Human herpesvirus 6A promotes glycolysis in infected T cells by activation of mTOR signaling

Human herpesvirus 6 (HHV-6) is an important immunosuppressive and immunomodulatory virus worldwide. However, whether and how HHV-6 infection influences the metabolic machinery of the host cell to provide the energy and biosynthetic resources for virus propagation remains unknown. In this study, we identified that HHV-6A infection promotes glucose metabolism in infected T cells, resulting in elevated glycolytic activity with an increase of glucose uptake, glucose consumption and lactate secretion. Furthermore, we explored the mechanisms involved in HHV-6A-mediated glycolytic activation in the infected T cells. We found increased expressions of the key glucose transporters and glycolytic enzymes in HHV-6A-infected T cells. In addition, HHV-6A infection dramatically activated AKT-mTORC1 signaling in the infected T cells and pharmacological inhibition of mTORC1 blocked HHV-6A-mediated glycolytic activation. We also found that direct inhibition of glycolysis by 2-Deoxy-D-glucose (2-DG) or inhibition of mTORC1 activity in HHV-6A-infected T cells effectively reduced HHV-6 DNA replication, protein synthesis and virion production. These results not only reveal the mechanism of how HHV-6 infection affects host cell metabolism, but also suggest that targeting the metabolic pathway could be a new avenue for HHV-6 therapy.

Understanding how virus infection manipulates cellular metabolism is an emerging and critical field for effective vaccine development and virus treatment. Increasing evidence suggests that both DNA viruses and RNA viruses can modulate the host cell metabolic profiles after their infection [15,16]. This mechanism also occurs in the infection of human herpesvirus family. Human cytomegalovirus (HCMV) can alter glycolysis, glutaminolysis, and fatty acid synthesis in infected cells, which is required for maximal virus production [17][18][19][20]. Kaposi's sarcoma-associated herpesvirus (KSHV) infection induces a Warburg effect in human endothelial cells (ECs) and lipogenesis in ECs and PEL cells, and these altered metabolic processes are required for maintaining KSHV latency [21,22]. Epstein-Barr virus-encoded LMP1 significantly increases glycolysis in nasopharyngeal carcinoma (NPC) cells through up-regulation of hexokinase 2 (HK2), and then facilitates proliferation by blocking apoptosis [23]. In addition, herpes simplex virus 1 (HSV-1) has been reported to impact tricarboxylic acid (TCA) cycle metabolic fluxes during infection [24]. However, whether and how HHV-6 regulates the metabolism of host cells to support its replication and/or persistence remains unknown.
In this study, we observed for the first time that HHV-6A infection dramatically promotes glycolysis in infected T cells with an increase in glucose uptake, glucose consumption and lactate production. The expression levels of glucose transporter 1 (Glut1), glucose transporter 3 (Glut3) and several key glycolytic enzymes are up-regulated in HHV-6A infected cells. Furthermore, we have demonstrated that HHV-6A infection activates AKT-mTORC1 signaling, which is involved in HHV-6A-mediated glycolysis activation in T cells. In addition, inhibition of glycolysis by 2-DG or pharmacologically inhibiting mTORC1 activity dramatically reduced viral DNA replication, protein synthesis and infectious virion production. cytopathic effects (CPE), immunofluorescence and western blot assay (S1A-S1C Fig). To determine if HHV-6A infection affects cell metabolism of host cells, we conducted a global RNA sequencing analysis of changes in whole genome expression in HHV-6 infected cells compared with mock-infected cells. We found that metabolism-related genes were significantly up-regulated in virus-infected cells. In particular, key glycolytic enzyme genes were markedly increased in HHV-6A infected cells, suggesting heightened glycolytic activity ( Fig  1A). To further confirm this result, we first examined the glucose uptake ability of HSB-2 cells infected with HHV-6A, using a fluorescent glucose analog 2-NBDG (2-(N-(7-Nitrobenz-2-oxa-1, 3-diazol-4-yl) Amino)-2-Deoxyglucose) labeling assay. We found that the uptake of 2-NBDG was increased in HHV-6A-infected HSB-2 cells compared with mock-infected cells at 72 h post-infection, suggesting that the glucose uptake was increased (Fig 1B). We then determined the glucose consumption in HHV-6A infected cells. As shown in Fig 1C, the glucose concentration decreased much faster over time in culture medium of HHV-6A-infected HSB-2 cells than that in mock-infected cells. To determine if the increased glucose uptake of HHV-6-infected HSB-2 cells leads to increased glycolysis, we further examined the production of glycolysis metabolite lactic acid. As shown in Fig 1D, there is a~50% increase (mock 7.57 ±0.11mmol/L versus HHV6 11.69±0.34 mmol/L) in lactic acid secretion in HHV-6A-infected cells compared with that in mock-infected cells at 72 h post-infection. Additionally, the extracellular acidification rate (ECAR), a marker of glycolysis were directly measured using the Seahorse XF extracellular flux analyzer. As shown in Fig 1E, HHV-6A-infected HSB-2 cells showed significantly increased ECARs following the addition of glucose compared with those of mock-infected cells. Then, oligomycin was added to inhibit mitochondrial ATP production and induce maximal glycolysis. Importantly, HHV-6-infected cells showed robust increases in the ECAR levels following oligomycin treatment, compared with mock-infected cells. Finally, ECARs were dramatically reduced both in HHV-6-infected cells and mock-infected cells following the administration of the glycolysis inhibitor 2-deoxy-D-glucose (2-DG). In conclusion, glycolysis, glycolytic capacity and glycolytic reserve were all substantially increased in HHV-6-infected cells in comparison with these parameters in mock-infected cells. Collectively, these results clearly demonstrate that HHV-6A infection increases glucose metabolism in HSB-2 cells.

