIKV African strain MR766 (EC 50 = 300 µg/mL) and reduced the production of ZIKV progeny by 3-log for the Asian strain (PF13) when tested at 1000 µg/mL. In relation to the mechanism of action, papaya pulp extract inhibited the adsorption of ZIKV to A549 cells. Papaya pulp polyphenol-rich extracts fermented with Lactobacillus plantarum, Leuconostoc pseudomesenteroïdes and Weissela cibaria were also tested, producing EC 50 of 1.5, 1.9 and 4.2 mg/mL, respectively. It was concluded that the lactic fermentation of papaya pulp affects its antiviral activity against ZIKV depending on the bacterial strain used [80].
The ethanolic extract of Ocimum basilicum leaves showed anti-ZIKV activity in a concentration dependent manner with EC 50 of 1:134 and reduced 97% of the ZIKV infectivity in Vero cells. The antiviral effect was ascribed to the ability of O. basilicum extract to inhibit ZIKV adsorption to Vero cells [81].
The ethanolic extracts of the fruits and barks of Schinus terebinthifolius, both at 100 µg/mL, showed anti-ZIKV activity, causing a significant reduction in the viral load. The results were obtained by RT-PCRq and flow cytometry in the HTR-8/SVneo cell line.  [84] assayed 1000 ethanolic extracts of plants against DENV and ZIKV. Among them, 21 extracts (2.1% of total) presented activity against ZIKV, with EC 50 values ranging from 10 to 100 µg/mL. In another screening, 37 plant extracts were tested in vitro against Vero cells infected with ZIKV, and significant anti-ZIKV activity was elicited by the extracts of Maytenus ilicifolia (4.5 log of inhibition), Maytenus rigida (1.7 log of inhibition), Terminalia phaeocarpa (3.7 log of inhibition), and Echinodorus grandiflorus (1.7 log of inhibition) [85].
The methanolic extracts of aerial parts from Stenocline ericoides and Stenocline inuloides showed anti-ZIKV activity in A549 cells. They lowered ZIKV progeny production by 2-log when assayed at non-cytotoxic concentrations below 200 µg/mL. S. ericoides has virucidal action, whereas S. inuloides inhibits the early steps of virus infection. Both extracts also inhibited DENV-2 infection. These extracts are rich in polyphenols and flavonoids and did not exert acute toxicity in Zebrafish [86].
The aqueous extracts of Artemisia capilaris and Hedyotis diffusa presented antivirus activity against JEV (Japanese encephalitis virus), ZIKV, and DENV. They inhibited virus replication and reduced viral RNA levels in a concentration-dependent manner in the range from 0.1 to 10 mg/mL in Vero cells. Both extracts presented anti-flavivirus activity in other human cell lines, including human glioblastoma (T98G), human chronic myeloid leukemia (K562), and human embryonic kidney (HEK-293T) cells [87].
Extracts of fresh tomato pomace prepared with ethanol, ethanol/ethyl acetate (1:1), and supercritical fluid extraction with CO 2 inhibited ZIKV infection up to 75% in a concentration-dependent manner, at 200 µg/mL, in A549 cells. These extracts act on the early stages of the viral cycle by blocking viral entry into the host cell [88].

NPs and Derivatives from Plants
NPs of different classes have been reported to inhibit ZIKV replication in different cell linages, being flavonoids, terpenes, and sesquiterpenes, among others, identified as the most promising compounds.
