Macromolecular Viral Entry Inhibitors as Broad-Spectrum First-Line Antivirals with Activity against SARS-CoV-2

Abstract

Inhibitors of viral cell entry based on poly(styrene sulfonate) and its core-shell nanoformulations based on gold nanoparticles are investigated against a panel of viruses, including clinical isolates of SARS-CoV-2. Macromolecular inhibitors are shown to exhibit the highly sought-after broad-spectrum antiviral activity, which covers most analyzed enveloped viruses and all of the variants of concern for SARS-CoV-2 tested. The inhibitory activity is quantified in vitro in appropriate cell culture models and for respiratory viral pathogens (respiratory syncytial virus and SARS-CoV-2) in mice. Results of this study comprise a significant step along the translational path of macromolecular inhibitors of virus cell entry, specifically against enveloped respiratory viruses.

Keywords: SARS-CoV-2; broad-spectrum antivirals; entry inhibitors; in vivo; macromolecules; polyanions; respiratory syncytial virus (RSV).

Figure 1

A) Chemical formula of the RAFT agent and the resulting polystyrene sulfonate polymer (PSS). B) Schematic illustration depicting association of the RAFT‐derived polymers (blue) with gold nanoparticles, due to affinity of the trithiocarbonate (red) to the gold surfaces. C) Photography images of the 10 nm gold nanoparticles, with or without the brush‐type corona comprised of PSS (100 kDa) in water, phosphate buffer, phosphate buffer saline, or DMEM cell culture medium: colloidally dispersed AuNP have a characteristic red color, whereas nanoparticle aggregation produces a characteristic blue color. D) UV/vis spectra corresponding to the images in panel C for AuNP with or without PSS corona in DMEM. E) Transmission electron microscopy image for AuNP (sized 5–40 nm) functionalized with PSS (3, 38, and 100 kDa), scale bar = 100 nm.

Figure 2

Broad‐spectrum antiviral activity of PSS polymers with variable length. A) PSS compounds with molar masses of 3, 38, and 100 kDa were dissolved and titrated in PBS before adding virus at 1:1 (v/v) dilution, resulting in the indicated concentrations. Following incubation at 37 °C for 30 min, mixtures were added to target cells (HIV‐1: TZM‐bl, ZIKV: VeroE6, HSV‐1: ELVIS, SARS‐CoV‐2: Caco‐2, RSV: A549, IAV: MDCK). Infection rates were determined one (HSV‐1, RSV) or two days postinfection (all other viruses) by quantification of reporter cell line activity (HIV‐1, HSV‐1), immunodetection of viral proteins (ZIKV, SARS‐CoV‐2), flow‐cytometric analysis of virally expressed reporter GFP (RSV) or neuraminidase activity assay (IAV). Raw values were baseline‐subtracted (uninfected cells) and normalized to cells infected without polymer pre‐treatment of virus. Shown are mean values obtained from two (n =2, for HIV‐1, ZIKV, SARS‐CoV‐2) or three (n = 3, for HSV‐1, IAV) independent experiments each performed in triplicates ± SEM. Data for RSV are shown as mean values from three independent experiments (n = 3) in duplicates ± SEM. B) IC₅₀ values for each virus and polymer are shown as log₁₀ transform to highlight differences, calculated from nonlinear regression curve fits (inhibitor vs. normalized response, GraphPad Prism).

Figure 3

Broad‐spectrum antiviral activity of AuNP‐coupled PSS. A) Compounds were dissolved and titrated in PBS before adding virus at 1:1 dilution, resulting in the indicated concentrations on virus. Following incubation at 37 °C for 30 min, mixtures were added to target cells at tenfold dilution (HIV‐1: TZM‐bl, ZIKV: VeroE6, HSV‐1: ELVIS, SARS‐CoV‐2: Caco‐2, RSV: A549, IAV: MDCK). Infection rates were determined one (HSV‐1, RSV) or two days after infection (all others) by measuring enzyme activity of the reporter cell line (HIV‐1, HSV‐1), immunodetection of viral proteins (ZIKV, SARS‐CoV‐2) or neuraminidase activity assay (IAV). Raw values were baseline‐subtracted (uninfected cells) and normalized to cells infected without polymer pre‐treatment of virus. Shown are mean values obtained from two (n = 2, for HIV‐1, ZIKV, SARS‐CoV‐2) or three (n = 3, for HSV‐1, IAV) independent experiments each performed in triplicates ± SEM. B) Activity of S3‐based AuNP‐coupled polymers against RSV measured by flow‐cytometric analysis of virally expressed reporter GFP. Data are shown as mean values from three independent experiments (n = 3) in duplicates ± SEM. C) IC₅₀ values for each virus and polymer are shown as log₁₀ transform to highlight differences, calculated from nonlinear regression curve fits (inhibitor vs normalized response, GraphPad Prism).

