Identification of inositol monophosphatase as a broad‐spectrum antiviral target of ivermectin

Ivermectin has broad‐spectrum antiviral activities. Despite the failure in clinical application of COVID‐19, it can serve as a lead compound for the development of more effective broad‐spectrum antivirals, for which a better understanding of its antiviral mechanisms is essential. We thus searched for potential novel targets of ivermectin in host cells by label‐free thermal proteomic profiling using Huh‐7 cells. Inositol monophosphatase (IMPase) was found among the proteins with shifted thermal stability by ivermectin. Ivermectin could inhibit IMPase activity and reduce cellular myo‐inositol and phosphatidylinositol‐4‐phosphate levels. On the other hand, inositol could impair the antiviral activity of ivermectin and lithium, an IMPase inhibitor with known antiviral activity. As phosphatidylinositol phosphate is crucial for the replication of many RNA viruses, inhibition of cellular myo‐inositol biosynthesis may be an important antiviral mechanism of ivermectin. Hence, inhibition of IMPase could serve as a potential target for broad‐spectrum antiviral development.

viruses, it is likely that its main broad-spectrum antiviral mechanism had not been identified.We reasoned that a broad-spectrum antiviral activity against various unrelated viruses is likely to target a host mechanism that is commonly used by many viruses, and therefore performed a search for cellular target of IVM.
Thermal proteomic profiling (TPP) is based on the principle that proteins denature and become insoluble when exposed to heat.The thermal stability can alter when protein interacts with small molecules such as drugs, metabolites, and nucleic acids.Mass spectrometry (MS) can be used to identify remaining soluble protein fractions after exposure to various temperatures, enabling the identification of drug targets. 10TPP has been successfully used to identify drug target interactions and drug off-targets. 11 this study, TPP was performed to identify IVM-targeting host cellular proteins in Huh-7 cells.Following the comparison of MSidentified proteins in the samples treated with IVM and the untreated control after heat treatment, a number of IVM-interacting protein candidates were identified, including inositol monophosphatase (IMPase).The enzyme IMPase catalyzes the dephosphorylation of inositol monophosphate and contributes significantly to maintain myo-inositol levels, which serve as the substrate for biosynthesis of inositol phosphates, phosphatidylinositol (PtdIns), and the phosphorylated derivatives of PtdIns, phosphoinositides (PPIn), 12 which are essential for replication of various RNA viruses. 13,14Functional validation was then performed using various assays to strengthen the role of IMPase as the target for IVM action.

| Chemicals, cells, and viruses
All chemicals, cells, and viruses are listed in Supporting Information: Methods 1.1−1.2.

| Drug treatment and heat treatment
The thermal proteomic protocol to identify potential IVM targets in uninfected hepatocyte cells was adapted from Franken et al. 10 Huh-7 cells were seeded at a density of 6 × 10 6 cells/flask.The cultured media were removed, and the cells were treated with 2% heat-inactivated fetal bovine serum (FBS)-DMEM containing 7.5 μM IVM or 0.5% DMSO as a vehicle control for 4 h (5% CO 2 , 37°C).After that, the cells were trypsinized, and the cell viability was assessed using trypan blue.The cells were washed twice with ice-cold phosphate-buffered saline (PBS) and centrifuged into pellets.Then the cell pellets were carefully resuspended with 640 μL of ice-cold PBS.Each sample was distributed into 0.2-mL PCR tubes for 100 μL/tube.The cell suspensions were pelleted again by centrifugation, and 80 μL of PBS were removed.
The cell pellets were kept on ice until heat treatment.The cell pellets were heated in a thermal cycler at 37°C or 47°C for 3 min.
After that, the tubes were incubated at room temperature for 3 min.Immediately after incubation, the cells were lysed using RIPA buffer (RIPA buffer pH7.4,RB447, Bio-basic).The cell lysates were put on ice for 10 min and vortexed three times.Then the cell lysates were centrifuged at 20 000g for 40 min at 4°C to precipitate aggregated proteins and cell debris. 11The supernatants (soluble protein fraction) were carefully transferred to new tubes.The protein concentrations were measured using the Bradford assay.Then the proteins in soluble fractions were identified using MS (the detailed methods for MS and data analysis are shown in Supporting Information: Methods 1.4−1.7).

