Potential Antivirals: Natural Products Targeting Replication Enzymes of Dengue and Chikungunya Viruses

Dengue virus (DENV) and chikungunya virus (CHIKV) are reemergent arboviruses that are transmitted by mosquitoes of the Aedes genus. During the last several decades, these viruses have been responsible for millions of cases of infection and thousands of deaths worldwide. Therefore, several investigations were conducted over the past few years to find antiviral compounds for the treatment of DENV and CHIKV infections. One attractive strategy is the screening of compounds that target enzymes involved in the replication of both DENV and CHIKV. In this review, we describe advances in the evaluation of natural products targeting the enzymes involved in the replication of these viruses.


Introduction
Neglected tropical diseases (NTDs) compose a group of infections that primarily affects the world's poorest populations; they continue to be the leading cause of morbidity and mortality among the poorest people in developing countries [1][2][3][4]. NTDs affect more than one billion individuals who have inadequate access to safe water, sanitation and appropriate housing. The NTDs are caused by helminths, protozoa and tropical bacteria. Because of their link to poverty in developing countries, diseases caused by several arboviruses, including dengue virus (DENV) and chikungunya virus (CHIKV), should be considered NTDs [3].
CHIKV is a mosquito-transmitted alphavirus that belongs to the Togaviridae family [5]. It is a reemergent arbovirus that is responsible for chikungunya fever (CHIKF), severe joint pain and rash [6]. CHIKF was first described in 1952 after an outbreak along the borders of Tanzania (former Tanganyika) and Mozambique. This alphavirus is transmitted by mosquitoes of the Aedes genus, including A. aegypti, A. furcifer and A. albopictus [6,7]. In 2006, CHIKV was responsible for an epidemic of unprecedented magnitude in the Indian Ocean, stressing the need for new therapeutic approaches. Since that time, researchers have acquired a better understanding of CHIKV biology, which should lead to the identification of new, active molecules against this reemergent pathogen [8].
DENV, like CHIKV, is a reemergent arbovirus. Within the last two decades, it has become the most important and frequently transmitted arbovirus worldwide [9]. The incidence of DENV

NS5
NS5 is the largest (104 kDa) and most conserved protein in DENV. Specifically, residues 320-368 are strictly conserved among the flaviviruses. These residues have also been implicated in the interaction between NS5 and NS3 [38,39]. The important role of NS5 in DENV replication makes these proteins interesting targets for virus inhibition [26]. Similar to NS3, NS5 possesses two major activities, an RNA-dependent RNA polymerase (RdRp; residues 320-900) at its C-terminal end and a methyltransferase (MTase; residues 1-296) at its N-terminal end [26,40]. The NS5 MTase is a surface polyprotein that is essential for the attachment of the virus to the host cell. Thus, the ligand-bound crystal structure of the N-terminal domain of the NS5 MTase has become a crucial tool for current drug discovery efforts. The crystal structure of the RdRp shows an active site with two zinc ion-binding motifs, which are ideal targets for designing novel RdRp inhibitors [4,41,42].
The NS5 RdRp plays a vital role in the replication of viral RNA. Following viral entry and translation of the viral genome, the NS5 RdRp performs de novo RNA synthesis, generating negative-sense RNA from the positive-sense viral RNA template. The latter then serves as a template for the synthesis of more positive-sense RNA strands, which are either used for protein translation or packaged into infectious virions [43][44][45].
The flavivirus MTase catalyzes two very distinct methylation reactions, but there is only a single binding site for the methyl donor. It is likely that the RNA substrate repositions itself after the first reaction so that the second reaction can proceed [45,46].
Natural compounds that display inhibitory effects against DENV enzymes are described below. The evaluation of a compound's activity against viral enzymes can be investigated in two ways. First, the effect of a substance can be assessed directly on the enzyme (enzymatic assay). Second, the evaluation can be carried out in replicon cell lines (which are unable to produce infectious particles, but are capable of RNA replication) using a virus strain (post-treatment assay). For the post-treatment assay, host cells are initially infected with viruses, which can attach to and enter the cell. Then, the viruses that did not attach to the host cells are removed. Finally, the compound under investigation is added. An observation of virus inhibition indicates that the compound acts on the virus replication process, which involves several enzymes. The compound is thus assumed to act on one or more viral enzymes. Molecular docking is another important approach that has been used to investigate inhibitors of viral enzymes. The results from molecular docking experiments and in vitro analyses help streamline the screening processes.

