Novel quinoline substituted autophagy inhibitors attenuate Zika virus replication in ocular cells

Highlights

• •Quinoline derivatives were synthesized to inhibit Zika virus (ZIKV) replication.
• •Lead compounds, GL-287 and GL-382 exhibited potent antiviral activity without causing cytotoxicity in ocular cells.
• •Novel quinoline derivatives acted via inhibition of autophagy and modulating innate inflammatory response.

Abstract

Zika virus (ZIKV) is a re-emerging RNA virus that is known to cause ocular and neurological abnormalities in infants. ZIKV exploits autophagic processes in infected cells to enhance its replication and spread. Thus, autophagy inhibitors have emerged as a potent therapeutic target to combat RNA viruses, with Hydroxychloroquine (HCQ) being one of the most promising candidates. In this study, we synthesized several novel small-molecule quinoline derivatives, assessed their antiviral activity, and determined the underlying molecular mechanisms. Among the nine synthesized analogs, two lead candidates, labeled GL-287 and GL-382, significantly attenuated ZIKV replication in human ocular cells, primarily by inhibiting autophagy. These two compounds surpassed the antiviral efficacy of HCQ and other existing autophagy inhibitors, such as ROC-325, DC661, and GNS561. Moreover, unlike HCQ, these novel analogs did not exhibit cytotoxicity in the ocular cells. Treatment with compounds GL-287 and GL-382 in ZIKV-infected cells increased the abundance of LC3 puncta, indicating the disruption of the autophagic process. Furthermore, compounds GL-287 and GL-382 effectively inhibited the ZIKV-induced innate inflammatory response in ocular cells. Collectively, our study demonstrates the safe and potent antiviral activity of novel autophagy inhibitors against ZIKV.

1. Introduction

Zika virus (ZIKV), an enveloped, positive-sense RNA virus, belongs to the family Flaviviridae, which gained global attention in 2015 due to an epidemic in South America that affected millions of people (Lanciotti et al., 2008; Leier et al., 2018; Samarasekera and Triunfol, 2016; Scaramozzino et al., 2001; Wikan and Smith, 2016). The primary mode of ZIKV transmission is through mosquito bites, but it can also spread through sexual contact, blood transfusion, breastfeeding, organ transplantation, and from pregnant mothers to their fetuses (Abrams et al., 2017; Basu and Tumban, 2016; Plourde and Bloch, 2016). While many ZIKV infections are asymptomatic or cause mild clinical symptoms like fever, arthralgia, maculopapular rash, headache, and conjunctivitis, infection during pregnancy can lead to severe complications such as congenital microcephaly, brain defects, Guillain-Barré syndrome, stillbirths, and miscarriages (Cao-Lormeau et al., 2016; Malkki, 2016; Mlakar et al., 2016; Oehler et al., 2014; Rasmussen et al., 2016). Additionally, ZIKV infection has been associated with severe ocular complications, including chorioretinal atrophy, hypoplasia, focal pigmented mottling, RPE mottling, retinal focal spots, severe retinal vessel attenuation, optic nerve atrophy, optic disc anomalies, and congenital glaucoma (Adebayo et al., 2017; De Moraes et al., 2018; de Paula Freitas et al., 2016; Singh et al., 2017; 2019; Singh and Kumar, 2018). Despite efforts to control its spread, no specific antiviral treatments or vaccines are currently available against ZIKV.

The replication of ZIKV involves intimate dependence on the host to successfully infect and establish viral replication, with several key pathways serving as pro- or antiviral platforms. One such important pathway is the lysosomal-dependent degradation pathway of autophagy (Chiramel and Best, 2018). The deviation of the autophagy machinery by viruses, including ZIKV, allows viral replication and spread. More specifically, ZIKV hijacks the autophagy machinery as an early step to increase the formation of viral replication complexes (Stoyanova et al., 2023). Previous studies have shown that ZIKV induces autophagy, and inhibition of autophagy attenuates viral replication (Liu et al., 2023; Stoyanova et al., 2023). Therefore, several autophagy inhibitors have been explored as antiviral targets to suppress ZIKV replication in different experimental models (Cao et al., 2017; Chiramel and Best, 2018; Liang et al., 2016; Liu et al., 2023).

More recently, repurposing pre-existing drugs has emerged as a promising strategy to find potential treatments for various infectious diseases (Das et al., 2022; Trivedi et al., 2020). One such drug of interest is hydroxychloroquine (HCQ), an analog of chloroquine (CQ), which has been a well-established anti-malarial drug for many years and is also used to treat autoimmune and inflammatory diseases such as Systemic Lupus Erythematosus (SLE) and Rheumatoid Arthritis (RA) (Martinez et al., 2020). During the COVID-19 pandemic, HCQ has garnered attention due to its potential antiviral properties, especially against RNA viruses such as SARS-CoV-2 (Boulware et al., 2020; Hennekens et al., 2022), and its activity has also been reported against ZIKV (Kumar et al., 2018).

HCQ is generally considered safe when used within recommended doses for approved indications such as malaria, RA, and SLE. However, there are concerns about potential side effects such as cardiotoxicity and cytotoxicity, mainly when used in high doses or for prolonged periods. These concerns include an increased risk of ventricular arrhythmias and decreased patient survival, ocular toxicity (retinopathy), and possible harm to the liver and kidneys (Achan et al., 2011; Ben-Zvi et al., 2012; Giudicessi et al., 2020; Nika et al., 2014). This led to several efforts to modify HCQ to reduce its toxic effects. For example, ROC-325, a new dimeric compound containing the modified HCQ and lucanthone, is a potent lysosomal-mediated autophagy inhibitor that induces lysosomal deacidification, autophagosome accumulation, and disruption of autophagic flux (Carew and Nawrocki, 2017). DC661 is another novel dimeric chloroquine, capable of deacidifying the lysosome and inhibiting autophagy significantly better than HCQ (Rebecca et al., 2019). Recently, GNS561 (Ezurpimtrostat) has emerged as a potent and novel small molecule inhibitor of late-stage autophagy that exhibits effective antiviral activity against SARS-CoV-2 through autophagy inhibition (Bestion et al., 2022). GNS561 is currently in Phase 2 clinical trials for the treatment of SARS-CoV-2 infection (NCT04637828).

