ABSTRACT
Cellular stress responses to energy insufficiency can impact virus reproduction. In particular, activation of the host AMP-activated protein kinase (AMPK) by low energy could limit protein synthesis by inhibiting mTORC1. Although many herpesviruses, including herpes simplex virus 1 (HSV-1), stimulate mTORC1, how HSV-1-infected cells respond to energy availability, a physiological indicator regulating mTORC1, has not been investigated. In addition, the impact of low-energy stress on productive HSV-1 growth and viral genetic determinants potentially enabling replication under physiological stress remains undefined. Here, we demonstrate that mTORC1 activity in HSV-1-infected cells is largely insensitive to stress induced by simulated energy insufficiency. Furthermore, resistance of mTORC1 activity to low-energy-induced stress, while not significantly influenced by the HSV-1 UL46-encoded phosphatidylinositol 3-kinase (PI3K)-Akt activator, was dependent upon the Ser/Thr kinase activity of Us3. A Us3-deficient virus was hypersensitive to low-energy-induced stress as infected cell protein synthesis and productive replication were reduced compared to levels in cells infected with a Us3-expressing virus. Although Us3 did not detectably prevent energy stress-induced AMPK activation, it enforced mTORC1 activation despite the presence of activated AMPK. In the absence of applied low-energy stress, AMPK activity in infected cells was restricted in a Us3-dependent manner. This establishes that the Us3 kinase not only activated mTORC1 but also enabled sustained mTORC1 signaling during simulated energy insufficiency that would otherwise restrict protein synthesis and virus replication. Moreover, it identifies the alphaherpesvirus-specific Us3 kinase as an mTORC1 activator that subverts the host cell energy-sensing program to support viral productive growth irrespective of physiological stress.
IMPORTANCE Like all viruses, herpes simplex virus type 1 (HSV-1) reproduction relies upon numerous host energy-intensive processes, the most demanding of which is protein synthesis. In response to low energy, the cellular AMP-activated protein kinase (AMPK) triggers a physiological stress response that antagonizes mTORC1, a multisubunit host kinase that controls protein synthesis. This could restrict virus protein production and growth. Here, we establish that the HSV-1 Us3 protein kinase subverts the normal response to low-energy-induced stress. While Us3 does not prevent AMPK activation by low energy, it enforces mTORC1 activation and overrides a physiological response that couples energy availability and protein synthesis. These results help explain how reproduction of HSV-1, a ubiquitous, medically significant human pathogen causing a spectrum of diseases ranging from the benign to the life threatening, occurs during physiological stress. This is important because HSV-1 reproduction triggered by physiological stress is characteristic of reactivation of lifelong latent infections.
INTRODUCTION
In addition to a critical role in nearly all fundamental biological processes, the availability of sufficient cellular energy resources sits poised to impact viral replication. The primary sensor of energy sufficiency conserved in most eukaryotes is the host AMP-activated protein kinase, AMPK (1). By detecting even small increases in intracellular AMP or ADP abundance, AMPK coordinates ATP production with consumption, effectively managing cellular energy homeostasis in response to physiological stress (1, 2). While AMPK controls numerous metabolic pathways, protein synthesis is particularly vulnerable to changes in energy sufficiency as it is among the most energy-demanding processes (3). Moreover, protein production is vital for viral replication as all viruses are completely reliant upon host ribosomes for their protein synthesis needs (4). Indeed, unrestricted activation of AMPK can limit productive replication of many viruses (5), including those from the medically important herpesvirus family. While AMPK activation reportedly restricts replication of Kaposi's sarcoma-associated herpesvirus, its impact on replication of the betaherpesvirus human cytomegalovirus (HCMV) is more complex (6). Although activation of AMPK shapes carbon and nucleotide metabolism to support HCMV replication, inhibitory effects of AMPK activation on protein synthesis are averted through the actions of specific viral factors (7–9). In contrast, precisely how AMPK activation is regulated in cells infected with the alphaherpesvirus subfamily member herpes simplex virus 1 (HSV-1) and how it might control productive viral replication remain unknown.
Following infection of epithelial cells at mucosal surfaces, HSV-1 establishes a lifelong, latent infection in neurons within peripheral nervous system ganglia. During latency, expression of the approximately 80 open reading frames (ORFs) associated with virus propagation is repressed, and infectious virus is not detected (10, 11). In response to physiological stress, episodes of productive viral replication resulting in infectious virus production and shedding at mucosal surfaces ensue. Many environmental and physiologic stresses associated with reactivation inhibit the activity of the mechanistic target of rapamycin, mTORC1 (12). A multisubunit cellular kinase, mTORC1 integrates fundamental physiological inputs, including energy sufficiency, to coordinate catabolic versus anabolic responses (2). Although inhibiting mTORC1 limits viral protein production and restricts HSV-1 replication (13), the capacity of HSV-1 to overcome physiological cues indicating energy insufficiency, which would limit mTORC1 activation, remains unknown.
By sensing increased AMP and ADP intracellular abundance, AMPK responds to changes in adenosine nucleotide levels (Fig. 1A). Activation of the heterotrimeric enzyme, comprised of catalytic alpha (isoform α1 or α2) and regulatory beta (isoform β1 or β2) and gamma (isoform γ1, γ2, or γ3) subunits, begins upon binding AMP by the gamma subunit (14, 15). The resulting conformational changes in the alpha subunit unmasks residue T172, which is phosphorylated by one of several AMPK activating kinases. While LKB1, calmodulin-dependent kinase kinase β (CaMKKβ), and TAK1 can all phosphorylate T172, LKB1 primarily activates AMPK in response to energy insufficiency (16–18). As T172 phosphorylation is required for full AMPK activation, AMP and ADP binding to the gamma subunit further enforces AMPK activation by inhibiting T172 dephosphorylation (19). Phosphorylation of the tuberous sclerosis complex (TSC) subunit TSC2 on residues T1227 and S1345 by AMPK activates the Rheb GTPase-activating protein (GAP) activity of TSC, which results in Rheb-GDP accumulation and prevents mTORC1 activation (20). AMPK can also inhibit mTORC1 directly by phosphorylating the raptor subunit (21). Thus, failure to restrict AMPK activation or suppress AMPK-dependent inhibition of mTORC1 might limit productive viral replication.