Direct Inhibitory Activities of Natural Compounds against DENV Enzymes
Boesenbergia rotunda (L.) Mansf. Kulturpfl (Br) is a member of the ginger family. A phytochemical investigation of yellow rhizomes of this species from Thailand led to the isolation of six compounds, which are shown in Figure 1.  These compounds were assayed against the DENV-2 NS3 protease [47]. The enzymatic assays were carried out using the fluorogenic peptide Boc-Gly-Arg-Arg-MCA, which is an active substrate of the DENV-2 NS3 protease. Initially, the activities of the six compounds against NS3 were evaluated at three different concentrations (120, 240 and 400 ppm); all compounds inhibited NS3 protease activity in a concentration-dependent manner. The most active compounds were 5 and 6. The activities of these compounds were further evaluated at 40, 80 and 160 ppm. The inhibitory activity of Compound 5 ranged from 27.1% (40 ppm)-99.8% (400 ppm), whereas the inhibitory activity of Compound 6 ranged from 52.0% (40 ppm)-99.8% (400 ppm). Individually, Compounds 2 and 4 were found to have low inhibitory activity. However, when these flavones were combined, an increase in inhibitory activity was noticed (synergistic effect). Kinetic studies carried out with Compounds 1, 4, 5 and 6 resulted in the determination of their Ki values (345 for 1; 377 for 4; 25 for 5; and 21 for 6). Compounds 1 and 4 appeared to function via a noncompetitive mechanism, whereas Compounds 5 and 6 displayed competitive inhibition [47].
Litaudon and co-workers biologically screened 1350 ethyl acetate extracts obtained from various parts of 650 New Caledonian plants. The extracts were screened at a concentration of 50 μg/mL in a DENV polymerase assay using the RdRp domain of DENV-2 NS5. A second screen of 320 active extracts These compounds were assayed against the DENV-2 NS3 protease [47]. The enzymatic assays were carried out using the fluorogenic peptide Boc-Gly-Arg-Arg-MCA, which is an active substrate of the DENV-2 NS3 protease. Initially, the activities of the six compounds against NS3 were evaluated at three different concentrations (120, 240 and 400 ppm); all compounds inhibited NS3 protease activity in a concentration-dependent manner. The most active compounds were 5 and 6. The activities of these compounds were further evaluated at 40, 80 and 160 ppm. The inhibitory activity of Compound 5 ranged from 27.1% (40 ppm)-99.8% (400 ppm), whereas the inhibitory activity of Compound 6 ranged from 52.0% (40 ppm)-99.8% (400 ppm). Individually, Compounds 2 and 4 were found to have low inhibitory activity. However, when these flavones were combined, an increase in inhibitory activity was noticed (synergistic effect). Kinetic studies carried out with Compounds 1, 4, 5 and 6 resulted in the determination of their K i values (345 for 1; 377 for 4; 25 for 5; and 21 for 6). Compounds 1 and 4 appeared to function via a noncompetitive mechanism, whereas Compounds 5 and 6 displayed competitive inhibition [47].
Litaudon and co-workers biologically screened 1350 ethyl acetate extracts obtained from various parts of 650 New Caledonian plants. The extracts were screened at a concentration of 50 µg/mL in a DENV polymerase assay using the RdRp domain of DENV-2 NS5. A second screen of 320 active extracts at a concentration of 10 µg/mL was then performed, which resulted in the selection of extracts exhibiting enzymatic inhibition of at least 80%. An extract from the bark of Cryptocarya chartacea displayed significant enzymatic inhibitory activity (90% at 10 µg/mL) and was subsequently selected for fractionation. Flavonoids 7-17 ( Figure 2) were isolated from this extract [48].  As can be seen in Figure 2, flavonoids 8-17 inhibited the DENV NS5 RdRp, with IC50 values ranging from 1.8 μM (Compound 15)-72.5 μM (Compound 13). Pinocembrin (7) was found to be inactive against the DENV-2 NS5 RdRp, suggesting that the aliphatic portion of the chartaceone structure plays an important role in the inhibition of the polymerase. In this investigation, the authors also evaluated the effects of the isolated compounds on nasopharyngeal carcinoma cells. No compounds were found to be toxic to this cell line at a concentration of 10 μg/mL [48].
The investigation of the ethyl acetate extracts of the bark and wood of Trigonostemon cherrieri, a rare endemic plant from New Caledonia, resulted in the isolation of a series of very complex, oxygenated diterpenoids, which included Compounds 18-20. These compounds were screened against the DENV NS5 RdRp (Figure 3).