In continuation of our efforts to develop more potent, effective, and safe autophagy inhibitors, in the current study, we synthesized several novel piperazine and piperidine substituted quinoline derivatives based on HCQ's quinoline scaffold and assessed their anti-ZIKV activity and toxicity profile in ocular cells.

2. Materials and methods

2.1. Cells, viruses, and drugs

Human retinal pigmented epithelial cell line, ARPE-19 were cultured using Dulbecco's Modified Eagle Medium F12 (DMEM/F12) media (Thermo Scientific, Rockford, IL), while primary human trabecular meshwork cells (Pr. HTMC) and Vero cells were cultured using Dulbecco's minimal essential medium (DMEM, Thermo Scientific, Rockford, IL), supplemented with 10 % Fetal bovine serum (FBS) and 1X penicillin-streptomycin (P/S) solution (Thermo Scientific, Rockford, IL). All cells were maintained at 37 °C with 5 % CO₂ and 95 % humidity in an incubator.

The ZIKV strain PRVABC59 (NR-50240), which was initially isolated from human blood in Puerto Rico in December 2015, was obtained from BEI Resources, National Institute of Allergy, and Infectious Diseases (NIAID), NIH. ZIKV was propagated in Vero cells, and titers were determined by standard plaque assays.

The drugs used in this study were hydroxychloroquine (HCQ) (Cayman Chemicals, #17911), Bafilomycin-A1 (Cayman Chemicals, #11038), ROC-325 (Ambeed, #A461073), GNS-561 (MedChemExpress, #HY-137978A), and DC-661 (Ambeed, #A1001762) as reference autophagy inhibitors.

2.2. Chemical synthesis and analysis

All the commercially available chemicals used for synthesis were purchased from Sigma Aldrich, Alfa Aesar, Acros, and Combi-Blocks Chemical Co. (USA) and used without further purification. Anhydrous solvents were obtained from Across Organics or Sigma Aldrich and used directly. Unless otherwise specified, reactions were performed under an inert atmosphere of argon and monitored by thin-layer chromatography (TLC). ¹H NMR spectra were recorded at 400 MHz using Bruker AV NMR spectrometer. ¹³C NMR spectra were recorded at 101 MHz using Bruker AV NMR spectrometer. Chemical shifts are expressed in parts per million (ppm, δ), relative to tetramethylsilane (TMS) as an internal reference. Signals are described as s (singlet), d (doublet), dd (doublet of doublets), dt (doubles of triplets), t (triplet), q (quartet), or p (pentet). Thin layer chromatography was performed using Merck silica gel 60 F-254 thin layer plates, which were developed using one of the following techniques: UV fluorescence (254 nm), alkaline potassium permanganate solution (0.5 % w/v) or ninhydrin (0.2 % w/v) and iodine vapors. All intermediates and final compounds were purified by flash column chromatography, and all final compounds submitted for biochemical and biological testing were confirmed to be ≥95 % pure by analytical HPLC. The detailed synthetic schemes, procedures, characterization of compounds, and HPLC analysis of the 9 final compounds are provided in supplementary data.

2.3. Virus infection and Plaque assay

The cells were plated to reach 70–80 % confluency and maintained at 37 °C within an incubator with 5 % CO₂ and relative humidity. After rinsing with 1X PBS, the cells were infected with ZIKV at a multiplicity of infection (MOI) of 1 or mock-infected using a serum-free medium. The virus was allowed to adsorb for 2 h, with intermittent agitation every 15 min. Subsequently, the media was replaced with the respective cell culture media supplemented with 2–5 % FBS and 1X P/S solution for the desired incubation time.

ZIKV plaque assay was performed per the protocol adapted from Vincent Racaniello's lab. Briefly, Vero cells were seeded to form a confluent monolayer in 12- or 24-well tissue culture plates. Once confluent, the cells were washed with 1X PBS. Serial dilutions of conditioned media containing viruses were prepared in serum-free DMEM medium and introduced to the cells for a 2-hour exposure. Following adsorption, the viral inoculum was removed, and the cell monolayer was overlayed with the first overlay media containing a 1:1 mixture of 2X EMEM, 4 % FBS, 2X P/S, 20 mM MgCl_2, and 1.6 % Noble Agar. Next day, a second overlay media containing DMEM, 1 mg/ml BSA, 40 mM MgCl₂, 0.2 % glucose, 2 mM sodium pyruvate, 4 mM L-glutamine, 4 mM oxaloacetic acid, 1X P/S, and 0.1 % sodium bicarbonate was added. The plates were incubated in a CO₂ incubator at 37 °C with humidity for five days.

To visualize the plaques, cells were fixed with 10 % Tricarboxylic Acid (TCA) for 20 min. After removing the overlay media, the plaques were stained with 0.2 % crystal violet for 20 min. Subsequently, the plaques were rinsed with distilled water, air-dried, and counted. The viral titer was quantified as log₁₀ PFU/mL. The plaque assay was performed in triplicates, and the results were presented alongside statistical analysis.

2.4. Immunofluorescence assay

For immunofluorescence assays, cells were seeded in four-well chamber slides (Thermo Scientific, Rockford, IL) and infected with ZIKV at MOI of 1. Following the desired incubation, cells were fixed with 4 % paraformaldehyde (PFA) in 1X PBS overnight at 4 °C, followed by three washes with 1X PBS. To enable permeabilization and block nonspecific binding, a blocking buffer containing 1 % BSA and 0.4 % Triton X-100 in 1X PBS was applied to the cells for 1 h at room temperature in a humidified chamber.