The expression levels of Glut1 and Glut3 are significantly increased in HHV-6A infected T cells
The glucose transporter (Glut) family is primarily responsible for the uptake of glucose into cells [25,26]. It was reported that Glut1-mediated glucose transport is required for efficient HIV-1 infection of CD4 + T cells and thymocytes [27]. We next determined whether the increase in glucose uptake and glucose consumption in HHV-6A-infected HSB-2 cells is due to alterations of Gluts. RNA sequencing analyses suggested apparent changes in Glut gene expression in HHV-6A-infected cells (Fig 2A). To confirm the RNA sequencing results, we examined the mRNA transcript abundance of the Glut transporter family in HSB-2 cells. It was found that mRNA levels of Glut1 and Glut3 were much higher than other Glut family members in HSB-2 cells (S2 Fig). We therefore examined the transcription levels of Glut1 and Glut3 in mock-infected and HHV-6A infected HSB-2 cells using real-time PCR. As expected, HHV-6A infection significantly increased the mRNA levels of Glut1 and Glut3 in HSB-2 cells at 24 h post-infection, and the expression level progressively increased at 48 h and 72 h post infection (more than a 10-fold increase at 72 h post infection compared with mock-infected cells) (Fig 2B). In addition, we examined the protein levels of Glut1 and Glut3 by the Western blot analysis. Consistent with the mRNA expression levels, the protein levels of Glut1 and Alterations of genes involved in glycolysis were identified and ranked in HSB-2 cells after infection with HHV-6A. Human HSB-2 cells were infected with HHV-6A for 72 hours. Total RNA was purified and pooled, and transcriptome analyses were performed by RNA sequencing. (B) HHV-6A infection significantly increased glucose uptake in HSB-2 cells. HHV-6A infected and mock infected HSB-2 cells were cultured for 72 hours, glucose uptake was determined Glut3 were also increased in HHV-6A-infected cells compared with mock-infected cells ( Fig  2C).
To be functional, the glucose transporters must be located to the cell membrane to uptake glucose into the cells [25,26]. Several reports have shown that virus infection could induce translocation of the Glut1 molecule to the cell membrane [28,29]. We thus hypothesized that HHV-6A infection may also alter Gluts translocation. We determined how Gluts localize and traffic in HHV-6A-infected HSB-2 cells using an immunofluorescence analysis. As shown in Fig 2D, we observed that HHV-6A-infected cells exhibited high fluorescence intensities in cell boundaries compared with mock-infected cells, suggesting a fairly dramatic relocalization of Glut1 and Glut3. Collectively, these results indicate that HHV-6A infection not only increases Glut expression but also promotes the transporters translocation.