| Molecular docking and molecular dynamics simulations
The molecular docking of human IMPase 1 (IMP1; PDB ID: 4AS4) and human IMPase 2 (IMP2; PDB ID: 2DDK) with inositol monophosphate or IVM was performed by Autodock Vina 15 using the CB-Dock server. 16The interaction between the IMP1-IVM and IMP2-IVM complexes was investigated using Desmond Molecular Dynamics System, version 6.9, D. E. Shaw Research. 17The details are shown in Supporting Information: Methods 1.8.

| IMPase activity
For the in vitro enzymatic assay, the cell lysate of Huh-7 cells was used as a source of IMPase.The cells were harvested and washed several times with normal saline.Then, the cells were resuspended in 600 μL of Tris buffer (50 mM Tris-HCl, 1 mM EDTA, 3 mM MgCl 2 , 150 mM KCl, 0.5 mg/ mL BSA, and 0.01% v/v Triton X pH 7.4) and lysed using sonication (on 10 s, off 20 s, frequency of 20 kHz for 2 min).The cell lysate was obtained after centrifugation at 10 000g, 4°C for 10 min.The amount of total protein in the cell lysate was measured using a Bradford assay.Total protein (50 μg) was incubated with or without various concentrations of IVM (ranged between 3-20 μM and 125-500 μM) and 250 μM of inositol monophosphate (D-myo-Inositol 1-monophosphate dipotassium salt, 59937; Sigma) in Tris buffer for 1 h at 37°C.LiCl, an IMPase inhibitor, was used as an inhibition control. 18All reactions contain 0.125% DMSO.
Following the incubation, each reaction was separated into two portions for phosphate and myo-inositol detection.The details of the statistical analysis used throughout the study are shown in Supporting Information: Methods 1.12.

| Phosphate detection
The reactions were filtered through a 10 kD molecular weight cut-off spin column to remove any large proteins to minimize the background in the colorimetric assay.Then the filtrates were mixed with malachite green solution (MAK308; Sigma) following the manufacturer's protocol.An absorbance was measured at 620 nm for both samples and phosphate standards.

| Myo-inositol quantification
Myo-inositol levels were determined using a myo-inositol assay kit (fluorometric) (ab252896; Abcam) following the manufacturer's protocol.The detailed method is shown in Supporting Information: Methods 1.9.

| Cellular myo-inositol levels
Vero or Huh-7 cells were treated with media containing various concentrations of IVM, inositol, mixtures of IVM and inositol, or LiCl for 48 h.After that, the cells were washed several times with PBS and detached by trypsinization.The number of cells in each well was determined.Then the cells were immediately lysed with ice-cold inositol assay buffer.The samples were centrifuged at 10 000g at 4°C for 5 min to remove cell debris.The supernatants were transferred to new tubes, and then myo-inositol levels were determined using a myo-inositol assay kit.Cell viability was assessed using the MTT assay (Supporting Information: Methods 1.3).

| Reversion of IVM antiviral activity by inositol
Huh-7 cells were inoculated with DENV-2 or ZIKV at a MOI of 1, or A549 cells were inoculated with ZIKV at a MOI of 1 for 2 h.After that, the cells were further maintained in media containing various concentrations of IVM, inositol, mixtures of IVM and inositol, or 0.5% DMSO as an untreated control.The virus supernatants were collected at 48 h postinfection (hpi).The virus titers were determined using focus immunoassay titration (the detailed method is shown in Supporting Information: Methods 1.10).

| Antiviral activity of lithium and reversion of lithium antiviral activity by inositol
Vero or Huh-7 cells were inoculated with DENV-2 at MOI 0.5 or 1, respectively, for 2 h.After that, the cells were further maintained in media containing various concentrations of LiCl for 48 h.For the reversion of lithium antiviral activity by inositol, the experiments were performed in Vero cells.Following DENV-2 infection, the cells were further incubated in 2% FBS-MEM containing various concentrations of LiCl, inositol, and mixtures of LiCl and inositol for 48 h.The virus titers were determined using focus immunoassay titration.