Compound
Polymerase Activity Inhibition a Cytotoxicity b 7 Not active 0 8 14.8 ± 2.4 3.2 9 15.3 ± 1.4 0 10 9.0 ± 3.5 11 11 27.0 ± 6.5 0 12 7.4 ± 2.  As can be seen in Figure 2, flavonoids 8-17 inhibited the DENV NS5 RdRp, with IC 50 values ranging from 1.8 µM (Compound 15)-72.5 µM (Compound 13). Pinocembrin (7) was found to be inactive against the DENV-2 NS5 RdRp, suggesting that the aliphatic portion of the chartaceone structure plays an important role in the inhibition of the polymerase. In this investigation, the authors also evaluated the effects of the isolated compounds on nasopharyngeal carcinoma cells. No compounds were found to be toxic to this cell line at a concentration of 10 µg/mL [48].
To identify new inhibitors of the DENV-2 NS5 RdRp, Litaudon and co-workers screened 820 ethyl acetate extracts from different parts of 400 plants collected in Madagascar. The bark extract of Flacourtia ramontchi was selected for fractionation, because of its pronounced effect on the DENV-2 NS5 RdRp. The compounds shown in Figure 4 were isolated from this extract and fully characterized [50]. The data displayed in Figure 4 reveal that among the isolated phenolic glycosides 21-31, Compounds 22, 27 and 30 were the most active. The cinnamic acid derivative Compound 31 showed significant inhibitory activity, although its activity is about four-times lower than the positive control. However, no structure-activity relationships could be established for any of these compounds.
Anacolosa pervilleana is a Madagascan plant whose leaves, bark and young shoots are used in traditional medicines for the treatment of schistosomiasis, syphilis and general weakness. Four acetylenic carboxylic acids (Compounds 32-35), two triterpenes (Compounds 36 and 37) and one aromatic compound (Compound 38) were isolated from the leaf ethyl acetate extract of this species ( Figure 5) [51].
The acetylenic acids were found to be the most active compounds, inhibiting the DENV NS5 RdRp with approximately the same efficacy. The observed IC50 values (<3 μM) were about 100-times lower than the reference compound, 3′-deoxy-GTP [51].
To identify new inhibitors of the DENV-2 NS5 RdRp, Litaudon and co-workers screened 820 ethyl acetate extracts from different parts of 400 plants collected in Madagascar. The bark extract of Flacourtia ramontchi was selected for fractionation, because of its pronounced effect on the DENV-2 NS5 RdRp. The compounds shown in Figure 4 were isolated from this extract and fully characterized [50]. The data displayed in Figure 4 reveal that among the isolated phenolic glycosides 21-31, Compounds 22, 27 and 30 were the most active. The cinnamic acid derivative Compound 31 showed significant inhibitory activity, although its activity is about four-times lower than the positive control. However, no structure-activity relationships could be established for any of these compounds.
Anacolosa pervilleana is a Madagascan plant whose leaves, bark and young shoots are used in traditional medicines for the treatment of schistosomiasis, syphilis and general weakness. Four acetylenic carboxylic acids (Compounds 32-35), two triterpenes (Compounds 36 and 37) and one aromatic compound (Compound 38) were isolated from the leaf ethyl acetate extract of this species ( Figure 5) [51].
The acetylenic acids were found to be the most active compounds, inhibiting the DENV NS5 RdRp with approximately the same efficacy. The observed IC 50 values (<3 µM) were about 100-times lower than the reference compound, 3 -deoxy-GTP [51].

Compound
Polymerase

Direct Inhibitory Activities of Plant Extracts against DENV Enzymes
The medicinal plants Vernonia cinerea, Hemigraphis reptans, Hedyotis auricularia, Laurentia longiflora, Tridax procumbens and Senna angustifolia were used to evaluate their abilities to inhibit the DENV NS3 protease. The highest inhibitory effects were observed for the ethanolic extract of S. angustifolia leaves, the methanolic extract of V. cinerea leaves and the ethanolic extract of T. procumbens stems (IC50 values: 30.1 ± 3.4, 23.7 ± 4.1 and 25.6 ± 3.8 μg/mL, respectively). The most active extracts were also tested in vitro against DENV-2-infected Vero cells, which were able to maintain a normal morphology without cytopathic effects. The percent viral inhibition of the extracts of V. cinerea (80.6% ± 6.1%) and T. procumbens (64.0% ± 9.4%) was significantly higher than that of S. angustifolia extract (26.3% ± 3.8%), as measured by a plaque-forming assay and RT-qPCR [52]. The authors of this investigation did not analyze which compounds were responsible for the observed activities.

Indirect Inhibitory Activities of Compounds and Extracts against DENV Enzymes
The plant Houttuynia cordata is a popular vegetable consumed in the northern and eastern regions of Thailand. Its aqueous extract was evaluated for inhibitory activity against DENV-2 within the 10-100-mg/mL concentration range. Both pre-and post-incubation of HepG2 cells with H. cordata extract resulted in a significant reduction in intracellular DENV-2 RNA production, correlating with a decrease in DENV-2 protein expression. The extract directly inhibited intracellular viral RNA

Direct Inhibitory Activities of Plant Extracts against DENV Enzymes
The medicinal plants Vernonia cinerea, Hemigraphis reptans, Hedyotis auricularia, Laurentia longiflora, Tridax procumbens and Senna angustifolia were used to evaluate their abilities to inhibit the DENV NS3 protease. The highest inhibitory effects were observed for the ethanolic extract of S. angustifolia leaves, the methanolic extract of V. cinerea leaves and the ethanolic extract of T. procumbens stems (IC 50 values: 30.1 ± 3.4, 23.7 ± 4.1 and 25.6 ± 3.8 µg/mL, respectively). The most active extracts were also tested in vitro against DENV-2-infected Vero cells, which were able to maintain a normal morphology without cytopathic effects. The percent viral inhibition of the extracts of V. cinerea (80.6% ± 6.1%) and T. procumbens (64.0% ± 9.4%) was significantly higher than that of S. angustifolia extract (26.3% ± 3.8%), as measured by a plaque-forming assay and RT-qPCR [52]. The authors of this investigation did not analyze which compounds were responsible for the observed activities.

