Characterization of Anti-Poliovirus Compounds Isolated from Edible Plants

Poliovirus (PV) is the causative agent of poliomyelitis and is a target of the global eradication programs of the World Health Organization (WHO). After eradication of type 2 and 3 wild-type PVs, vaccine-derived PV remains a substantial threat against the eradication as well as type 1 wild-type PV. Antivirals could serve as an effective means to suppress the outbreak; however, no anti-PV drugs have been approved at present. Here, we screened for effective anti-PV compounds in a library of edible plant extracts (a total of 6032 extracts). We found anti-PV activity in the extracts of seven different plant species. We isolated chrysophanol and vanicoside B (VCB) as the identities of the anti-PV activities of the extracts of Rheum rhaponticum and Fallopia sachalinensis, respectively. VCB targeted the host PI4KB/OSBP pathway for its anti-PV activity (EC50 = 9.2 μM) with an inhibitory effect on in vitro PI4KB activity (IC50 = 5.0 μM). This work offers new insights into the anti-PV activity in edible plants that may serve as potent antivirals for PV infection.


Introduction
Poliovirus (PV) is a small non-enveloped virus with a positive-sense single-stranded RNA genome of about 7500 nt belonging to the family Picornaviridae, including poliovirus (PV, species Enterovirus C) [1]. PV is the causative agent of poliomyelitis, which mainly affects children under 5 years of age, and is a target of global eradication by the World Health Organization (WHO). Through vaccination programs of the Global Polio Eradication Initiative beginning in 1988 with a live oral PV vaccine (OPV) and/or an inactivated PV vaccine (IPV), type 2 and 3 wild-type PVs (WPVs) have been eradicated (declared in 2015 and 2019, respectively), and only Pakistan and Afghanistan remain as endemic countries of type 1 WPV as of 2022. However, circulating vaccine-derived PV (cVDPV) remains a substantial threat against the eradication (724 cases in 2022), especially type 2 cVDPV that emerged after the global cessation of type 2 OPV in 2016 [2], as well as type 1 WPV (30 cases in 2022) [3]. Transmission of PV could be re-established by the importation of the following strains: polio cases by type 1 WPV in Malawai in 2021 and Mozambique in 2022, a case by type 2 cVDPV, and the silent circulation in the United States of America and [4] the United Kingdom in 2022 [5]. To interrupt PV circulation, only an OPV campaign for a potentially susceptible population in the area could serve as the effective mean at present (in case of an outbreak response in Israel, the target population was children under 10 years of age who have received at least one dose of IPV) [6]. In addition to conventional OPV, novel OPV type 2 (nOPV2), which was designed to have more genetic stability than type 2 OPV and to decrease the risk of VDPV [7], is available for the outbreak response under Emergency Use Listing of the WHO authorized in 2020. Currently, nOPV2 has mainly been used in African countries since 2021 (reviewed in [8]).
Target population of PV is mainly for children under five years of age as well as other enteroviruses (EVs) [33,34]; therefore, safety is one of the major challenges for the antiviral development [10]. In a previous study, a highly active anti-EV compound (EC 50 = 2.0 µM) was isolated from avocado [35], suggesting that edible plants provide a promising source for potent anti-EV compounds. Here, we isolated anti-PV compounds from edible plants, Rheum rhaponticum and Fallopia sachalinensis, and analyzed the potency and the mechanism of action of their isolated compounds.