Subsequently, the cells were incubated overnight at 4 °C with primary (mouse anti-Flavivirus 4G2 and rabbit anti-LC3, 1:100 dilution) antibodies in a dilution buffer containing 1 % BSA and 0.4 % Triton X-100. After incubation with primary antibodies, the cells were washed three times with 1X PBS. The cells were then incubated with anti-mouse/rabbit Alexa Fluor 488/594 secondary antibodies (1:200) (Thermo Scientific, Rockford, IL) for 1 h at 37 °C within a humidified chamber. Following three washes, cells were mounted using Vectashield anti-fade mounting medium with DAPI (Vector Laboratories, Burlingame, CA). Cells were visualized and imaged using a Keyence fluorescence microscope (Keyence, Itasca, IL).

2.5. Western blotting

For immunoblotting, cells were lysed using RIPA lysis buffer with a protease inhibitor cocktail (Thermo Scientific, Rockford, IL). The total protein concentration was determined using a BCA assay kit (Thermo Scientific, Rockford, IL) as per the manufacturer's instructions. Denatured whole-cell proteins were resolved on a 10 %−16 % sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto a nitrocellulose or PVDF membranes with a pore size of 0.45 μm or 0.2 μm (Bio-Rad, Hercules, CA) depending on the molecular weight of the target proteins.

Following protein transfer, the membrane was subjected to blocking using 5 % non-fat milk and then washed with 1X TBST (Tris-glycine buffer containing 0.5 % Tween 20). The blots were incubated with respective primary antibodies (1:1000), diluted in 5 % BSA, overnight at 4 °C with gentle agitation. After incubation with the primary antibodies, the membranes were washed three times using 1X TBST and incubated with anti-mouse/rabbit HRP-conjugated secondary antibodies (1:2000) at room temperature for two hours. Subsequently, after three washes with 1X TBST, blots were developed using Supersignal West Femto chemiluminescent substrate and imaged using iBright FL1500 imager (Thermo Scientific, Rockford, IL).

2.6. Cell viability assay

Cells were seeded in a 96-well culture plate and treated with various drugs at different concentrations, followed by an incubation for 48 h. Following incubation, cell viability was determined using an MTT (3-(4,5-dimethylthiazol-2-yl)−2,5-diphenyl tetrazolium bromide) assay. Briefly, the MTT reagent at a concentration of 0.5 mg/mL, suspended in the cell culture medium, was added to the cells, and incubated for 4 h. Subsequently, the cells were lysed using a 20 % Sodium dodecyl sulfate (SDS) solution in 50 % Dimethylformamide (DMF). The plates were read at a wavelength of 570 nm using a microplate reader, and the resulting values were plotted relative to an untreated control to express cell viability as a percentage.

2.7. RNA isolation and qPCR

Cellular RNA was extracted using Trizol reagent as per the manufacturer's instructions (Thermo Scientific, Rockford, IL). The cDNA was prepared using 1 μg of total RNA, employing the Maxima First Strand cDNA Synthesis kit (Thermo Scientific, Rockford, IL), adhering to the manufacturer's guidelines. Subsequently, quantitative real-time PCR (qRT-PCR) was performed with gene-specific primers utilizing the StepOnePlus Real-time PCR (Thermo Scientific, Rockford, IL). The quantification of gene expression was ascertained through the comparative 2^ΔΔCT method and presented as relative fold change expression.

2.8. Statistical analysis

The data used in this study have been expressed as mean ± standard deviation (SD). Statistical differences between the experimental groups were analyzed using Graph Pad Prism V10 software (GraphPad Software, La Jolla, CA). A P value of <0.05 was considered statistically significant. All experiments were performed at least three times unless mentioned otherwise.

3. Results

3.1. Chemical synthesis of novel autophagy inhibitors

The synthetic approach developed for the preparation of piperazine and piperidine-substituted quinoline derivatives is illustrated in Fig. 1. The synthetic route for the compounds GL-287, GL-289, GL-290, and GL-292 involves, the initial conversion of Boc-protected alcohol (1) to the corresponding mesylate (2) in basic condition. The conversion of alcohol to a mesylate prevents the alcohol from acting as an acid or nucleophile, or from undergoing other undesirable reactions. Subsequently, we treated the compound (2) with N-methyl piperazine (3) and deprotected Boc group in the presence of 4 N HCl in moderate yield. Lastly, N-alkylation at the 4th position of 4,7-dichloroquinoline (5) by replacing the chloro group with amine (4) resulted in the target compounds (Fig. 1A).

Fig. 1.

General synthesis approach of novel piperazine, and piperidine substituted quionoline derivatives. The steps involved in the synthesis of indicated quinoline derivatives, GL-287, GL-289, GL-290, GL-292, GL-296, GL-374, GL-377, GL-380, and GL-382 with their chemical properties. LogP, CLogP, and Topological polar surface area (tPSA) calculated using ChemBioDraw Ultra 14.0 (CambridgeSoft). The lead candidates in the study have been highlighted with a rectangular box.

The synthetic route for the compounds GL-296, GL-374, GL-377, GL-380, and GL-382 is depicted in Fig. 1B. First, the Boc-protected amino acids (6 or 7) were employed for acid amine coupling with substituted amine (8) using EDCI/HOBt amide synthesis strategy followed by Boc deprotection in the presence of 4 N HCl in dioxane. The target compounds were obtained by N-alkylation of amine (9 or 10) with either 4,7-dichloroquinoline (5) or 4‑chloro-7-fluoroquinoline (11) in the presence of phenol solvent at 150 °C. The elaborated chemical structures and the chemical properties of the target compounds are shown in Fig. 1C.