Indirect Inhibitory Activities of Compounds and Extracts against DENV Enzymes
The plant Houttuynia cordata is a popular vegetable consumed in the northern and eastern regions of Thailand. Its aqueous extract was evaluated for inhibitory activity against DENV-2 within the 10-100-mg/mL concentration range. Both pre-and post-incubation of HepG2 cells with H. cordata extract resulted in a significant reduction in intracellular DENV-2 RNA production, correlating with a decrease in DENV-2 protein expression. The extract directly inhibited intracellular viral RNA replication, with an effective concentration (EC 50 ) of 0.8 mg/mL. Within the 10-40-mg/mL concentration range, the H. cordata extract also exhibited a protective effect on virion release from infected LLC-MK2 cells (Rhesus Monkey Kidney Epithelial Cells). The reduction in RNA production and decrease in DENV-2 protein expression suggest the inhibition of viral enzymatic activity by the aqueous extract. High-performance liquid chromatography of the aqueous extract revealed that hyperoside (Compound 39) (Figure 6) is the major component of the extract, suggesting that this flavonoid plays an important role in DENV-2 inhibition. The aqueous extract had no toxic effects on human blood cells [53].  The fractionation of the ethanolic extract of Zoanthus spp. (sea anemone) collected in Taiwan resulted in the isolation of 14 ecdysones; the inhibitory activities of these compounds against DENV-2 were evaluated. Among the isolated ecdysonoids, Compounds 40-43 were the most potent. The results shown in Figure 7 suggest that ajugasterone C (Compound 42) is equipotent to 2′-C-methylcytidine (positive control), with a better Seletive Index (SI). In addition, the most active ecdysone, Compound 42, was also tested against other DENV serotypes. Ecdysone 42 was found to be active against all DENV serotypes, with the following EC50 values: DENV-1 (15.70 ± 2.36 μM), DENV-3 (9.48 ± 0.24 μM) and DENV-4 (12.15 ± 1.22 μM) with a high CC50 (the 50% cytotoxic concentration). After analyses of both the structure-activity relationship and molecular docking, the authors proposed that ecdysone 42 impairs DENV RNA replication by blocking the viral polymerase channel [54].   The fractionation of the ethanolic extract of Zoanthus spp. (sea anemone) collected in Taiwan resulted in the isolation of 14 ecdysones; the inhibitory activities of these compounds against DENV-2 were evaluated. Among the isolated ecdysonoids, Compounds 40-43 were the most potent. The results shown in Figure 7 suggest that ajugasterone C (Compound 42) is equipotent to 2 -C-methylcytidine (positive control), with a better Seletive Index (SI). In addition, the most active ecdysone, Compound 42, was also tested against other DENV serotypes. Ecdysone 42 was found to be active against all DENV serotypes, with the following EC 50 values: DENV-1 (15.70 ± 2.36 µM), DENV-3 (9.48 ± 0.24 µM) and DENV-4 (12.15 ± 1.22 µM) with a high CC 50 (the 50% cytotoxic concentration). After analyses of both the structure-activity relationship and molecular docking, the authors proposed that ecdysone 42 impairs DENV RNA replication by blocking the viral polymerase channel [54].   Figure 7 suggest that ajugasterone C (Compound 42) is equipotent to 2′-C-methylcytidine (positive control), with a better Seletive Index (SI). In addition, the most active ecdysone, Compound 42, was also tested against other DENV serotypes. Ecdysone 42 was found to be active against all DENV serotypes, with the following EC50 values: DENV-1 (15.70 ± 2.36 μM), DENV-3 (9.48 ± 0.24 μM) and DENV-4 (12.15 ± 1.22 μM) with a high CC50 (the 50% cytotoxic concentration). After analyses of both the structure-activity relationship and molecular docking, the authors proposed that ecdysone 42 impairs DENV RNA replication by blocking the viral polymerase channel [54].  β-Carbolines are widespread plant and animal alkaloids that possess important biological activities [55]. Quintana and co-workers evaluated the effects of natural and synthetic β-carbolines; they found that the natural product harmol (Compound 44) and its synthetic derivative 9-N-methylharmine (Compound 45) were the most potent alkaloids, with inhibitory activities against DENV-2 and SI values of 56. β-Carbolines are widespread plant and animal alkaloids that possess important biological activities [55]. Quintana and co-workers evaluated the effects of natural and synthetic β-carbolines; they found that the natural product harmol (Compound 44) and its synthetic derivative 9-N-methylharmine (Compound 45) were the most potent alkaloids, with inhibitory activities against DENV-2 and SI values of 56.2 and 61.3, respectively (Figure 8). Vero cells were infected with DENV at a multiplicity of infection (MOI) of 1 or 0.1 PFU/cell. At 48 h post-infection, the cells were lysed; and the supernatants were harvested; virus yields were quantified by plaque assay. The results indicate that the compounds likely act via viral enzyme inhibition. Harmol (Compound 44) and 9-N-methylharmine (Compound 45) were also found to possess inhibitory effects against DENV-1, DENV-3 and DENV-4, albeit with lower efficiencies than against DENV-2 [56].

