Antidepressant Sertraline Is a Broad-Spectrum Inhibitor of Enteroviruses Targeting Viral Entry through Neutralization of Endolysosomal Acidification

Abstract

Enterovirus 71 (EV71) is an etiological agent of hand foot and mouth disease and can also cause neurological complications in young children. However, there are no approved drugs as of yet to treat EV71 infections. In this study, we conducted antiviral drug screening by using a Food and Drug Administration (FDA)-approved drug library. We identified five drugs that showed dose-dependent inhibition of viral replication. Sertraline was further characterized because it exhibited the most potent antiviral activity with the highest selectivity index among the five hits. The antiviral activity of sertraline was noted for other EV serotypes. The drug's antiviral effect is not likely associated with its approved indications as an antidepressant and its mode-of-action as a selective serotonin reuptake inhibitor. The time-of-addition assay revealed that sertraline inhibited an EV71 infection at the entry stage. We also showed that sertraline partitioned into acidic compartments, such as endolysosomes, to neutralize the low pH levels. In agreement with the findings, the antiviral effect of sertraline could be greatly relieved by exposing virus-infected cells to extracellular low-pH culture media. Ultimately, we have identified a use for an FDA-approved antidepressant in broad-spectrum EV inhibition by blocking viral entry through the alkalization of the endolysosomal route.

Keywords: antidepressant sertraline; broad-spectrum antiviral; drug repurposing; enterovirus; host-cell targets; viral entry.

Figure 1

Fluorescence resonance energy transfer (FRET)-based biosensor for the drug screening of EV71 inhibitors. (A) Diagram illustrating the principle of FRET-based drug screening by using HeLa-G2AwtR cells where the green fluorescence protein (GFP²)–red fluorescence protein (DsRed2) tandem fluorophores with the cleavage site (CS) of virus 2A protease in between is stably expressed. FRET occurs due to the close proximity of the fluorophore pair. FRET is reduced upon the viral protease cleavage in the context of EV71 replication, whereas FRET remains intact when viral replication is significantly blocked. (B) Flowchart of the FRET-based biosensor for drug screening with the BML-2843-0100 drug library. (C) Scatter plot of the inhibition rate resulted from the action of test drugs. A total of 24 drugs from the screen yielded an inhibition rate of >50%, as indicated. The average FRET ratios from EV71-infected HeLa-G2AwtR cells and mock-infected cells were arbitrarily set as 0% and 100% inhibition rates, respectively.

Figure 2

Dose–response relationships of novel anti-EV71 hits identified from the screen. (A) RD and (B) HeLa cells were pretreated for 1 h at increasing concentrations (0.1, 0.3, 1, 3, and 10 μM) of the indicated drugs and then infected with EV71 stocks at a multiplicity of infection (MOI) of 0.1, with test drugs maintained throughout the infection. At 12 post-infection (p.i.), the cells were analyzed using an immunofluorescence assay (IFA). For each condition, the percentage of infection was calculated as the ratio of the number of infected cells stained for viral VP1 to the number of cells stained with DAPI. Drug cytotoxicity was determined by treating HeLa or RD cells with increasing concentrations of the indicated drugs (1, 3, 10, 30, and 100 μM) for 12 h. Cell viability was measured using a cell proliferation assay and expressed as the percentage of drug-free cells. IC₅₀ and CC₅₀ values were calculated using GraphPad Prism5 software. The solid circle and empty circle represented the inhibition rate (%) and cell viability (%), respectively. The Selective Index (SI) value was calculated by value of CC₅₀ divided by that of IC₅₀. Data represented the means of triplicated experiments and the SEM. Drug-free wells contained 0.05% DMSO.