Materials and Methods
Cells. Cells were cultured as monolayers in Dulbecco's modified Eagle medium (DMEM, FUJIFILM Wako Pure Chemical Corporation, Osaka, JPN, 044-29765) supplemented with 10% foetal calf serum (FCS). RD cells (human rhabdomyosarcoma cells) were used for the titration of PV and evaluation of anti-PV activity of plant extracts. HEK293 cells (human embryonic kidney cells) were used for the production of type 1 PV pseudovirus (PV1 pv ). A PI4KB-knockout RD cell line (RD[∆PI4KB]) was used to evaluate the antiviral effects of plant extracts targeting PI4KB/OSBP-independent viral replication [36].
General methods for molecular cloning. Escherichia coli strain XL10gold (Agilent Technologies, Inc., Santa Clara, CA, USA) was used for the preparation of plasmids. PCR was performed using KOD Plus DNA polymerase (TOYOBO CO., LTD., Osaka, Japan). DNA sequencing was performed using a BigDye Terminator v3.1 cycle sequencing ready reaction kit (Thermo Fisher Scientific Inc., Waltham, MA, USA) and then analyzed with a 3500xL genetic analyzer (Thermo Fisher Scientific Inc., Waltham, MA, USA).
Plasmids: Rupintrivir-resistant PV replicon. A resistant mutation to rupintrivir was introduced into a plasmid encoding the cDNA of a PV replicon (pPV-Fluc mc) [43], by PCR with primer set 1.
Primer set 2: 5 -CAGCTCCAGCCCGGTTCTCGGTACCGTATG-3 5 -CATACGGTACCGAGAACCGGGCTGGAGCTG-3 Edible plant extract library. A plant extract library was prepared in the Research Center for Medicinal Plant Resources (NIBIOHN), by the methanol extraction of dried and pulverized plant materials, with evaporation, dissolution in DMSO, and filtration. The concentration of the final DMSO solution was adjusted to 40 mg/mL for all extracts. This plant extract library is a collection of extracts from a wide range of wild plants in Japan, which can be widely used not only in drug discovery but also in the life sciences field, such as health food development. All the original plants in the library are annotated with information on whether or not they have been eaten by humans. The presence of food experience in the original plant reflects the safety of the extract for human beings, which is useful information for drug discovery. We selected samples from this library derived from the original plants with food experience and used them in this study.
Measurement of inhibitory effect of compounds on in vitro PI4KB activity. The in vitro activity of purified GST-PI4KB (PV5277, Thermo Fisher Scientific Inc., Waltham, MA, USA) was evaluated by using an ADP-Glo Lipid Kinase Systems kit (Promega Corporation, Madison, WI, USA) as previously described. In a total 5.5 µL reaction solution, the PI4KB activity of 32 ng of purified GST-PI4KB (final concentration of 48 nM) with lipid substrates (0.025 mg/mL of phosphatidylinositol and 0.075 mg/mL or phosphatidylserine) and 25 µM of ATP was measured in the presence or the absence of compounds. The net signals of the mock-treated samples were taken as 100% of the PI4KB activity. The 50% inhibitory concentration (IC 50 ) values of the compounds were determined by nonlinear regression analyses of the dose-response curves.
Statistical analysis. The results of the experiments are shown as means with standard deviations. Values of p < 0.05 by one-tailed t tests were considered to indicate a significant difference, and were indicated by asterisks (* p < 0.05, ** p < 0.01, *** p < 0.001).

Screening of Edible Plant Extracts for Anti-PV Activity
We screened a total of 6032 edible plant extracts for anti-PV activity in RD cells as that previously performed for the screening for anti-EV-D68 activity [35] (Figure 1). The plant extracts were added to the RD cells (final concentration of 0.2 mg/mL), and then infected with PV1(Sabin 1) at a multiplicity of infection (MOI) of 0.1. The extracts that completely protected the cells from the viral infection after 1 day post-infection (p.i.) were identified as initial hit extracts. We identified 8 hit extracts, which consisted of 7 plant species. In this study, we focused on the identification of the identity of the antiviral effects in R. rhaponticum and F. sachalinensis, among the hits for availability of the plant materials.

Purification and Structure Determination of Anti-PV Compound in F. Sachalinensis
The methanolic extracts of the F. sachalinensis (Polygonaceae) root was purified by silicagel column chromatography with chloroform-methanol as an eluent to give 20 fractions ( Figure 3A). Fractions eluted with 50% methanol/chloroform were combined and subjected to HPLC separation to obtain vanicoside B (VCB, 1.1 mg). The chemical structure was determined via comparison with NMR data from the literature [53].

