Heparan sulfate proteoglycans remodel SARS-CoV-2 spike conformation to allow integrin interaction and infection of endothelial cells

Abstract

SARS-CoV-2 infects ACE2-negative primary HL-mECs through the interaction of an RGD motif, included in all spike proteins, up to the Omicron BA.1 subvariant, with α_vβ₃ integrin. Following its entry, SARS-CoV-2 remodels ECs phenotype and promotes angiogenesis in the absence of productive viral replication. Moreover, lack of spike/α_vβ₃ interaction, occurring in Omicron BA.5 which contains the D405N mutation in the RGD motif, inhibits HL-mECs infection and dysfunction. It is worth noting that anti-spike antibodies do not impact SARS-CoV-2 entry into HL-mECs. This data highlights the fact that i) the RGD motif is not exposed in the entire spike protein and ii) the need of a cofactor favoring spike/α_vβ₃ interaction. HSPGs are used by different viruses as receptors and coreceptors for their entry into host cells. Here, we use different approaches to scrutinize the role exerted by HSPGs in favoring SARS-CoV-2 infection of ECs. We highlight HSPGs as key molecules responsible for RGD exposure allowing its binding to the α_vβ₃ integrin as the first step toward viral entry by endocytosis. Indeed, SPR analysis showed lack of spike/α_vβ₃ interaction in the absence of heparin. This data was further corroborated by immunofluorescence and infectivity assays. Interestingly, the use of Heparinase III or sodium chlorate counteracts the release of proangiogenic molecules and inhibits signaling pathways induced by SARS-CoV-2 infection. Thus, HSPGs may represent a target for preventing SARS-CoV-2 infection of ECs and EC dysfunction-related COVID-19 severity.

Keywords: FAK-Src and Erk signaling pathway; SARS-CoV-2; angiogenesis; heparan sulphates; integrins; von Willebrand factor.

Figure 1

Heparin mediates SARS-CoV-2 Omicron BA.1 interaction with α_vβ₃ integrin. (A) BA.1 (blue line) and XBB.1.5 (purple line) RBDs or BA.1 (red line) and XBB.1.5 (black line) spike proteins were injected onto α_vβ₃ surface at a single concentration. (B) Sensogram overlay showing the binding of increasing amounts of BA.1 (range from 6.25 to 100 nM, left panel) and XBB.1.5 (range from 1.56 to 25 nM, right panel) spike proteins to heparin surface. The response, in resonance units (RU), was recorded as a function of time. Inset of left and right panels: equilibrium binding data generate the saturation curve and scatchard plot regression. (C) BA.1 (red line, left panel) and XBB.1.5 (black line, left panel) spikes were co-injected with heparin onto the α_vβ₃ surface. Antibody against α_vβ₃ (black line, right panel) and soluble heparin alone (green line, right panel) were injected as control.

Figure 2

HSPGs mediate SARS-CoV-2 Omicron BA.1 interaction with α_vβ₃ integrin on HL-mECs. HL-mECs were incubated with recombinant Omicron SARS-CoV-2 BA.1 (A, left panel) or XBB.1.5 (B, left panel) spike proteins (rBA.1 and rXBB.1.5, respectively) for 2 h at 4°C, then washed and fixed with 4% PFA in PBS. For staining, cells were incubated for 1 h at 4°C with a human serum containing IgG to SARS-CoV-2 or anti-α_vβ₃ antibody followed by Alexa Fluor 488-conjugated anti-human IgG or Alexa Fluor 594-conjugated anti-mouse IgG, respectively. Nuclei were counterstained with DAPI. Images display SARS-CoV-2 spike signal in green, α_vβ₃ signal in red and cell nuclei in blue. HL-mECs were pretreated with sodium chlorate, incubated with rBA.1 (A, middle panel) and rXBB.1.5 (B, middle panel) and decorated as described above. After all, pretreated sodium chlorate-HL-mECs were incubated with a mix solution of rBA.1 (A, right panel) or rXBB.1.5 (B, right panel) spike proteins and soluble heparin (preincubation at 37°C for 1 h) and decorated as above. Scale bar, 20 µm.

Figure 3

HSPGs are indispensable to mediate Omicron BA.1-HL-mECs infection. HL-mECs were pretreated with Heparinase III and sodium chlorate for 1 h and 48 h at 37°C, respectively. After treatment, the cells were infected with Omicron BA.1 and Omicron XBB.1.5 at a MOI of 1 for 1 h at 37°C, then washed and cultured until day 1 p.i. (A) The graph shows quantification of SARS-CoV-2 genomes at the intracellular level by qRT-PCR. At least three replicates were performed. Data are representative of two independent experiments with similar results. Statistical analysis was performed by one-way ANOVA (**p = 0.0023). (B) HL-mECs were fixed, permeabilized, and saturated with BSA. For staining, cells were incubated for 1 h at 37°C with a human serum containing IgG to SARS-CoV-2 and followed by Alexa Fluor 488-conjugated anti-human IgG. Nuclei were counterstained with DAPI. Images display SARS-CoV-2 signal in green and cell nuclei in blue (scale bar, 20 µm). (C) Spike-related fluorescent area was quantified in 20-30 Omicron BA.1-infected cells pretreated or not with sodium chlorate or Heparinase III and subtracted from not-infected cells fluorescence signal. Decrease of spike-related fluorescence in pretreated cells with sodium chlorate or Heparinase III Omicron BA.1 infected-cells was related to Omicron BA.1 signal and was expressed as fold change. Statistical analysis was performed by unpaired t-test (****p < 0.0001).