HHV-6A infection increases key glycolytic enzymes expression
Considering that the increase in glucose consumption and lactate production were observed in HSB-2 cells during HHV-6A infection, we reasoned that the expression of glycolytic enzymes might be up-regulated by infection. Our transcriptome analyses demonstrated that HHV-6A infection significantly increased the gene expression levels of enzymes involved in glycolysis in HHV-6A infected cells (Fig 1A). We then analyzed the mRNA expression levels of a panel of key glycolytic enzymes using quantitative real-time PCR, including glycolysisrelated enzymes hexokinase 2 (HK2), glucose-6-phosphate isomerase (GPI), phosphofructokinase 1 (PFK1), triosephosphate isomerase 1 (TPI1), enolase 1 (ENO1), pyruvate kinase muscle 2 (PKM2) and lactate dehydrogenase A (LDHα) (Fig 3A). We found that HHV-6A infection markedly increased the expression of the glycolytic enzymes in HHV-6A infected cells compared with mock-infected cells in a time-dependent manner (Fig 3B). We next examined the protein levels of those glycolytic enzymes using the Western blot analysis. As shown in

HHV-6A infection activates AKT-mTORC1 signaling in HSB-2 cells
Mammalian target of rapamycin (mTOR) signaling plays a critical role in the regulation of energy metabolism, cell growth and proliferation [30,31]. RNA sequencing analyses showed that HHV-6A infection induced significant alterations in 16 genes involved in mTOR by flow cytometry after addition of 2-NBDG for 15 min. Results in the right histogram are mean ± SD of fluorescence intensity (MFI) quantifications from three independent experiments. �� p < 0.01, compared with mock infection groups. (C) HHV-6A infection decreased glucose levels in the culture medium of in HSB-2 cells. HHV-6A infected and mock infected HSB-2 cells were cultured for 24 h, 48 h and 72 h, and glucose levels in the culture medium were determined using a Glucose Oxidation Assay Kit. Results shown in histogram are mean ± SD from three independent experiments. � p<0.05 and �� p<0.01, compared with the respective mock infection groups. (D) HHV-6A infected HSB-2 cells produce more lactate than mock-infected cells. The lactate levels in the culture supernatants were determined by the lactate assay kit. Results shown are mean ± SD from three independent experiments. �� p<0.01, compared with the levels in mock infected cells. (E) ECAR in mock-infected and HHV-6A-infected cells was measured using extracellular flux analysis. Mock-and HHV-6-infected HSB-2 cells were cultured for 24 h. The ECAR over time was measured at a basal level and after the injection of glycose, oligomycin (an inhibitor of ATP synthase) and 2-DG (an inhibitor of glycolysis). Statistical analysis of glycolysis, glycolytic capacity and glycolytic reserve are shown on the right. Data are mean ± SD of triplicates � p < 0.05, �� p < 0.01, compared with the levels in mock infected cells. https://doi.org/10.1371/journal.ppat.1008568.g001

