Herpes Simplex Virus Requires VP11/12 To Induce Phosphorylation of the Activation Loop Tyrosine (Y394) of the Src Family Kinase Lck in T Lymphocytes

Abstract

Herpes simplex virus (HSV) tegument proteins are released into the cytoplasm during viral entry and hence are among the first viral proteins encountered by an infected cell. Despite the implied importance of these proteins in the evasion of host defenses, the function of some, like virion protein 11/12 (VP11/12), have not been clearly defined. Previously, we reported that VP11/12 is strongly tyrosine phosphorylated during the infection of lymphocytes but not in fibroblasts or an epithelial cell line (G. Zahariadis, M. J. Wagner, R. C. Doepker, J. M. Maciejko, C. M. Crider, K. R. Jerome, and J. R. Smiley, J. Virol. 82:6098-6108, 2008). We also showed that tyrosine phosphorylation depends in part on the activity of the lymphocyte-specific Src family kinase (SFK) Lck in Jurkat T cells. These data suggested that VP11/12 is a substrate of Lck and that Lck is activated during HSV infection. Here, we show that HSV infection markedly increases the fraction of Lck phosphorylated on its activation loop tyrosine (Y394), a feature characteristic of activated Lck. A previous report implicated the immediate-early protein ICP0 and the viral serine/threonine kinases US3 and UL13 in the induction of a similar activated phenotype of SFKs other than Lck in fibroblasts and suggested that ICP0 interacts directly with SFKs through their SH3 domain. However, we were unable to detect an interaction between ICP0 and Lck in T lymphocytes, and we show that ICP0, US3, and UL13 are not strictly required for Lck activation. In contrast, VP11/12 interacted with Lck or Lck signaling complexes and was strictly required for Lck activation during HSV infection. Thus, VP11/12 likely modulates host cell signaling pathways for the benefit of the virus.

The virions of all members of the Herpesviridae, including herpes simplex virus type 1 (HSV-1), contain a layer of proteins known as the tegument that resides between the capsid and the envelope. Tegument proteins are released into the cytoplasm of the newly infected cell upon virus entry, where they are able to rapidly manipulate host cell functions for the benefit of the virus. Examples include the HSV-1 tegument proteins VP16, which promotes the rapid transcription of the viral immediate-early genes (reviewed in reference 63), and virion host shutoff protein, which contributes to the early inhibition of host protein synthesis through the degradation of RNA (reviewed in reference 50).

The function of HSV-1 virion protein 11/12 (VP11/12), one of the most abundant tegument proteins, currently is unknown (68). Transient transfection assays suggest that VP11/12 modulates the transcriptional activity of VP16 to enhance the expression of immediate-early genes (67); consistent with this hypothesis, VP11/12 and VP16 have been shown to physically interact (18, 59). However, viral mutants bearing null mutations in the UL46 open reading frame encoding VP11/12 are viable in cultured cells and display no detectable defect in immediate-early gene expression (67, 69), suggesting that the primary function of VP11/12 still needs to be discovered.

The subcellular localization of VP11/12 has been examined in efforts to gain insight into its functions during infection. The bulk of VP11/12 localizes to perinuclear cytoplasmic foci, which are thought to be sites of virion assembly (18, 33, 61). Subcellular fractionation experiments indicate that the VP11/12 present within the infected cell associates with membranes, while the VP11/12 that is packaged into mature virions does not (31), suggesting that membrane association is reversible and occurs before the final stages of virion assembly. There also is evidence that VP11/12 delivered by the infecting virion localizes to the plasma membrane shortly following viral entry (61). Thus, VP11/12 may function at membranes, either upon viral entry or at later stages prior to virion assembly.

We previously reported that VP11/12 is tyrosine phosphorylated in a cell-type-specific fashion (67). Strong tyrosine phosphorylation was seen during the infection of several lymphocyte cell lines, but phosphorylation was greatly reduced during the infection of human embryonic lung (HEL) fibroblasts or an epithelial cell line (67). This cell type specificity implied that a cell-type-specific protein kinase is involved; consistent with that idea, we found that the tyrosine phosphorylation of VP11/12 was greatly reduced in a T-cell line that lacks the lymphocyte-specific Src family kinase (SFK) Lck (67). Despite the well-established and critical role of Lck in T-cell receptor (TCR) signaling, the Lck-mediated phosphorylation of VP11/12 appeared to be independent of this pathway, as the tyrosine phosphorylation of VP11/12 occurred in the absence of TCR ligation (67). Moreover, HSV-1 infection is known to block TCR signaling (48), and VP11/12 is not required for this effect (67).