| Immunofluorescence assay
The cells were mock-infected or infected with ZIKV or SARS-CoV-2 and treated with IVM or LiCl for the indicated time periods.After that, the cells were fixed and subjected to an immunofluorescence assay.The details of the immunofluorescence assay are shown in Supporting Information: Methods 1.11.

| Identification of IVM-targeting proteins using label-free TPP
The comparison of TPP-identified proteins in samples treated with IVM and vehicle control with heat treatment at 37°C or 47°C is shown in Figure 1.There are 44 proteins present in both vehicle controls (at 37°C and 47°C) and IVM-treated cells at 37°C, but are absent in IVM-treated cells at 47°C (Supporting Information: Table S1).And 62 proteins were only present in IVM-treated cells at 47°C (Supporting Information: Table S2).The conditions for drug treatment and heat treatment were optimized before performing TPP (Supporting Information: Figure S1).Both 44 and 62 identified proteins were subjected to enrichment analysis (Supporting Information: Figure S2).The TPP-identified proteins were enriched in various cellular compartments, including mitochondria, endoplasmic reticulum (ER), ER-to-Golgi transport, and ribosomes.The enriched biological processes and pathways related to ribonucleotide biosynthesis, ER-Golgi vesicle-mediated transport, the tricarboxylic cycle, protein transport, translation, fatty acid metabolism, and the biosynthesis of unsaturated fatty acids.We also detected previously identified targets of IVM, including importin.The identification of mitochondria as a potential target is also in agreement with previous reports showing that IVM interfered with mitochondrial function. 19,20Going through each candidate in the list, we looked for potential targets with known mechanisms clearly related to replication of viruses or with data showing involvement in viral replication.We identified phosphoribosylformylglycinamidine synthase (PFAS), YTH domaincontaining family protein 2, ATP-dependent RNA helicase DDX39A, and IMPase 2 as potential targets.PFAS is a key enzyme in de novo purine biosynthesis, and its inhibition may deplete cellular purine, which is required for viral RNA replication. 21YTH domain-containing family protein 2 regulates stability of mRNA containing N(6)-methyl adenosine and a closely related protein, YTH domain-containing family protein 3, was previously shown to suppress interferon-dependent antiviral responses. 22Many proteins in the DDX family were shown to be involved either as mediators of antiviral innate defense or as tools for viruses to evade the defense. 23Because of the possibility to test the candidates by reversing the antiviral activity of IVM by supplying downstream products of the candidates, that is, inosine and hypoxanthine for PFAS and inositol for IMPase, we decided to test these two candidates first.Inosine and hypoxanthine can be used in salvage pathway to generate purine nucleotides and should bypass inhibition of PFAS in purine biosynthesis.We found that adding inosine or hypoxanthine had no effect on IVM antiviral activity (Supporting Information: Figure S3).PFAS as IVM target was therefore not further pursued and we focused our further studies on IMPase.

| Molecular docking and molecular dynamics simulations
To elucidate whether IVM exert its activity through IMPase, interactions between the drug and the enzyme were examined using molecular docking.Two IMPase enzymes (IMP1 and IMP2) were retrieved from the database.The structures of inositol monophosphate, a substrate of IMPase, and IVM are shown in Supporting Information: Figure S4.Based on sequence alignment, IMP1 and IMP2 shared 53.65% sequence identity; however, the overall crystal structures of these two IMPase isoforms are very well superimposed (Figure 2A).Docking inositol monophosphate and IVM to IMP1 resulted in affinity binding scores of −7.3 and −7.7 kcal/mol, respectively (Figure 2B).A higher negative score indicates a stronger binding.Thus, docking results indicated a strong affinity between IVM and IMP1, similar to the enzyme and its substrate, inositol monophosphate.For IMP2, molecular docking of inositol monophosphate and IVM resulted in affinity binding scores of −7.6 and −9.3 kcal/mol, respectively (Figure 2C), suggesting that IVM has great binding affinity toward this protein.
The stability and interactions of IMP2-IVM and IMP1-IVM complexes were analyzed using root-mean-square deviation (RMSD), solvent-accessible surface area (SASA), and contact frequencies from molecular dynamics simulations (Figure 2D-G).
The RMSD of IVM in the IMP2 complex stabilized at approximately 5 Å after 15 ns, indicating its high stability.This stability is supported by the SASA data, which remained between 450 and 600 Å², suggesting low solvent exposure and consistent contact during the 100 ns simulation (Figure 2D).Moreover, specific interactions between IVM and IMP2 residues, such as GLN224 and HIS228, were maintained for over 30% of the simulation, demonstrating strong binding efficiency (Figure 2E).The IVM-IMP1 complex exhibited a relatively weaker interaction, with RMSD values increasing within the first 30 ns and fluctuating between 10 and 16 Å thereafter (Figure 2F).In addition, the SASA plot and contact maps for IMP1 indicate shorter contact durations with the residues and higher solvent exposure, suggesting lower stability compared to the IMP2 complex (Figure 2G).