PLOS PATHOGENS
HHV-6A promotes glycolysis during infection  (Fig 4A). Knowing that PI3K/AKT is a well-documented upstream site of mTORC1 signaling, which is involved in activation of mTORC1[32, 33], we next determined AKT expression and phosphorylation levels in HHV-6A infected cells. As shown in Fig 4B, HHV-6A infection effectively induced phosphorylation of AKT at Thr308 and Ser473 in virus-infected cells. In contrast, the phosphorylation of AKT expression was hardly detected in the mock-infected cells at various times. Consistent with our results, it was reported that HCMV infection is able to activate AKT signaling, which was shown by the increase in AKT phosphorylation at Ser473 [34]. Given that tuberous sclerosis complex (TSC), a negative regulator of mTORC1, is located downstream of AKT [35,36], we investigated TSC2 (a major component of TSC) expression and its inhibitory phosphorylation levels in HHV-6A infected cells. And we found that HHV-6A infection significantly induced TSC2 phosphorylation at Thr1462 (the major AKT-dependent phosphorylation site), indicating inactivation of TSC during HHV-6 infection (Fig 4C).
We next determined whether mTOR signaling is also activated following HHV-6A infection. We analyzed the expression and phosphorylation status of mTOR and its downstream substrates p70S6K and 4E-BP1 in HHV-6A infected T cells. As expected, we found that HHV-6A infection significantly induced phosphorylation of mTOR (Ser2448) and mTORC1 downstream targets p70S6K (Thr389) in HHV-6A infected cells compared with the mock-infected cells. Furthermore, we found that both total 4E-BP1 and phosphorylated 4E-BP1 (Thr37/46) were induced in cells by HHV-6A infection, while there was only minimal expression in the mock-infected cells (Fig 4D). In support of the above results, we also found that the phosphorylation of AMPK, a kinase that inhibits mTORC1 signaling [37,38], was significantly downregulated following HHV-6A infection (S4 Fig). These data clearly indicate that mTORC1 signaling is significantly activated following HHV-6 infection.

Inhibition of mTORC1 signaling blocks HHV-6A-induced glycolytic activation
To further determine whether mTORC1 signaling is important for HHV-6A-mediated activation of glycolysis, we treated HHV-6-infected cells with rapamycin, a specific inhibitor of mTORC1, and then determined glucose metabolism and lactate production. Consistent with the results shown in Fig 1B, HHV-6A-infected cells had a significantly higher glucose uptake than that of mock-infected cells (Fig 5A). The glucose uptake was markedly decreased in HHV-6-infected cells after rapamycin treatment. Furthermore, rapamycin treatment only had a slight impact on glucose uptake in mock-infected HSB-2 cells (Fig 5A). Next, the involvement of mTORC1 signaling in glucose consumption and lactate production in HHV-6A infected HSB-2 cells was explored. We found that rapamycin treatment significantly inhibited glucose consumption in HHV-6A-infected cells. However, in mock-infected HSB-2, there was no obvious difference in glucose consumption with or without rapamycin treatment (Fig 5B).
In addition, treatment with rapamycin dramatically reduced lactate production in both mockinfected cells and HHV-6-infected cells. However, the inhibition effect was much stronger in HHV-6-infected HSB-2 cells than mock-infected HSB-2. Specifically, in mock-infected cells, expressions of Glut1 and Glut3 in HSB-2 cells. Whole-cell extracts were prepared from mock-infected and HHV-6A-infected infected HSB-2 cells at 24, 48 and 72 hpi. Glut1, Glut3 and β-actin protein levels were determined by Western analysis. (D) HHV-6A infection induced Glut1 and Glut3 expression and translocation. Glut1 and Glut3 expression were determined by an indirect immunofluorescence assay. Mock and HHV-6A-infected HSB-2 cells were stained for nuclear DNA (DAPI, blue) and Glut1 and Glut3 (green) antibodies at 48 h post-infection. Scale bar, 10 μm. The results in the right of the histograms are mean ± SD of fluorescence intensity (MFI) quantifications of Gluts from three independent experiments. � p<0.05, �� p<0.01, compared with the mock-infected group. https://doi.org/10.1371/journal.ppat.1008568.g002