3.2. HCQ attenuates ZIKV replication but exerts cellular cytotoxicity

HCQ exhibits antiviral properties against several RNA viruses through diverse mechanisms, including hindrance of viral entry and autophagy inhibition (Kumar et al., 2018; Marti-Carvajal et al., 2017; Meo et al., 2020; Romanelli et al., 2004; Wang et al., 2015). However, recent scrutiny of HCQ usage stems from its cardiac toxicity in COVID-19 patients and known ocular adverse effects (Doyno et al., 2021; Izcovich et al., 2022; Nika et al., 2014; Paniri et al., 2020). Here, we utilized cell culture models to assess the toxicity and antiviral activity of HCQ on Vero cells and human ocular cells (ARPE-19 and Pr. HTMC), which are highly permissive to ZIKV infection (Singh et al., 2017; 2019; 2018). First, we evaluated the cytopathic effects of HCQ by exposing cells to various concentrations of HCQ. Our data showed dose-dependent cytotoxicity of HCQ wherein ≤10 µM of HCQ was found to cause minimal cell death in Vero (Fig. 2A), ARPE-19 (Fig. 2D), and Pr. HTMC cells (Fig. 2G). Thus, 10 µM HCQ was used to test its antiviral activity against ZIKV. The plaque assay was performed from the culture supernatant of ZIKV-infected Vero (Fig. 2B), ARPE-19 (Fig. 2E), and Pr. HTMC (Fig. 2H) cells showed a significant reduction in viral replication in HCQ-treated cells. Notably, the HCQ treatment was found to be most effective in reducing viral progeny in ARPE-19 and Pr. HTMC cells. The antiviral activity of HCQ was further confirmed by immunofluorescence detection of the ZIKV envelope (E) protein in these cells (Fig. 2C, F, & I). The antiviral activity of HCQ correlated with a profound accumulation of LC3 puncta in ZIKV-infected cells, indicating the role of HCQ in the inhibition of autophagy. Together, these results indicate antiviral properties of HCQ potentially via inhibition of autophagy.

Fig. 2.

Cellular toxicity and antiviral activity of Hydroxychloroquine (HCQ) against ZIKV. Vero (A-C), ARPE-19 (D-F), and Pr. HTMC (G-I) cells were grown to confluency and treated with HCQ at varying concentrations and cell viability was assessed by MTT at 24 h (A, D, G). In another experiment, cells were infected with ZIKV at MOI of 1 and treated with HCQ followed by plaque assay (B, E, H) and immunofluorescence assay (C, F, I)) to visualize ZIKV 4G2 antigen (red), LC3 puncta (green), and counterstained cell nuclei using DAPI (blue). The images were captured at 200x and 600x magnification. The data represented are the culmination of three independent experiments and are shown as means ± SD. Statistical analysis was performed using an unpaired t-test. ns; non-significant, **p < 0.005, ****p < 0.0001.

3.3. Current HCQ analogs inhibit ZIKV replication but exhibit cytotoxicity

HCQ is known for its ability to impede autophagy by disrupting lysosomal function (Schrezenmeier and Dorner, 2020). Building upon this aspect of its mechanism, researchers have explored novel compounds (e.g., ROC-325, GNS561, and DC661) incorporating modified core elements from HCQ to discover new inhibitors of autophagy that offer improved effectiveness, tolerability, and potency (Brun et al., 2022; Jones et al., 2019; Xu et al., 2022). We evaluated the effect of these three established autophagy inhibitors and observed significant cytotoxicity. Notably, GNS561 and DC661 exhibited the highest cytotoxicity across all cell types, while ROC-325 demonstrated considerable cytotoxicity in ARPE-19 cells (Fig. 3C) with minimal cytotoxicity in Vero (Fig. 3A) and Pr. HTMC cells (Fig. 3E) at 5μM concentration. Interestingly, treatment with all these compounds caused a noticeable abundance of LC3B puncta compared to mock-treated and ZIKV-infected cells in immunofluorescence studies, reaffirming decreased degradation of LC3B and inhibition of autophagic process (Fig. 3B, D, & F). The immunofluorescence data also revealed significant cell death following treatment with GNS561 and DC661. These results underscore that the existing chemically modified HCQ derivatives can inhibit autophagy and viral replication but exert cytotoxicity in ocular cells. This emphasizes the importance of exploring new compounds with reduced cytotoxicity in mammalian cells in the quest for innovative and safe antiviral drugs.

Fig. 3.

Cellular toxicity and antiviral activity of known autophagy inhibitor drugs against ZIKV. Vero (A-B), ARPE-19 (C-D), and Pr. HTMC (E-F) cells were grown to confluency and treated with 5μM of ROC-325, GNS561, and DC661 and cell viability was assessed by MTT at 24 h (A, C, E). The values were expressed as cell percent viability normalized to mock-treated control cells. In another experiment, cells were infected with ZIKV at MOI of 1 and treated with indicated drugs, and immunofluorescence assay (B, D, F) was performed to visualize ZIKV 4G2 antigen (red) and LC3 puncta (green), and DAPI (blue) was used to counterstain cell nuclei. The images were captured at 200x and 600x magnification.

3.4. Novel quinoline derivatives are safer and exert potent antiviral activity

The absence of effective therapies for ZIKV infection, combined with the notable side effects of previously investigated antiviral drug candidates (HCQ, ROC-325, DC661, and GNS561), led us to engineer chemically modified quinoline derivatives as novel antivirals. These compounds were designed to exhibit minimal cytotoxicity, enhanced cell permeability, and increased antiviral potency, and their hypothesized mode of action is centered on autophagy inhibition. Here, we undertook the evaluation of nine distinct in-house synthesized chemically modified novel analogs, each assigned the numerical designations as GL-287, GL-289, GL-290, GL-292, GL-296, GL-374, GL-377, GL-380, and GL-382 (Fig. 1C).