Control of mTORC1 signaling by AMPK in response to energy insufficiency. (A) In uninfected cells, accumulation of AMP or ADP relative to ATP signals energy insufficiency and activates AMPK. The binding of AMP to AMPK exposes Thr172, which is phosphorylated by an AMPK-activating enzyme, such as LKB1. Once activated, AMPK stimulates TSC Rheb-GAP activity by phosphorylating TSC2 on residues T1227 and S1345. Besides promoting Rheb-GDP accumulation, AMPK phosphorylates the mTORC1 subunit raptor, and both inhibit mTORC1 activation. (B) HSV-1 UL46-encoded VP11/12 stimulates PI3K activity, which activates Akt. Both the host Akt and the HSV-1 Us3 Ser/Thr kinases phosphorylate TSC2 on residues S939 and T1462. This inhibits TSC Rheb-GAP activity and allows Rheb-GTP accumulation, which subsequently activates mTORC1. Thus, differential phosphorylation of TSC2 residues by either AMPK or Akt/Us3, respectively, stimulates or inhibits TSC Rheb-GAP activity, which in turn inhibits or activates mTORC1.
Activation of mTORC1 in cells productively infected with HSV-1 is dependent upon the viral Us3 Ser/Thr kinase (13), an enzyme unique to alphaherpesviruses (Fig. 1B). Although Us3 lacks detectable primary sequence homology with the host Akt kinase outside an identifiable ATP binding motif, it directly phosphorylates many Akt substrates on sites targeted by Akt, including TSC2 (13). In contrast to TSC2 phosphorylation by AMPK, phosphorylation of TSC2 on residues S939 and T1462 by Akt or Us3 inhibits TSC Rheb-GAP activity. The resulting Rheb-GTP accumulation, in turn, activates mTORC1 (13). Once activated, mTORC1 subsequently stimulates viral protein synthesis by phosphorylating and inactivating the translational repressor 4E-BP1 (13). In addition to Us3, which acts downstream of Akt to activate mTORC1, the UL46-encoded gene product VP11/12 interacts with the phosphatidylinositol 3-kinase (PI3K) regulatory subunit p85 to stimulate Akt (22). While AMPK activation reportedly only modestly reduces virus replication in neurons, the capacity of HSV-1-encoded function(s) to antagonize AMPK activity, the identity of the function(s), and the mechanisms of action remain undefined (23, 24).
Here, we demonstrate that HSV-1 infection alters the responsiveness of mTORC1 to physiological stress induced by energy insufficiency. Unlike the case in uninfected cells, where mTORC1 is inhibited by activated AMPK, mTORC1 activation is sustained in HSV-1-infected cells exposed to simulated energy-insufficient conditions. Resistance of HSV-1-infected cells to low-energy-induced stress, while not primarily influenced by VP11/12, was dependent upon Us3 Ser/Thr kinase activity. By sustaining mTORC1 activity during energy insufficiency, Us3 facilitated infected cell protein synthesis and productive replication. Significantly, we show that Us3 did not prevent energy stress-induced AMPK activation. Instead, stress-induced AMPK activation in cells infected with a Us3-deficient virus inhibited mTORC1 signaling in a manner dependent upon TSC2, a shared substrate of both AMPK and Us3. Together, these results establish that HSV-1 not only enforces mTORC1 activation but also subverts the host cell energy-sensing program to support viral productive growth irrespective of physiological stress.
RESULTS
Responses of mTORC1 to energy insufficiency are subverted in HSV-1-infected cells.To investigate how mTORC1 responds to physiological cues indicative of energy insufficiency in HSV-1-infected cells, phosphorylation of the mTORC1 substrates S6K1 and the translational repressor 4E-BP1 were measured after exposure to the AMPK activator 5-aminoimidazole-4-carboxamide 1-1-β-d-ribofuranoside (AICAR). An analog of AMP, AICAR activates AMPK by simulating a high ratio of AMP to ATP (25). As AMPK is a multifunctional kinase that regulates several metabolic processes in addition to mTORC1 activity, care was taken to avoid pretreatment of cells with AICAR. Instead, primary normal human dermal fibroblasts (NHDFs) mock infected or infected with wild-type (WT) HSV-1 were either left untreated or were supplemented with AICAR at 12 h postinfection (hpi). By this time in the virus productive growth cycle, HSV-1 is especially reliant upon mTORC1 activity to stimulate viral mRNA translation. After limited exposure of cells to AICAR for 3 h, total protein was isolated at 15 hpi and analyzed by immunoblotting (Fig. 2). AICAR treatment activated AMPK in both mock- and WT HSV-1-infected cells as indicated by the presence of AMPK phosphorylation on residue Thr172 (p-T172). While the AMPK p-T172 signal was greater in mock-infected, AICAR-treated cells than in infected cells, overall AMPK abundance was also greater in mock-infected cultures (Fig. 2, compare lanes 2 and 4), making precise comparisons of the extent of AMPK activation difficult.