PLOS PATHOGENS
HHV-6A promotes glycolysis during infection rapamycin treatment only slightly decreased lactate production from 4.87±0.199 mmol/L to 3.27±0.031 mmol/L. However, in HHV-6A-infected cells, rapamycin treatment significantly decreased lactate production from 11.02±0.145 mmol/L to 6.38±0.091mmol/L. (Fig 5C). In addition, the ECAR levels was also measured following rapamycin treatment both in mock-  Mock infected and HHV-6A infected cells were lysed and analyzed by western blotting using specific antibodies against Akt, p-Akt (Thr308) and p-Akt (Ser473). Phosphorylated AKT protein levels were quantitatively analyzed and were compared with β-actin expression with a densitometer. Results are means ± SD from three independent experiments. �� p<0.01, compared with the mock-infected group. (C) HHV-6A infection induced phosphorylation of TSC2 in infected HSB-2 cells. Mock infected and HHV-6A infected cells were lysed and analyzed by western blotting using specific antibodies against TSC2 and p-TSC2 (Thr1462). Phosphorylated TSC2 protein levels were quantitatively analyzed and were compared with β-actin expression with a densitometer. Results are means ± SD from three independent experiments. � p<0.05, �� p<0.01, compared with the mock-infected group.  Fig 1E, HHV-6Ainfected cells had a significantly higher glycolysis and glycolytic capacity than that of mockinfected cells (Fig 5D), and these two parameters were both markedly decreased in HHV-6-infected cells after rapamycin treatment. We also found that the glycolytic capacity decreased in mock-infected cells with rapamycin treatment, which might be due to the toxic effect of rapamycin on these cells.
To investigate further the mechanism of HHV-6A-mediated glycolytic activation, we assessed the impact of mTOR inhibition on virally-induced up-regulations of glucose transporters and glycolytic enzymes. Consistent with the results above, we found that the mRNA levels of Glut1, Glut3, HK2, ENO1, LDHα were dramatically increased in HHV-6A infected cells. However, rapamycin treatment significantly reduced these molecules expression in HHV-6A infected cells (Fig 5E). These results clearly suggest that the mTORC1 activation is critical to regulating HHV-6A-mediated glycolytic activation in the infected T cells.

Pharmaceutical inhibition of glycolysis or mTORC1 activity attenuates HHV-6A viral replication
Based on our findings that HHV-6A infection increases glycolytic rate in infected HSB-2 cells, we wanted to understand if direct blockage of glycolysis could attenuate viral replication. We performed quantitative PCR and Western blot analyses to determine viral DNA replication and protein expression. The inhibition of glycolysis in HSB-2 cells was achieved with 2-Deoxy-D-glucose (2-DG), a glucose analog that inhibits the conversion of glucose to glucose-6-phosphate by competing with glucose for binding to HK2. We observed that 2-DG treatment significantly reduced glucose uptake, glucose consumption and lactate production in HHV-6A infected cells (Fig 5A-5C). Furthermore, treatment of HHV-6A infected cells with 2-DG markedly attenuated viral DNA replication and immediate-early and late protein expression (Fig 6A and 6B). In addition, we tested if 2-DG treatment could reduce virion progeny release by analysis of viral DNA contents in cell supernatant using real-time PCR. As shown in Fig  6C, 2-DG treatment significantly reduced the viral DNA content in the cell culture medium compared with the DMSO treated group.
Given that mTORC1 signaling activation is important for HHV-6A-mediated enhanced glycolysis, we determined if mTORC1 signaling inhibition could attenuate virus production similar to that observed with 2-DG treatment. As shown in Fig 6D and 6E, there was a significant decrease in viral DNA levels and protein expression in HHV-6A infected cells after rapamycin treatment. We tested if rapamycin treatment could reduce virion progeny release by analysis of viral DNA contents in cell supernatants using real-time PCR. As shown in Fig 6F, rapamycin treatment significantly reduced the viral DNA content in cell culture medium compared with the DMSO treated group. We also observed that inhibition of mTOR signaling in HSB-2 cells with Torin1, another mTOR inhibitor that blocks both mTORC1 and mTORC2 complexes, dramatically blocked virus DNA replication, protein expression and virus virion production, which were similar to rapamycin treatment (Fig 6G-6I).These results collectively suggest that both glycolysis and mTORC1 activity is important for HHV-6A efficient replication in host cells. activation of the mTORC1 signaling pathway in infected HSB-2 cells. Mock and HHV-6A-infected cells were lysed and analyzed by Western blotting using antibodies against various proteins involved in mTORC1 signaling. P-mTOR, p-p70S6K and p-4E-BP1 protein levels were quantitatively analyzed and were compared with β-actin expression with a densitometer. Results are means ± SD from three independent experiments. �� p<0.01, compared with the mockinfected group. https://doi.org/10.1371/journal.ppat.1008568.g004