The cytotoxicity and antiviral properties of the novel quinoline derivatives and their functioning as autophagy inhibitors were tested on Vero cells. Our results showed that quinoline derivatives alone did not cause significant toxicity (Supplementary Fig. S1). However, ZIKV infection led to roughly 25 % cell mortality by 24hpi and GL-287, GL-377, GL-380, and GL-382 at 10 μM demonstrated noteworthy protection against ZIKV-induced cell death, with cell mortality rescued back to >98 %. The novel quinoline derivatives surpassed the degree of rescue against ZIKV-induced cell death when compared to HCQ-treated-ZIKV-infected cells (Fig. 4A). The antiviral activity of these four lead compounds was further confirmed by immunofluorescence and plaque assay to detect ZIKV E protein and viral progeny, respectively. Our results show that these compounds at 1 and 10 µM doses markedly inhibited the viral protein NS3 levels (Fig. 4B), as well as viral progeny production (Fig. 4D).

Fig. 4.

Toxicity and antiviral properties of quinoline derivatives against ZIKV infection. (A) Vero cells were grown to confluency and infected with ZIKV at MOI of 1 followed by treatment with nine in-house synthesized novel quinoline analogs at 10 μM concentration and cell viability was assessed at 24hpi. (B) The level of viability of control cells was marked with a bold dotted line whereas the cut-off for the maximum death induced by ZIKV infection alone was marked by a maroon dotted line to assess the antiviral protective role of the drugs. The cell viability was expressed as percent viability compared to mock-treated-uninfected cells. In another set of experiments, Vero cells were grown to confluency and infected with ZIKV, and treated with compounds GL-287, GL-377, GL-380, and GL-382 at 1 and 10 μM concentrations followed by immunofluorescence staining for flaviviral antigen 4G2 (red) at 24hpi. The cell nuclei were counterstained using DAPI, and images were captured at 200x magnification. (C) The whole cell lysates of Vero cells infected with ZIKV (Z) and treated with Bafilomycin (B), HCQ, GL-287, GL-377, GL-380, and GL-382 were immunoblotted to detect ZIKV NS3, LC3B-I/II, and β-actin. (D) The culture supernatant from the experiment was used to perform a plaque assay and the viral progeny was expressed as viral titer (PFU/mL) on a logarithmic scale. The data represented are the culmination of three independent experiments and are shown as means ± SD. Statistical analysis was performed using one-way ANOVA. ns; non-significant, ****p < 0.0001.

Subsequently, to study the cardiotoxicity of our two lead compounds, we performed in-vitro hERG assay and the results indicate that two of our lead compounds, GL-287 and GL-382 do not bind to the hERG ion channel at concentrations of up to 20 μM (hERG IC_50s for chloroquine and HCQ are 2.5 μM and ∼5.6 μM, respectively (Wan et al., 2020)). We used E-4031, a known hERG ligand, as a positive control in the assay.

To understand the autophagy inhibitory characteristics of our lead candidates, we investigated the accumulation of LC3B that undergoes lipidation and incorporates into autophagosomes during autophagy, resulting in the formation of LC3B-II, a widely accepted autophagosome formation indicator (Yoshii and Mizushima, 2017). Our analysis unveiled an increased accumulation of LC3B-II upon ZIKV infection, indicating activation of autophagy. Treatment with drugs GL-287 and GL-382, in addition to standard autophagy inhibitors such as HCQ and Bafilomycin-A1 (B), accumulated more autophagosome and LC3B-II (halted degradation), a hallmark of autophagic arrest (Fig. 4C). Interestingly, drugs GL-377 and GL-380 also exhibited antiviral activity against ZIKV infection in Vero cells but without significant LC3B-II accumulation, suggesting minimal perturbation of the autophagic pathway by these two drugs. Based on these findings, we primarily focused on these two compounds, GL-287 and GL-382, and evaluated their antiviral and autophagy inhibition activity in both Vero and ocular cells.

3.5. GL-287 and GL-382 inhibit ZIKV replication via autophagy inhibition

To investigate the antiviral activity of lead compounds GL-287 and GL-382, first, we evaluated their cytotoxicity on Vero cells. Our findings confirmed that compounds GL-287 and GL-382 exhibited minimal cytotoxicity, with noticeable effects only from concentrations exceeding 25 μM (Fig. 5A & 5B). Subsequently, we investigated the dose-dependent antiviral potential and LC3B-II puncta formation in response to treatment with these compounds in ZIKV-infected Vero cells. Co-immunostaining for ZIKV E antigen and LC3B protein revealed that compounds GL-287 and GL-382 exhibited concentration-dependent antiviral effects, coupled with reduced viral E antigen detection and augmented LC3B puncta formation (Fig. 5C). These outcomes were in congruence with plaque assay (Fig. 5D) and western blot detection of viral NS3 protein (Fig. 5E) indicating a concentration-dependent antiviral activity of both drugs. Notably, their antiviral efficacy at 5 μM concentration surpassed that of HCQ (10 μM) while not inducing cytotoxicity.

Fig. 5.

Quinoline derivatives GL-287 and GL-382 inhibit ZIKV replication in Vero cells. Vero cells were grown to confluency and treated with varying concentrations of compounds GL-287 (A) and GL-382. (B) MTT assay was performed for cell viability and expressed as percent viability relative to untreated control cells. (C) Cells were pre-treated with 1, 5, and 10 μM of compounds GL-287 and GL-382 for one hour and infected with ZIKV at MOI of 1, followed by immunofluorescence staining for flaviviral antigen 4G2 (red), LC3B (green), and DAPI nuclear stain (blue). The microscopic images were captured at 200x and 600x magnification. (D) Plaque assay was performed from culture supernatant and the viral titers were expressed as PFU/mL on a logarithmic scale. (E) Lysates from ZIKV-infected and drug-treated cells were immunoblotted to detect NS3 protein. (F) In another set of experiments, Vero cells were pre-treated with 10 µM of GL-287, GL-382, and HCQ for an hour and infected with ZIKV, and whole cell lysates were subjected to western blot for LC3B-I/II, p62/SQSTM1, and β-actin. The data represented are the culmination of three independent experiments and are shown as means ± SD. Statistical analysis was performed using one-way ANOVA. ns; non-significant, ***p < 0.0005.