Figure 4

HSPGs inhibitors-treated HL-mECs do not acquire an angiogenic phenotype and do not release angiogenic factors upon Omicron BA.1 infection. HL-mECs were not-infected (NI) or infected with SARS-CoV-2 belonging to Omicron BA.1 or XBB.1.5 sublineages at MOI 1, for 1 h at 37°C and then washed and cultured until day 3 p.i. When indicated HL-mECs were pretreated with sodium chlorate (50 mM) or Heparinase III (5 mU/ml). (A) Sprouting of spheroids generated with NI, Omicron BA.1, and HSPGs inhibitors pretreated-HL-mECs. Pictures are representative of one out of three independent experiments with similar results (scale bar, 10 μm). FGF-2 was used as a positive control. (B) Values reported in the graph are the mean ± SD of one representative experiment out of three independent experiments with similar results performed in triplicate. Statistical analysis was performed by one-way ANOVA and Bonferroni’s post-test was used to compare data (****p < 0.0001). (C) Sodium chlorate and Heparinase III pretreated- or not HL-mECs, were infected with SARS-CoV-2 Omicron BA.1 lineage at MOI 1, for 1 h at 37°C and then washed and cultured until day 3 p.i. Clarified supernatants were evaluated for the presence of angiogenic molecules by a human proteome array. The results are expressed as mean values ± SD of duplicates given as fold increase as compared to NI cells. Data are representative of one out of two independent experiments with similar results.

Figure 5

SARS-CoV-2 induces vWF accumulation in WPBs and subsequently degradation in ECs. HL-mECs were nucleofected with a mCherry-vWF-expressing plasmid and 24 h after nucleofection, cells were infected with Omicron BA.1 virus and analyzed at different time points (1, 3, 6 and 24 h). (A, left panel) The images display vWF signals in red and cell nuclei in blue (scale bar, 10 μm). Red-positive punctate structures were counted in order to quantify the levels of WPBs. (A, right panel) Values reported for vWF positive structures are the mean ± SD of 3 independent experiments with similar results. (B) Amount of vWF in SARS-CoV-2 infected HL-mECs supernatants are the mean ± SD of 3 independent experiments with similar results. Statistical analysis was performed by one-way ANOVA, and the Bonferroni post-test was used to compare data (**p = 0.022, ***p = 0.0008). (C) Infected HL-mECs pretreated or not with HSPGs inhibitors were analyzed at 3 h p.i. Values reported for vWF positive structures are the mean ± SD of 3 independent experiments with similar results. Statistical analysis was performed by one-way ANOVA, and the Bonferroni post-test was used to compare data (***p = 0.0008, ****p < 0.0001).

Figure 6

SARS-CoV-2 Omicron BA.1 induces FAK/Src and ERK signaling after α_vβ₃ interaction. HL-mECs were not infected (NI) or infected with Omicron SARS-CoV-2 belonging to BA.1 and only XBB.1.5 sublineages (Omicron BA.1 and Omicron XBB.1.5, respectively) at a MOI of 1 for 1 h at 37°C. After 1 h p.i., cell lysates were analyzed by western blotting with anti-pFAK, anti-pSRC and anti-pERK antibodies (A-C, respectively). Blots are representative of three independent experiments with similar results. Quantification was carried out by densitometric analysis and plotting of the pFAK, pSrc and pERK normalized on GAPDH, tSrc or tERK as indicated in the graph. Values reported are the means ± the SD of three independent experiments. Statistical analysis was performed by one-way ANOVA, and the Bonferroni post-test was used to compare data (*p = 0.05, **p = 0.022, ***p = 0.0008). When indicated HL-mECs were pretreated with a mAb against α_vβ₃ integrin (mAb anti- α_vβ₃) or HSPGs inhibitors and analyzed for FAK/Src and ERK phosphorylation status (D-F).

Figure 7

Phylogenetic and temporal analysis of SARS-CoV-2 sequences carrying D405D and its mutations. (A) Maximum likelihood phylogenetic tree showing the distribution of sequences with D405D (red) and its alternative substitutions: D405N (green), N405S (yellow), and S405D (blue). (B) Temporal distribution of SARS-CoV-2 sequences by substitution type.

Figure 8

Proposed mechanisms for SARS-CoV-2-induced angiogenesis on ACE2-negative HL-mECs. HSPGs mediate spike/α_vβ₃ interaction enhancing phosphorylation of FAK, Src and ERK, which mediates Ang-2 and vWF accumulation in WPBs of SARS-CoV-2-infected HL-mECs. Then, Ang-2 secretion and vWF degradation induces infected-HL-mECs toward an angiogenic phenotype.