The α-dystroglycan N-terminus is a broad-spectrum antiviral agent against SARS-CoV-2 and enveloped viruses

Abstract

The COVID-19 pandemic has shown the need to develop effective therapeutics in preparedness for further epidemics of virus infections that pose a significant threat to human health. As a natural compound antiviral candidate, we focused on α-dystroglycan, a highly glycosylated basement membrane protein that links the extracellular matrix to the intracellular cytoskeleton. Here we show that the N-terminal fragment of α-dystroglycan (α-DGN), as produced in E. coli in the absence of post-translational modifications, blocks infection of SARS-CoV-2 in cell culture, human primary gut organoids and the lungs of transgenic mice expressing the human receptor angiotensin I-converting enzyme 2 (hACE2). Prophylactic and therapeutic administration of α-DGN reduced SARS-CoV-2 lung titres and protected the mice from respiratory symptoms and death. Recombinant α-DGN also blocked infection of a wide range of enveloped viruses including the four Dengue virus serotypes, influenza A virus, respiratory syncytial virus, tick-borne encephalitis virus, but not human adenovirus, a non-enveloped virus in vitro. This study establishes soluble recombinant α-DGN as a broad-band, natural compound candidate therapeutic against enveloped viruses.

Keywords: Broad-range antiviral; Coronaviruses; Enveloped viruses; Extracellular matrix; SARS-CoV-2; α-dystroglycan.

Fig. 1. α-DGN and its inhibitory activity against pseudotyped human coronaviruses.

(A) Left: schematic representation of the dystrophin-glycoprotein complex (DGC). This multiprotein complex, of which dystroglycan (DG) is a central element, anchors the extracellular matrix (ECM) to actin and other components of the cytoskeleton. Extracellular α-DG binds different ECM proteins, such as laminin, while transmembrane β-DG binds the actin cytoskeleton via direct interaction with dystrophin. Also depicted are other intracellular molecules associated with the DGC. The N-terminal of α-DG (α-DGN), highlighted in the dashed circle, is liberated in circulation following cleavage by furin. α-Ct: α-DG C-terminal, α-Nt: α-DG N-terminal. Right: cartoon representation of the 3D-structure of α-DGN (PDB: 1U2C). Shown are the two domains (N-term Ig-like, in cyan and C-term S6-like, in magenta and enclosed into the full circle) and the position of the flexible undefined loop (FUL) connecting them is indicated. (B) Alignment of the amino acid sequences of mα-DGN (Uniprot: Q62165, top) and hα-DGN (Uniprot: Q14118, bottom), showing a degree of identity of ~93 % (~97 % similarity). The sequences of the Ig-like and S6-like domains are highlighted in cyan and magenta boxes, respectively. Amino acid conservation code: (*) identity, (.) strong similarity, (.) weak similarity. (C) SDS-PAGE of the recombinant α-DGN proteins (indicated over each lane) used in this study, as final products of the purification procedure. (D) Recombinant α-DGN unfolding curves measured by intrinsic tryptophan fluorescence spectroscopy, as fitted to a two-state linear extrapolation model. The mid-point of the transitions (Cm), 2.8M for mα-DGN, 2.7M for hα-DGN and 1.8M for the shortened version mS6, are indicative of stable polypeptides. (E-H) Dose-response curves for mα-DGN, hα-DGN or mS6 pre-treated Caco-2 cells infected with VSV pseudotyped particles expressing the spike protein from E) SARS-CoV-2, F) SARS-CoV, G) MERS-CoV and H) HCoV-229E. Infection was quantified by measuring GFP-positive cells and data are expressed as % inhibition normalized to an unrelated protein control (mean ± SEM). IC₅₀ values (from n = 3 independent experiments) were calculated using nonlinear regression calculations. (I) Representative immunofluorescence images of infected cells treated with either mα-DGN, hα-DGN, mS6 (10 μM), or vehicle control, as indicated (infected, GFP-positive cells in green and Hoechst for nuclei in blue). Scale bar: 200 μm.

Fig. 2. α-DGN blocks SARS-CoV-2 in epithelial cells and human primary gut organoids.