Lck and other SFKs are membrane-associated nonreceptor tyrosine kinases that have been studied extensively for their roles in cancer, phagocytosis, cell adhesion, and signaling through growth factor receptors and immune receptors (reviewed in references 5, 7, and 23). On a biochemical level, SFK activity is enhanced as the enzyme switches from a closed inactive conformation to an open active configuration. This conformational change is regulated by the phosphorylation state of a C-terminal inhibitory tyrosine residue and by the binding of ligands to the Src homology 2 (SH2) and SH3 domains of the SFK (reviewed for Src in reference 42). Crystal structures of inactive SFKs indicate that the closed conformation is characterized by intramolecular interactions between the SH2 domain and the C-terminal inhibitory phosphotyrosine motif and between the SH3 domain and the linker region connecting the SH2 domain and the catalytic domain (47, 65). The dephosphorylation of the inhibitory tyrosine suffices to trigger the open conformation and initiate enzyme activation (2, 8, 20, 64). Kinase activity also can be triggered by signaling adaptors and SFK substrates that bind to the SFK SH2 and/or SH3 domains and disrupt their interactions with the intramolecular inhibitory motifs (reviewed in references 7 and 10). The affinity of SH2 domains for phosphotyrosine depends on the amino acid sequence surrounding the phosphotyrosine residue (51); hence, the high-affinity motifs found in a variety of signaling proteins are able to outcompete the relatively low-affinity SFK C-terminal inhibitory tyrosine for binding to the SH2 domain, leading to enzyme activation (7). Similarly, high-affinity ligands of the SH3 domain are able to disrupt the intramolecular SH3-SH2-kinase linker interaction (10).

In the open and active conformation, the key regulatory tyrosine residue in the SFK activation loop is exposed for phosphorylation (47, 64, 65), and this modification, while symptomatic of activation, also significantly enhances kinase activity by reinforcing the open conformation (9, 66). The phosphorylation of the activating tyrosine has been shown to play a dominant role in enzyme activation in vitro, as it blocks the negative effect of phosphorylating the C-terminal inhibitory tyrosine (55). The activation loop tyrosine also appears to be dominant in vivo, as documented by the observations that SFKs activated by hydrogen peroxide or the SH3 ligand HIV-1 Nef display the enhanced phosphorylation of the activating tyrosine and no measurable decrease in phosphorylation at the inhibitory tyrosine (14, 21).

Many viruses activate SFKs to initiate a variety of cellular signaling processes that benefit the virus. Polyoma middle T antigen stimulates Src to trigger cell cycle entry and DNA replication, thereby providing the environment needed for viral genome replication (reviewed in references 16 and 45). Vaccinia virus produces cell-associated progeny virus particles that initiate outside-in signaling to activate Src. Src then phosphorylates the vaccinia protein A36R, triggering the formation of actin tails that facilitate virus spread by thrusting the cell-associated virus particles toward neighboring uninfected cells (reviewed in reference 30). In the gammaherpesvirinae, genes next to the terminal repeats of the genome encode a family of proteins called terminal membrane proteins (TMPs). TMPs physically interact with SFKs, activate SFKs, and are SFK substrates (reviewed in reference 6). TMPs orchestrate signaling pathways that can regulate lytic or latent infection and alter cell survival and/or immune function in infected lymphocytes (6). For example, the TMP LMP2A of Epstein-Barr virus interacts with the SFK Lyn in infected B cells and promotes its degradation (reviewed in reference 36). This impairs signaling through the B-cell receptor (28), suppressing viral reactivation from latency (11, 28, 29).

HSV-1 also has been reported to alter SFK activity (22). Src isolated from infected fibroblasts was shown to have slightly increased activity in vitro (22). Consistent with activation, SFKs from infected cells demonstrated a small but distinct decrease in electrophoretic mobility through sodium dodecyl sulfate-polyacrylamide gel electrophoresis and an increase in phosphorylation at the activating tyrosine (22). The change in electrophoretic mobility was influenced by the viral proteins US3 and UL13, while the increase in activating phosphorylation required the viral immediate-early protein ICP0 (22). ICP0 also was shown to interact with the isolated SH3 domains of the SFKs Src, Yes, Fyn, and Fgr, suggesting that it directly influences SFK activity (22).

Our previous report demonstrated that the efficient tyrosine phosphorylation of HSV-1 VP11/12 in Jurkat T cells requires the lymphocyte-specific SFK Lck (67), suggesting that VP11/12 is an Lck substrate. Furthermore, the tyrosine phosphorylation of VP11/12 occurred in the absence of TCR ligation, raising the possibility that Lck is activated by HSV infection. We therefore asked if the HSV-1 infection of T lymphocytes activates Lck, and if so, whether VP11/12 is required. Our results implicate VP11/12 in the regulation of SFK activity in T cells.

MATERIALS AND METHODS

Cells and viruses.