Autophagy, a pivotal cellular process, involves the degradation and recycling of damaged cellular components to maintain cellular equilibrium. LC3B and p62 (SQSTM1) represent key players in the autophagy cascade, where autophagy inhibition leads to concomitantly elevated levels of both, SQSTM1/p62 and LC3B-II (Yoshii and Mizushima, 2017). Our data showed that ZIKV infection increased LC3B-II and p62 protein levels in Vero cells, a trend further accentuated by the treatment with GL-287 and GL-382, indicating inhibition of autophagy. Notably, compound GL-287 displayed the most pronounced elevation in LC3B-II and p62 levels, surpassing GL-382 and HCQ (Fig. 5F).

The long-term HCQ treatment has been associated with ocular toxicity, occasionally leading to HCQ-induced retinopathy and potential vision loss (Nika et al., 2014; Paniri et al., 2020). Hence, we evaluated the effect of novel quinoline derivatives on ocular cells, i.e., ARPE-19 and Pr. HTMC, representing cells from posterior and anterior eye segments, respectively. The cytotoxicity assay revealed minimal toxic effects of both drugs GL-287 and GL-382 in ARPE-19 (Fig. 6) or Pr. HTMC (Fig. 7) cells, even at concentrations up to 25 μM. Similar to the trends observed in Vero cells, both GL-287 and GL-382 exhibited concentration-dependent antiviral effects in ocular cells as assessed by immunofluorescence staining for ZIKV E protein (Fig. 6C & 7C), plaque assay (Fig. 6D & 7D), and western blot for ZIKV NS3 protein (Fig. 6E & 7E). Moreover, LC3B-II and p62 levels were augmented in cells treated with these drugs (Fig. 6F & 7F). Importantly, the extent of ZIKV suppression achieved with compounds GL-287 and GL-382 mirrored that of HCQ at 10 μM concentration. Collectively, these results indicate the non-cytotoxic and potent antiviral properties of our novel quinoline derivatives via autophagy inhibition.

Fig. 6.

Antiviral and autophagy inhibition properties of GL-287 and GL-382 in ZIKV-infected ARPE-19 cells. ARPE-19 cells were grown to confluency and treated with varying concentrations of drugs GL-287 (A) and GL-382 (B). MTT assay was performed for cell viability and expressed as percent viability relative to untreated cells. (C) ARPE-19 cells were infected with ZIKV at MOI of 1 and treated with 1, 5, and 10μM concentrations of GL-287 and GL-382, followed by immunofluorescence staining for flaviviral antigen 4G2 (red), LC3B (green), and DAPI nuclear stain (blue). The microscopic images were captured at 200x and 600x magnification. (D) Plaque assay was performed from culture supernatant, and the viral titers were expressed as PFU/mL on a logarithmic scale. (E) Whole cell lysate from ARPE-19 cells infected with ZIKV and treated with different concentrations (1, 5, and 10 µM) of GL-287, GL-382, and 10 µM of HCQ were immunoblotted to detect NS3 protein. (F) In another set of experiments, ARPE-19 cells were infected with ZIKV in the presence and absence of 10 µM of GL-287, GL-382, and HCQ, and whole cell lysates were subjected to western blot for LC3B-I/II, p62/SQSTM1, and β-actin. The data represented are the culmination of three independent experiments and are shown as means ± SD. Statistical analysis was performed using one-way ANOVA. ns; non-significant, ***p < 0.0005; ****p < 0.0001.

Fig. 7.

Antiviral and autophagy inhibition properties of GL-287 and GL-382 in ZIKV infected Pr. HTMC cells. Pr. HTMC cells were grown to confluency and treated with varying concentrations of compounds GL-287 (A) and GL-382 (B). MTT assay was performed for cell viability and expressed as percent viability relative to untreated cells. (C) Pr. HTMC were infected with ZIKV at MOI 1 and treated with 1, 5, and 10μM concentrations of compounds GL-287 and GL-382, followed by immunofluorescence staining for flaviviral antigen 4G2 (red), LC3B (green), and DAPI nuclear stain (blue). The microscopic images were captured at 200x and 600x magnification. (D) Plaque assay was performed from culture supernatant, and the viral titers were expressed as PFU/mL on a logarithmic scale. (E) Whole cell lysate from Pr. HTMC infected with ZIKV and treated with different concentrations (1, 5, and 10 µM) of GL-287, GL-382, and 10 µM of HCQ were immunoblotted to detect ZIKV-NS3 protein. (F) In another set of experiments, Pr. HTMC were infected with ZIKV in the presence and absence of 10 µM of GL-287, GL-382, and HCQ, and whole cell lysates were subjected to western blot for LC3B-I/II, p62/SQSTM1, and β-actin. The data represented are the culmination of three independent experiments and are shown as means ± SD. Statistical analysis was performed using one-way ANOVA. ns; non-significant, **p < 0.005; ****p < 0.0001.