Chloroquine did not affect cell viability at concentrations ≤ 50 µM. Chloroquine treatment of Vero cells infected with a Brazilian strain of ZIKV at concentrations ranging from 12.5 to 50 µM increased cell viability from 55% to 100% (CC 50 = 134.54 ± 16.76 µM, EC 50 = 9.82 ± 2.79 µM, SI = 13.70). Furthermore, treatment with 25 µM of chloroquine led to a 16-fold reduction in the level of viral RNA detected in the supernatant by RT-qPCR. Chloroquine interferes in the early stages of the ZIKV replication cycle. Other experiments were performed with an African strain in human brain microvascular endothelial cells hBMECs: chloroquine reduced by 45% and 50% the number of ZIKV-infected cells at 25 and 50 µM, respectively, and protected approximately 80% of these cells against infection. In neural stem cells, treatment with 50 µM chloroquine decreased by 57% the number of ZIKV-infected cells and protected 70% of those cells during infection without cytotoxicity effects. In mouse neurospheres, ZIKV infection decreased after treatment with 12.5 µM of chloroquine, according to phase contrast microscopy analysis [92]. Furthermore, hydroxychloroquine tested in vitro by an enzymatic assay inhibited the activity of ZIKV NS2B-NS3 protease [29]. In addition, hydroxychloroquine at 80 µM significantly decreased ZIKV infection in JEG3 cells, reducing both the number and size of foci, according to immunofluorescence microscopy analysis [29].
The flavonoid isoquercitrin showed anti-ZIKV activity in cell lineages A549 (CC 50 [93]. The flavanone naringenin, widely found in citrus fruits, tomatoes, cherries, grapefruits, and cocoa, showed activity against distinct ZIKV strains (Brazilian, Asian and African strains) in a concentration-dependent manner in A549 human cells. Narigenin was shown to be active against the Brazilian strain (CC 50 = 693.60 µM, EC 50 = 58.79 µM and SI = 11.79 as well as against the Asian (~4 fold reduction in infection) and African lineages (~2 fold reduction in infection), which were both tested at 125 µM. The antiviral activity of naringenin was also observed in primary human monocyte-derived dendritic cells infected with ZIKV. The mechanism of action indicated that naringenin acts on viral replication or viral particle assembly [37].
Two other flavones, baicalein (EC 50 = 0.004 µM, CC 50 = 420.000 µM, SI = 105) and baicalin (EC 50 = 14.0 µM, CC 50 = 553.0 µM, SI = 40), decrease ZIKV replication significantly. Baicalein was more effective against post-entry replication of ZIKV, while baicalin acted majorly at the stage of virus entry into the host cell [46]. Some semisynthetic derivatives of baicalein were also described to possess antiviral activity. Hence, a derivative obtained by the introduction of a thiocarbonyl group at the C-4 position and another prepared by setting a halophenylamino group at C-4 showed potent antiviral activities against CHIKV, WNV (West Nile virus) and ZIKV, and were shown to be 5-10 times more potent than baicalein. Both derivatives showed CC 50 of 100.0 µM and EC 50 = 3.8 µM and EC 50 = 1.9 µM, respectively, in Huh7 cells [94].
The dihydrochalcone phloretin, found in apple and pear trees, decreased the infectious titers of ZIKV African strain MR766 (EC 50 = 22.85 ± 5.20 M, SI = 4.40 ± 0.80) and PRVABC59 ZIKV clinical isolate from Puerto Rico (EC 50 = 9.31 ± 2.50 M, SI = 10.70 ± 2.20). Phloretin acts in several stages of ZIKV replication, including RNA production, and in a later stage of propagation, such as assembly or exit. Phloretin decreased the apoptotic activity of caspase-3 and caspase-7 and reduced the phosphorylation of Akt/mTOR pathways. In addition, phloretin acts inhibiting the glucose uptake, thus reducing the spread of ZIKV [95].
Ellagic acid, isolated from Punica granatum, was active against the Asian strain HPF2013 of ZIKV (CC 50 19). When tested at concentrations between 12.5 and 25 µM, the viral surface antigen expression was eliminated in Vero cells [106].
Glycyrrhetinic acid is a triterpene isolated from Glycyrrhiza glabra and Glycyrrhiza uralensis. This compound and some semisynthetic derivatives were tested in vitro against SF268 cells infected with ZIKV. The best results were obtained for the glycyrrhetinic acid  [44].
