Broad-spectrum synthetic carbohydrate receptors (SCRs) inhibit viral entry across multiple virus families

Abstract

Viral pandemics continue to threaten global health and economic stability. Despite medical advances, the absence of broad-spectrum antivirals (BSAs) prevents rapid responses to emerging viral threats. This is largely due to the lack of universal drug targets across diverse viral families and high variability among viral proteins. In this study, we evaluated 57 synthetic carbohydrate receptors (SCRs) for antiviral activity in cellulo using pseudotyped virus particles (PVPs) from six high-risk viruses across three families: Paramyxoviridae, Filoviridae, and Coronaviridae. Four SCRs inhibited all tested PVPs, and their efficacy was confirmed against live viruses including SARS-CoV-2, MERS-CoV, EBOV, MARV, NiV, and HeV. Notably, SCR005 and SCR007, which exhibited minimal toxicity, significantly reduced SARS-CoV-2 infection in a severe animal model with a single dose. Mechanistic studies suggested that SCRs bind viral envelope N-glycans, blocking viral attachment and/or fusion. These results identify conserved viral N-glycans as promising BSA targets and establish SCRs as candidate prophylactic agents against enveloped viruses with pandemic potential.

Fig. 1.. SCRs are inhibitors of SARS-CoV-2, SARS-CoV-1, MERS-CoV, NiV, HeV, EBOV, and MARV entry.

(A) SCRs can inhibit viral entry and membrane fusion. Created in BioRender. H. Aguilar-Carreno (2025) https://BioRender.com/lbjkacr. (B) Molecular structures of SCRs with broad-spectrum activity and control compound SCR013. (C) VERO-TMPRSS2 cells were pretreated with 10 μM indicated SCRs, followed by pseudotyped virus infections. Postinfection, levels of intracellular infection were evaluated via a high-content analysis of luciferase activity, which was then quantified in comparison to control cells treated with 0.1% dimethyl sulfoxide (DMSO) (based on three independent studies). (D and E) IC₅₀ and CC₅₀ values were determined using GraphPad Prism 10 by fitting dose-response curves with a four-parameter logistic (4PL) nonlinear regression model in Vero cells. This was accomplished using the PRNT test and IFA with spike protein immunostaining in cells that received pretreatment (SCR005, SCR007, SCR0064, SCR0065, and SCR0013, compared to 0.1% DMSO, n ≥ 3). The intermittent line indicates 50% toxicity. Data were reported as means ± SD from at least three independent experiments.

Fig. 2.. SCR005 and SCR007 mitigate morbidity and mortality in the hACE2 mouse model of SARS-CoV-2 infection with minimal toxicity.

(A) Mice (n = 8 to 10 per treatment group) were treated with SCRs at day 0 upon infection. Surviving mice were euthanized at the study end point. (B to D) Weight change and probability of survival of mice treated with SCR013 (B), SCR005 (C), or SCR007 (D) are depicted compared to control. (E and F) Quantitative evaluations of viral RNA copies (virus genome copies per nanogram RNA) were performed via RT-qPCR on lung (E) and brain (F) tissues from infected mice (n = 8 to 10 per group) at the time of death or at the study end point. Notably, in the lung, the viral loads in the SCR005 and SCR007 groups were significantly reduced compared to the control group, with *P = 0.0199 and ***P = 0.0005, respectively. Similarly, in the brain, SCR007 led to a substantial decrease in viral burden compared to the control mice (***P = 0.0002). The significance of these observations was confirmed using a one-way analysis of variance (ANOVA) multiple comparison test. Data presented as median. (G and H) Quantification of the SARS-CoV-2 nucleocapsid (N) protein presence per square millimeter, determined by IHC staining in lung and brain sections at the time of death or at the study end point. The reduction in N-positive events per square millimeter was significant in the tissues of the lung (G) and brain (H) of mice treated with SCR007 compared to the control group, with *P = 0.0283 and ***P = 0.0005, respectively. An unpaired t test was used for the statistical analysis. The data are shown as means ± SEM, with each data point representing an entire lung or brain section from each mouse (n = 8 to 10 per treatment).

Fig. 3.. SCR007 and SCR005 treatment reduces SARS-CoV-2–induced morbidity, viral burden, and weight loss in hACE2 transgenic mice.

(A and B) Mice received a single treatment upon SARS-CoV-2 infection, with euthanasia at 5 dpi, and less weight loss was observed in SCR007-treated group (****P < 0.0001). (C) Viral titers (PFU per gram tissue) from lungs showed significant reduction in SCR005 and SCR007 groups versus control (n = 8 to 10; *P = 0.0211 and *P = 0.0197, respectively). (D) Brain viral load was significantly reduced in SCR007 (***P = 0.0005); SCR005 showed nonsignificant decrease. One-way ANOVA; median values. (E and F) RT-qPCR of viral RNA in lung tissue showed significantly lower levels in SCR005 (**P = 0.0039) and SCR007 (****P < 0.0001) versus control. In brain, *P = 0.433 (SCR005) and ***P = 0.0007 (SCR007). (G and H) Counts of nucleocapsid-positive cells/mm² in lungs showed significant reduction with SCR005 (***P = 0.0009) and SCR007 (P = 0.0002). Brain nucleocapsid-positive cells were lower in SCR007-treated mice (***P = 0.0002). Unpaired t test, median values (n = 8 to 10). (I and J) Lung pathology scores were comparable except SCR005, which was elevated (*P = 0.0241). Brain pathology improved in both treatments, with SCR007 significant (****P < 0.0001). (K and L) H&E staining showed perivascular inflammation [(K) and (L) arrowheads, (ii), (iv), (vi), (viii)]. SCR005 lungs showed alveolar cell increase and type II pneumocyte hyperplasia [(K) and (L) arrow, (v), (vi)]. IHC showed strong nucleocapsid signal in control [(K) and (L), (ix) and (x)] and SCR013 [(K) and (L), (xi) and (xii)] and minimal in SCR005 and SCR007 [(K) and (L), (xiii) to (xvi)]. Brain pathology improved in SCR005 and SCR007 [(K) and (L), (v) to (viii)] and perivascular cuffs in control and SCR013 [(K) and (L), (i) to (iv)]. IHC in brain showed immunolabeling in control [(K) and (L), (ix) and (x)] and SCR013 [(K) and (L), (xi) and (xii)], reduced in SCR005 [(K) and (L), (xiii) and (xiv)], and rare/absent in SCR007 [(K) and (L), (xv) and (xvi)]. Scale bars, 1 mm [(K) and (L), (i), (iii), (v), (vii), (ix), (xi), (xiii), and (xv)] and 50 μm [(K) and (L), (ii), (iv), (vi), (viii), (x), (xii), (xiv), and (xvi)]. H&E, hematoxylin and eosin.