Human embryonic lung (HEL) fibroblasts and Vero cells were obtained from the ATCC, while Jurkat 6.8 and JCAM1.6 cells (13) were donated by H. L. Ostergaard (University of Alberta). These cells were maintained as described previously (67). CD8⁺ T cells (donated by R. C. Bleakley [University of Alberta]) were derived from human peripheral blood lymphocytes isolated as described previously (3) and then selected based on CD8 expression using the RosetteSep human CD8⁺ T-cell enrichment cocktail (Stem Cell Technologies). The CD8⁺ T cells were maintained as described previously for unsorted human peripheral blood lymphocytes (3). Human U2OS osteosarcoma cells were maintained in Dulbecco's modified eagle's medium (DMEM; Gibco) supplemented with 10% fetal bovine serum (FBS), 100 mM sodium pyruvate, and 100 U/ml penicillin-streptomycin. The HSV-1 strain KOS ICP0 mutant n212 was a kind gift from Priscilla Schaffer. HSV-1 strain F and the F-derived mutants R7356 (UL13⁻), R7041 (US3⁻), and R7306 (US3 repair virus) were generously provided by Bernard Roizman. HSV-1 strain KOS37 was derived from the KOS37 Bacmid (12) provided by David Leib. The KOS37-derived mutants ΔUL46galK (VP11/12⁻), ΔUL46 (VP11/12⁻), and RUL46 (VP11/12 repaired) have been described previously (67). Stocks of n212 and KOS were prepared from infected U20S cells. All other virus stocks were prepared from infected Vero cells.

Infection and preparation of cell extracts.

For experiments involving n212, Jurkat cells were infected for 1 h at a high multiplicity of infection (MOI), as indicated in the figure legends. For all other experiments, Jurkat and JCAM cells were infected in serum-free RPMI medium for 1 h at an MOI of 10. CD8⁺ T cells were infected via exposure to HSV-1-infected HEL fibroblasts. HEL fibroblasts were infected at an MOI of 10 in complete medium. At 12 h postinfection, the HEL fibroblasts were incubated with CD8⁺ T cells in the T-cell medium for 8 h. Cell extracts were prepared at 4°C by a 20-min incubation in lysis buffer (1% Nonidet P-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM EGTA, 1 mM NaF, 1 mM Na₃VO₄, 50 mM Tris-HCl, pH 7.4) supplemented with a complete protease inhibitor cocktail (Roche). Infected Jurkat and JCAM cells were harvested at 7 to 9 h postinfection, while CD8⁺ T cells and/or HEL fibroblasts were harvested at 20 h after the infection of the fibroblasts.

Immunoprecipitation.

All immunoprecipitation steps were carried out at 4°C. Cell lysates were cleared of debris by centrifugation at 16,000 × g for 15 min and then were preabsorbed to protein G-Sepharose during a 30-min incubation. The appropriate amount of antibody was incubated with protein G-Sepharose for 2 h in phosphate-buffered saline (PBS), and the resulting antibody-protein G conjugates were washed twice with lysis buffer. Lysate then was incubated with the protein G-Sepharose-conjugated antibody and 0.1% bovine serum albumin (BSA) for 1 to 2 h. Precipitates were washed twice with lysis buffer.

Antibodies.

Immunoprecipitation was performed using anti-Lck (3A5; 1 μg; Santa Cruz Biotechnology Inc.), anti-phospho-Src family (Tyr416) (20 μl; Cell Signaling), and anti-green fluorescent protein (GFP) (4 μl; kind gift of L. Berthiaume). Antibodies used for primary detection by Western blotting were directed against Lck (1/1,000), phospho-Src family (Tyr 416) (1/2,000), GFP (1/3,000), phospho-Lck (Tyr505) (1/1,000; Cell Signaling), HSV-1 ICP27 (no. 1113; 1/2,500; Virusys Corporation), HSV-1 VP16 (LP1; 1/30,000; provided by Tony Minson), ICP0 (no. 1112; 1/1,000; Goodwin Institute), and actin (1/1,000; Sigma). For the detection of the Lck protein via Western blotting following immunoprecipitation, rabbit immunoglobulin G TrueBlot (1/1000; eBioscience) and mouse TrueBlot Ultra (1/1,000; eBioscience) secondary antibodies were used. Other secondary antibodies included donkey anti-mouse IR800 (1/15,000; Rockland inc.), goat anti-rabbit Alexa Fluor680 (1/150,000; Invitrogen), goat anti-mouse Alexa Fluor680 (1/150,000; Invitrogen), goat anti-rabbit horseradish peroxidase (HRP) (1/3,000; Promega), goat anti-mouse HRP (1/5,000; Promega), and donkey anti-goat HRP (1/15,000; Jackson ImmunoResearch).

Western blotting.

Western blotting was performed as described previously (67), except that each lane contained the equivalent of 400,000 cells or 1.6 million cells for cell lysate and immunoprecipitation samples, respectively, and the membranes prepared for enhanced chemiluminescence detection were blocked by incubation with 4% BSA.

RESULTS

HSV infection enhances phosphorylation of the activation loop tyrosine residue of a 56-kDa SFK in Jurkat T cells.