A set of glycyrrhizic acid derivatives conjugated with amino acids was tested against ZIKV-infected SF268 cells and other cell lines. The derivatives bearing two amino acid methyl ester residues in the sugar unit, along with those possessing two amino acid ethyl ester residues in the carbohydrate chain, showed values of EC 50 between 0.09 ± 0.08 µM and 2.23 ± 0.72 µM and induced a concentration-dependent reduction in the cytopathic effect in infected cells. Of note, glycyrrhizic acid conjugated with amino acids and their esters aspartate (methyl ester)-methyl ester (EC 50 = 0.23 ± 0.04 µM) and tyrosine methyl ester (EC 50 = 0.09 ± 0.02 µM) showed the best inhibitory activity against ZIKV infection. A time-of-addition/removal assay carried out with both derivatives indicated that they may act inhibiting the entry stage and blocking the post-entry stage, respectively. Moreover, docking studies revealed that both compounds interact with the active pocket of NS5 MTase [108].
Some compounds herein described and/or those enclosed in Table 1 do not act solely at one stage of the ZIKV replication cycle but may target different stages. For example, (-)epigallocatechin gallate elicits virucidal effect by targeting both viral entry and NS2B-NS3 protease. This bifunctional feature could be advantageous for drug development, since dual effects are achieved for one single compound [110]. It should be remembered that NPs offer a vast and unexplored diversity of chemical structures unmatched by even the biggest combinatorial databases [111]. Jointly, these features reinforce the relevance of searching NPs to identify new bioactive antiviral compounds that are potentially useful for drug development. Table 1 summarizes some selected NPs and derivatives that showed in vitro activity against ZIKV for which studies to investigate the mechanism of action were performed, and the antiviral activity was evaluated using different assays.

NPs from Bacteria
Although less numerous than NPs from plants, several compounds obtained from different bacterial strains have been reported to inhibit ZIKV replication. Duramycin is a cyclic peptide produced by Streptomyces cinnamoneus, which is capable of binding to phosphatidylethanolamine found in enveloped virions, thus preventing TIM1 binding. It reduced ZIKV infection by >50% in primary human placental cells and by >95% in primary cells from chorionic villi when assayed respectively at 0.1 and 0.2 µM. Furthermore, duramycin at 1 µM reduced by 2-log the viral load in amniochorionic membranes [138,139].
SH-SY5Y and A549 cells pre-treated with bafilomycin A1 (10 nM), a macrolide antibiotic produced by different Streptomycetes species, reduced viral RNA of African and Asian ZIKV strains, as determined by RT-qPCR. The mechanism of action investigated in A549 cells showed that bafilomycin A1 inhibits ZIKV adsorption to the host cell, preventing the spread of infection and interfering with viral maturation [141].
Antimycin . They were also shown to inhibit different ZIKV strains, including African, Asian, and American strains, thus characterizing a broad spectrum of action [142].

NPs from Fungi
Fungi are recognized sources of bioactive compounds potentially useful to be developed as antivirals. Although promising, the research of fungal compounds with antiviral activity is still an emerging field of investigation [111].
The fermentation products of Penicillium brevicompactum, mycophenolate mofetil (CC 50  Interestingly, these compounds were capable of inhibiting different ZIKV strains, including African, Asian, and American strains, thus characterizing a broad spectrum of action [142]. Cavinafungin is a lipopeptide isolated from Colispora cavincula. This compound was reported to be a potent inhibitor of ZIKV multiplication in A549 cells (EC 50 = 150 ± 23 nM, CC 50 = 1650 ± 49 nM, SI = 11), acting as a selective inhibitor of viral peptidase signaling [148]. It also inhibited in vitro the replication of four DENV serotypes. Brefeldin A is a secondary metabolite isolated from the ethanolic extract of the culture broth of Penicillium sp. FKI-7127. When tested against ZIKV in Vero cells, brefeldin A induced a strong inhibition in a concentration-dependent manner (EC 50 = 54.8 ± 0.4 nM, CC 50 = 2000.0 nM, SI = 36.5). Brefeldin A was also active against four serotypes of dengue and Japanese encephalitis virus [149].