HSV-1 sustains mTORC1 activity in cells exposed to energy insufficiency. NHDFs growth arrested by serum deprivation were mock infected or infected with wild-type (WT) HSV-1 (strain F). Cultures were either untreated (−) or treated (+) with AICAR at 12 hpi. Total protein was isolated at 15 hpi, fractionated by SDS-PAGE, and analyzed by immunoblotting using the indicated antibodies. M, mock.
In asynchronous NHDF cultures, the translational repressor 4E-BP1 exists as a heterogenous collection of phosphorylated isoforms within the population. They are readily fractionated by SDS-PAGE on high-percentage gels and detected by immunoblotting. Dividing cells with activated mTORC1 are enriched with hyperphosphorylated isoforms, whereas nondividing cells accumulate hypophosphorylated isoforms, reflecting the natural coordination of mTORC1 signaling with proliferation in this cell type. Thus, at the time of sampling, the control 4E-BP1 population in mock-infected, untreated cultures is predominately but not completely hypophosphorylated (Fig. 2). Activation of AMPK resulted in accumulation of the mTORC1 substrate 4E-BP1 in its hypophosphorylated form and did not detectably stimulate p70S6K phosphorylation (p-S6K1), consistent with enforced mTORC1 inhibition under these conditions (Fig. 2, compare lanes 1 and 2). In contrast, hyperphosphorylated 4E-BP1 and p-S6K1 accumulated in HSV-1-infected cells (Fig. 2, compare lanes 1 and 3). Furthermore, S6K1 and 4E-BP1 phosphorylation was less sensitive to AMPK activation in HSV-1-infected cells than in mock-infected cells (Fig. 2, compare lanes 2 and 4). This demonstrates that phosphorylation of key mTORC1 substrates 4E-BP1 and S6K1 in HSV-1-infected cells is relatively insensitive to AMPK activation. It further suggests that mTORC1 activation and signaling in HSV-1 infected cells are resistant to simulated energy insufficiency and raises the possibility that this phenotype results from subversion of cellular energy-sensing cues or responses by a viral function.
Identification of HSV-1 genes required to maintain mTORC1 signaling during low-energy stress.Energy insufficiency activates AMPK and inhibits mTORC1, whose activity is regulated in part by PI3-kinase-Akt signaling in uninfected cells. Since HSV-1 Us3 and UL46 gene products commandeer control of the cellular PI3-kinase-Akt signaling axis (Fig. 1B) (22, 26), how these viral functions might subvert infected cell responses to energy insufficiency was investigated. NHDFs were either mock infected or infected with WT HSV-1 or Us3-deficient (ΔUs3), UL46-deficient (ΔUL46), or Us3 and UL46 doubly deficient (ΔUs3ΔUL46) HSV-1. After 12 h, cells were left untreated, or energy insufficiency was simulated by AICAR treatment. Cell lysates were prepared at 15 hpi, and mTORC1 activation was analyzed by immunoblotting for phosphorylated mTORC1 substrates p70S6K1 and 4E-BP1 (Fig. 3A). AICAR exposure effectively activated AMPK in infected cells as indicated by similar T172 phosphorylation (p-T172) levels, irrespective of the presence or absence of Us3 and/or UL46 gene product. Interestingly, limited AMPK p-T172 accumulation under unstressed, energy-replete conditions was greater in cells infected with either of the two Us3-deficient (ΔUs3 or ΔUs3ΔUL46) viruses in this experiment (Fig. 3A, compare lanes 5 and 9 to lanes 3 and 7) and in subsequent experiments (see Fig. 5A and B and Fig. 6, compare lanes 3 and 7 to lanes 5 and 9; and see Fig. 7, compare lanes 1 and 2 to lanes 5 and 6). This suggests that HSV-1 infection activates AMPK, perhaps by inducing low-energy stress, and this is primarily countered by Us3. In cells infected with ΔUL46 or WT HSV-1 (Fig. 3A, compare lanes 3 and 4 to lanes 7 and 8, respectively), both of which express Us3, phosphorylation of p70S6K1 and 4E-BP1 was relatively insensitive to AICAR-induced AMPK activation compared to effects in mock-infected cells. In contrast, hypophosphorylated 4E-BP1 accumulated, and p70S6K1 phosphorylation was reduced in cells infected with either Us3-deficient virus (ΔUs3 or ΔUs3ΔUL46) upon AICAR exposure (Fig. 3A, compare lanes 5 and 6 and lanes 9 and 10). Furthermore, appreciable differences in S6K1 phosphorylation or 4E-BP1 hyperphosphorylation were not detected upon comparing cells infected with ΔUs3 to those infected with ΔUs3ΔUL46 and exposed to AICAR-induced energy insufficiency. Thus, the presence or absence of the UL46 gene in a Us3-deficient isogenic background did not detectably impact S6K1 phosphorylation or 4E-BP1 hyperphosphorylation in response AICAR-induced energy insufficiency. Although we are unable to detect a role for UL46 under these conditions, these results demonstrate that the HSV-1 Us3 gene encodes a function responsible for enforcing mTORC1 activation under low-energy stress in primary fibroblasts. Moreover, they suggest that Us3 is the primary virus-encoded genetic determinant capable of countering AMPK activation and physiological stress resulting from energy insufficiency.