As reviewed in the introduction, the kinase activity of Lck and other SFKs is regulated by the phosphorylation status of two key tyrosine residues (43). The dephosphorylation of the C-terminal inhibitory tyrosine (Y505 in Lck) partially activates Lck (2, 25), while the phosphorylation of the activation loop tyrosine (Y394 in Lck) stabilizes the active conformation and serves as a signature of the activated state irrespective of the phosphorylation status of Y505 (1, 9, 14, 25). We therefore examined the phosphorylation of these regulatory tyrosine residues during HSV-1 infection. Jurkat T cells were mock infected or were infected with HSV-1 strain F for 9 h. Cell lysates then were analyzed by Western blotting using antibodies that detect total Lck and Lck phosphorylated at the two regulatory tyrosine residues. We used an antibody specific for Lck phosphorylated at Y505 to evaluate the phosphorylation status of the C-terminal inhibitory tyrosine residue (anti-LckPY505) (Fig. 1A). An antibody specific for the phosphorylated activation loop tyrosine residue of Lck (Y394) was not available. Therefore, we used an antibody that detects any SFK phosphorylated at the activating tyrosine (phospho-Y416 for the prototypical SFK, Src) to examine this residue (anti-active SFK) (Fig. 1B).

FIG. 1.

HSV-1 infection enhances the phosphorylation of the activation loop tyrosine of a 56-/60-kDa SFK in Jurkat T cells. Jurkat T cells were mock infected or infected with HSV-1F for 9 h. Cell lysates then were analyzed by Western blotting with the indicated antibodies using two-channel infrared detection. (A) Western blot analysis using antibodies that detect Lck phosphorylated at the inhibitory tyrosine residue (Y505) and total Lck. (B) The infrared signal obtained with the anti-Lck phospho-Y505 antibody for panel A was divided by the signal obtained with the anti-Lck antibody, and the ratios were normalized to yield a value of 1 for the mock-infected sample. (C) Western blot using antibodies that detect SFKs phosphorylated on the activating tyrosine residue (the equivalent of phospho-Y416 of Src, anti-active SFK) and total Lck. (D) The infrared signal obtained with the anti-active SFK antibody for panel C was divided by the signal obtained with the anti-Lck antibody, and the ratios were normalized to yield a value of 1 for the mock-infected sample.

Two forms of Lck with apparent molecular masses of ca. 56 and 60 kDa were detected using the anti-Lck antibody in both infected and uninfected cells (Fig. 1A and B). Although HSV-1 infection had no significant effect on total Lck levels, in some, but not all, experiments the proportion of Lck displaying reduced mobility (ca. 60 kDa) was markedly increased in infected cells (Fig. 1A and B). A similar shift from ∼56 to ∼60 kDa has been described previously following T-cell activation through stimulation with mitogens, interleukin-2, antigen-presenting cells, or anti-TCR antibodies (15, 26, 57, 58). The precise modification(s) responsible for this change in apparent molecular size has not been defined, but it has been linked to a number of phosphorylation events, including the compulsory phosphorylation of two serine residues in the N-terminal region (15, 24, 52, 57, 58, 60, 62). Using two-channel infrared detection, we found a perfect overlap in the migration patterns of Lck and the proteins detected by the anti-phospho-Lck Y505 and the anti-active SFK antibodies, with an ∼56- and ∼60-kDa species reacting with each antibody. Thus, the mobility shift from ∼56 to 60 kDa is not caused by the phosphorylation of either of these tyrosine residues.

We quantified the 56- and 60-kDa signals obtained with all three antibodies to determine the effects of infection on the phosphorylation of the inhibitory and activating tyrosines (Fig. 1A to D). The ratio of phospho-Y505 Lck to total Lck did not change, indicating that HSV infection does not alter the fraction of Lck phosphorylated at the inhibitory Y505 (Fig. 1B). In contrast, the proportion of the 56-/60-kDa SFK phosphorylated on the activation loop tyrosine increased in HSV-1-infected samples compared to that for mock-infected samples (Fig. 1D). This increase was apparent when the active SFK signal was quantified relative to total Lck (Fig. 1D) or β-actin (data not shown). The increase was not readily detected until after 3 h postinfection, and it peaked at ca. 9 h (Fig. 2A), indicating that one or more newly synthesized viral proteins are required. Consistent with this interpretation, the effect was abolished when the infecting virus stock was inactivated by UV irradiation (Fig. 2B). Taken together, these data suggested that one or more SFKs that precisely comigrate with Lck display a more-than fourfold increase in the levels of the phosphorylation of the activation loop tyrosine during HSV-1 infection. However, additional experiments were required to positively identify Lck as the active 56-/60-kDa phospho-SFK detected in these experiments.

FIG. 2.

Time course of Lck Y394 phosphorylation. Jurkat cells were either mock infected or infected with HSV-1 F or UV-inactivated HSV-1F as described for Fig. 1. Cell lysates then were prepared at the indicated times (hours postinfection [hpi]). (A) Western blot analysis using the indicated antibodies. (B) The signals obtained with the anti-active SFK antibody were divided by those obtained with the anti-Lck antibody, and the ratios obtained were normalized to yield a value of 1 for the mock-infected sample.

HSV increases the fraction of Lck phosphorylated at Y394.