α-DGN inhibits infection of cells from different SARS-CoV-2 variants. VeroE6-TMPRSS2 cells were pre-treated with 10 μM mα-DGN before infection with SARS-CoV-2 (A) Wuhan, (B) Gamma, (C) Delta, (D) Omicron BA.1 or (E) Omicron BA.2 (MOI = 0.1). Infection was quantified using viral N protein staining (n = 3 independent experiments). Bars are means + SEM. Scale bar, 200 μm. p-values were determined using an unpaired t-test. p*<0.1. (F) Representative immunostaining images of control or mα-DGN treated VeroE6-TMPRSS2 cells infected with the indicated variants of SARS-CoV-2 (viral N protein in green and Hoechst for nuclei in blue). (G) mα-DGN inhibits SARS-CoV-2 binding to cells. HeLa-ACE2 cells were pre-incubated with 10 μM mα-DGN or vehicle control, after which SARS-CoV-2 was added for 60 min on ice, and cells were fixed. SARS-CoV-2 binding per cell after treatment with mα-DGN (blue) or control (green) was quantified for >100 cells per condition. Significance was determined using a Mann-Whitney test (n = 3 independent experiments), p****<0.0001. (H) Representative immunostaining images of extracellular SARS-CoV-2 particles on HeLa-ACE2 cells after pre-incubation with vehicle control or mα-DGN. Cells were stained with Hoechst for nuclei (blue), wheat germ agglutinin (WGA) for cell outline (grey) and for SARS-S protein (green). Scale bars, 10 μm. (I) mα-DGN blocks SARS-CoV-2 infection of human primary gut organoids. Organoids were treated with 1 or 10 μM of mα-DGN or buffer control and infected with SARS-CoV-2 at MOI =1 for 48 h before fixation. The graph shows the percentage of infected (double stranded RNA-positive) organoids (mean + SD, n = 4). Significance was determined using an unpaired t-test (n = 3 independent experiments), p****<0.0001 (J) Representative immunostaining images of human gut organoids treated with indicated concentrations of mα-DGN or vehicle (PBS) control and infected with SARS-CoV-2 (dsRNA in magenta and nuclei are stained with Hoechst in cyan). Images represent the max Z-projection of 5 optical slices (out of 180 for each image stack) showing the mid-section of the organoids. Scale bars, 30 μm.

Fig. 3. α-DGN inhibits SARS-CoV-2 infection in K18-hACE2 mice.

(A) mα-DGN or buffer control was administered to K18-hACE2 mice at 7.5 μg/mouse or 0.75 μg/mouse at the same time as the lethal dose of the challenging virus (SARS-CoV-2 SG12-B) via the intranasal (IN) route. Mice were observed daily for (B) survival (n = 5 per group) and (C) weight loss (n = 5 per group). (D) Lung viral loads as determined at end point (n = 4 per group). Statistical significance was determined using a Mann Whitney test, p*<0.05. ns: not significant. (E) Buffer control, mα-DGN or hα-DGN at 7.5 μg/mouse were administered via the IN route 2 h prior to virus challenge (SARS-CoV-2 SG12-L) and then daily for 3 days. Mice were then observed daily for (F) survival (n = 5 per group) and (G) weight loss (n = 5 per group). (H) Lung viral loads as determined at endpoint (n = 5 per group). Statistical significance was determined using a Mann Whitney test, p** <0.01.

Fig. 4. mα-DGN inhibits a broad range of enveloped viruses.

(A-F) Inhibitory activity of mα-DGN in infection assays against (A) IAV, (B) RSV, (C) SFV, (D) TBEV, (E) VSV and (F) hAdV5. Infection was analysed by immunostaining and results are represented as percentage of infected cells. (G-J) Inhibitory activity of mα-DGN against the 4 Dengue Virus serotypes (G) DENV1, (H) DENV2, (I) DENV3 and (J) DENV4 in cell infection assays. DENV RNA was analysed by qRT-PCR. DENV copies were normalized to the internal control protein TBP. Values are means + SD. Statistical significance was determined using an unpaired t-test. ns = not significant, p* <0.1, p** <0.01, p*** <0.001.