Figure 3

Antiviral activity of sertraline against five other EV serotypes. (A) RD and (B) HeLa cells were pretreated for 1 h at increasing concentrations (0.1, 0.3, 1, 3, and 10 μM) of sertraline and subsequently infected with CVA16, CVB1, CVB2, Echo9, and Echo30 stocks at an MOI of 0.1 for 12 h, with the corresponding concentration of sertraline maintained throughout the infection. The IFA was conducted using the aforementioned protocol. Primary antibodies (Abs) were the mouse anti-EV71 Ab for EV71 and CVA16, mouse anti-coxsackie virus B blend Ab for CVB1 and CVB2, and mouse anti-echovirus blend Ab for Echo9 and Echo30. The secondary Ab was the FITC-conjugated goat anti-mouse Ab. The inhibition rate (%) was measured as described in Figure 2. Triplicated experiments were conducted with the mean and the SEM shown. Drug-free wells contained 0.05% DMSO.

Figure 4

Timing of EV71 inhibition by sertraline. (A) Time-of-addition assay with the schematic representation. Each compound was added to RD cells at the indicated times relative to viral infection at an MOI of 0.1 at 37 °C, with the initiation (−1 h) and completion (0 h) of viral adsorption indicated. Solid and dashed lines indicate the periods with and without compound treatment, respectively. Whole: −2 to 12 h; entry: −2 to 2 h; post-entry: 2–12 h. Sertraline (SER), 10 μM; chlorpromazine (CPZ), 20 μM; ribavirin (RIB), 100 μM. (B) The binding and internalization assay with the schematics. EV71 at an MOI of 5 was allowed to bind to the cells on ice for 1 h in the presence of 5 or 10 μM of sertraline or without drug treatment. For the binding assay, unbound virus was extensively washed off. For the internalization assay, the cells were warmed to 37 °C for 1 h after incubation on ice, allowing viruses to enter cells. Cell surface–bound viruses were removed using trypsin. Quantitative reverse transcription polymerase chain reaction (RT-qPCR) was conducted using total RNA prepared from each condition in (A,B), and shown as relative levels to drug-free control containing 0.05% DMSO. These data were from three independent experiments. ** p < 0.01; *** p < 0.001.

Figure 5

The alkalizing effect of selective serotonin reuptake inhibitors tested was associated with the antiviral potency. (A) RD cells were treated with chloropromazine (CPZ), fluvoxamine (FLX), paroxetine (PAR), sertraline (SER), and chloroquine (CQ) at 5 and 10 μM for 2 h. The cells were labeled with 150 nM Lysotracker (LTR) for 1 h. Flow cytometry was performed using the indicated mean fluorescence intensity value. Histogram overlays of LTR-free control (in the top panel) and LTR-labeled cells in the absence or presence of indicated drugs are shown. (B) RD cells were pretreated with the indicated drugs at 5 or 10 μM for 1 h. Culture media were replaced with EV71 stock at an MOI of 1 and the corresponding drugs of the same concentrations for 1 h. Culture media were removed and replaced with media containing the corresponding drugs of the same concentrations for 6 h. Cell lysates were prepared and subjected to Western blot analysis using the mouse anti-EV71 VP1 antibody, using the antitubulin antibody as the internal control. The percentage shown below each lane represents the intensity of viral VP1 relative to that of tubulin.

Figure 6

Exposure of the cells to low pH alleviated the antiviral activity of sertraline. The low-pH exposure assay with the schematic representation. (A,C) RD and (B,D) HeLa cells were pretreated with 10 μM sertraline for 1 h. EV71 was used to inoculate the pretreated cells at an MOI of 1 in the presence of 10 μM sertraline on ice for 1 h. The supernatant containing unbound virus was removed, and the cells were incubated with medium containing 10 μM sertraline at 37 °C for 1 h. The cells were incubated with media with different pH values (7.4, 6.5, 5.5, and 5.0) that contained 10 μM sertraline for 10 min and then replaced with 10 μM sertraline-containing medium (pH 7.4). (A,B) At 6 h after the addition of the media with the different pH values, levels of viral VP1 protein and tubulin (the internal control) were measured by Western blot. The percentage shown below each lane represents the intensity of viral VP1 relative to that of tubulin. (C,D) At 12 h after the addition of the media with the different pH values, viral titers from the supernatants were titrated by plaque assay. Triplicated experiments were conducted (* p < 0.05; ** p < 0.01; *** p < 0.001).