In Silico Investigations of DENV Enzyme Inhibitors
Galiano and co-authors conducted an in silico study to identify new inhibitors of the NS5 RdRp in the four DENV serotypes. They used a chemical library of 372,792 non-nucleotide compounds to perform molecular docking experiments at the binding site of the RNA template tunnel of the polymerase. After the screening process was completed, 39 compounds were identified as leading DENV RdRp inhibitor candidates. The selected compounds had a highly negative free energy variation (∆G) when docked to the binding site of the RNA template tunnel in the four DENV serotypes. In addition, the majority of the selected compounds had favorable druggability and optimal ADMET (absorption, distribution, metabolism, excretion, and toxicity) properties [57]. Among the 39 selected compounds, 10 (Compounds 46-55 in Figure 9) were natural products.

In Silico Investigations of DENV Enzyme Inhibitors
Galiano and co-authors conducted an in silico study to identify new inhibitors of the NS5 RdRp in the four DENV serotypes. They used a chemical library of 372,792 non-nucleotide compounds to perform molecular docking experiments at the binding site of the RNA template tunnel of the polymerase. After the screening process was completed, 39 compounds were identified as leading DENV RdRp inhibitor candidates. The selected compounds had a highly negative free energy variation (∆G) when docked to the binding site of the RNA template tunnel in the four DENV serotypes. In addition, the majority of the selected compounds had favorable druggability and optimal ADMET (absorption, distribution, metabolism, excretion, and toxicity) properties [57]. Among the 39 selected compounds, 10 (Compounds 46-55 in Figure 9) were natural products.

Structural Biology of CHIKV
The genome of CHIKV is a positive-sense, single-stranded RNA (ssRNA) genome, consisting of two open reading frames: The 5 end encodes nonstructural proteases necessary for the formation of viral replicase complexes, and the 3 end encodes structural proteins necessary for receptor binding and cell membrane fusion [4]. Alphaviruses have a positive-sense ssRNA genome that is approximately 11 kb in length and encodes nine proteins. Four nonstructural proteins (nsP1-nsP4) are encoded by the 5 end of the genome. The 3 end of the genome encodes a polyprotein precursor containing three structural proteins: PE2 (the precursor of E3 and E2), E1 and the capsid protein (C) [59].
The nonstructural proteins nsP1, nsP2 and nsP4 possess enzymatic activities. They are involved in viral RNA genome synthesis, methylation of the 5 cap of the viral genome and protein processing, among other activities [59].

nsP1
CHIKV nsP1 is a palmitoylated protein, consisting of 535 amino acids. The N-terminal region of nsP1 is responsible for the MTase and guanylyltransferase activities involved in capping and methylation of the newly-formed viral genomic and subgenomic RNAs. Recently, this protein was reported to play a crucial role in the downregulation of bone marrow stromal antigen-2 (BST-2), suggesting that nsP1 would be a potential target of BST-2-mediated therapeutics targeting CHIKV [7,59]. Elimination of nsP1 abolishes CHIKV replication. nsP1 genes are highly conserved in CHIKV strains and are important for virus proliferation in host cells. Therefore, nsP1 represents a rational target for antiviral therapies [60].

nsP2
The nonstructural protein nsP2 possesses numerous enzymatic activities and functional roles. This viral protein consists of an N-terminal domain with RNase and NTPase activities and a protease domain and an MTase-like region of unknown function at its C-terminus. In addition to its role as a cofactor of the viral polymerase complex, nsP2 is a virulence factor [4].
Similar to DENV NS3, CHIKV nsP2 cleaves viral polyproteins into four nonstructural proteins via the thiol protease complex at its C-terminus. The proteolytic activity of nsP2 plays an important role in the cleavage of nonstructural polyproteins, which are critical for viral replication. On the other hand, the RNA triphosphatase activity essential for RNA capping is found in the N-terminus of nsP2. The recently-solved crystal structure of CHIKV nsP2 might function as a crucial starting point for the development of novel antivirals targeting CHIKV [4,7,61,62]. The complete nucleotide sequence of the CHIKV genome revealed that nsP2 is the largest nonstructural protein: 798 amino acids long with a large, net positive charge [63].