HSV-1-enforced mTORC1 activation during energy insufficiency requires Us3 kinase activity but does not detectably depend upon VP11/12. (A) NHDFs growth arrested by serum deprivation were mock infected or infected with wild-type (WT) HSV-1 Kos37, a Us3 deletion mutant (ΔUs3), a VP11/12 deletion mutant (ΔUL46), or a virus doubly deleted for Us3 and VP11/12 (ΔUs3ΔUL46) (MOI of 5). Cultures were either untreated (−) or treated (+) with AICAR at 12 hpi. Total protein was isolated at 15 hpi, fractionated by SDS-PAGE, and analyzed by immunoblotting using the indicated antibodies. (B) The same experiment as described for panel A except that NHDFs were infected with F-strain HSV-1 expressing WT Us3 (HSV1ΔUs3 Repair), a Us3-deficient virus (HSV-1 ΔUs3), or a virus containing a Us3 kinase-deficient Us3 allele (HSV-1 K220A).
To determine if the relative resistance of mTORC1 activation in HSV-1-infected cells to energy-insufficient cues required Us3 kinase activity, NHDFs were mock infected or infected with a Us3 deletion mutant (HSV-1 ΔUs3), a virus that expresses a catalytically inactive Us3 mutant protein (HSV-1 K220A), or HSV-1 expressing WT Us3 (HSV1ΔUs3 Repair). At 3 or 12 hpi cells were untreated or supplemented with AICAR. Total protein was collected at 15 hpi, and 4E-BP1 phosphorylation was evaluated by immunoblotting (Fig. 3B). Compared to levels in cultures infected with WT HSV-1, hyperphosphorylated 4E-BP1 accumulated to reduced levels in cultures infected with HSV-1 ΔUs3 or HSV-1 K220A, as we reported previously (13). However, activation of AMPK by AICAR addition at 3 hpi suppressed 4E-BP1 hyperphosphorylation in all cultures infected with HSV-1 (Fig. 3B). While hyperphosphorylation of the mTORC1 substrate 4E-BP1 in cultures infected with WT HSV-1 was comparatively insensitive to AMPK activation at 12 hpi, phosphorylation of 4E-BP1 in cultures infected with either HSV-1 ΔUs3 or HSV-1 K220A Us3 remained sensitive to energy insufficiency simulated by AICAR (Fig. 3B, compare lane 2 to lanes 3 and 4). This shows that mTORC1 activation in HSV-1-infected cells becomes increasingly resistant to energy insufficiency simulated by AICAR between 3 and 12 hpi. This time frame correlates with the synthesis and accumulation of the Us3 delayed early protein. Furthermore, the insensitivity of mTORC1 signaling in HSV-1-infected cells to low-energy-induced AMPK activation was dependent upon the Ser/Thr kinase activity of the Us3 gene product. Finally, the dependence upon Us3 was not strain specific as it was readily observed using mutant viruses independently produced in different genetic backgrounds (strain KOS in the experiment shown in Fig. 3A and strain F in that shown in B).
Us3 promotes protein synthesis and productive replication during energy stress.To evaluate how the capacity of Us3 to stimulate mTORC1 during energy insufficiency impacts productive HSV-1 growth, the influence of Us3 upon protein synthesis and virus reproduction and spread was measured. NHDFs infected with WT HSV-1 or HSV-1 ΔUs3 were either untreated or subjected to low-energy-induced stress simulated with AICAR at 12 hpi. After metabolic pulse-labeling with 35S-labeled amino acids, total protein was isolated and quantified by liquid scintillation counting of acid-insoluble radioactivity (Fig. 4A). Whereas AMPK activation did not significantly reduce ongoing protein synthesis in cells infected with WT HSV-1, energy insufficiency significantly and selectively reduced protein synthesis by approximately 40% in cells infected with a Us3-deficient virus. This establishes that the capacity of HSV-1-infected cells to sustain normal rates of protein synthesis under conditions of energy insufficiency is Us3 dependent.
Hypersensitivity of Us3-deficient HSV-1 to energy insufficiency. (A) NHDFs growth arrested by serum deprivation were infected with wild-type (WT) HSV-1 strain F or a Us3 deletion mutant (HSV-1 ΔUs3) (MOI of 5). Cultures were either untreated or treated with AICAR at 12 hpi. Cultures were metabolically pulse-labeled for 45 min with 35S-labeled amino acids at 14 hpi. Protein synthesis was quantified by measuring acid-insoluble radioactivity using liquid scintillation counting. Incorporation of 35S was normalized to that in untreated cells. Error bars represent standard errors of the means. P values were determined by one-sample t tests (n = 5). (B) The same experiment as described for panel A except NHDFs were infected at a low MOI (MOI of 2 × 10−4) to measure virus reproduction and spread through the culture. After 4 days, cell-free lysates were prepared by freeze-thawing, and infectious virus was quantified by plaque assay using Vero cells. (P values were determined by two-way analysis of variance with multiple comparisons; n = 3).
To define the corresponding impact of energy insufficiency on productive virus replication, NHDFs infected with WT HSV-1 or HSV-1 ΔUs3 were either untreated or stimulated with AICAR at 12 hpi to induce low-energy stress. The experiment was performed at a low multiplicity of infection (MOI) as this is likely a more physiological condition than high-MOI infections and allows virus replication and spread through the culture to be assessed. Cultures were collected at 4 days postinfection (Fig. 4B), and the amount of infectious virus produced was quantified by plaque assay in Vero cells. While activating AMPK did not significantly reduce productive replication of WT HSV-1, productive growth of HSV-1 ΔUs3 was reduced by 15-fold. Together, these results show that the Us3-deficient virus is hypersensitive to energy stress-induced AMPK activation. This demonstrates that the capacity of HSV-1 to productively replicate to wild-type levels during AMPK activation induced by simulated energy insufficiency is Us3 dependent. Taken together, these results are consistent with the Us3 gene product acting as a virus-encoded determinant that counters stress responses associated with energy insufficiency and sustains productive virus growth under stress.