| IVM treatment reduces cellular myo-inositol levels through inhibition of IMPase activity
To investigate whether IVM affects IMPase function, the cells were treated with IVM at various concentrations, followed by cellular myoinositol level quantification.The results showed a dose-dependent reduction of cellular myo-inositol in both IVM-treated Vero (Figure 3A) and Huh-7 cells (Figure 3B).LiCl, an IMPase inhibitor, was used as an inhibition control.In Vero cells, a significant reduction  The root-mean-square deviation (RMSD), solvent accessible surface area (SASA) plots, and ligand-protein interactions over a 100 ns molecular dynamic simulation.RMSD plots for the protein backbone (cyan) and ivermectin (maroon) relative to the starting position of the protein and the corresponding SASA plot over the simulation time for ivermectin complexes of (D) IMP2 and (F) IMP1.The interaction maps detail contact frequency and contact interaction types between ivermectin and the residues of (E) IMPA2 and (G) IMPA1.IMP1, IMPase 1; IMPA2, IMPase 2; IMPase, inositol monophosphatase; IVM, ivermectin.
To verify that the reduction of myo-inositol occurs via the binding of IVM to IMPase and that IVM can specifically inhibit IMPase activity, an in vitro enzymatic assay was performed.
Following the incubation, myo-inositol and the hydrolyzed phosphate were measured.The results showed a dose-dependent inhibitory effect of IVM on myo-inositol (Figure 3C) and phosphate levels (Figure 3D) in both the reaction with and without the added IMPase substrate, inositol monophosphate.The relatively lower myo-inositol levels were observed in the reactions with 3 and 10 μM IVM  A similar trend was observed in phosphate levels (Figure 3D).The relatively lower phosphate levels showed in the reactions with 3, 10, and 20 μM IVM compared to the untreated control (p = 0.547, 0.316, and 0.162, respectively).A significant reduction in phosphate levels was observed in the reactions with 125, 250, 500 μM IVM, and 30 mM LiCl compared to the untreated control (p = 0.017, 0.011, 0.007, and 0.013, respectively).The phosphate levels were significantly increased in the reaction with the added IMPase compared to untreated control (p = 0.002).The levels of phosphate were restored in the reactions treated with IVM and added inositol monophosphate.