PLOS PATHOGENS
HHV-6A promotes glycolysis during infection Discussion HHV-6 was initially isolated from patients with lymphoproliferative disorders [39]. Like other herpesviruses, HHV-6 is capable of persisting in the host after primary infection and hardly cleared from its host. Reactivation of HHV-6 frequently occurs in the immunocompromised patients suffering from HIV infection, cancer or following transplantation, which may cause severe consequences [2]. Both HHV-6A and HHV-6B can infect several types of immune cells, either productively or nonproductively. The primary target for productive infection is the activated CD4 + T lymphocytes [40], which causes significant phenotypic and functional alterations in the infected cells following HHV-6 infection. In infected CD4 + T cells, HHV-6 induces apoptosis [41,42], inhibits IL-2 synthesis [43] and suppresses T-cell proliferation by cell cycle arrest [44], as well as down-regulates the expression of TCR and MHC-I molecules [45,46]. So far, no study relating to the effects on cellular metabolism regulation by HHV-6 infection has been reported. In the current study, we investigated for the first time the molecular process by which HHV-6A induces cell metabolism perturbation in the infected T cells, which may be one of the key mechanisms responsible for HHV-6 successful infection and latency establishment in host cells.
Viruses are obligate parasites that depend on host cells to provide energy and molecular precursors for their successful infection. Manipulation of host cell metabolism is a common strategy utilized by many viruses for their infection and persistence [47,48]. The first landmark study of the viral effect on host metabolism was performed in HCMV infected cells. This study found changes in many metabolic pathways in HCMV infected cells including glycolysis and fatty acid synthesis [17]. Subsequently, many other viruses have been widely studied, including HSV-1, HCV, KSHV, EBV, and Dengue virus [15,16,49]. In this study, we demonstrate that HHV-6 infection significantly increases glucose uptake and lactic acid production in infected T cells, indicating glycolysis activation. Furthermore, we found the expression levels of glucose transporters (Glut1 and Glut3) and glycolytic enzymes (such as HK2, PFK1 and LDHα) were significantly increased in HHV-6 infected cells. In support of our findings, several studies with other viruses have shown that virus infection could activate glycolysis using different manners. It has been reported that HSV-1 infection activates glycolysis through up-regulating the expression and activity of phosphofructokinase-1 (PFK-1) in infected cells [50]. EBV LMP1 prompts aerobic glycolysis by activation of the FGFR1 signaling pathway [51]. More recently, it was shown that KSHV employs specific viral microRNAs targeting the key regulators of glucose metabolism and mitochondrial biogenesis in the infected cells to induce glycolytic activity [52]. Adenovirus E4ORF1 can induce MYC activation and subsequently result in elevated