Figure 4

Inhibition of SARS‐CoV‐2 Spike pseudoparticles by A) PSS and AuNP‐PSS and B) specific mAbs. Compounds or mAbs were serially titrated in PBS before adding pseudoviruses containing Spike of indicated SARS‐CoV‐2 VOC 1:1, resulting in indicated concentrations on virus. After 30 min at 37 °C, mixtures were added to Vero E6 cells. Pseudoparticle entry was evaluated by Firefly luciferase measurement 16–18 h post‐transduction. Values were background‐subtracted (uninfected cells) and normalized to cells infected in absence of inhibitors. Shown are mean values obtained from three independent experiments (n = 3) each performed in triplicates ± SEM for (A), one experiment in triplicates (n = 1) for mAbs tested as conformation of VOC immune‐evading behavior in (B).

Figure 5

Inhibition of common cold coronavirus HCoV‐229E and‐OC43 by PSS and AuNP‐PSS. Compounds were titrated in PBS and mixed with virus at 1:1 dilution (v/v), resulting in indicated concentrations (for PSS). After incubation at 37 °C for 30 min, mixtures were added to Huh7 cells. After two (229E) or three (OC43) days, infection was quantified by immunodetection of viral N protein. Values of infected cells were subtracted and infection rates normalized to cells infected in the absence of compounds. Data from three independent experiments (n = 3) in triplicates, means ± SEM.

Figure 6

Effect of cell treatment versus virus treatment with polymers on SARS‐CoV‐2‐Spike driven entry. Compounds were titrated and added to VeroE6 cells, which were then incubated for 60 min at 37 °C and subsequently infected with Wuhan Spike‐VSV‐pseudoparticles (“Cell treatment”‐grey circles). Alternatively, in the “Virus treatment” (yellow circles) conditions, virions and compounds were premixed, incubated for 30 min at 37 °C and then the mixture was added to cells at tenfold dilution. Shown are mean values obtained from three independent experiments (n = 3) each performed in triplicates ± SEM.

Figure 7

Nanoparticle tracking analysis reveals biophysical changes in virus‐like particles following incubation with AuNP‐PSS and free PSS. Fluorescent lentiviral VLPs were used to observe the binding of AuNP‐PSS and PPS to viral particles by fluorescent nanoparticle tracking analysis. After incubating VLPs with 100 µg mL⁻¹ BS3, S3 or B for 60 min at 37 °C, A) the median VLP diameter was determined. PBS was used as a negative control, while peptide nanofibrils aggregating virions (SEVI, 100 µg mL⁻¹) were used as a positive control. B) Additionally, pre‐incubated VLPs were diluted in water to determine the surface charge by measuring zeta potential. Polycation polybrene and polyanion enoxaparin were used as controls to induce strong changes in virion surface charge in a positive and a negative direction, respectively. Both parameters were acquired by F‐NTA at 488 nm using a ZetaView TWIN, thus excluding scattering signals from polymers, AuNP or non‐VLP particles contained in the medium. Five acquisitions (n = 5) were performed for each compound, values show mean values ± SEM. One‐way‐ANOVA with Dunnett's post‐test compares median VLP diameter and Zeta potential to PBS/water treated VLPs. * = p < 0.05, *** = p < 0.001, all without indication non‐significantly different from PBS/water only.

Figure 8

Evaluation of membrane‐lytic/virucidal activity of AuNP‐PSS and PSS. (A) Membrane‐disrupting activity of BS3, S3 and B evaluated by liposome leakage assay. Virus‐like‐liposomes (DOPC/SM/Chol; 45/25/30 mol%) filled with carboxyfluorescein (50 × 10⁻³ m) and purified by SEC, were added to 96‐well black side‐clear bottom plates and baseline fluorescence recorded for 5 min once per minute. Compounds were then added at indicated concentration (first vertical line) and fluorescence increase (indicating membrane disruption) recorded for 30 min once per minute. Triton X‐100 was then added to all wells at 1% final concentration to induce total lysis (second vertical line). Percent values shown are background‐subtracted (signal of liposome prior to addition of compounds) and normalized to fluorescent signal achieved in each well after complete liposome leakage induced by addition of 1% (v/v) Triton X‐100. Magnitude of relative leakage prior to addition of Triton X‐100 thus reveals the membrane‐disrupting activity of the tested compounds. Two independent experiments (n = 2) in triplicates, means ± SEM. B) Membrane‐disrupting activity by BS3, S3 and B evaluated by red blood cell lysis assay. 3×10⁶ RBC were incubated with indicated concentrations or 1% Triton X‐100 (colors as in A) for 30 min, samples then centrifuged for 5 min at 500 × g, supernatants transferred to new plates and released hemoglobin measured via absorbance at 450 nm. Blank values (PBS only) subtracted, normalized to values after addition of 1% Triton X‐100. One experiment (n = 1) in triplicates, means ± SEM. C) SARS‐CoV‐2 virus stocks were incubated with 500 µg mL⁻¹ of compounds for 30 min at 37 °C before residual infectivity was determined by TCID₅₀ endpoint titration according to Reed & Muench. Two independent experiments (n = 2) in triplicates, means ± SEM.

