Tyrosine phosphorylation of bovine herpesvirus 1 tegument protein VP22 correlates with the incorporation of VP22 into virions

Abstract

Tyrosine phosphorylation has been shown to play a role in the replication of several herpesviruses. In this report, we demonstrate that bovine herpesvirus 1 infection triggered tyrosine phosphorylation of proteins with molecular masses similar to those of phosphorylated viral structural proteins. One of the tyrosine-phosphorylated viral structural proteins was the tegument protein VP22. A tyrosine 38-to-phenylalanine mutation totally abolished the phosphorylation of VP22 in transfected cells. However, construction of a VP22 tyrosine 38-to-phenylalanine mutant virus demonstrated that VP22 was still phosphorylated but that the phosphorylation site may change to the C terminus rather than be in the N terminus as in wild-type VP22. In addition, the loss of VP22 tyrosine phosphorylation correlated with reduced incorporation of VP22 compared to that of envelope glycoprotein D in the mutant viruses but not with the amount of VP22 produced during virus infection. Our data suggest that tyrosine phosphorylation of VP22 plays a role in virion assembly.

FIG. 1

Tyrosine-phosphorylated BHV-1 structural proteins. (A) BHV-1 proteins were subjected to SDS-PAGE, immunoblotted, and probed with an antiphosphotyrosine antibody (PY54). Molecular mass markers are noted at the left in kilodaltons. (B) Coomassie blue staining of the purified BHV-1 structural proteins using the same sample preparation as that used for panel A. The protein band isolated and identified as VP22 is marked by arrows.

FIG. 2

LC-mass spectroscopy of trypsin-digested VP22 illustrating residues 23 to 42 of the peptide sequence ENSLYDYESGSDDHVYEELR, which contains tyrosines 27, 29, and 38. A shift in mass from 2,420 to 2,500 m/z correlates with the phosphorylation of one of the three tyrosines in the peptide. amu, atomic mass units.

FIG. 3

Mapping of the VP22 phosphorylation sites. Bovine fibroblast cells were transfected with VP22 or its tyrosine-to-phenylalanine mutant constructs for 2 days. The His-tagged VP22 proteins were purified using Ni²⁺ columns. The partially purified proteins were analyzed by immunoblotting and probed with polyclonal antiphosphotyrosine (A) or anti-VP22 (B) antibody. Note that the tyrosine 38-to-phenylalanine mutation abolished the phosphorylation of VP22. (C) The blotted signals were quantified using NIH Image software, and the ratios of phosphorylation were normalized to the amount of VP22. Molecular masses are noted at the left in kilodaltons.

FIG. 4

VP22 is still tyrosine phosphorylated in vY38F mutant virus. VP22 and two degradation peptides are indicated by arrows. The N-terminal peptide (N-term) has a molecular mass of 22 kDa, and the C-terminal peptide has a molecular mass of 14 kDa. The purified BHV-1 and vY38F mutant viruses were analyzed by immunoblotting and probed with anti-VP22 antibody (A), anti-C-terminal-peptide antibody 114_235–258 (B), or polyclonal antiphosphotyrosine antibody (C). The two lanes of vY38F represent two isolates. Note that the N terminus was phosphorylated in wild-type virus but that the C terminus was phosphorylated in vY38F. The ∼60- and 50-kDa bands observed in panel B may represent multimers of VP22 and the C-terminal peptide or nonspecific bands. Molecular masses are noted at the left in kilodaltons.

FIG. 5

The loss of tyrosine phosphorylation correlates with the decreased incorporation of VP22 into virions but not in VP22 production from infected cells. Purified viruses were analyzed by immunoblotting and probed with polyclonal antiphosphotyrosine (A) or anti-VP22 and anti-gD (B) antibodies. Note the gradually decreased phosphorylation and incorporation of VP22 in vY38F and vYNull mutant viruses. The shift in VP22 in the vYNull lane likely represents the non-tyrosine-phosphorylated form of VP22. (C) The ratio of incorporation of VP22 normalized to the amount of gD was plotted using NIH Image software. (D) MDBK cells infected with BHV-1, vY38, vYNull, or vdUL49 virus or mock infected for 24 h. Cell lysates were collected, and SDS-PAGE was performed. Proteins were analyzed by Western blotting using anti-gD and anti-VP22 antibodies simultaneously. Note in panel D that, regardless of tyrosine mutations (vY38 or vYNull), the amounts of VP22 in the infected cell lysates were similar; in contrast, the amounts of VP22 in the virions (B and C) were reduced depending on the loss of tyrosines or phosphorylated tyrosines, suggesting the inability of the virus to incorporate tyrosine mutant VP22. Molecular masses are noted at the left of the gels in kilodaltons.

FIG. 6

The loss of tyrosine phosphorylation does not affect the subcellular localization of VP22 or mutant virus replication in vitro. MDBK cells were infected at an MOI of 0.01 with BHV-1 (A) or vYNull (B) for 18 h and then fixed with 4% paraformaldehyde, labeled with anti-VP22 antibody, and analyzed by indirect immunofluorescence microscopy. Both forms of VP22 localized in cell nuclei. (C) Levels of virus growth in vitro were compared using a single-step growth curve. The dephosphorylation of VP22 resulted in only a slight decrease of virus titer.

FIG. 7

Comparison of levels of tyrosine phosphorylation in infected and uninfected cells. Equal amounts of protein from total cell lysates of mock-infected and virus-infected MDBK cells were analyzed by SDS-PAGE and Coomassie blue staining (A) or immunoblotted and probed with polyclonal antiphosphotyrosine antibody (B). Note that, in infected cells (B), tyrosine phosphorylation increased and that the phosphorylated protein bands have patterns and molecular masses similar to those of the purified BHV-1 phosphorylated structural proteins (C). Molecular mass markers are indicated at the left in kilodaltons.