Characterization of Anti-PV Activity of VCB
To evaluate the potency of VCB as an anti-PV compound, we determined the 50% cytotoxic concentration (CC 50 ) in human RD cells and 50% effective concentration (EC 50 ) for the PV infection of VCB ( Figure 3B). The cytotoxicity of VCB was not observed when the cells were treated with 100 µM VCB for 7 h, but the CC 50 of VCB after 2 days of treatment was 27 µM. The EC 50 of VCB for PV1 pv infection was 9.2 µM. The selectivity index (SI) of VCB for anti-PV activity in RD cells was 2.9, suggesting a low specificity for the anti-PV activity of VCB (e.g., SI of PI4KB inhibitors for the anti-PV activities could be >1000) [25]. VCB protected the RD cells from the infection of PV1(Sabin 1), EV-A71(Nagoya), or EV-D68(Fermon) only at 20 µM, suggesting that the potential therapeutic window of VCB is quite narrow.

Mechanism of Anti-PV Activity of VCB
To evaluate the specificity of the anti-PV activity for VCB, we analyzed the antiviral activity with a panel of drug-resistant PV mutants in parental (wild-type) RD cells       To further dissect the target of VCB, we analyzed the effect of VCB on the subcellular localization of host OSBP. OSBP relocalizes to the Golgi in the presence of OSBP inhibitors via the lipid-transfer domain [30,54,55]. HEK293 cells overexpressing C-terminally EGFP-fused OSBP were treated with an OSBP inhibitor T-00127-HEV2 or VCB ( Figure 4B). While treatment of the cells with T-00127-HEV2 caused the relocalization of the ectopically expressed OSBP to the Golgi, the treatment with VCB did not affect the subcellular localization, suggesting that OSBP is not the target of VCB. Next, we analyzed the effect of VCB on the PI4KB activity ( Figure 4C). VCB showed an inhibitory effect on the in vitro PI4KB activity albeit with low potency (IC 50 = 5000 nM) compared to a PI4KB inhibitor T-00127-HEV1 (IC 50 = 34 nM), which possibly targets the ATP-binding site of PI4KB, similar to its analogue [26]. We also analyzed a potential allosteric effect of VCB on the PI4KB activity. VCB inhibited the activity of a PI4KB variant (C646S) as well as T-00127-HEV1, in contrast to MDL-860 that has an allosteric inhibitory effect on PI4KB via a covalent modification of the Cys646 residue [27,56]. These results suggested that PI4KB is the direct target of VCB for the anti-PV activity.
The petiole of R. rhaponticum is edible, called rhubarb, and is used mainly as a jam. On the other hand, the roots of many Rheum spp. are considered medicinal and laxative because they contain many anthraquinones. F. sachalinensis is a large herbaceous plant whose grass can grow up to 2 m tall. The Ainu tribe of Japan traditionally eats its young stems and sprouts. The rhizome of F. sachalinensis has antibacterial, antitussive, and diuretic properties, as well as an improvement in its laxative effects. The Japanese name for this plant ("ooitadori") is derived from the fact that when bruised, its leaves can be applied to the affected area to relief pain. Approximately 360 g of the root of R. rhaponticum contained 0.2 g of chrysophanol (about 0.05%w/w). The content of VCB in the roots of F. sachalinensis dry was estimated to be about 0.001%w/w, based on the weight of the isolate and its presence in other fractions. Although the roots of these two plant species are not edible parts, other parts have been traditionally consumed. Therefore, a certain degree of safety is considered to be assured.
Chrysophanol is a well-known purgative component of the roots of the Rheum spp. and Senna leaf (the leaves of Cassia angustifolia, C. acutifolia). It has also been reported as a constituent of Fallopia japonica, a closely related plant to F. sachalinensis. Chrysophanol is also known as an anti-PV compound [52]. Emodin, a compound structurally related to chrysophanol, has anti-EV activity targeting viral protein synthesis or virion maturation [62,63]. Chrysophanol targeted a site of a PV capsid protein similar to the capsid-binding uncoating inhibitors; however, the inhibitory effect of the chrysophanol on PV1 pv infection was rather weak even at a high concentration (about an 8-fold reduction at 