HSV-1 infection does not preclude energy stress-induced AMPK activation.Although AMPK is activated by T172 phosphorylation (p-T172), the cellular kinase Akt can limit LKB1-mediated AMPK activation by phosphorylating the AMPK catalytic subunits α1/α2 on residues S485/491 (27, 28). Given that Us3 and Akt share some substrates (13) and that both Us3 and UL46 can influence Akt activity (22, 26), AMPK phosphorylation levels in cells infected with WT HSV-1, a Us3-deficient virus, a UL46-deficient virus, or a virus doubly deficient for Us3 and UL46 were compared. NHDFs mock infected or infected with WT, ΔUs3, ΔUL46, or ΔUs3ΔUL46 were left untreated or exposed to AICAR to activate AMPK at 12 hpi. Total protein was collected at 15 hpi, and the phosphorylation status of AMPK was evaluated by immunoblotting (Fig. 5A). AICAR exposure increased AMPK T172 and S485/491 phosphorylation in mock-infected cells. Similar levels of T172 and S485/491 AMPK phosphorylation were likewise observed in cells infected with WT, ΔUs3, ΔUL46, or ΔUs3ΔUL46 virus and treated with AICAR (Fig. 5A, lanes 2, 4, 6, 8, and 10). These results suggest that Us3 and UL46 do not detectably regulate AMPK phosphorylation during simulated energy insufficiency. While analogous levels of p-S485/491 phosphorylation were detected in infected and in mock-infected cells in the absence of acute energy stress (Fig. 5A, lanes 1, 3, 5, 7, and 9), greater AMPK p-T172 accumulation was consistently observed in cells infected with Us3-deficient HSV-1-infected than in WT HSV-1- or ΔUL46-infected cells (Fig. 3A, Fig. 5A and B, and Fig. 6, compare lanes 3 and 7 to lanes 5 and 9; and Fig. 7, compare lanes 1 and 2 to lanes 5 and 6). This suggests that Us3 impairs AMPK T172 phosphorylation and naturally limits AMPK activation in the absence of AICAR-induced energy insufficiency.
Acute energy stress-induced AMPK activity in HSV-1-infected cells is independent of Us3 and UL46. NHDFs growth arrested by serum deprivation were mock infected, infected with wild-type (WT) HSV-1 Kos37, a Us3 deletion mutant (ΔUs3), a UL46 deletion mutant (ΔUL46), or a virus doubly deleted for Us3 and VP11/12 (ΔUs3ΔUL46) (MOI of 5). Cultures were either untreated (−) or treated (+) with AICAR at 12 hpi. Total protein was isolated at 15 hpi, fractionated by SDS-PAGE, and analyzed by immunoblotting to determine the phosphorylation status of AMPK (A) or ACC (B). Accumulation of ICP0, a key HSV-1 immediate early regulatory protein, was monitored as a marker of infection.
Inhibition of AMPK catalytic activity in HSV-1-infected cells is Us3 dependent. Growth-arrested NHDFs were mock infected, infected with wild-type (WT) HSV-1 Kos37, a Us3 deletion mutant (ΔUs3), a UL46 deletion mutant (ΔUL46), or a virus doubly deleted for Us3 and UL46 (ΔUs3ΔUL46) (MOI of 5). Cultures were either untreated (−) or treated (+) with compound C at 12 hpi. Total protein was isolated at 15 hpi, fractionated by SDS-PAGE, and analyzed by immunoblotting using the indicated antibodies.
Hypersensitivity of mTORC1 to energy insufficiency in ΔUs3-infected cells is relieved by TSC2 depletion. NHDFs were treated with TSC2 siRNA (siTSC2; +) or a nonsilencing, control siRNA (−), growth arrested by serum deprivation, and mock infected (M) or infected with WT HSV-1 (strain F) or a Us3 deletion mutant (HSV-1 ΔUs3). Cultures were either untreated (−) or treated (+) with AICAR at 12 hpi. Total protein was isolated at 15 hpi, fractionated by SDS-PAGE, and analyzed by immunoblotting using the indicated antibodies. Accumulation of ICP0, a key HSV-1 immediate early regulatory protein, was monitored as a marker of infection.
To investigate whether Us3 and UL46 influence AMPK activity, phosphorylation of acetyl-coenzyme A (CoA) carboxylase (ACC), an AMPK substrate, was measured. Phosphorylation of ACC Ser79 by AMPK inhibits ACC catalytic activity to downregulate fatty acid biosynthesis (29) (Fig. 1). NHDFs mock infected or infected with WT, ΔUs3, ΔUL46, or ΔUs3ΔUL46 virus were exposed to AICAR or left untreated at 12 hpi. Total protein was collected at 15 hpi, and ACC Ser79 phosphorylation was evaluated by immunoblotting (Fig. 5B). AICAR-induced AMPK p-T172 phosphorylation resulted in a corresponding accumulation of phosphorylated ACC (p-ACC) in both mock-treated and infected cells, indicating an increase in AMPK catalytic activity (Fig. 5B, lanes 2, 4, 6, 8, and 10). These results show that HSV-1 does not prevent AMPK activation or antagonize AMPK catalytic activity, at least with respect to its substrate ACC in response to energy stress. In the absence of acute AICAR-induced energy stress, a small increase in AMPK p-T172 and p-ACC was observed in cells infected with Us3-deficient viruses (ΔUs3 and ΔUs3ΔUL46) compared to levels in cells infected with Us3-expressing viruses (WT and ΔUL46) (Fig. 5B, compare lanes 5 and 9 with lanes 3 and 7). Together, these results indicate that while HSV-1 does not prevent phosphorylation of AMPK or its substrate ACC in response to acute energy stress, Us3 may control AMPK under energy-replete conditions.