nsP4
CHIKV nsP4 is an RdRp that is believed to participate in protein unfolding in the host cell, which helps replicate genomic RNA via a negative-strand RNA. CHIKV nsP4 also helps transcribe the 26S subgenomic mRNA, which encodes structural proteins [64,65].
Mature CHIKV nsP2 has been found to interact with both nsP1 and nsP4. Newly-synthesized simple strand RNAs may then be capped by nsP2 (RNA triphosphatase activity) and nsP1 (MTase and guanylyltransferase activities), suggesting that the interactions between nsP1, nsP2 and nsP4 are indispensable, as observed in a study of CHIKV nsPs [66,67]. Based on the crucial role of the CHIKV enzymes in viral replication, nsP1, nsP2 and nsP4 are important antiviral drug targets.

Direct Inhibitory Activities of Compounds against CHIKV Enzymes
Lucas-Hourani and collaborators [8] conducted a phenotypic assay to identify CHIKV inhibitors that target nsP2 [8]. In their investigation, the inhibitory activities of 3040 compounds, at a final concentration of 10 µg/mL (average concentration of 30 ± 13 µM), were tested; the natural compound derivative 60, shown in Figure 11, partially blocks nsP2 activity. This compound was then tested for its potential impact on CHIKV replication in vitro: HEK-293T cells were infected with CHIKV-Luc (MOI = 0.2) in the presence of the compound at 7.75, 15.5 or 31 µM (5, 10 or 20 µg/mL, respectively). A 50% reduction in Renilla luciferase activity was observed in the presence of Compound 60 at 31 µM (the IC 50 value), thus demonstrating a weak, but statistically-significant inhibition of viral replication [8].
its potential impact on CHIKV replication in vitro: HEK-293T cells were infected with CHIKV-Luc (MOI = 0.2) in the presence of the compound at 7.75, 15.5 or 31 μM (5, 10 or 20 μg/mL, respectively). A 50% reduction in Renilla luciferase activity was observed in the presence of Compound 60 at 31 μM (the IC50 value), thus demonstrating a weak, but statistically-significant inhibition of viral replication [8].

Indirect Inhibitory Activities of Compounds and Extracts against CHICKV Enzymes
The diterpenoids shown in Figure 3 were not evaluated against a specific CHIKV enzyme; however, all of them inhibited virus replication in the post-treatment assay of Vero cells, as seen in Table 1. Even though a specific target was not identified, the inhibitory effects suggest that enzymes involved in replication were affected. From the data presented in Table 1, the selective indices for Compounds 18, 19 and 20 correspond to 23, 2.4 and 2.7, respectively. Compounds 18-20 were found to be more potent than chloroquine, which was used as a positive control [49].

Indirect Inhibitory Activities of Compounds and Extracts against CHICKV Enzymes
The diterpenoids shown in Figure 3 were not evaluated against a specific CHIKV enzyme; however, all of them inhibited virus replication in the post-treatment assay of Vero cells, as seen in Table 1. Even though a specific target was not identified, the inhibitory effects suggest that enzymes involved in replication were affected. From the data presented in Table 1, the selective indices for Compounds 18, 19 and 20 correspond to 23, 2.4 and 2.7, respectively. Compounds 18-20 were found to be more potent than chloroquine, which was used as a positive control [49]. 1.9 ± 0.6 4.6 ± 0.8 20 3.9 ± 1.0 10.5 ± 0.1 Chloroquine 11.0 ± 2.1 100 ± 25 Using a cell-based bioassay, the natural products shown in Figure 5 were also evaluated for inhibitory activities against CHIKV. Although Compounds 33-35 had antiviral effects (Table 2), no selectivity was observed. On the contrary, terpenes 36 (EC 50 = 77 µM) and 37 (EC 50 = 86 µM) presented lower, but selective activity against CHIKV (an SI of about three) [51]. The flavonoids shown in Figure 12 were investigated for their inhibitory activities against CHIKV. qRT-PCR, immunofluorescence assay and western blot analyses all indicated that baicalein (Compound 61), fisetin (Compound 62) and quercetagetin (Compound 63) affect CHIKV RNA production and viral protein expression, with IC 50 values of 1.89 µg/mL (6.99 µM), 8.44 µg/mL (29.5 µM) and 13.85 µg/mL (43.52 µM), respectively, and minimal cytotoxicity. These data provide the first evidence of the intracellular anti-CHIKV activities of these compounds [68].
Molecules 2017, 22,505 14 of 20 (29.5 μM) and 13.85 μg/mL (43.52 μM), respectively, and minimal cytotoxicity. These data provide the first evidence of the intracellular anti-CHIKV activities of these compounds [68]. The CHIKV replicon cell line, BHK-CHIKV-NCT, was validated and used for screening 356 compounds, including 123 natural compounds and 233 clinically-approved drugs and other pharmaceutical compounds. After a 48-h exposure of the cell line to the compounds at a concentration of 50 mM, cellular EGFP appeared red, as the endpoint of the primary screen. The limit of the screen was set at a greater than 75% reduction of the EGFP signal, and the antiviral activities of all active compounds were confirmed in a second experiment that examined both EGFP and Rluc marker levels. The dose-dependent suppression of marker genes in the replicon vector after the 48-h exposure was observed for natural compounds apigenin (Compound 64), chrysin (Compound 65), naringenin (Compound 66) and silybin (Compound 67). The IC50 values for EGFP and Rluc are shown in Figure 13. All compounds showed low cytotoxicity with CC50 values greater than 200 μM, except for naringenin (Compound 85), which had a CC50 value of 122.1 μM [69].
Harringtonine (Compound 68) (Figure 14), a cephalotaxine alkaloid, potently inhibited CHIKV infection (EC50 = 0.24 μM), with minimal cytotoxicity (cell viability >90%). Harringtonine treatment resulted in a significant dose-dependent reduction in both negative-and positive-sense RNAs in both CHIKV strains, suggesting that it inhibited a phase in the CHIKV replication cycle that occurs before RNA production. Additionally, results also suggest that this alkaloid may inhibit CHIKV protein production, leading to a decrease in infectious virus titers [70].