β-1,3-β-1,6-Glucan isolated from Agaricus subrufescens fruiting bodies (FR) and its sulfated derivative (FR-S) were evaluated as potential anti-ZIKV agents. FR-S was able to inhibit ZIKV replication in human monocytic cells when added simultaneously with viral infection (CC 50 > 500 µg/mL, EC 50 = 188.70 ± 12.87 µg/mL, SI > 2.78), whereas FR was inactive. Additionally, no inhibitory effect was observed when either FR-S or FR were added post-infection. The effect of these compounds on permeability was assayed in vitro in human pulmonary microvascular endothelial cells (HPMECs) to assess their capacity of protecting against endothelial barrier dysfunction induced by ZIKV. FR and FR-S, tested at 0.12 µg/mL, promoted 100% inhibition of ZIKV NS1-induced hyperpermeability, as measured by trans-endothelial electrical resistance [150].

NPs from Marine Organisms
The marine environment offers a unique ecosystem for identifying new compounds as potential therapeutic drugs, although it is currently an under-exploited resource. Com-pounds obtained from marine organisms may induce antiviral activity with unique mechanisms of action [151].
The dichloromethane extract of the red seaweed Bryotamnion triquetrum showed inhibitory activity on the ZIKV replication in Vero cells with values of CC 50 = 400.00 ± 13.50 µg/mL, EC 50 = 1.38 µg/mL and SI = 289.85. The investigation of its mechanism of action revealed a moderate virucidal effect and moderate inhibition at the attachment stage of the virus in the host cell [153].
Roseotoxin A and B, destruxin A, F, Ch1 and Br1, along with chlorohydrin, are cyclohexadepsipeptides obtained from the extract of the marine fungus Beauveria felina SX-6-22. These compounds, tested at 10 µM, were able to inhibit ZIKV total RNA replication and NS5 production levels in virus-infected A549 cells and showed low cytotoxicity, which was observed by qRT-PCR [150]. The authors reported that felinotoxin E, roseotoxin B, and roseotoxin A exert antiviral activity during the early stages of infection based on a timecourse detection assay. Moreover, it was proposed that the cyclohexadepsipeptides may act in the endosomal fusion, according to an assay that tracks the organelles acidification [154].
Four semisynthetic derivatives of dolabellane diterpenes from species of Eunicea soft corals showed antiviral activity against ZIKV in a plaque reduction assay using Vero cells.

NPs from Animals
The peptide lycotoxin-An1a (Av-LCTX-An1a) was isolated from the venom of the spider Alopecosa nagpag. It inhibited ZIKV replication in HUVEC and A549 cells, at 10 µM, observed by a real-time fluorescence assay, and acted as a competitive inhibitor of the NS2B-NS3 protease of ZIKV with a Ki value of 12.54 ± 1.88 µM determined by the method of Dixon. It was also able to inhibit DENV-2 infection. No SI value was reported in this study [157].
Quinoline has been so far reported as a natural product from the Peruvian stick insect Oreophoetes peruana [158]. Barbosa et al. (2017) [159] evaluated the activity of quinoline derivatives against ZIKV on Vero cells. Expressive activity against ZIKV was observed for the synthetic quinoline derivatives N1- (2,8- The peptide Yodha, initially isolated from the skin of the frog specie Indosylvirana aurantiaca, and then obtained by synthesis, showed antiviral activity against different strains of ZIKV in Vero cells evaluated through qRT-PCR and focus-forming assay. The peptide showed no toxicity to human red blood cells up to 1000 µM. The cells treated with Yodha before the infection showed only 20% ZIKV RNA in comparison to the controls. In addition, we observed a decreased expression of the ZIKV envelope protein in samples treated with the Yodha peptide, according to the results of an immunofluorescence assay. Moreover, through electron microscopy, it was possible to observe that ZIKV treated with the Yodha peptide showed significant structural disruptions [160]. The peptide brevinin-2GHk (BR2GK) was isolated from the skin of the amphibian Fejervarya limnocharis. BR2GK reduced the virus-induced cytopathic effect in ZIKV-infected Vero cells and diminished the expression of ZIKV genomic RNA, E protein, and NS1. A time-addition assay was carried out and indicated that BR2GK showed significant inhibitory activity in the early and middle stages of ZIKV infection. In addition, a cellbased immunodetection assay revealed that BR2GK blocked ZIKV E protein expression with an EC 50 of 3.40 ± 0.73 µM. BR2GK inactivated ZIKV by disrupting the integrity of the envelope and may also penetrate the host cell membrane to inhibit the middle stage of ZIKV infection [161].  [162].