Suppression of AMPK activity in HSV-1-infected cells is Us3 dependent.To directly investigate the ability of Us3 and UL46 to control AMPK, phosphorylation of ACC was measured during HSV-1 infection in the absence of AICAR treatment. Sensitivity to compound C, a selective small-molecule AMPK inhibitor, was used to verify that any observed changes in phosphorylation reflected altered AMPK activity (30). Compound C is an ATP-competitive inhibitor that has been shown to impede the phosphorylation of AMPK T172 in HCMV-infected cells (9). NHDFs were mock infected or infected with WT HSV-1, ΔUs3, ΔUL46, or ΔUs3ΔUL46 virus and then treated with compound C at 12 hpi or left untreated. Total protein isolate was collected at 15 hpi and analyzed by immunoblotting. As expected, AMPK p-T172 abundance was greater in cells infected with ΔUs3 and ΔUs3ΔUL46 than in cultures mock infected or infected with WT HSV-1 or ΔUL46 (Fig. 6, compare lanes 5 and 9 to lanes 1, 3, and 7). The increase in p-T172 levels in response to infection with ΔUs3 or ΔUs3ΔUL46 was sensitive to compound C (Fig. 6, compare lanes 5 and 6 and lanes 9 and 10). Compound C also induced a modest increase in phosphorylated S6K1 abundance in ΔUs3- and ΔUs3ΔUL46-infected NHDFs (Fig. 6, compare lanes 5 and 6 and lanes 9 and 10), demonstrating that inhibiting AMPK increased mTORC1 activity. Significantly, p-ACC accumulated to a greater degree in cells infected with Us3-deficient viruses (ΔUs3 and ΔUs3ΔUL46) than in cells infected with a Us3-expressing (WT and ΔUL46) virus (Fig. 6, compare lanes 5 and 9 to lanes 1, 3, and 7). p-ACC accumulation was sensitive to compound C, consistent with it being dependent upon AMPK catalytic activity (Fig. 6, compare lanes 5 and 6 and lanes 9 and 10). Together, these results demonstrate that AMPK is activated in cells infected with ΔUs3 or ΔUs3ΔUL46 virus and indicate that the suppression of AMPK activity in HSV-1-infected cells is Us3 dependent. Thus, although Us3 and UL46 do not detectably prevent AMPK activation in response to acute low-energy stress simulated by AICAR, Us3 effectively limits AMPK activation during infection in the absence of acute, AICAR-induced energy stress.
Inhibition of mTORC1 by AMPK is TSC2 dependent in ΔUs3-infected cells.Since mTORC1 was insensitive to energy insufficiency and since AMPK activation induced by low-energy-induced stress was not suppressed in HSV-1-infected cells, it was possible that Us3 interfered with AMPK signaling to mTORC1. AMPK reportedly restricts mTORC1 both directly and indirectly, in the latter case through TSC2 (2, 21) (Fig. 1). Both Us3 and AMPK directly phosphorylate TSC2, albeit on different residues and with different outcomes. TSC2 phosphorylation by Us3 inhibits TSC Rheb-GAP, stimulating mTORC1. In contrast, TSC2 phosphorylation by AMPK activates TSC Rheb-GAP, repressing mTORC1 (2). Thus, whether TSC contributed to the insensitivity of mTORC1 to low-energy stress in HSV-1-infected cells was investigated. To address this, NHDFs treated with a control, nonsilencing (NS) small interfering RNA (siRNA) or a TSC2-specific siRNA were mock infected or infected with WT HSV-1 or HSV-1 ΔUs3. Cultures were left untreated or treated with AICAR at 12 hpi, and mTORC1 activation was analyzed at 15 hpi by immunoblotting (Fig. 7). While phosphorylated S6K1 and hyperphosphorylated 4E-BP1 in WT HSV-1-infected cells were relatively insensitive to AICAR, S6K1 phosphorylation was not detected, and hypophosphorylated 4E-BP1 accumulated in AICAR-treated cultures that were infected with HSV-1 ΔUs3 (Fig. 7, compare lanes 1 and 3 with lanes 5 and 7). In agreement with previously reported results, S6K1 phosphorylation and 4E-BP1 hyperphosphorylation were stimulated in HSV-1 ΔUs3-infected cultures upon TSC2 depletion, indicating that TSC2 restricts mTORC1 activation in the absence of Us3 (Fig. 7, compare lanes 5 and 6). Compared to levels in cells infected with WT HSV-1, a modest increase in AMPK p-T172 levels was observed in cells infected with HSV-1 ΔUs3 (Fig. 7, compare lanes 1 and 5). Significantly, whereas S6K1 phosphorylation was reduced and hypophosphorylated 4E-BP1 accumulated in response to AICAR in cultures infected with HSV-1 ΔUs3, S6K1 phosphorylation and 4E-BP1 hyperphosphorylation were relatively insensitive to AICAR in HSV-1 ΔUs3-infected cultures treated with TSC2 siRNA (Fig. 7, compare lanes 7 and 8). Thus, TSC2 depletion allowed sustained mTORC1 signaling in cells infected with HSV-1 ΔUs3 when AMPK activation was induced by AICAR. This established that AMPK activated in response to simulated energy insufficiency limits mTORC1 activation by targeting TSC2, a direct substrate of Us3.