Compound
EGPF (in μM) Rluc (in μM) 64 22    52 μM), respectively, and minimal cytotoxicity. These data provide the first evidence of the intracellular anti-CHIKV activities of these compounds [68]. The CHIKV replicon cell line, BHK-CHIKV-NCT, was validated and used for screening 356 compounds, including 123 natural compounds and 233 clinically-approved drugs and other pharmaceutical compounds. After a 48-h exposure of the cell line to the compounds at a concentration of 50 mM, cellular EGFP appeared red, as the endpoint of the primary screen. The limit of the screen was set at a greater than 75% reduction of the EGFP signal, and the antiviral activities of all active compounds were confirmed in a second experiment that examined both EGFP and Rluc marker levels. The dose-dependent suppression of marker genes in the replicon vector after the 48-h exposure was observed for natural compounds apigenin (Compound 64), chrysin (Compound 65), naringenin (Compound 66) and silybin (Compound 67). The IC50 values for EGFP and Rluc are shown in Figure 13. All compounds showed low cytotoxicity with CC50 values greater than 200 μM, except for naringenin (Compound 85), which had a CC50 value of 122.1 μM [69].
Harringtonine (Compound 68) (Figure 14), a cephalotaxine alkaloid, potently inhibited CHIKV infection (EC50 = 0.24 μM), with minimal cytotoxicity (cell viability >90%). Harringtonine treatment resulted in a significant dose-dependent reduction in both negative-and positive-sense RNAs in both CHIKV strains, suggesting that it inhibited a phase in the CHIKV replication cycle that occurs before RNA production. Additionally, results also suggest that this alkaloid may inhibit CHIKV protein production, leading to a decrease in infectious virus titers [70].

Compound
EGPF (in μM) Rluc (in μM) 64 22    Harringtonine (Compound 68) (Figure 14), a cephalotaxine alkaloid, potently inhibited CHIKV infection (EC 50 = 0.24 µM), with minimal cytotoxicity (cell viability >90%). Harringtonine treatment resulted in a significant dose-dependent reduction in both negative-and positive-sense RNAs in both CHIKV strains, suggesting that it inhibited a phase in the CHIKV replication cycle that occurs before RNA production. Additionally, results also suggest that this alkaloid may inhibit CHIKV protein production, leading to a decrease in infectious virus titers [70]. Inhibitory activities of the aqueous, aqueous-ethanolic and ethanolic extracts of the roots of Decalepis hamiltonii and the leaves of Vitex negundo and Hyptis suaveolens were assessed against two strains (Asian and East Central South African lineages) of CHIKV in a post-treatment assay. Both the V. negundo ethanolic extract and the H. suaveolens aqueous-ethanolic extract were found to effectively inhibit the Asian strain, with both exhibiting 50% virus inhibition at a concentration of 15.62 μg/mL. The selectivity index of H. suaveolens (16) was higher than that of V. negundo (2) against the Asian CHIKV strain. No extracts were able to inhibit the East Central South African CHIKV strain, presumably because of the higher virulence and replicative features of the East Central South African strain compared to the Asian strain [71].

