Enterovirus A71 Promotes Exosome Secretion by the Nonstructural Protein 3A Interacting with Rab27a

Abstract

Exosomes are small membrane-bound vesicles which are intraluminal vesicles (ILVs) secreted to the extracellular space after multivesicular bodies (MVBs) fuse with the plasma membrane. Although it is known that exosomes play a multitude of roles during viral infection, the mechanism that regulates their secretion during viral infection is unknown. Here, we found that enterovirus A71 (EV-A71) infection increased exosome secretion both in vivo and in vitro. Importantly, the expression of nonstructural protein 3A was sufficient to promote exosome secretion, while a mutation affecting the amino acid 18 position abrogated this effect, without changing the size of exosomes in vivo or in vitro. Transmission electron microscopy (TEM) analysis revealed that 3A decreases the number of MVBs and ILVs in vivo and in vitro, which suggested 3A may boost the fusion between MVBs and the plasma membrane. Furthermore, we demonstrated that an interaction between 3A and the small GTPase protein, Rab27a, protected Rab27a from ubiquitination, resulted in increasing exosome release. Data indicated a novel mechanism by which EV-A71 3A modifies exosome secretion during viral infection. IMPORTANCE Research has shown that viral infection impacts exosome secretion, but its regulation mechanisms remain poorly understood. Nonstructural protein 3A of EV-A71 interacts with many host factors and is involved in the remodeling of cellular membranes. In this investigation, we applied exogenous expression of 3A protein for exploring its regulation on exosome secretion and utilized immunoprecipitation combined with proteomics approaches to identify 3A-interacting factors. Our results demonstrate that 3A protein upregulates the release of the exosomes and that the 3A mutant strain of EV-A71 induce less exosome release compared with the EV-A71 wild type. Viral 3A protein interacts with the host factor Rab27a to prevent it from being ubiquitinated, which in turn improves exosome secretion both in vitro and in vivo. EV-A71 3A protein is a novel viral factor in the control of exosome production.

Keywords: 3A; Rab27a; enterovirus; exosome secretion; nonstructural protein.

FIG 1

EV-A71 infection increased exosome secretion in vitro and in vivo. Exosomes were isolated and purified by serial centrifugation from supernatants of EV-A71- and or mocked-infected HT-29 cells (MOI, 1; postinfection time, 24 h). (A) Representative TEM images of exosomes. Scale bar = 100 nm. (B) Exosomal and cellular marker proteins detected by WB analysis. (C) Quantification of exosomal protein levels using ImageJ software. (D) The total exosomal proteins as detected by BCA assay. (E) Quantitation of exosomes using the Exo ELISA kit sensitive for CD63. (F and G) Neonatal ICR mice were injected with 10⁶ TCID₅₀/mL of virus by intraperitoneal injection (n = 20) and sacrificed on day 5. Total exosomal proteins and abundance from brain, intestine, lung, and heart tissues were analyzed by BCA assay (F) or CD63 ELISA (G). Figures represent three independent experiments containing 3 samples in each group. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

FIG 2

Exogenous 3A-enhanced exosome secretion. HeLa cells were transfected by a pcDNA3.1 construct carrying the full-length 3A cDNA (pcDNA3.1-3A) or the empty pcDNA3.1 vector (pcDNA3.1), and exosomes were isolated from the supernatants. (A) Total protein content of the exosomes detected by BCA assay. (B) Detection of exosomal markers using WB analysis. (C) The exosomal protein levels were quantified using the ImageJ software. (D and E) NTA analysis was carried out to compare the abundance (D) and the size (E) of exosomes. Results were derived from three independent experiments with 3 samples in each group. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 3

Effect of EV-A71 3A mutations on exosomes production. (A) Schematic diagram of the 3A protein highlighting the position of amino acids where mutations were introduced. (B) The level of RNA expression of the WT or mutated viruses in infected HeLa cells was determined by qRT-PCR analysis. (C) The total protein content of exosomes was detected by BCA assay. (D) The abundance of exosomes derived from WT or mutated virus-infected cells was detected using the ExoELISA kit. (E) WB analysis of exosomal markers in the supernatants of cells infected by the WT virus or the 3A-P18A mutant strain. (F) Quantitation of exosomal proteins with ImageJ software analysis of the image in panel E. (G and H) The exosome abundance (G) and exosomal diameter (H) were determined by NTA analysis. All results are from three independent experiments containing 3 replicates in each group. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

FIG 4

Exogenous 3A decreased the number of MVBs and the number of ILVs inside them. (A) Representative electron microscopic (EM) images of pcDNA3.1-3A and control pcDNA3.1 transfectants, showing MVBs and ILVs inside them. Scale bar = 1 μm. Enlarged sections (panels on the right) are magnified 2.6-fold. (B and C) The numbers of MVBs (B) and ILVs (C) were counted in 5 randomly selected images. (D) Representative confocal microscopic IF images of two cells stained with CD63 antibody. Red corresponds to CD63; blue corresponds to 4′,6-diamidino-2-phenylindole (DAPI) staining of the nucleus. Scale bar = 10 μm. (E) WB analysis of intracellular CD63 abundance in the pcDNA3.1- and pcDNA3.1-3A-transfected cells. (F) Graphic representation of CD63 abundance after ImageJ analysis. Expression was normalized using the abundance of the endogenous protein, GAPDH. The data shown represent three independent experiments with 3 samples in each group. *, P < 0.05; ***, P < 0.001.

