Rubella virus capsid associates with host cell protein p32 and localizes to mitochondria

Abstract

Togavirus nucleocapsids have a characteristic icosahedral structure and are composed of multiple copies of a capsid protein complexed with genomic RNA. The assembly of rubella virus nucleocapsids is unique among togaviruses in that the process occurs late in virus assembly and in association with intracellular membranes. The goal of this study was to identify host cell proteins which may be involved in regulating rubella virus nucleocapsid assembly through their interactions with the capsid protein. Capsid was used as bait to screen a CV1 cDNA library using the yeast two-hybrid system. One protein that interacted strongly with capsid was p32, a cellular protein which is known to interact with other viral proteins. The interaction between capsid and p32 was confirmed using a number of different in vitro and in vivo methods, and the site of interaction between these two proteins was shown to be at the mitochondria. Interestingly, overexpression of the rubella virus structural proteins resulted in clustering of the mitochondria in the perinuclear region. The p32-binding site in capsid is a potentially phosphorylated region that overlaps the viral RNA-binding domain of capsid. Our results are consistent with the possibility that the interaction of p32 with capsid plays a role in the regulation of nucleocapsid assembly and/or virus-host interactions.

FIG. 1

Capsid binds to p32 in vitro. Sepharose beads coated with GST (lanes 2 and 6), GST-p32 (lanes 3 and 7), or glucose oxidase (GOD) (lanes 4 and 8) were mixed with ³⁵S-labeled capsid (lanes 2 to 4) or GFP (lanes 6 to 8). The beads were washed, and bound proteins were eluted and visualized by SDS-PAGE and fluorography. Five percent of the capsid and GFP in vitro translation reaction mixtures were loaded in lanes 1 and 5, respectively.

FIG. 2

Capsid and p32 associate in vivo. (A) COS cells were transfected (Tfxn) with expression vectors encoding the RV structural proteins (24S) and p32 (lanes 1 to 3), the RV envelope proteins (E2-E1) and p32 (lanes 4 to 6), or 24S alone (lanes 7 to 9) or were mock transfected (lanes 10 to 12). At 24 h posttransfection the cells were biosynthetically labeled with [³⁵S]methionine-cysteine, lysed and then subjected to immunoprecipitation with preimmune serum (PI) (lanes 1, 4, 7, and 10), or antibodies (Ab) specific for capsid (lanes 2, 5, 8, and 11) or p32 (lanes 3, 6, 9, and 12). Immune complexes were subjected to SDS-PAGE and fluorography. (B) Vero cells were infected or mock infected with RV at a multiplicity of infection of 1. At 48 h postinfection, the cells were biosynthetically labeled with [³⁵S]methionine-cysteine. Immunoprecipitations were performed using lysates from infected cells (lanes 1 to 4) and mock-infected cells (lanes 5 to 8) as described above, using antibodies specific for capsid (lanes 1 and 5), p32 (lanes 2 and 6), E1 (lanes 3 and 7), or preimmune serum (lanes 4 and 8). The positions of E1, capsid, and p32 are indicated to the left of the gels. E2 which coprecipitates with E1 (lane 3) is indicated with an arrowhead. The asterisk indicates an unknown protein that comigrates with E2 in lanes 1, 2, and 4.

FIG. 3

Capsid colocalizes with p32 at the mitochondria. Vero cells were transfected with expression vectors encoding the RV structural proteins (24S) and p32 (A to F) or 24S alone (G and H). Cells were incubated with antibodies to capsid (A and G), p32 (B, D, F, and H) or E1 (E). For samples shown in panels E and F, mitochondria were labeled prior to fixation with Mitotracker Red (C) followed by staining with anti-p32. Primary antibodies were detected with Texas Red-conjugated donkey anti-mouse IgG and FITC-conjugated donkey anti-rabbit IgG. The Texas Red channel is shown on the left (A, C, E, and G), and the FITC channel is shown on the right (B, D, F, and H). Areas of colocalization are shown by arrows, whereas the arrowhead indicates a perinuclear pool of capsid which does not overlap with p32. Bar, 20 μm.

FIG. 4

Capsid is associated with the cytoplasmic side of the mitochondria. Vero cells were transfected with expression vectors encoding the RV structural proteins (24S) and p32, and fixed with 3% paraformaldehyde. (A to D) The plasma membranes were permeabilized with digitonin (A and B). To permeabilize intracellular membranes, cells (C and D) were treated with Triton X-100 (+TX100). Cells were double labeled with antibodies specific for capsid (A and C) and p32 (B and D). Primary antibodies were detected with Texas Red-conjugated donkey anti-mouse IgG and FITC-conjugated donkey anti-rabbit IgG. The Texas Red channel is shown on the left (A and C), and the FITC channel is shown on the right (B and D). Bar, 20 μm. (E to G) Confocal images of Triton X-100-treated cells. The arrowheads in panel E indicate capsid staining at the periphery of mitochondria which does not coincide with p32 staining (G). Panel F is a merge of the images shown in panels E and G.

FIG. 5

The p32-binding site is located in the amino-terminal region of capsid. AH109 yeast cells were transformed with pGAD-p32 and GAL4 DNA binding domain plasmids encoding full-length capsid (pGBKT7-capsid) or portions of capsid. The capsid constructs are named according to the amino acids in capsid which they encode. For example, 1–45 is pGBKT7 plus the coding region for amino acids 1 to 45 of capsid. The transformants were plated onto medium lacking tryptophan and leucine (-Trp-Leu) or medium lacking tryptophan, leucine, histidine, and adenine (-Trp-Leu-His-Ade). The positive control (+) is a Clontech system control that utilizes two strongly interacting proteins, p53 (in pGBKT7) and simian virus 40 large T antigen (in pGAD). As a negative control (−), AH109 was transformed with two plasmids which do not interact, pGBKT7-p53 and pGAD-ABP280. Growth on the -Trp-Leu-His-Ade plates is indicative of a strong interaction between the two proteins being tested.