A set of 24 nucleoside analogue ProTides, obtained by synthesis, was tested against ZIKV in Vero cell culture and afforded the nucleoside analogue 2 -C-methylcytidine as a promising anti-ZIKV compound (CC 50 > 100 µM, EC 50 = 0.30 ± 0.12, and SI > 333). The compound was tested in a luminescence-based cell survival assay in a human fetal neural stem cell model of ZIKV infection. It protected the cells from virus-induced cell death with an EC 50 of 8.56 µM and a CC 50 of >100 µM [164].

NPs Tested in Animal Models of ZIKV Infection
A reduced number of NPs have been evaluated in laboratory animals infected with ZIKV. The reviewed literature indicates that so far, 12 compounds have been reported to induce anti-ZIKV activity in animal models, of which only 25-hydroxycholesterol was assayed in mice and in non-human primates ( Table 2). Twelve compounds showed promising antiviral effect, reducing the viremia caused by ZIKV infection. However, the obtained results cannot be directly translated into humans [166].   Figure 2 shows the chemical structures of some selected NPs with potential anti-ZIKV activity discussed in this review. These compounds were selected based on their selectivity index values above 500 and/or because their potential antiviral effect has been demonstrated in more than one assay (in silico, in vitro or in vivo assays).

Discussion
The present revision covers the scientific literature on natural products tested against ZIKV published from January 1997 to December 2022. It comprises 157 original articles ( Figure S1 in Supplementary Material), which is a reduced number of publications taking into account the impact of ZIKV infection on public health. This small number of articles demonstrates the slow progress on the research and development of bioactive NPs potentially useful as anti-ZIKV agents. In addition, it reinforces the current scenario that ZIKV infection is still a neglected tropical disease. In most countries, the epidemic outbreaks occurred in areas of low basic sanitation, which contributes to the proliferation of Aedes spp. and other ZIKV-transmitting mosquitoes [167].
The majority of studies herein reviewed-totalizing 82.4%-were carried out with natural products from plants and derivatives. The remaining publications report data obtained for NPs from fungi (3.1%), bacteria (7.6%), animals (1.2%), and marine organisms (1.9%) along with miscellaneous compounds (3.8%). Therefore, the vast majority of natural products so far investigated with anti-ZIKV activity were performed with plant secondary metabolites, mostly alkaloids, tannins, flavonoids, terpenoids, steroids, polyphenols, saponins, and stilbenes, among others. Phenolic compounds, especially flavonoids, comprise the largest number of natural products reported to possess anti-ZIKV activity.
The reported NPs mostly act at the stages of viral adsorption and internalization, and they present a virucidal effect. The reviewed data demonstrate their potential for developing new anti-ZIKV agents and also highlight the importance of additional studies addressing the molecular mechanisms of action of promising compounds such as epigallocatechin gallate (SI ≥ 25,000) [101], anisomycin (SI ≥ 11,900) [145], dolastane (SI = 1246) [153], and myricetin (SI ≥ 862) [101], among others. Some of the active compounds against Zika virus have been shown by in silico and in vitro assays to be active against other clinically important viruses and are promising candidates for further investigations aiming to develop broad-spectrum antivirus agents.
The chemical diversity of NPs offers the possibility of targeting different stages of the ZIKV infection and replication processes. For example, epigallocatechin-3-gallate and resveratrol show a virucidal effect and inhibit ZIKV adsorption to the host cell [114,122], whereas digitonin and conessine act on the stages of viral adsorption and internalization [116]. Furthermore, natural products tend to have more favorable physicochemical properties and lower cytotoxicity than synthetic compounds [110]. On the other hand, the intensive and time-consuming processes required for the extraction and isolation of NPs may limit or retard their application in drug development [168].