In Silico Investigations of CHIKV Enzyme Inhibitors
Setzer and collaborators conducted a molecular docking investigation to identify alphavirus protease inhibitors from natural compounds [72]. The inhibitory activities of 2174 natural substances were screened against several proteases, including nsP2 CHIKV. Several compounds presented notable docking energies and selectivities when screened against CHIKV nsP2, including the alkaloids 2,3-dehydrosarsalignone (Compound 69) and sarachine (Compound 70) (Figure 15). Both compounds occupy the same position in their lowest-energy docked position with the CHIKV protease.  Inhibitory activities of the aqueous, aqueous-ethanolic and ethanolic extracts of the roots of Decalepis hamiltonii and the leaves of Vitex negundo and Hyptis suaveolens were assessed against two strains (Asian and East Central South African lineages) of CHIKV in a post-treatment assay. Both the V. negundo ethanolic extract and the H. suaveolens aqueous-ethanolic extract were found to effectively inhibit the Asian strain, with both exhibiting 50% virus inhibition at a concentration of 15.62 µg/mL. The selectivity index of H. suaveolens (16) was higher than that of V. negundo (2) against the Asian CHIKV strain. No extracts were able to inhibit the East Central South African CHIKV strain, presumably because of the higher virulence and replicative features of the East Central South African strain compared to the Asian strain [71].

In Silico Investigations of CHIKV Enzyme Inhibitors
Setzer and collaborators conducted a molecular docking investigation to identify alphavirus protease inhibitors from natural compounds [72]. The inhibitory activities of 2174 natural substances were screened against several proteases, including nsP2 CHIKV. Several compounds presented notable docking energies and selectivities when screened against CHIKV nsP2, including the alkaloids 2,3-dehydrosarsalignone (Compound 69) and sarachine (Compound 70) (Figure 15). Both compounds occupy the same position in their lowest-energy docked position with the CHIKV protease. Inhibitory activities of the aqueous, aqueous-ethanolic and ethanolic extracts of the roots of Decalepis hamiltonii and the leaves of Vitex negundo and Hyptis suaveolens were assessed against two strains (Asian and East Central South African lineages) of CHIKV in a post-treatment assay. Both the V. negundo ethanolic extract and the H. suaveolens aqueous-ethanolic extract were found to effectively inhibit the Asian strain, with both exhibiting 50% virus inhibition at a concentration of 15.62 μg/mL. The selectivity index of H. suaveolens (16) was higher than that of V. negundo (2) against the Asian CHIKV strain. No extracts were able to inhibit the East Central South African CHIKV strain, presumably because of the higher virulence and replicative features of the East Central South African strain compared to the Asian strain [71].

In Silico Investigations of CHIKV Enzyme Inhibitors
Setzer and collaborators conducted a molecular docking investigation to identify alphavirus protease inhibitors from natural compounds [72]. The inhibitory activities of 2174 natural substances were screened against several proteases, including nsP2 CHIKV. Several compounds presented notable docking energies and selectivities when screened against CHIKV nsP2, including the alkaloids 2,3-dehydrosarsalignone (Compound 69) and sarachine (Compound 70) (Figure 15). Both compounds occupy the same position in their lowest-energy docked position with the CHIKV protease.  The amide glucoside drodrenin (Compound 71) was also selective for the CHIKV nsP2 protease; the glucose moiety of drodrenin occupies a position near the active site (close to amino acids Cys 1013 and His 1083 ).

Concluding Remarks
The exploitation of the natural product pool has afforded compounds and extracts that present important inhibitory effects on enzymes involved in DENV and CHIKV replication and protein expression. This is an important approach since inhibition of these enzymes can disrupt the virus life cycle. Most of the studies herein described were conducted in vitro using enzymatic assays [8,[47][48][49][50][51][52] or viral strain/replicon on cells evaluating replication, inhibition and/or protein expression [49,[51][52][53][54]56,[68][69][70][71]. Three studies were conducted utilizing the in silico docking approach [57,58,72]. In vivo investigations of the described compounds and extracts were not reported. Most of the described enzymatic assays were carried out on DENV enzymes, and just one used CHIKV enzyme [8]. For CHIKV, assays performed on cells with a viral strain are one important option. Further investigations with heterologous expression of CHIKV enzymes for antiviral analyses are necessary so that a more profound evaluation of compounds against viral enzymes could be accomplished. In order to obtain a deeper understanding of the potential of the described natural products and extracts as antivirals for DENV and CHIKV, it is essential to carry out detailed in vitro and in vivo studies. It is important to mention that subsequent clinical trials are necessary, since efficiency in vitro is indicative, but it does not represent a guarantee of a good clinical trial. Considering the natural compounds described in this review, it is noticed that flavonoids are the ones that have been more extensively investigated.
Most of the investigations described in this review were performed with natural products from plant extracts. However, other sources of natural products, such as animals, microorganisms and marine organisms, should be explored towards the discovery of antivirals against DENV and CHIKV. It is also important to mention the fact that structures of natural compounds can be optimized, in terms of biological activities, via synthetic transformations. Thus, the compounds described in this review might be useful as starting points for synthetic endeavors towards the synthesis of more potent compounds against DENV and CHIKV. Finally, it is noticed from this review that, to date, there are few reports on the effect of natural products on dengue and chikungunya viral enzymes. In other words, the natural product pool is relatively unexplored in this regard. Considering the vast array of compounds available from nature, it is possible to anticipate that more investigations in this field will be described in the near future.