The data displayed in Fig. 1 demonstrate that HSV infection enhances the phosphorylation of the activation loop tyrosine of an SFK that comigrates with Lck. To directly address whether Lck represents some or all of this 56-/60-kDa phospho-SFK population, we examined the effects of HSV-1 infection on SFKs in the Lck-deficient Jurkat derivative JCAM1.6 (Fig. 3A). The active SFK detected in Jurkat cells was not present in JCAM1.6 cells. To verify comparable viral gene expression in both cell lines, we examined the accumulation of the immediate-early gene product ICP27 and the late gene product VP16. The robust expression of these proteins was observed following the infection of both Jurkat and JCAM1.6 cells, indicating that Lck is not required for viral protein synthesis. These data indicated that Lck is the main component of the 56-/60-kDa SFK population detected by the anti-active SFK antibody in Jurkat cells.

FIG. 3.

HSV enhances the phosphorylation of Lck residue Y394. Jurkat and JCAM1.6 cells were mock infected or infected with HSV-1F for 9 h, and cell extracts were prepared. (A) Western blot (WB) analysis using the indicated antibodies. (B) Western blot analysis of immunoprecipitates (IP) obtained with the anti-active SFK antibody. Lysate controls were analyzed directly. (C) Western blot analysis of immunoprecipitates obtained with the anti-Lck antibody. Lysate controls were analyzed directly.

To further corroborate that Lck undergoes phosphorylation at the activating tyrosine (Y394) during the HSV infection of Jurkat cells, we asked if Lck reacts with the anti-active SFK antibody. SFK members phosphorylated at the activating tyrosine were immunoprecipitated using the anti-active SFK antibody, and the immunoprecipitates then were analyzed by Western blotting using the anti-Lck antibody (Fig. 3B). The results demonstrated that Lck was precipitated by the anti-active SFK antibody. Conversely, the anti-Lck antibody precipitated an SFK that reacts with anti-active SFK antibody (Fig. 3C). Thus, Lck was phosphorylated at Y394 in all samples tested. Although these assays were not structured for quantitative analysis, the proportion of Lck that reacts with the anti-active SFK antibody appeared to be higher in HSV-1F-infected samples than in mock-infected samples (Fig. 3B and C). Taken together, these data indicate that HSV-1 infection triggers an increase in the fraction of Lck phosphorylated at the activation loop tyrosine, residue Y394.

ICP0 is not required for enhanced phosphorylation of Lck at Y394.

As described in the introduction, a previous report presented evidence that the HSV-1 immediate-early protein ICP0 is required for the virus-induced phosphorylation of the activating tyrosine residue of SFKs in HEL fibroblasts and in a human kidney cell line, HEK 293 cells (22). This conclusion was based on assays of an ICP0 null virus, R7910. However, cells infected with R7910 displayed severely reduced expression of the viral thymidine kinase (22), raising the possibility that the failure of this mutant to stimulate the phosphorylation of SFKs was due to the reduced expression of one or more viral proteins other than ICP0 rather than the loss of ICP0 per se. The defect in viral gene expression displayed by ICP0 null viruses can be overcome by using a sufficiently high MOI (44, 54). Therefore, we examined the effects of increasing MOI on the ability of an ICP0 null virus to enhance the phosphorylation of the activation loop tyrosine of Lck.

Jurkat cells were mock infected or were infected with the parental wild-type HSV-1 KOS and the ICP0 null mutant n212 at an MOI of 1, 10, and 100 (Fig. 4A). At all MOIs tested, n212 produced less of the immediate-early gene product ICP27 and the late gene product VP16 than the wild-type KOS virus. Similarly, the signal obtained with the anti-active SFK antibody was reduced in the n212-infected samples relative to that of KOS at each MOI. However, n212-infected samples displayed an MOI-dependent increase in both viral protein expression (ICP27 and VP16) and the anti-active SFK signal (Fig. 4B). These data demonstrate that the enhanced phosphorylation of Lck at Y394 does not depend strictly on the presence of ICP0 but rather correlates with the production of viral proteins during n212 infection.

FIG. 4.

ICP0 is not directly required for the enhanced phosphorylation of Lck Y394. Jurkat cells were infected with HSV-1 KOS and the ICP0 mutant n212 at the indicated MOI (PFU/cell), and cell lysates were prepared for Western blotting. (A) Western blot analysis using the indicated antibodies. (B) The signals obtained with the anti-active SFK antibody were divided by those obtained with anti-Lck, and the ratios were normalized to yield a value of 1 for the mock-infected sample.

UL13 may play a role in posttranslational modifications of Lck.

Like ICP0, the viral serine/threonine kinases US3 and UL13 have been implicated in SFK activation during HSV-1 infection (22). Specifically, these two proteins were shown to be necessary for the posttranslational modifications that result in an increase in the apparent molecular mass of SFKs in fibroblasts. We therefore examined HSV-1 mutants null for US3 or UL13 to assess the role of these proteins in HSV-1-induced posttranslational modifications of Lck in Jurkat T cells.

