Characterization of the Interaction between the Matrix Protein of Vesicular Stomatitis Virus and the Immunoproteasome Subunit LMP2

Abstract

The matrix protein (M) of vesicular stomatitis virus (VSV) is involved in virus assembly, budding, gene regulation, and cellular pathogenesis. Using a yeast two-hybrid system, the M globular domain was shown to interact with LMP2, a catalytic subunit of the immunoproteasome (which replaces the standard proteasome catalytic subunit PSMB6). The interaction was validated by coimmunoprecipitation of M and LMP2 in VSV-infected cells. The sites of interaction were characterized. A single mutation of M (I96A) which significantly impairs the interaction between M and LMP2 was identified. We also show that M preferentially binds to the inactive precursor of LMP2 (bearing an N-terminal propeptide which is cleaved upon LMP2 maturation). Furthermore, taking advantage of a sequence alignment between LMP2 and its proteasome homolog, PSMB6 (which does not bind to M), we identified a mutation (L45R) in the S1 pocket where the protein substrate binds prior to cleavage and a second one (D17A) of a conserved residue essential for the catalytic activity, resulting in a reduction of the level of binding to M. The combination of both mutations abolishes the interaction. Taken together, our data indicate that M binds to LMP2 before its incorporation into the immunoproteasome. As the immunoproteasome promotes the generation of major histocompatibility complex (MHC) class I-compatible peptides, a feature which favors the recognition and the elimination of infected cells by CD8 T cells, we suggest that M, by interfering with the immunoproteasome assembly, has evolved a mechanism that allows infected cells to escape detection and elimination by the immune system.

Importance: The immunoproteasome promotes the generation of MHC class I-compatible peptides, a feature which favors the recognition and the elimination of infected cells by CD8 T cells. Here, we report on the association of vesicular stomatitis virus (VSV) matrix protein (M) with LMP2, one of the immunoproteasome-specific catalytic subunits. M preferentially binds to the LMP2 inactive precursor. The M-binding site on LMP2 is facing inwards in the immunoproteasome and is therefore not accessible to M after its assembly. Hence, M binds to LMP2 before its incorporation into the immunoproteasome. We suggest that VSV M, by interfering with the immunoproteasome assembly, has evolved a mechanism that allows infected cells to escape detection and elimination by the immune system. Modulating this M-induced immunoproteasome impairment might be relevant in order to optimize VSV for oncolytic virotherapy.

FIG 1

The matrix protein of VSV interacts with LMP2. (A) cDNAs encoding pre-LMP2, mature LMP2, and the 20-amino-acid N-terminal propeptide were constructed in pGAD and used in Y2H assays against a pLEX VSV M construct. The empty pGAD (AD) or pLEX (LEX BD) vector was used as a negative control. The interaction was evaluated by quantification of β-galactosidase activity in liquid yeast cultures as described in the Materials and Methods section. Bars represent standard deviations of the mean from three independent experiments performed in triplicate. *, P < 0.05 compared to the control. (B) Coimmunoprecipitation of M and LMP2. N2A cells were transfected or not transfected (Non TF) for 24 h with plasmids carrying pre-LMP2-myc, myc-pre-LMP2, myc-mature-LMP2, or myc-dystonin as a negative control and then infected with VSV (4 h; MOI, 3) before preparation of cell lysates. Following immunoprecipitation with an anti-M antibody or anti-myc antibody 9E10, cell extracts (inputs) and immune complexes (immunoprecipitates [IP]) were separated by SDS-PAGE and analyzed by Western blotting (WB) using anti-myc antibody 9E10 or anti-M antibody to reveal the presence of the different myc-LMP2 and myc-dystonin constructs and M. Non Inf, noninfected. (C) N2A cells were transfected with plasmids carrying myc-tagged versions of VSV M, CHAV M, SVCV M, and LMP2-HA, followed by immunoprecipitation with an anti-HA antibody. The immunoprecipitates were analyzed by Western blotting using a myc-specific antibody to reveal the presence of VSV M.