3.6. GL-287 and GL-382 treatment modulates ZIKV-induced innate immune responses

Autophagy regulates multiple components of innate immunity during viral infection (Pradel et al., 2020; Wang et al., 2015). As both drugs attenuated ZIKV replication in ocular cells, we postulated that they would also alter the innate inflammatory and antiviral responses. To test this, ARPE-19 and Pr. HTMC cells were pretreated with GL-287 and GL-382 for 1 h, followed by ZIKV infection in the presence of the drugs for 48 h. The qPCR analysis was performed to measure the relative mRNA expression of the classical inflammatory and antiviral genes, as reported in our prior studies (Singh et al., 2020; 2024). ZIKV infection elevated mRNA expression of pathogen-recognition receptors (PRRs) (TLR3, RIG-I, and MDA5), IFNs (IFNα1, IFNβ2, and IFNγ), IFN-inducible genes (ISG15, OAS2, and MX1), and inflammatory mediator genes (TNF-α, IL1β, IL-6) in both cells. Overall, GL-287 and GL-382 treatments in ARPE-19 cells decreased the ZIKV-induced expression of PRRs, inflammatory mediators, as well as IFNs, and ISGs (Fig. 8A). Furthermore, the novel quinoline analogs alone did not exhibit any effect on cytokine (IL1β, IL-6, and TNFα) expression while HCQ treatment increased their expression (Supplementary Fig. S2). However, in Pr. HTMC cells, GL-287 increased the expression of TLR3 with no significant induction for RIG-I and MDA5, while treatment with GL-382 significantly reduced all three of these PRRs (Fig. 8B). There was an increase in transcript levels of TNF-α, IFNγ, and OAS2 in the Pr. HTMC treated with either GL-287 or GL-382. None of the other inflammatory mediators, IFNs, and ISGs were significantly upregulated upon treatment with GL-287, while there was a decrease in their expression upon treatment with GL-382. Treatment with HCQ on the ZIKV-infected ocular cells reduced the PRRs, IFNs (except IFN-β2), and ISGs, whereas there was a significant increase in the expression of inflammatory mediators. Taken together, these results indicate that the novel quinoline analogs GL-287 and GL-382 inhibit PRRs-mediated signaling, which plausibly resulted in reduced expression of IFNs/ISGs and an overall decrease in pro-inflammatory responses in the cells that eventually culminated in better control of viral replication with minimal cellular damage and death.

Fig. 8.

Modulation of innate immune response by quinoline derivatives GL-287 and GL-382 in ZIKV-infected ARPE-19 and Pr. HTMC cells. (A) ARPE-19 and (B) Pr. HTMC were pretreated with GL-287, GL-382, and HCQ (10 μM) for 1 h followed by ZIKV infection at MOI of 1. After 48 h, total RNA was extracted and subjected to qPCR analysis of PRRs (TLR3, RIGI, and MDA5), inflammatory mediators (TNFα, IL-1β, and IL-6), IFNs (IFN-α2, IFN-β1, and IFN-γ), and IFN-induced genes (OAS2, ISG15, and MX1). The data represented are the culmination of three independent experiments and are shown as means ± SD. Statistical analysis was performed using one-way ANOVA. ns; non-significant, *p < 0.05; **p < 0.005; ***p < 0.0005; ****p < 0.0001.

4. Discussion

ZIKV has emerged as a global public health threat due to its ability to cause microcephaly and congenital malformations, and its human-to-human transmissibility. Currently, no vaccine or specific antivirals are available for treatments, emphasizing the necessity for effective anti-ZIKV drugs (Cao-Lormeau et al., 2016; Malkki, 2016; Miner and Diamond, 2017; Miranda et al., 2016; Mlakar et al., 2016; Oehler et al., 2014; Rasmussen et al., 2016). Repurposing existing drugs, like hydroxychloroquine (HCQ), has emerged as a promising approach to expedite antiviral therapeutic development against several RNA viruses. Modification of HCQ chemical structure, pharmacokinetic properties, cellular uptake, and affinity for autophagy-related targets have allowed controlling viral replication and spread efficiently. Numerous studies have demonstrated the effectiveness of modified HCQ derivatives in inhibiting RNA virus replication both in-vitro and in-vivo, exhibiting promising results in reducing viral load and mitigating virus-induced pathogenesis (Choi et al., 2018; Liang et al., 2016; Liu et al., 2023). However, HCQ is reported to exert toxicity when used at higher doses and for long durations.

Here, we synthesized nine distinct chemically modified novel quinoline analogs and assessed their potential as antiviral agents against ZIKV by manipulating the autophagy process. Autophagy, a conserved cellular process, plays a dual role during viral infections by promoting or inhibiting viral replication in disease and viral-specific contexts. Many flaviruses, including ZIKV and DENV, manipulate autophagy to facilitate their replication and evade the immune response (Choi et al., 2018; Metz et al., 2015). Therefore, targeting the autophagic process offers a promising avenue for antiviral drug development. Among the nine synthesized quinoline analogs, four compounds displayed the most potent antiviral properties against ZIKV, with two lead compounds, GL-287 and GL-382, exhibiting minimal cytotoxicity. Because ZIKV has been shown to cause pathological manifestations in both the anterior segment (glaucoma) and the retina (retinal atrophy); therefore, we used Pr. HTMC and ARPE-19 cells to represent the anterior and posterior segments of the eye, respectively (Singh et al., 2017; 2019; 2018) to investigate the anti-ZIKV properties of the lead compounds, GL-287 and GL-382.

While the exact antiviral mechanism of HCQ remains elusive, it is believed to interfere with viral entry, replication, and immune modulation (Chandler et al., 2020; Hashem et al., 2020). HCQ's impact on cellular compartment acidity, particularly endosomes, can affect viral fusion and replication (Kumar et al., 2018; Martinez et al., 2020; Romanelli et al., 2004; Wang et al., 2015). Additionally, HCQ's modulation of autophagy, critical for viral replication and host immunity, has been explored previously. In a mouse model of pregnancy, HCQ has been shown to inhibit ZIKV replication by targeting the NS2B-NS3 serine protease (Cao et al., 2017; Kumar et al., 2018). In this model, HCQ also disrupted the ZIKV-induced autophagic process, leading to reduced viral replication. Furthermore, HCQ has exhibited antiviral effects against DENV via increased activation of innate immune signaling and reactive oxygen species (ROS) production (Wang et al., 2015). Similarly, we found that HCQ is effective in reducing ZIKV replication in Vero and ocular cells via inhibition of autophagy but also exerted toxicity. HCQ-mediated ocular toxicity and other side effects include an increased risk of arrhythmic heart, blood, and lymph disorders, kidney injury, liver problems, and failure (Ben-Zvi et al., 2012; Doyno et al., 2021; Izcovich et al., 2022; Nika et al., 2014). Furthermore, HCQ could not inhibit autophagy potently enough in humans at doses that could be well tolerated to recapitulate the results observed in preclinical studies (Mahalingam et al., 2014; Perez-Hernandez et al., 2019).