The UL13 (R7356) and US3 (R7041) null viruses were comparable to the parental HSV-1 F and the US3-repaired virus (R7306) in their ability to enhance the phosphorylation of the activating tyrosine of Lck, Y394 (Fig. 5A and B). The US3 null virus and its US3-repaired derivative also were indistinguishable from HSV-1F in their ability to enhance the fraction of Lck displaying reduced electrophoretic mobility (most readily visible using the anti-active SFK antibody) (Fig. 5A). The UL13 null virus R7356 also gave rise to a slowly migrating form of Lck, but the apparent molecular mass of this form was less than ∼60 kDa (Fig. 5A), implying that some of the posttranslational modifications that give rise to the ∼60-kDa form of Lck were impaired during this infection. However, the UL13 null virus also demonstrated the defective expression of the late viral protein VP16 (Fig. 5A), likely the consequence of the cell-type-dependent attenuation of infection that has been described previously for HSV UL13 null mutants (40). Further experimentation is required to determine whether the altered posttranslational modifications stem directly from the loss of UL13 or indirectly from the reduced expression of another viral protein. Overall, these data indicated that UL13 and US3 are not strictly required for the virus-induced posttranslational modifications that give rise to a high-molecular-weight form of phospho-Y394 Lck.

FIG. 5.

US3 and UL13 are not required for the enhanced phosphorylation of Lck Y394. Jurkat cells were infected with HSV-1 F, the UL13 null mutant R7356, the US3 null mutant R7041, and the US3-repaired virus R7306, and cell lysates were prepared. (A) Western blot analysis using the indicated antibodies. (B) The signals obtained with the anti-active SFK antibody were divided by those obtained with anti-Lck, and the ratios were normalized to yield a value of 1 for the mock-infected sample.

VP11/12 is necessary for enhanced phosphorylation of Lck residue Y394.

The tyrosine phosphorylation of HSV VP11/12 in T lymphocytes is dependent largely on Lck (67). Because some SFK substrates can activate the cognate kinase, we asked whether VP11/12 is necessary for the enhanced phosphorylation of Lck residue Y394 during HSV infection (Fig. 6). Jurkat cells infected with two mutants lacking VP11/12, ΔUL46galK and ΔUL46, displayed severely reduced levels of phospho-Y394 Lck compared to those of cells infected with the parental HSV-1 KOS37 or the VP11/12-repaired virus RUL46 (Fig. 6A and B). Indeed, the level of phospho-Y394 Lck during infection for a VP11/12 null virus was indistinguishable from that of mock-infected cells. Both mutants directed the accumulation of normal levels of ICP27 and VP16 (Fig. 6A), arguing that the defect does not stem from the reduced expression of viral proteins other than VP11/12. These data indicate that VP11/12 is necessary for the HSV-induced enhanced phosphorylation of Lck residue Y394 in Jurkat T cells.

FIG. 6.

VP11/12 is required for the enhanced phosphorylation of Lck Y394. Jurkat cells were infected with HSV-1 KOS37, the VP11/12 null mutants ΔUL46galK and ΔUL46, and the VP11/12-repaired virus RUL46. (A) Western blot analysis using the indicated antibodies. (B) The signals obtained with the anti-active SFK antibody were divided by those obtained with anti-Lck, and the ratios were normalized to yield a value of 1 for the mock-infected sample.

Jurkat cells are an immortalized CD4⁺ T line derived from an Epstein-Barr virus-negative T-cell leukemia (46). To determine if HSV infection enhances the phosphorylation of Lck residue Y394 in normal T cells, we examined CD8⁺ T cells isolated from human peripheral blood. Others have observed that primary human T cells are resistant to infection by cell-free HSV, and that the efficiency of infection can be enhanced by exposing the T cells to infected fibroblasts or by activating the T cells with mitogens (4, 19, 38). We therefore exposed CD8⁺ T cells to HEL fibroblasts that previously had been infected with HSV-1, using a 5:1 ratio of T cells to fibroblasts. The presence of SFKs phosphorylated at the activating tyrosine was examined in HEL fibroblasts alone and in CD8⁺ T-cell/HEL fibroblast cocultures. The anti-active SFK antibody displayed little or no reactivity with samples of mock- and HSV-1 KOS37-infected HEL fibroblasts (which lack Lck), indicating that signals obtained in cocultures are derived from SFKs specific to the T cells. In the samples containing CD8⁺ T cells, HSV enhanced the Y394 phosphorylation of Lck and reduced the electrophoretic mobility of phospho-Y394 Lck in a VP11/12-dependent manner (Fig. 7A and B).

FIG. 7.

VP11/12 is required for the enhanced phosphorylation of Lck Y394 in primary human CD8⁺ T cells. HEL fibroblasts were either mock infected or infected with HSV-1 KOS37, the VP11/12 null mutants ΔUL46galK and ΔUL46, and the VP11/12-repaired virus RUL46. Twelve hours later, human CD8⁺ T cells were added to some of the cultures at a 5:1 ratio. Cell lysates were prepared 8 h later and analyzed by Western blotting. (A) Western blot analysis using the indicated antibodies. CD8⁺ T cells indicates HEL cells overlaid with a fivefold excess of CD8⁺ T cells; HEL indicates HEL cells incubated in the absence of T cells. (B) The signals obtained with the anti-active SFK antibody in the samples containing CD8⁺ T cells were divided by those obtained with anti-Lck, and the ratios were normalized to yield a value of 1 for the mock-infected sample.