It should be remembered that the prior evaluation of activity prediction of natural products by in silico approaches represents a "starting point" for further in vitro and in vivo testing. The coumarin-derived antibiotic novobiocin obtained from Streptomyces niveus is an example of an NP that presented promising results on in silico studies, which were further confirmed by results from in vitro and in vivo assays (Figure 3). Figure 3. Nobobiocin, isolated from Streptomyces niveus, is an anti-ZIKV compound, whose activity has been predicted by in silico data and demonstrated by in vitro and in vivo assays. Median inhibitory concentration (IC 50 ), selectivity index (SI), and in vivo data from [43].
The knowledge about the antiviral activity of flavones dates to the 1990s, when it was demonstrated that the simultaneous administration of apigenin with acyclovir amplified the antiviral effect in vitro against Herpes virus types 1 and 2 (HSV-1 and HSV-2) [169]. Moreover, the antiviral activity of different flavonoids has been described by using in vitro and in vivo models, such as apigenin against adenovirus and hepatitis B, baicalein and isorhamnetin against Influenza virus, and chrysosplenol C against poliovirus type 3, among others, to cite a few examples [169,170]. Another polyphenol, epigallocatechin-3-gallate, has been reported to exert a broad spectrum of action against several viruses such as Hepatitis B virus (HBV), HSV, Epstein Barr virus (EBV), Adenovirus, Human Immunodeficiency virus (HIV), Hepatitis C virus (HCV), Influenza virus, DENV, Japanese encephalitis virus (JEV), Borne encephalitis virus (TBEV), CHIKV, and also ZIKV, among others [171]. According to the authors, epigallocatechin-3-gallate can be regarded as a nucleophilic reagent, since the phenolic hydroxyl groups found in the pyrogallol and galloyl moieties provide a higher number of lone electron pairs than other catechins. Therefore, it can react or combine with different molecules under appropriate conditions, thus inducing a broad antiviral effect.
According to Newman and Cragg (2020) [23], NPs still offer the best strategy for finding new bioactive compounds that may lead to effective agents for treating a variety of human diseases, including antiviral candidates. Different strategies have been adopted for this purpose, including the repositioning of drugs (e.g., mefloquine, hydroxychloroquine, chloroquine, and silymarin) and the screening of libraries of bioactive NPs and derivatives. Among these, chloroquine stands out as a promising compound based on the results from in vitro and in vivo tests. In addition, chloroquine has a defined mechanism of action, as it blocks the virus internalization in the host cell, which may favor its development as an antiviral drug. It can be therefore expected that one of the active NPs might be further investigated and developed as a drug or an herbal preparation for preventing or treating ZIKV infection. However, there is limited progress in this direction, and only one preclinical study has been so far reported in non-human primates [129]. To the best of our knowledge, there is no report of a clinical trial conducted with an NP or derivatives.

Conclusions
The vast majority of the studies performed to investigate the anti-ZIKV effect of extracts, natural products, and derivatives were carried out using in vitro assays. The ethanolic extract of the seaweed Osmundaria obtusiloba, along with the aqueous extracts of aerial parts from the plants Psiloxylon mauritianum and Aphloia theiformis, were identified as the most active ones. In its turn, epigallocatechin gallate, anisomycin, dolastane, and myricetin were disclosed as the most promising compounds for developing new anti-ZIKV agents based on results from in vitro and in vivo assays. Of note, the number of investigations so far performed using animal models is limited, especially using nonhuman primates. Therefore, it is mandatory to prioritize the pre-clinical investigation of the promising compounds disclosed by in vitro assays, aiming to select candidates for clinical studies. To the best of our knowledge, no data on the clinical investigation of NPs or derivatives with an anti-ZIKV effect have been so far reported. The development of antiviral agents that act both in prophylaxis and the treatment of ZIKV infection is urgently demanded, taking into account the severe outcomes of ZIVK infections to the population of several countries, especially to individuals living in low sanitation areas. Furthermore, a global effort is needed to control not only ZIKV but also other endemic species of Flavivirus and possible future outbreaks. Importantly, compounds showing high selectivity for ZIKV are considered potentially active against other members of the Flavivirus family. Therefore, the identification of bioactive NPs and derivatives and their mechanisms of extra-and intra-cellular action constitute a suitable strategy for the development of new antiviral drugs.

Data Availability Statement:
No new data were created or analyzed in this study. Data sharing is not applicable to this article.