| Reversion of IVM antiviral activity by inositol
To demonstrate that one of the IVM antiviral mechanisms truly occurs via the inhibition of IMPase, the cells were infected with DENV-2 or ZIKV and then treated with IVM or IVM in the presence of inositol.The concentrations of IVM used in this experiment (0.75, 1.5, and 3 μM) corresponded to approximately 20%-80% inhibition of DENV-2 and ZIKV (Supporting Information: Figure S7 and Tables S4-S6).
For ZIKV infection in Huh-7 cells (Figure 5B), the addition of 2-, 5-, and 10-mM inositol to 0.75 μM IVM resulted in a significant increase in percent infectivity (p = 0.041, 0.006, and 0.006, respectively).A significant increase in percent infectivity was observed in the cells treated with a mixture of 1.5 μM IVM and 2 mM inositol compared to 1.5 μM IVM (p = 0.018).The presence of 2-, 5-, and 10-mM inositol in 3 μM IVM also resulted in a significant increase in percent infectivity compared to 3 μM IVM (p = 0.016, 0.008, and 0.002, respectively).For ZIKV infection in A549 cells (Figure 5C), there is a significant increase in percent infectivity when treated with mixtures of 0.75 μM IVM and 5 or 10 inositol compared to 0.75 μM IVM (p = 0.033 and 0.002, respectively).Only the treatment with 1.5 μM IVM in the presence of 10 mM inositol showed a significant increase in percent infectivity (p = 0.001).A significant increase in percent infectivity was observed in the cells treated with mixtures of 3 μM IVM and 5-or 10-mM inositol compared to 3 μM IVM (p = 0.018 and 0.001, respectively).The addition of inositol alone resulted in increased virus titers for both DENV-2 and ZIKV.
Here we demonstrated that IVM treatment resulted in reduced PI4P expression in Huh-7 cells (Figure 6A).The results showed that LiCl treatment also reduced PI4P expression.The expression of PI4P in SAR-CoV-2 or ZIKV-infected cells was investigated upon IVM treatment.From Figure 6B control cells, similar to that observed in Huh-7 cells (Figure 6A).However, the obvious increase PI4P expression was not observed upon SARS-CoV-2 infection (Figure 6B).The IVM-treated SARS-CoV-2-infected cells showed a reduction in both PI4P and SARS-CoV-2 NP protein expressions.
The expression of PI4P was investigated in ZIKV-infected Huh-7 cells at various time periods.It was shown that PI4P expression increased upon ZIKV infection at 48 h and 72 hpi (Figure 7).And the expression of PI4P gradually decreased after IVM treatment at 48 and 72 hpi.ZIKV infection was almost completely inhibited by 2 μM IVM.