Fig. 4.. Mechanistic studies of the viral life cycle, SCR binding by STD NMR spectroscopy, and MD analysis.

(A) Temperature-dependent assay to investigate the phase of the SARS-CoV-2 and NiV PVPs life cycle, receptor binding, or membrane fusion that is inhibited by SCR007. The data are shown as means ± SEM. (B) STD-NMR Spectroscopy of SCR007 (1 mM) and RBD–SARS-CoV-2 S glycoprotein. The chemical shifts of the protons that belong to the SCR are shown. The signals in the STD difference spectrum (×5) indicate close contact between the SCR and the glycoprotein. (C) Average distance between the protons of the SCR007 and the center-of-mass of carbohydrates (numbering shown in fig. S109) of the fucosylated triantenary glycan computed for the dominant structure observed in the MD simulations shown in (D). The specific interactions between the SCR007 and the glycan are highlighted in red. (D) Dominant structure of an N-glycan•SCR007 complex observed in MD simulations. The specific contacts between the arms of the receptor and the GlcNAc₂Man₃ core (1 and 2) and antennae (3 and 4) are highlighted in red. (E). Example of SCR007 binding to the RBD-SARS-Cov-2 S glycoprotein (RBD-M2) during one trajectory. The glycosylation sites, N331 and N343, are occupied by triantennary glycans, and the extent of their flexibility is shown using 20 overlapping simulation frames sampled with 5-ns intervals. The red surface indicates the space near glycans occupied by SCR007 in the same 20 frames. (F) Average probability of finding at least one SCR007 molecule within 1.5 nm from the center-of-mass an N-glycan of a specific model of the RBD throughout five independent simulations. The gray bars indicate the SDs calculated from five independent stimulations. ppm, parts per million.

Fig. 5.. Binding of SCRs to N-glycan microarrays.

(A) Binding of SNA to S₂G₂A₂-functionalized brushes. (B) Optical microscopy image of S₂G₂A₂-functionalized polymer brushes. Scale bar, 800 μm. (C) Profilometry trace of features denoted by the black box in (B). (D) XPS of patterns containing S₂G₂A₂-functionalized PETT/EGDMA brushes and PETT-EGDMA brush features (“control”), which were printed without S₂G₂A₂. (E) ToF-SIMS of (bottom) areas of the surface without polymer and (top) S₂G₂A₂-functionalized PETT/EGDMA brushes. Red arrows at m/z = 113.04 [C₅H₅O₃⁺], m/z = 115.05 [C₅H₇O₃]⁺, and m/z = 147.05 [C₅H₇O₃]⁺. Scale bar, 100 μm. (F) Binding of SCR073 to S₂G₂A₂-functionalized PETT/EGDMA brushes. NF (orange) and height (blue) of the features in white box (inset). Scale bar, 50 μm. (G) Binding of SCR073 to control, where PETT/EGDMA brush features were printed without S₂G₂A₂. NF (orange) and height (blue) of the features in white box (inset). Scale bar, 50 μm. (H) NF versus height for binding of SCR073 (100 μM) to S₂G₂A₂-functionalized PETT/EGDMA brushes and to control. (I) (i) Binding of SNA to S₂G₂A₂-functionalized PETT/EGDMA brushes. Scale bar, 200 μm. (ii) Binding of SCR007 (25 μM) followed by SNA (100 nM) to S₂G₂A₂-functionalized PETT/EGDMA brushes. Scale bar, 200 μm. (iii) Binding of SCR007 (150 μM), followed by SNA (100 nM) solution to S₂G₂A₂-functionalized PETT/EGDMA brushes. Scale bar, 200 μm. (iv) Binding of SNA (100 nM) and SCR007 (150 μM) solution to S₂G₂A₂-functionalized PETT/EGDMA brushes. Scale bar, 200 μm. (v) Binding of ECA (25 μM) to S₂G₂A₂-functionalized PETT/EGDMA brushes. Scale bar, 200 μm. (J) % Intensity normalized to SNA versus height for binding of SNA (100 nM) in [(I), (i)], SCR007 (25 μM) + SNA (100 nM) in [(I), (ii)], simultaneous addition of SCR007 (150 μM) + SNA (100 nM) in [(I), (iii)], SNA (100 nM) + SCR007 (150 μM) in [(I), (iv)], and ECA in [(I), (v)].