VP11/12 interacts with Lck in infected T cells.

The data presented in the preceding sections establish that VP11/12 is necessary for HSV-induced enhanced phosphorylation of the activation loop tyrosine of Lck, raising the possibility that VP11/12 interacts with Lck or Lck signaling complexes in vivo. To determine if this is the case, Jurkat T cells were infected with wild-type HSV-1 KOS or GHSV-UL46, a KOS-derived mutant virus expressing a GFP-tagged VP11/12 (61). Potential interactions between VP11/12-GFP and Lck then were assessed by immunoprecipitation/Western blot assays, using an antibody directed against GFP to precipitate and detect VP11/12-GFP. Detectable levels of Lck phosphorylated on Y394 coprecipitated with VP11/12-GFP (Fig. 8A). Conversely, VP11/12-GFP (Fig. 8B) was coprecipitated by the anti-Lck antibody. Taken together, these data indicate that VP11/12-GFP complexes with phospho-Y394 Lck in infected Jurkat T cells.

FIG. 8.

Lck interacts with GFP-tagged VP11/12 during HSV infection. Jurkat cells were mock infected or infected with HSV-1 KOS or GHSV-UL46, a KOS mutant encoding a VP11/12-GFP fusion protein. Cell extracts were used for immunoprecipitation (IP) and Western blot (WB) analysis. (A) Western blot analysis of immunoprecipitates obtained with an anti-GFP antibody. Lysate controls were analyzed directly. (B) Western blot analysis of immunoprecipitates obtained with the anti-Lck antibody. Lysate controls were analyzed directly.

A previous report suggested that ICP0 interacts with the SH3 domain of Src through a putative SH3-binding motif (22). We therefore asked if ICP0 detectably interacts with Lck during infection. Immunoprecipitates of Lck from KOS and GHSV-UL46-infected Jurkat T cells did not contain detectable amounts of ICP0 (Fig. 8B). Therefore, our data provide no indication that ICP0 interacts with either Lck or Lck signaling complexes in Jurkat T cells.

DISCUSSION

The results presented in this report show that HSV-1 infection enhances the phosphorylation of the activation loop tyrosine (Y394) of the lymphocyte-specific SFK Lck in T lymphocytes, and that the abundant tegument protein VP11/12 is required for this effect. The phosphorylation of Y394 is a diagnostic signature of the active conformation of Lck, as the SFK activation loop tyrosine is buried in the inactive conformation and hence cannot be phosphorylated (47, 64, 65). Moreover, the phosphorylation of this residue significantly enhances SFK activity (9) and overrides the inhibitory effects of phosphorylating the C-terminal inhibitory tyrosine Y505 (14, 21, 55). Thus, our results indicate that HSV-1 infection activates Lck in infected cells, even though no change in the phosphorylation state of Y505 is observed (Fig. 1A). However, we have not yet been able to detect enhanced activity in in vitro kinase assays of infected cell lysates (unpublished data), perhaps suggesting that only a subset of Lck is activated. Consistent with this interpretation, in many experiments the majority of Lck had an apparent molecular mass of 56 kDa, while phospho-Y394 Lck was distributed equally between the ∼56- and ∼60-kDa species (Fig. 4, 6, and 7). Alternatively, it is possible that all of the activated Lck is sequestered by VP11/12 (likely a substrate) and consequently is not able to act on exogenous substrates. In this context, we note that activation peaks at late times postinfection, and thus newly synthesized VP11/12 likely contributes to the effect.

Liang and Roizman have implicated ICP0, US3, and UL13 in the activation of SFKs other than Lck in non-lymphoid cells (22). In contrast, our data indicate that these proteins are not required for the activation of Lck, although UL13 does appear to contribute to the altered electrophoretic mobility of a subset of the activated molecules (Fig. 5A). However, the UL13 null mutant accumulated lower levels of VP16, indicating a defect in late gene expression. Therefore, further experiments are required to determine if the effects of deleting UL13 are due to the loss of UL13 function per se or stem from the reduced accumulation of one or more additional viral proteins.

In contrast to our findings with ICP0, US3, and UL13, the deletion of VP11/12 abrogated Lck activation but had no apparent effect on viral protein expression in T cells (Fig. 6). Moreover, VP11/12 null mutations do not affect viral replication in other cell types (69). The HSV-1 infection of Jurkat cells is abortive (data not shown), and hence the effects of VP11/12 in this cell type cannot be easily determined. These data imply that VP11/12 acts directly to enhance Lck activity. Consistently with this hypothesis, VP11/12 interacted with Lck or an Lck-signaling complex (Fig. 6). Further studies will be required to determine whether VP11/12 is sufficient to activate Lck in the absence of other viral proteins.