| DISCUSSIONS
IVM also possesses anticancer and antiviral activity. 24,25IVM was shown to inhibit several RNA viruses; however, the mechanisms of its antiviral activity have been proposed for only some specific viruses.
In addition to importin, previously identified IVM-interacting proteins include Cytochrome P450 3A4, which is the major enzyme responsible for the metabolism of IVM by human liver microsomes. 26Most of the IVM-interacting proteins discovered previously belong to ATPbinding cassette multidrug transporters, including P-glycoprotein (Pgp), 27 multidrug resistance proteins (MRPs; MRP1, 2, and 3), 28 and human breast cancer resistance protein. 29IVM also interacted with and blocked liver-specific organic anion transporters 1 and 2 (OATP1B1 and OATP1B3), which are expressed on the sinusoidal membrane of hepatocytes. 30Additionally, IVM interacts with Farnesoid X receptor resulted in decrease serum glucose and cholesterol levels in mice. 31Supporting Information: Table S3 shows previously identified IVM-interacting proteins.
The enrichment of TPP-identified proteins revealed various biological processes related to metabolism and biosynthesis.Viruses rely on the metabolic processes of the host cells for viral replication.
Several viruses increased glycolysis and fatty acid biosynthesis. 32us, the binding of IVM to these proteins may contribute to the inhibition of viral replication.
IMPase has a crucial role in converting inositol monophosphate to myo-inositol, which is a molecule that serves as the basic structure for various forms of inositol phosphates, PtdIns and PPIn.These molecules play important roles in a wide array of cell functions, including cell growth, apoptosis, secretion, and information The expression of PI4P in response to IVM treatment or viral infection.Huh-7 cells (A) were treated with 2 μM IVM, 30 mM LiCl, or 0.5% DMSO (CC) for 48 h.Vero cells (B) were mock-infected or infected with SARS-CoV-2 at MOI 0.02 for 2 h.After removing the virus inoculum, the cells were further maintained in the medium containing 2 μM IVM or 20 mM LiCl for 48 h.The cells were collected, fixed, and investigated using confocal microscopy using specific antibodies against PI4P (green) and SARS-CoV-2 NP protein (red).The nuclei were represented in blue.The merged images were shown in the last panel.CC, untreated control cells; IVM, ivermectin.
processing, and they are essential for replication of many viruses, especially in the formation of cytoplasmic membrane replication organelles (ROs) of many RNA viruses, which requires PPIn as integral phospholipid components of the ROs. 33There are two isoforms of the human IMPase: IMP1 and IMP2.Different tissues, including the brain, small intestine, pancreas, liver, heart, kidney, spleen, and testis, express both isoforms in varied amounts. 34Despite their only 53.65% sequence identity, the crystal structures of IMP1 and IMP2 are superimposed.From molecular docking, both IMP1 and IMP2 bind their substrates, inositol monophosphate, with similar affinity.
The binding of IVM with IMP1 showed similar binding affinity as that of inositol monophosphate, but greater binding affinity was found between IVM and IMP2.The molecular dynamics simulations also showed greater binding efficiency between IVM and IMP2.This might be the reason why only IMP2 was detected in the TPPidentified proteins.However, IVM might inhibit both IMP1 and IMP2.
Although high concentrations of inositol were added to the cells, only a moderate increase of cellular myo-inositol levels was observed (Figure 4B).This suggested that there was a limitation in cellular inositol uptake.This may explain the incomplete reversion of IVMsuppressed viral replication by inositol, together with the fact that there may be other IVM mechanisms involved.Myo-inositol is uptake into the cells via sodium-ion-coupled and proton-coupled inositol transporters. 35Two types of sodium/myo-inositol transporters (SMIT) were found in mammalian cells: SMIT1 and SMIT2, which are mainly regulated by osmotic pressure. 36ruses exploit various host cellular pathways to favor their replication cycle. 37Especially, RNA viruses were found to remodel the membrane of organelle for establishing the site of genome replication.
For instance, flaviviruses, including DENV, WNV, and ZIKV, induce invaginations of the negative curvature of the ER membrane as convoluted membranes and vesicle packets. 38[41] The PPIn enriched in the viral ROs of these viruses was PI4P. 144P is required to alter membrane curvature and regulate viral RNA synthesis.It also acts as the binding site of some viral proteins. 13In HCV infection, viral NS5A protein recruited the PtdIns-4-kinases (PI4Ks) to generate PI4P-enriched cellular membranes of viral ROs.
The inhibition of PI4K activity using specific inhibitors such as wortmannin resulted in a decrease in virus titers. 33The ZIKV NS1 protein binds to PI4P of the ER membrane via its R31 residue.The inhibition of PI4K also reduced ROs formation and ZIKV virus production. 42Additionally, it was found that PI4K IIIβ is required for SARS-CoV entry. 43PI4Ks can be used as a target for viral inhibition, 44 however, different viruses or even the same virus in different study models revealed that they are dependent on different types of PI4Ks. 33,45Since IMPase is a key enzyme required for both de novo myo-inositol biosynthesis and the breakdown of PPIn to myoinositol; inhibiting it might be a better antiviral approach against various viruses.

F I G U R E 7
The expression of PI4P upon ZIKV infection and IVM treatment.Huh-7 cells were mock-infected or infected with ZIKV at MOI 1 for 2 h.After removing the virus inoculum, the cells were further maintained in the medium containing 2 μM IVM for 24, 48, and 72 h.The infected cells were collected, fixed, and investigated using confocal microscopy.The infected cells were probed with specific antibodies against PI4P (green) and ZIKV NS1 protein (red).The nuclei were represented in blue (first panel).The merged images were shown in the last panel.CC, untreated control cells; IVM, ivermectin; ZIKV, Zika virus.

F I G U R E 1
Figure S5.

F
I G U R E 2 Molecular docking of IMPs and their substrates.(A) Superimposed crystal structures of IMP1 (PDB ID: 4AS4) and IMP2 (PDB ID: 2DDK).The structures of IMP1 and IMP2 are shown in cyan and pink, respectively.(B) Docking of inositol monophosphate and IVM to IMP1.(C) Docking inositol monophosphate and IVM to IMP2.Discovery Studio Visualizer was used to visualize the model.The 3-dimensional structures of IMP1 and IMP2 were shown in solid ribbon.Hydrogen bond donors are shown in pink, and hydrogen bond acceptors are shown in green.
compared to the untreated control (p = 0.19 and 0.108, respectively).The significant reduction of myo-inositol levels was shown in the treatments with 20, 125, 250, 500 μM IVM, and 30 mM LiCl compared to the untreated control (p = 0.02, 0.043, 0.014, 0.019, and 0.007, respectively).The myo-inositol level significantly increased with the addition of 250 μM of inositol monophosphate compared to the untreated control (p = 0.001).Similar to what was observed in Huh-7 cells (Figure3B), myo-inositol levels were restored by the addition of inositol monophosphate to the reactions.