DISCUSSION
While HSV-1 reproduction triggered in response to physiological stress is characteristic of reactivation (31), the capacity of HSV-1 to productively replicate under stress is poorly understood. Likewise, specific underlying genetic determinants enabling productive virus replication under physiological stress have not been systematically defined. By integrating signals indicating availability of fundamental physiological cues, the cellular kinase mTORC1 coordinates specific anabolic or catabolic responses (2, 32). In addition, HSV-1 Us3 Ser/Thr kinase stimulates mTORC1 (13). However, the responsiveness of mTORC1 in HSV-1-infected cells to critical indicators, like energy availability, which naturally regulate mTORC1 and the potential of physiological signals that regulate mTORC1 to control productive HSV-1 replication have not been explored. Here, we establish that mTORC1 activity in HSV-1-infected cells is largely insensitive to stress induced by energy insufficiency. Furthermore, we show that resistance of mTORC1 activity to low-energy-induced stress is dependent upon the kinase activity of Us3. Compared to cells infected with WT virus, protein synthesis and productive virus replication in cells infected with a Us3-deficient virus were hypersensitive to low-energy-induced stress induced by AICAR. Although Us3 restricted AMPK activity in HSV-1-infected cells that were not exposed to AICAR, Us3 did not detectably prevent acute energy stress-induced activation of AMPK. Instead, Us3 effectively prevented AMPK activated in response to AICAR-induced energy stress from inhibiting mTORC1. Significantly, inhibition of mTORC1 in ΔUs3-infected cells by stress-induced AMPK activation was dependent upon TSC2, a direct substrate of AMPK and Us3. Thus, not only does Us3 stimulate mTORC1 activity, but it also sustains mTORC1 signaling under conditions of energy insufficiency that would otherwise inhibit protein synthesis and interfere with productive virus replication. In this capacity, Us3 represents a virus-encoded mTORC1 activator that sustains mTORC1 function irrespective of AMPK activation in response to low-energy stress.
Productive virus replication is dependent upon numerous energy-intensive macromolecular processes. Activation of AMPK by the ATP turnover resulting from viral infection could stimulate catabolic processes and inhibit mTORC1. Through its substrate p70S6K1 and the translational repressor 4E-BP1, mTORC1 coordinates protein synthesis output in response to a variety of physiological insults, including energy insufficiency (2). Failure to enforce mTORC1 activation reduces synthesis of proteins needed to manufacture progeny virus and restricts HSV-1 reproduction (13, 33). The capacity to sustain mTORC1 activation during the productive growth cycle is conserved among representative herpesvirus subfamily members (4). Significantly, both HSV-1, an alphaherpesvirus, and HCMV, a betaherpesvirus, can sustain mTORC1 activation even though AMPK is activated (8). This is unusual as inhibiting mTORC1 in response to energy insufficiency represents an important means of conserving energy for cell survival (34). Indeed, TSC-deficient cells, which constitutively activate mTORC1, die upon low-energy-induced stress induced by glucose starvation, accumulating p53 and activating the unfolded protein response (UPR) (20, 35, 36). Loss of LKB1, an AMPK-activating kinase, allows some tumor cells where mTORC1 is constitutively activated to avoid this fate by preventing AMPK activation (37). The more limited time scale of survival needed to complete a viral reproductive cycle and a variety of virus-encoded functions that prevent cell death are likely contributing factors that allow activation of both AMPK and mTORC1 without catastrophic effects on virus replication.
Despite stimulating Akt, VP11/12 did not significantly influence mTORC1 activation in HSV-1-infected NHDFs during low-energy-induced stress under these experimental conditions. Consistent with this finding, PI3-kinase activation by insulin in uninfected cells also did not preclude AMPK activation by low-energy stress (28). Nevertheless, our observations do not preclude a more pronounced role for VP11/12 in other cells types, including neurons, or in mitigating mTORC1 responses to other forms of stress.
Sustained mTORC1 activation in HSV-1-infected cells exposed to acute energy insufficiency was dependent upon the Us3 Ser/Thr kinase. By directly phosphorylating TSC2 S939 and T1462, Us3 enforced constitutive mTORC1 activation (13). We now show that insensitivity of mTORC1 signaling to stress-induced AMPK activation in HSV-1-infected cells is dependent upon the Us3 kinase activity. While we cannot exclude a direct role of AMPK acting on the mTORC1 raptor subunit, inhibition of mTORC1 by activated AMPK in ΔUs3-infected cells was dependent upon TSC2, a direct substrate of both AMPK and Us3. This suggests that an underlying conflict between AMPK, which stimulates TSC, and Us3, which inhibits TSC, determines the sensitivity of mTORC1 to low-energy-induced AMPK activation. Although precisely how Us3 dominates and interferes with AMPK action on TSC2 remains to be determined, molecular conflicts of this nature can have a powerful impact upon viral replication (38). Significantly, targeting TSC by HSV-1 Us3 to sustain mTORC1 activity during low-energy-induced stress in infected cells mirrors findings in uninfected cells, where inhibition of mTORC1 activity by AICAR is compromised in TSC2-deficient cells (39). This allows HSV-1 to efficiently enforce and sustain mTORC1 activation during energy insufficiency via a single viral function that targets a single host gene product. In addition, it enables AMPK to be activated in response to acute energy insufficiency without concomitant mTORC1 inhibition. A distinct but related strategy operates in HCMV-infected cells whereby AMPK activation stimulates productive HCMV replication (7, 9). Instead of the Us3 kinase, which is specific to alphaherpesviruses, HCMV encodes the TSC2 binding protein UL38 to antagonize TSC activity and stimulate mTORC1 (40). This precludes AMPK-mediated inhibition of mTORC1 while supporting AMPK-dependent metabolic changes conducive to virus reproduction (7, 9). Compared to betaherpesviruses like HCMV, alphaherpesviruses have a very fast productive growth cycle that may not require AMPK activation. While AMPK is not naturally activated by HSV-1 infection in fibroblasts, it could play a more prominent role in other cell types like neurons, which are particularly reliant on glucose, metabolically highly active, and sensitive to energy insufficiency (41–43). Significantly, AMPK activation and activity in HSV-1-infected cells were restricted in a Us3-dependent manner. This might potentially protect the infected cell from the ATP turnover associated with infection, precluding subsequent AMPK activation in response to a rise in AMP concentration. Alternatively, Us3 might prevent AMPK activation through unconventional, noncanonical means.