Given the limitations of HCQ usage, we next evaluated the effect of three promising autophagy inhibitors, ROC-325, GNS561, and DC661, used extensively for treating cancer, RA, and SARS-CoV-2 (Bestion et al., 2022; Brun et al., 2022; Carew et al., 2017; Jones et al., 2019; Nawrocki et al., 2019; Xu et al., 2022). ROC-325 is a dimeric molecule containing HCQ's core and lucanthone, demonstrating tenfold higher autophagic inhibition than HCQ (Jones et al., 2019). Similarly, DC661, a selective PPT1, and a late-stage autophagy inhibitor, exhibited robust lysosomal membrane permeabilization compared to other chloroquine derivatives (Xu et al., 2022). GNS561/Ezurpimtrostat is a basic lipophilic molecule that targets PPT1 to induce lysosomal degradation and inhibit autophagic flux. Bestion et al. has demonstrated anti-SARS-CoV-2 activities of GNS561 with potent autophagic inhibition (Bestion et al., 2022). However, in our study, ROC-325, DC661, and GNS561 induced significant cytotoxicity in the ocular cells, possibly due to a cell-line-specific response. Consequently, our focus shifted to developing novel autophagy inhibitors possessing reduced cytotoxicity while maintaining antiviral efficacy against ZIKV. We synthesized nine different modified analogs with piperazine and piperidine-substituted quinoline derivatives with superior efficacy and safety profiles. Upon screening these compounds for toxicity, antiviral properties, and protection against ZIKV-induced cytopathy in Vero cells, we found four compounds with significant antiviral activity against ZIKV. Among these four lead compounds, two compounds, GL-287 and GL-382, demonstrated potent autophagic inhibitory properties with ZIKV infection. Most importantly, both lead compounds, GL-287 and GL-382, do not show any cardiotoxicity up to 20 μM concentrations in an in-vitro hERG assay and inhibited viral replication in a dose-dependent manner via inhibition of autophagy. Although our results demonstrated the antiviral activities of GL-287 and GL-382 via autophagic inhibition, the additional potential mechanism of antiviral activity of these drugs, such as modulation of cell surface distribution of heparan sulfate affecting the viral entry and egress along with the different steps of viral infection/replication (Agelidis and Shukla, 2020), cannot be ruled out and needs further investigation. Currently, beyond the scope of this study, the other two compounds, GL-377 and GL-380, with potent antiviral properties against ZIKV, also need further exploration.

In response to viral infection, the host cells employ intracellular PRRs, such as TLRs and RIG-I-like receptors, to trigger an innate immune response, predominantly with the production of type I and II IFNs and antiviral ISGs. These innate immune responses promote inflammation, immune cell activation, and viral clearance. Importantly, our data show that ZIKV infection in ARPE-19 and Pr. HTMC cells induced the expression of PRRs-TLR3, RIG-I, and MDA5, along with increased expression of IFNs, ISGs, and inflammatory mediators, including cytokines and chemokines. Interestingly, our results show that treatment with the quinoline derivatives attenuated the innate inflammatory response by suppressing the ZIKV-induced expression of PRRs, inflammatory mediators, IFNs, and ISGs, which could be owed to the reduced viral infectivity as well as the interference with TLR signaling pathways by HCQ. HCQ can hinder TLR3, TLR7, TLR8, and TLR9 processing, potentially preventing TLR activation by altering endosomal pH (Ewald et al., 2008; Hacker et al., 1998; In 't Veld et al., 2021; Kuznik et al., 2011). HCQ's direct binding to nucleic acids could also block TLR7 and TLR9 signaling by inhibiting ligand interactions (Chandler et al., 2020). HCQ's inhibition of cGAS activity through ligand binding interference is another proposed mechanism (Zhang et al., 2014). Moreover, HCQ has been shown to inhibit TLR-mediated cytokine response, including IFNs, in several autoimmune and inflammatory diseases (van den Borne et al., 1997; Wallace et al., 1994; 1993). Whether novel quinoline analogs GL-287 and GL-382 also exert their anti-inflammatory response via inhibiting TLR-mediated innate immune response warrants further investigation.

In summary, our study synthesized and evaluated novel quinoline analogs with potent antiviral and immunomodulatory properties. Our two lead compounds, (GL-287 and GL-382) exhibited superior safety and antiviral activities against ZIKV-infected ocular cells. Although the current study demonstrates the antiviral properties of our novel quinolone derivatives against ZIKV in in-vitro models, their evaluation in in-vivo models would be essential to determine the therapeutic efficacy of these compounds in ZIKV-induced ocular complications, which is beyond the scope of the current study. Overall, our study provides the rationale for drug design and rapidly deploying these platforms to develop newer antiviral agents with the goal of pandemic preparedness. Moreover, these novel quinoline analogs can replace conventional HCQ-based treatment regimens for other autoimmune diseases, including RA and SLE.

CRediT authorship contribution statement

Sneha Singh: Writing – review & editing, Writing – original draft, Visualization, Software, Methodology, Formal analysis, Data curation. Faraz Ahmad: Writing – original draft, Visualization, Software, Methodology, Formal analysis, Data curation. Hariprasad Aruri: Writing – original draft, Validation, Software, Methodology, Formal analysis, Data curation. Susmita Das: Writing – review & editing, Validation, Resources, Methodology, Formal analysis, Data curation. Prahlad Parajuli: Writing – original draft, Validation, Software, Resources, Methodology, Data curation. Navnath S. Gavande: Writing – review & editing, Supervision, Project administration, Funding acquisition, Formal analysis, Conceptualization. Pawan Kumar Singh: Writing – review & editing, Writing – original draft, Supervision, Project administration, Funding acquisition, Investigation, Formal analysis. Ashok Kumar: Writing – review & editing, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

• All relevant data are within the manuscript, figures, and supplementary data files.