F
I G U R E 3 Myo-inositol levels upon IVM treatment.The cellular myo-inositol levels in Vero (A) and Huh-7 cells (B) were determined.The cells were treated with various concentrations of IVM, inositol, or mixtures of IVM and inositol for 48 h.LiCl, an IMPase inhibitor, was used as an IMPase inhibitory control.After that, the cells were collected and subjected to cellular myo-inositol quantitation using a fluorometric assay.The data on myo-inositol levels relative to the untreated controls are shown as the means of triplicate experiments.The in vitro enzymatic assay determining IMPase activity was performed.The cell lysates of Huh-7 cells were incubated with various concentrations of IVM in the presence or absence of the IMPase substrate, inositol monophosphate.The levels of myo-inositol (C) and phosphate (D) were measured and expressed as pmoles.The data are shown as the means of triplicate experiments.Error bars represent the standard deviation (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001).IMPase, inositol monophosphatase; IP, inositol monophosphate; IVM, ivermectin.
LiCl inhibited DENV-2 virus production.The half-maximal inhibitory concentration (IC 50 ) values in Vero and Huh-7 cells were 21.81 and 50.72 mM, respectively (Figure 4A,B).A reversion of LiCl antiviral activity using inositol was performed to find out whether the suppression of IMPase activity contributed to the antiviral mechanism.The results (Figure 4C) showed that inositol partially reversed the LiCl antiviral activity.A significant increase in percent infectivity was observed in the cells treated with mixtures of 15 mM LiCl and 2-or 5-mM inositol compared to 15 mM LiCl alone (p = 0.016, 0.001, respectively).Similar results were seen in the cells treated with the mixtures of 20 mM LiCl and 2-or 5-mM inositol (p = 0.025, 0.001, respectively).The treatments with mixtures of 1-, 2-, or 5-mM inositol and 25 or 30 mM LiCl significantly increased the percent infectivity when compared to LiCl alone (p < 0.001, p = 0.001, and 0.016, respectively, compared to 25 mM LiCl; p = 0.001 compared to 30 mM LiCl for both added 1-, 2-, and 5-mM inositol).The cell viability of reversion of LiCl antiviral activity by inositol in Vero cells showed in Supporting Information: Figure S6.
LiCl treatment and reversion of LiCl antiviral activity by inositol.Vero (A) or Huh-7 cells (B) were infected with DENV-2 at MOI 0. 5 or 1, respectively.After that, the cells were treated with various concentrations of LiCl.For the reversion of LiCl antiviral activity by inositol (C), Vero cells were infected with DENV-2 at MOI 0.5 and treated with various concentrations of LiCl, inositol, or the mixtures of LiCl and inositol.The infected cells were further maintained in the media containing indicated chemicals for 48 h.Then virus titers were determined using the focus immunoassay titration.The data are shown as mean of triplicate experiments.Error bars show the standard deviation (*p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001).DENV, dengue virus.
, a reduction in PI4P expression was observed in IVM or LiCl-treated Vero cells compared to untreated F I G U R E 5 Reversion of IVM antiviral activity by inositol.Vero cells were infected with DENV-2 at MOI 0.5 (A) or ZIKV (B), and A549 cells were infected with ZIKV (C) at MOI 1 for 2 h.After removing the virus inoculum, the cells were treated with media containing various concentrations of IVM, inositol, mixtures of IVM and inositol, or 0.5% DMSO for 48 h.The culture media were collected, and the virus titers were determined using focus immunoassay titration.The data are shown as mean of triplicate experiments.Error bars represent the standard deviation (*p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001).DENV, dengue virus; IVM, ivermectin; ZIKV, Zika virus.