The capacity to sustain productive replication under physiological stress, including energy insufficiency, is a critical adaptation for herpesvirus family members. Virus reproduction from latent reservoirs requires episodic productive replication triggered in response to stress, and the emerging evidence suggests that discrete stress-responsive host cell signaling pathways regulate entry into the productive growth cycle (12). In the case of HSV-1, changes in neuronal homeostasis are sensed by mTORC1 signaling, which integrates fundamental environmental and physiologic cues to control entry into the productive growth cycle (31). Physiological stress interrupts mTORC1 signaling and stimulates latently infected neurons to commence the virus reproductive cycle (12). Thus, hijacking mTORC1 as a sensor of neuronal homeostasis allows the virus to respond to a variety of physiological insults (31). In the case of limited oxygen, which, like energy insufficiency, also inhibits mTORC1 signaling, the productive replication cycle is not only triggered but can also be completed (12). To complete the virus productive growth cycle under physiological stress, HSV-1 and other herpesvirus family members likely rely upon an array of viral functions akin to Us3 that mitigate the potentially detrimental inhibitory effects of physiological stress, including energy insufficiency, on mTORC1 activation.
MATERIALS AND METHODS
Cell culture, viruses, and chemicals.Normal human dermal fibroblasts (NHDFs; purchased from Lonza, Walkersville, MD) were grown in Dulbecco's modified Eagle medium (DMEM) supplemented with 5% fetal bovine serum (FBS) and growth arrested by serum deprivation in 0.2% fetal bovine serum as described previously (44). Vero cells (ATCC) were grown in DMEM supplemented with 5% calf serum. WT HSV-1, HSV-1 ΔUs3 (strain R7041), HSV-1 K220A Us3 mutant (VRR1204), and HSV-1 ΔUs3 repair (VRR1202repair) F-strain viruses were described previously (45, 46). Kos37 strain WT HSV-1, ΔUs3, ΔUL46, and ΔUs3ΔUL46 viruses were described in Eaton et al. (26). AICAR (catalog number 9944) was purchased from Cell Signaling Technology. Compound C (catalog number 171260) was purchased from Millipore.
Antibodies and siRNAs.S6K1 (catalog number 9202), S6K1-pThr389 (9234), AMPK (2532), phospho-AMPKα1 (Ser485)/AMPKα2 (Ser491) (4185), AMPK-pT172 (2531), ACC (3662), ACC-pS79 (11818), and TSC2 (3612) were purchased from Cell Signaling Technology. The tubulin antibody (catalog number T5168) was from Sigma, anti-HSC70 (10011384) was from Cayman, and anti-4E-BP1 (A300-501A) was from Bethyl Laboratories. The TSC2 siRNA (13) was chemically synthesized by Sigma: CCAAUGUCCUCUUGUCUUU. AllStars negative-control siRNA was purchased from Qiagen. To deplete TSC2, NHDFs seeded in a 12-well dish were transfected with an siRNA using Lipofectamine RNAi Max (Invitrogen) according to the manufacturer's instructions. After 24 h, the transfection protocol was repeated. The following day, cells were growth arrested by serum deprivation in 0.2% FBS for 72 h and then infected with HSV-1.
Multicycle virus growth assay.NHDFs were seeded in a six-well dish; upon reaching confluence (4.4 × 105 cells per well), cells were growth arrested in 0.2% FBS for 72 h as described above. Cells were infected with 100 PFU virus in 0.5 ml of medium per well for 1.5 h, after which the virus inoculum was removed and replaced with fresh medium. At 12 hpi, infected cells were treated with 1 mM AICAR or left untreated. Infected cultures were collected at 96 hpi and freeze-thawed three times, and infectious virus production was quantified by plaque assay on Vero cells.
Quantification of protein synthesis.Infected NHDFs were metabolically labeled for 45 min with methionine- and cysteine-free DMEM supplemented with 70 μCi/ml 35S Express (Perkin-Elmer). Total protein was isolated, boiled in SDS sample buffer, and then precipitated in 10% trichloroacetic acid (TCA) on ice for 30 min. Acid-insoluble protein was collected onto Whatman GF-C filters which were subsequently washed twice in 10% TCA and twice in 100% ethanol. Radioactivity collected on the filters was quantified by liquid scintillation counting.
ACKNOWLEDGMENTS
We thank members of the Mohr lab and Angus Wilson for stimulating discussions and Christopher Bianco for critically reading the manuscript.
This work was supported by grants from the NIH to I.M. (R01GM056927 and AI073898) and grants from the Canadian Institutes for Health Research (FRN 12172) and Alberta Innovates—Health Solutions to J.R.S. J.R.S. holds a Canada Research Chair in Molecular Virology (Tier I). E.I.V. was supported in part by NIH grant 5T32AI7647 and an American Cancer Society Postdoctoral fellowship (grant PF-16-048-01-MPC).
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
FOOTNOTES
- Received 21 February 2017.
- Accepted 24 April 2017.
- Accepted manuscript posted online 3 May 2017.
- Copyright © 2017 American Society for Microbiology.