FIG 5

3A interacts with Rab27a directly. (A) HeLa cells were stably transfected with the pcDNA3.1-flag-3A vector, encoding a flag-tagged 3A protein. Cells were fixed and stained with anti-3A (green) and anti-flag (red) antibodies. Confocal laser scanning microscopy was performed to detect the fluorescence signal. Bar = 20 μm. (B) Samples were subjected to WB analysis using antibodies against 3A, flag peptide, and GAPDH. (C) Biological process proteins interacting with 3A were involved according to their GO classification. (D) Cells stably transfected with the pcDNA3.1-flag-3A vector were fixed and stained with anti-Rab27a (red) and anti-flag (green) antibodies. Confocal laser scanning microscopy was performed to image fluorescence signal. (E) Cells lysate was immunoprecipitated with an anti-3A antibody and immunoblotted with an anti-Rab27a antibody. These results were from 3 independent experiments with 3 samples in each group.

FIG 6

The interaction of 3A with Rab27a prevented the ubiquitination of Rab27a. (A and B) qRT-PCR (A) and WB (B) analysis of Rab27a expression in pcDNA3.1-3A cells after transfection with si-Rab27a or si-RNA-NC. (C and D) Total protein concentration in exosomes (C) and the abundance of exosomes (D) in pcDNA3.1 and pcDNA3.1-3A cells transfected with si-Rab27a or si-RNA-NC. (E) Rab27a protein levels in cell lysates using WB analysis. Normalization of Rab27a to endogenous GAPDH protein levels was carried out using ImageJ software. (F) Rab27a mRNA levels were measured by qRT-PCR analysis. The endogenous gene GAPDH was utilized as a housekeeping gene to normalize Rab27a expression. (G) WB analysis of Rab27a levels in cells treated with CHX for 12 h. In some experiments the medium was also supplemented with the proteasome inhibitor Lac or the lysosome inhibitor CHD. NC, nontreated control cells. (H) Normalization of Rab27a with endogenous protein GAPDH was carried out with ImageJ software. (I) pcDNA3.1 and pcDNA3.1-3A cells treated with Lac for 12 h before IP using anti-Rab27a or a control IgG antibody. The ubiquitinated Rab27a level was detected by WB using an anti-ubiquitin antibody. All immunoblot data represent 3 independent experiments. *, P < 0.05; ***, P < 0.001; ns, not significant.

FIG 7

3A-P18 is important in the facilitation of exosome release via a Rab27a-dependent mechanism. (A to D) Exosomes were isolated by serial centrifugation from brain tissues of animals infected with wild-type EV-A71 virus (WT) or a mutant strain carrying an amino acid substitution at position 18 of the 3A protein (3A-MT). (A) Total protein concentration of purified exosomes detected by BCA assay. (B) The abundance of exosomes as detected by the ExoELISA kit. (C and D) NTA was performed to quantify exosomes in three independent experiments, with 10 samples in each treatment group. (C) The number of exosome particles. (D) Comparison of exosomal diameter between the two treatment groups. Particle numbers and percentage distribution are shown from representative NTA traces. (E) Representative TEM images of MVBs and ILVs. Scale bar = 1 μm. Enlarged sections (panels on the right) are enlarged 2.6-fold. (F and G) The number of MVBs and ILVs in MVBs. MVBs and ILVs in 5 typical randomly selected images were counted. (H) WB analysis of Rab27a protein expression in the brain, intestine, lung, and heart of 3A-MT mice and WT mice. (I) Rab27a protein expression normalized to the abundance of the GAPDH internal control. (J) IF staining showing Rab27a expression in the brain of 3A-MT mice and WT mice. (K) qRT-PCR analysis of Rab27a mRNA abundance in various tissues of 3A-MT and WT mice. These results represent three independent sets of experiments with 3 samples in each group. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

FIG 8

The mutation of 3A causing reduced exosome release interfered with the infectivity of EV-A71 in vitro and in vivo. (A) The segmented PCR product of viral RNA in exosomes was identified by agarose gel electrophoresis. The EV-A71 virion was used as a positive control. (B) RD cells were treated with exosomes derived from mock-infected cells (Exo-mock) or Exo-EV-A71 for 24 h. CPE in RD cells caused by the infection of Exo-EV-A71 was detected. (C and D) RD cells were treated with infectious rescued viral strains EV-A71-WT, 3A-P18A, 3A-L24A, and 3A-K80A for 24 h. (C) CPE in RD cells caused by the infection of different viral strains. (D) Viral RNA abundance level in RD cells was detected by qPCR. (E) qRT-PCR analysis of the abundance of viral RNA in the brain, intestinal, lung, and heart tissues of 3A-MT mice infected with the WT virus or the strain carrying the P18A mutated version of the 3A protein (3A-MT). (F) The weight of mice recorded on the fifth day after infection. Each group consisted of 5 animals. *, P < 0.05; **, P < 0.01; ns, not significant.