FIG 2

The matrix protein of VSV does not interact with PSMB6. (A) cDNAs encoding pre-LMP2 as well as pre-PSMB6 and mature PSMB6 were constructed in pGAD and used in Y2H assays against a pLEX VSV M construct. The empty pGAD (AD) or pLEX (LEX BD) vector was used as a negative control. The interaction was evaluated by quantification of β-galactosidase activity in liquid yeast cultures (see the Materials and Methods section). Bars represent standard deviations of the means from three independent experiments performed in triplicate. (B) Coimmunoprecipitation of M and PSMB6. N2A cells were transfected or not transfected (Non TF) for 24 h with plasmids carrying myc-pre-LMP2 (positive control), myc-pre-PSMB6, myc-mature-PSMB6, or myc-dystonin as a negative control and then infected with VSV (4 h; MOI, 3) before preparation of cell lysates. Following immunoprecipitation with an anti-M antibody or anti-myc antibody 9E10, cell extracts (inputs) and immune complexes (immunoprecipitates [IP]) were separated by SDS-PAGE and analyzed by Western blotting using anti-myc antibody 9E10 or M-specific antibody to reveal the presence of the different myc-LMP2, myc-PSMB6, and myc-dystonin constructs and M.

FIG 3

Mapping of M-protein-binding domain with pre-LMP2. (A) Interaction between the flexible domain (amino acids 1 to 54) and the globular domain (amino acids 55 to 229) of M protein expressed in LexA and pre-LMP2 constructed in pGAD assayed by the Y2H assay. wt, wild type. (B) Surface representation of the crystal structure of VSV M (35). All the surface residues which were mutated are either in red or in yellow (residue I96). (C) The interaction between mutant M I96A and pre-LMP2 was assayed by the Y2H assay and quantification of β-galactosidase activity in liquid yeast cultures. Bars represent standard deviations of the means from three independent experiments performed in triplicate, *, P < 0.05 compared to the control. (D) N2A cells were transfected or not with plasmids carrying M, M I96A, pre-LMP2-HA, and HA-pre-LMP2 for 16 h before preparation of cell lysates. Following immunoprecipitation with an anti-HA antibody, cell extracts (inputs) and immune complexes (immunoprecipitates [IP]) were separated by SDS-PAGE and analyzed by Western blotting using anti-myc antibody 9E10 to reveal the presence of M and M I96A.

FIG 4

M binds specifically to pre-LMP2. HeLa cells were stimulated with gamma interferon for 8 h before transfection for 16 h with pmyc-M or pmyc-M I96A, followed by immunoprecipitation with an LMP2- or M-specific antibody. The immunoprecipitates were analyzed by Western blotting using an LMP2- or myc-specific antibody to reveal the presence of LMP2 or VSV M, respectively.

FIG 5

Mapping of the LMP2 binding site with M. (A) Sequence alignment of the first 60 residues of mature LMP2 and mature PSMB6. (B) The interactions of the pre-LMP2 T1A, D17A, K33A, S129A, V20T, L45R, V20T/L45R, and D17A/L45R mutants with M were assessed by the Y2H assay and quantification of β-galactosidase activity in liquid yeast cultures. Bars represent standard deviations of the means from three independent experiments performed in triplicate, *, P < 0.05 compared to the control. (C) N2A cells were transfected or not for 24 h with plasmids carrying myc-pre-LMP2 and myc-pre-LMP2 D17A, myc-pre-LMP2 L45R, or myc-pre-LMP2 D17A/L45R and infected or not with VSV for 4 h before preparation of cell lysates. Following immunoprecipitation with an anti-M antibody, cell extracts (inputs) and immune complexes (immunoprecipitates [IP]) were separated by SDS-PAGE and analyzed by Western blotting using anti-myc antibody 9E10 or an M-specific antibody to reveal the presence of LMP2 and M, respectively.

FIG 6

Growth kinetics of VSV on cells endogenously expressing LMP2, LMP7, and MECL-1. (A) B8 cells expressing LMP2 or the empty vector and B27M.2 cells expressing LMP2, LMP7, and MECL-1 were infected with VSV at MOIs of 3 and 0.3. Samples were harvested at 1, 4, 6, 8, 10, and 12 h postinfection, and the titer of progeny virus on BHK-21 cells was determined. Bars represent standard deviations of the means from three independent experiments. (B) Infected cell lysates were separated by SDS-PAGE and analyzed by Western blotting using anti-VSV M, anti-VSV G, and antitubulin antibodies. (C) PANC-1 cells constitutively expressing high levels of endogenous immunoproteasome were infected with VSV (MOI, 3) for the indicated times. Total cell lysates were examined by Western blotting with antibodies specific for the PSMA3 and PSMB6 subunits of the proteasome, LMP2, and M. hpi, hours postinfection.

FIG 7

LMP2 residues which are important for the interaction with VSV M are buried within the immunoproteasome. Ribbon representation of the crystal structure of the mouse immunoproteasome (41). (Inset) Magnification of the M-binding region seen from within the immunoproteasome. Red, LMP2; blue, residues T1, D17, and L45. These residues face inwards in the immunoproteasome.