790 µM, Figure 2B). The infection cycle of PV1 pv includes viral binding, uncoating, and replication, but not virion production (assembly, encapsidation, maturation, and egress) [37]; therefore, chrysophanol may have additional targets after the replication step, such as virion maturation, similar to emodin. The partial resistance of a brefeldin A-resistant mutant and a rupintrivir-resistant mutant against chrysophanol suggest the effects on the replication step that could have functional coupling to virion production [64,65]. Pocapavir (viral capsid-binding uncoating inhibitor) [66,67] and V-7404 (3C protease inhibitor) [68][69][70] have been considered as candidate antivirals in the polio eradication program [71,72]. The availability of chrysophanol in a broad plant species would allow further evaluation of the potency of the extracts, possibly in combination with other drugs/extracts with different antiviral mechanisms.
VCB was first isolated from nature in 1994 as a protein kinase C inhibitor from Polygonum pensylvanicum (Polygonaceae) together with vanicoside A [73]. Vanicosides have inhibitory effects on the viral proteases of the human immunodeficiency virus or SARS-CoV-2 [74,75], but its antiviral effects have yet to be evaluated. Unexpectedly, resistant mutants (the enviroxime-resistant mutant and ∆PI4KB-resistant [−2C] mutant) suggested that the target of VCB for the anti-PV activity is the host PI4KB/OSBP pathway rather than viral proteases (Figure 3). The inhibitory effect on in vitro PI4KB activity (IC 50 = 5.0 µM) suggested that VCB is a novel PI4KB inhibitor. PI4KB is a host factor required for the replication of EV identified by Hsu et al. [22]. The subsequent analysis on a group of anti-EV compounds (designated enviroxime-like compounds), which have PI4KB and an unknown factor as the targets for the anti-EV activity [23,29], revealed the host oxysterol-binding protein (OSBP) family I (OSBP and OSBP2/ORP4) as another target of this compound group [30,31]. PI4KB and OSBP form an inseparable functional axis for the formation of a viral replication complex by providing phosphatidylinositol 4-monophosphate (PI4P) for the recruitment of OSBP on viral replication organelles (ROs), and the accumulation of unesterified cholesterol on the ROs by OSBP [76]. This process enhances the cleavage of the viral 3AB protein and development of the RO for the synthesis of viral plus-strand RNA [36,57,[77][78][79]. In addition to the 3AB protein, the viral 2B protein is essential to complement the functional axis [36,65], while the functional role of the 2B protein remains largely unknown. PI4KB inhibitors generally show low cytotoxicity to cultured cells [20,21,23,28,43,80,81]; however, the antiproliferative effect in lymphocytes [80] and lethality in a mouse line [81] raised concerns on the safety in vivo. A recent study revealed a protective effect in PI4KB heterozygous kinase-dead mice against EV infection and the therapeutic potency of a specific PI4KB inhibitor in vivo [82], supporting the potential safety of PI4KB inhibitors in clinical use as opposed to earlier findings. Therefore, the PI4KB inhibitors isolated from edible plants could have a more important role than ever thought.
The limitations of this study include elucidation of the mechanism of the inhibitory effect of VCB on PI4KB activity and its off-target effect against clinical applicability. Most of the identified PI4KB inhibitors target the ATP-binding site of PI4KB [26,83], with MDL-860 as the exception. The inhibitory effect of VCB on in vitro PI4KB activity might suggest the direct interaction with PI4KB ( Figure 3C), but the target site remained to be determined. While the specificity to the PI4KB/OSBP pathway in terms of the anti-PV activity was clear, we could not exclude the potential contribution of the off-target effect of VCB (CC 50 = 27 µM) to the observed anti-PV activity, which had quite a narrow therapeutic window (complete protection of the cells from PV1[Sabin 1] infection at 20 µM). Data Availability Statement: Raw data sets not included in this paper are available from the corresponding authors upon request.