Future experimentation also is needed to determine if VP11/12 binds directly to Lck, and if so, which motifs in VP11/12 facilitate the interaction. We previously showed that VP11/12 is tyrosine phosphorylated in an Lck-dependent fashion, suggesting that it is a substrate of Lck (67). As noted in the introduction, SFK substrates can induce kinase activation through high-affinity binding to the SH2 and/or SH3 domain (reviewed in references 7 and 10). In this context, it is interesting that the Scansite algorithm (34) identifies two putative Lck SH2 binding motifs in VP11/12 that are conserved between HSV-1 and HSV-2 (located at Y613 and Y624 in VP11/12 of HSV-1 strain 17; data not shown). The motif at Y624 has the sequence YEEI, which corresponds to the peptide sequence preferred by the SH2 domains of Lck, Fgr, Fyn, and Src (51). Indeed, the affinity of the Lck SH2 domain for YEEI is so high that a mutant Lck with YEEI in the place of the native C-terminal inhibitory motif cannot be activated by the TCR (32). Thus, an attractive model is that the YEEI motif of VP11/12 binds the SH2 domain of Lck with high affinity, displacing the C-terminal inhibitory motif. Such a mechanism would be analogous to the SH2-YEEI interactions used by polyoma middle T antigen to activate Src signaling (7, 45).

Although our data suggest that VP11/12 binds and activates Lck, as noted above it seems likely that only a fraction of total cellular Lck is affected. Consistent with this view, several considerations suggest that VP11/12 does not globally alter all Lck-dependent signaling events in the infected cell. For example, although HSV infection remodels the TCR signaling pathway, the virus-induced alteration lies downstream of the Lck substrate Zap-70 (48), and Zap-70 activation requires TCR ligation even in HSV-infected cells (data not shown). Thus, VP11/12 does not trigger Zap-70 activation, and conversely, VP11/12 does not prevent Lck from participating in the first steps of TCR signaling. Furthermore, the HSV-induced TCR signaling blockade does not require VP11/12 (67). We therefore consider it likely that VP11/12 activates only a subset of Lck-mediated signaling pathways, perhaps by localizing to a subcellular location that is associated with components of a specific pathway, or by forming a complex with Lck and one or more downstream signaling protein(s).

Lck expression is largely restricted to lymphocytes (27). Thus, our data suggest a lymphocyte-specific function for VP11/12. Accumulating evidence indicates that HSV infects T lymphocytes and alters their function as an immune evasion strategy (reviewed in reference 17). HSV is able to infect primary T cells in vitro (4, 19, 37, 38, 41, 48) and also infects peripheral T cells in the intact human host (4). Infected T cells display diminished target cell killing, enhanced fratricide, and altered cytokine profiles (38, 39, 41, 48, 49), at least in part due to HSV-induced modifications to the TCR signaling pathway (48, 49). Although VP11/12 is not required for at least some of the HSV-induced changes in TCR signaling (67), it is possible that it modulates one or more additional signaling pathways that contribute to altered lymphocyte function.

It also is possible that VP11/12 exerts effects on Lck-dependent signaling pathways in cell types other than lymphocytes. Although Lck conventionally is considered a lymphocyte-specific kinase, immunocytochemistry has been used to suggest that Lck also is present in mouse and rat neurons (35, 56). However, the neural expression of Lck has not been widely confirmed in the literature, and the reported expression level in the mouse is 10-fold lower in brain than in the thymus (35); moreover, one report indicates that Lck cannot be detected in normal human brain (53). If Lck is in fact present in neurons, then VP11/12-Lck signaling could play a role in latency, reactivation, or the neural spread of HSV.

A key question when considering possible biological functions of VP11/12 is whether it is capable of similarly activating additional SFKs in lymphocytes and/or other cell types. We have reported previously that the tyrosine phosphorylation of VP11/12 can be detected in two B-cell lines; in addition, we occasionally observed a much weaker phosphotyrosine signal in human embryonic lung fibroblasts and Vero cells (an epithelial cell line) (67). None of these cells likely express Lck, suggesting that VP11/12 binds and activates other SFKs. Thus, it is possible that VP11/12 contributes to the previously reported ability of HSV-1 to activate SFKs other than Lck in non-lymphoid cells (22). Further studies are required to test this possibility and define which SFKs are activated. The possibility is intriguing, because SFKs have been linked to a variety of biological processes. Thus, it is possible that VP11/12 regulates many aspects of the virus-host interaction in a fashion that varies by cell type.

Overall, our data show that HSV induces an active phenotype in at least a fraction of the Lck present in T cells. Activation requires VP11/12 and likely depends on the in vivo interactions that we have documented between VP11/12 and Lck. Although the biological functions of VP11/12 have yet to be fully defined, our findings suggest that it is a viral signaling protein that directly regulates the activity of Lck and possibly other SFKs for the benefit of the virus.