Figure 9

In vivo tolerability study of PSS and AuNP‐PSS in BALB/c mice. A) Experimental setup for tolerability study. BALB/c mice (5 male and 5 female/group; n = 5) received indicated amounts of compound once daily by intranasal application, with the total dose split to both nostrils. B,C) Animals were weighed daily. At 24 h after the final application, animals were sacrificed and nasal cavity, olfactory bulb, trachea, lung, esophagus, and stomach were analyzed by histology. Data presented as means ± SEM, n = 5 (lower for timepoints where some animals were already sacrificed). Significance in weight differences between vehicle and PSS‐treated mice for each time point tested using ordinary one‐way ANOVA with Dunnett's post‐test, * = p < 0.05. All without indication are not significantly different from vehicle. Schematic in (A) created with BioRender.com.

Figure 10

Histopathological evaluation in respiratory organs following administration of PSS (A–J) and core–shell PSS‐AuNP nanoformulations (K–V). Images depict nasal epithelium (A–C; K–N), trachea (D–F; O–R), and lung tissue (G–J; S–V) for vehicle controls (saline, panels A, D, G, K, O, S), 100 kDa PSS dosed at 250 µg (highest feasible dose, HFD, panels B, E, H) or 2.5 µg (highest tolerable dose, panels C, F, J); AS3 (L, P, T); BS3 (M, Q, U) and CS3 (N, R, V) dosed at 2.5 µg. For PSS and core–shell nanoformulations dosed at 2.5 µg, no histopathological changes, as compared to vehicle, were evident.

Figure 11

In vivo inhibition of RSV infection by PSS and its core–shell formulations with AuNP sized 5, 10, and 20 nm. A) Female BALB/c mice (n = 5 per group) were infected by intranasal administration of rHRSV‐Luc (10⁵ pfu) pre‐incubated with 0.1 g L⁻¹ polymers. Mice were treated at two dpi before visualization and B) quantification by the luminescence of rHRSV‐Luc replication in mice at four dpi. Values are shown as means ± SEM. Ordinary one‐way ANOVA with Dunnett's post‐test compares mean radiance of treated animal groups to PBS‐treated control animals. *** = p < 0.001. pfu, plaque‐forming units; dpi, days post‐infection. Schematic in (A) created with BioRender.com.

Figure 12

In vivo inhibition of SARS‐CoV‐2 by PSS. A) In a prophylactic application scenario, K18‐hACE2 mice (6 male mice per group) received BS3 or S3 (5 µg i.n., half per nostril, 50 µL volume in total) 60 and 10 min prior to infection. For infection, SARS‐CoV‐2 Wuhan Hu‐1 was inoculated i.n. (300 FFU). All animals were sacrificed 2 dpi and lungs were harvested. After tissue homogenization, viral RNA levels B) and nucleoprotein levels C,D) were analyzed by RT‐qPCR and western blot, respectively. E) In a pre‐incubation scenario, SARS‐CoV‐2 Wuhan Hu‐1 (300 FFU) was mixed with polymers (100 µg mL⁻¹ concentration on virus) and the mixture was applied i.n. (50 µL total volume). At seven hpi, mice were further treated i.n. with 5 µg polymer in 50 µL volume. Mice were sacrificed and lung tissue was analyzed for viral RNA and protein two dpi (F–H). Gray values of WB images for tubulin (cellular control) and SARS‐CoV‐2‐N bands were quantified by Fiji. Error bars show means ± SEM of six animals (n = 6). Ordinary one‐way ANOVA with Dunnett's post‐test compares mean viral loads and viral protein levels of treated animal groups to PBS‐treated control animals. ** = p < 0.01. Schematics in (A) and (E) created with BioRender.com