A Single Mutation in the VP1 Gene of Enterovirus 71 Enhances Viral Binding to Heparan Sulfate and Impairs Viral Pathogenicity in Mice

Figure 1. Generation and in vitro characterization of the L929-adapted EV71-GZCII virus. (A) EV71-GZCII virus was passaged 30 times in L929 cells to generate the GZCII-P30 strain. (B) L929 and Neuro2a cells were infected with GZCII or GZCII-P30 (MOI = 0.1) or mock infected for 24 h. Scale bar: 200 μm. (C) Multistep growth curve of GZCII and GZCII-P30 in L929 and Neuro2a (MOI = 1) cells (D). Viral titers were determined by measuring TCID50.

Figure 2. Adaptive mutations in the GZCII-P30 genome revealed by Next-Generation Sequencing were mapped onto the EV71 capsid. (A) Genomic RNA of the GZCII and GZCII-P30 viruses was sequenced by NGS, and the mutation frequency and distribution were analyzed according to the reference sequence (cGZCII). The open symbols represent nonsynonymous mutations, and the closed symbols are synonymous mutations. Red: GZCII-P30; Blue: GZCII. (B) A 3D structure of the GZCII-P30 virus was predicted by SWISS-MODEL homology modeling based on PDB:3VBF.1. VP1-E92K and E98K reside at the 5-fold axis, while VP2-I149V is located in the E-F loop on VP2. Blue: VP1; red, VP2; green, VP3; yellow, VP4.

Figure 3. Replication kinetics of the recombinant GZCII viruses in cells. (A) Growth curves of GZCII variants in L929 or (B) Neuro2a cells (Multiplicity of infection = 0.1). (C) Vero, RD, HeLa, and SK-N-SH cells infected with cGZCII or cGZCII-98K virus (MOI = 0.1) were harvested at 24 h post-infection for virus titration in Vero cells (***: p < 0.001). (D) L929 cells in 96-well plates were infected with GZCII mutants (MOI = 1 or 3), fixed, and stained using crystal violet to assess cell cytotoxicity.

Figure 4. Viral replication in SCARB2-knockout L929 cells. (A) Construction of a SCARB2-knockout cell line. Both alleles of L929-SCARB2KO cells were genotyped by Sanger sequencing to confirm SCARB2 gene disruption. (B) Expression of SCARB2 in wild-type L929 or L929-SCARB2KO cells was detected by western blotting. GZCII mutants were used to infect L929 and L929-SCARB2KO cells (MOI = 0.1). Viral titer (C) at 24 h post-infection were detected by TCID50 method. (**: p < 0.01; ***: p < 0.001).

Figure 5. Characterization of the binding efficiency and sensitivity of EV71-GZCII variants to heparin sulfate (HS). (A) Viruses were incubated with detached L929 cells (A) or an HS-sepharose column (B) to allow attachment, and the ratio of adsorbed to input virus was determined by qPCR. Viruses pretreated with soluble HS sodium were used to infect L929 (MOI = 0.1), and the viral RNA copy numbers were determined at 24 h post-infection (C). The percentage of viral RNA change is shown in (D). (***: p < 0.001).

Figure 6. Replication of EV71-GZCII variants in wildtype and EXT1-knockout L929 cells. Cloned L929-EXT1KO cells were genotyped to confirm gene disruption or analyzed by flow cytometry to detect HS expression (A). Normal mouse IgM was used as an isotype control. Replication and cytotoxicity of GZCII mutants in HS-depleted cells. L929 and L929-EXT1KO cells (B) were infected with GZCII variants (MOI = 0.1). Viral RNA copy numbers and percent inhibition were determined at 24 h post-infection by qRT-PCR. Viral RNA copy numbers and percent inhibition by EXT1KO were determined (C). Images of infected wild-type or EXT1 knockout cells were acquired at 24 h post-infection (D) to observe the CPE induced by GZCII variants. Scale bar: 200 μm. (**: p < 0.01; ***: p < 0.001).

Figure 7. Virulence and pathogenesis of GZCII variants in 2-week-old BALB/c mice. Mice were infected with (A,B) GZCII or GZCII-P30 virus (104 or 105 TCID50 units, respectively, n = 6) or (C,D) cGZCII-98K, cGZCII-149V92K, and cGZCII-VKK viruses (105 TCID50 units, n = 5 or 6). PBS was used as a control. Mouse survival and disease development of each group were observed and scored until 15 dpi. The following scoring was used: healthy, 0 points; lethargy, inactivity, and limb weakness, 1 point; less exercise and limb paralysis, 2 points; quadriplegic and moribund, 3 points; and death, 4 points. (E) Pathological changes in different tissues from mice infected with cGZCII-98K, cGZCII-98K, and cGZCII-149V98K viruses (2 × 105 TCID50) or PBS (as control) using the i.p. route. Mouse brains, lungs, livers, and hearts were harvested and subjected to histopathological examination at 3 dpi. Scale bar: 200 μm.

Figure 8. Expression of HS and SCARB2 proteins in different tissues of mouse. The liver, lung, intestine, and brain tissues were harvested from BALB/c mice. Sections incubated anti-HS and anti-SCARB2 antibody, or not incubated with primary antibodies (reaction control), were stained with fluorescent secondary antibodies to detect the expression of both receptors. Red: HS signal. Green: SCARB2 signal. Cell nuclei were stained with DAPI (blue). Scale bar: 50 μm.

Figure 9. Binding efficiency and replication of GZCII variants in vivo. (A) Homogenates of mouse brain, kidneys, liver, or intestines (n = 3) were incubated with ~106 copies of GZCII variants for attachment, and the binding efficiency was determined by qPCR. (B) To monitor the viral load in mice, blood was collected at 1 dpi or 2 dpi, and mouse heart (C), kidney (D), brain (E), liver (F), lung (G), and muscle (H) tissues were harvested at 1 dpi or 3 dpi (n = 3). Viral RNA copy number in 1 μg of total RNA was determined by qPCR (*: p < 0.05; **: p < 0.01; ***: p < 0.001).