PLOS PATHOGENS
HHV-6A promotes glycolysis during infection expression of specific glycolytic enzymes in infected cells, which is sufficient to promote glucose metabolism of host cells for virus replication [53]. These studies suggest that glycolysis may be a common metabolic pathway explored by many viruses to support virus propagation [20,49,53,54].
Glucose transporters mediate the transport of monosaccharides, polyols and other small carbon compounds across the membranes of eukaryotic cells [55]. In this study, we found that Glut1 and Glut3 expression is much higher than other glucose transporters, and further increased after HHV-6 infection. However, the mRNA transcripts of Glut2 and Glut4 are hardly detected in transcriptional levels in HSB-2 cells. Unlike HHV-6 infection, HCMV infection was shown to dramatically increase the expression of the adipose tissue-specific Glut4 and eliminate the expression of the ubiquitously distributed Glut1 in infected fibroblasts [20]. Glut3 is generally considered as a neuronal glucose transporter, which has been detected in the peripheral cells [56,57]. Glut3 has both a higher affinity for glucose and greater transport capacity than other the facilitative glucose transporters [58]. In this study, there is an obviously elevated expression of Glut1 and Glut3 in HHV-6A-infected HSB-2 cells, thus enhancing the cell capacity to transport glucose and meet increased energy demands following virus infection.
The glycolytic pathway consists of a series of catalytic reactions which controlled by a number of enzymes, e.g. hexokinase, phosphofructokinase and pyruvate kinase. In this study, we found that HHV-6 enhances glycolysis through upregulation of key glycolytic enzymes HK2 and PFK1 which are the first and third enzymes along the pathway and also the key rate-limiting enzymes that limit the overall glycolytic rate [59,60]. We also found that HHV-6A have no obvious effect on the expression of ENO1 and PKM2 which control the final steps of glycolysis. PKM2 is the rate-limiting enzyme in the last stage of the glycolytic pathway and when its catalytic rate is lower than the upstream glycolytic rate, phosphoenolpyruvate (PEP) and its upstream glycolytic intermediates may accumulate for the synthesis of macromolecules [61]. Thus, we speculate that during HHV-6 infection there will be an accumulation of glycolytic intermediates which can be used for the biosynthesis of lipids, nucleotides, and proteins required for rapidly virus reproduction. mTOR is a highly conserved serine/threonine kinase that controls cell growth, cell proliferation and energy metabolism [62,63]. In this study, we demonstrated that HHV-6A infection dramatically promotes the phosphorylation levels of AKT, TSC2, mTORC1, P70S6K and 4E-BP1, all indicating mTORC1 activation. Additional studies demonstrated that inhibition of mTORC1 activity using a specific pharmaceutical inhibitor, rapamycin, not only decreases glycolysis in the infected T cells but also inhibits virus replication. In molecular levels, it was shown that the key glycolytic genes (Glut1, Glut3, HK2, ENO1 and LDHα) expression are down-regulated in HHV-6A infected cells after rapamycin treatment. From these data, we concluded that HHV-6A-induced glycolysis activation is mainly depended on AKT-mTORC1 signaling activation, resulting in up-regulation of glucose transporters and glycolytic enzymes expression, and subsequently elevating glycolysis in infected cells. In support of our assertion, several studies from other groups have also shown that the mTORC1 pathway plays an important role in virus infection. It has been reported that HCMV infection activates the mTORC1 signaling in the infected cells [64]. EBV LMP1 upregulates Glut1 transcription to control h post infection, respectively. (C), (F) and (I) Real-time PCR quantification of viral progeny production in the supernatant of HHV-6A infected cell treated with 2-DG, rapamycin or Torin1. At 72 h post infection, the viral DNA from supernatant was extracted and the amount of viral U22 gene was determined by real-time PCR. Expression levels of U22 gene in HHV-6A infected cells served as 1.Values are means ± SD from three independent experiments. �� p<0.01, compared with the untreated group. https://doi.org/10.1371/journal.ppat.1008568.g006

PLOS PATHOGENS
HHV-6A promotes glycolysis during infection aerobic glycolysis and tumorigenic growth of NPC cells through activation of mTORC1/NF-kappaB signaling [65]. Avian reovirus structural protein sigmaA activates the mTOC1/eIF4E/ HIF-1α pathway to enhance glycolysis and the TCA cycle in the infected cells for virus replication [66]. In this study, we demonstrated that the phosphorylation of AMPK, a kinase that inhibits mTORC1 signaling [37,38], was down-regulated following HHV-6A infection (S4  Fig), further supporting HHV-6 mediated-mTORC1 activation. In contrast, it has been reported that HCMV infection increases AMPK activity which was found to be necessary for glycolytic activation and high titer replication [67,68]. These contradictory results suggest that the regulatory pathway and mechanism induced by HHV-6 infection might be different from other herpesvirus members.
Metabolism is a significant factor determining immune cells survival, proliferation, and the ability to perform its specific immune functions [69,70]. Since viruses also rely on some similar nutrients and metabolic pathways used by immune cells, understanding the effects of modulating metabolic pathways by virus will be important in assessing the overall immune function of the host cells. Viruses may potentially take advantage of metabolism to alter immune responses to evade clearance [71]. HHV-6 is an important immunosuppressive virus and can utilize a variety of strategies to modulate or suppress host immune responses to facilitate virus spread and persistence [72,73]. We speculate that manipulation of cell metabolism maybe another important strategy used by HHV-6 to escape antivirus immune responses and maintain the latency in host. Future studies should be aimed at identifying the interplay between metabolism and immune system during HHV-6 infection, which is critical to the development novel therapeutic strategies to restrain viral replication and enhance antiviral immune responses.

Cells and viruses
The human T-lymphoblastoid cell line HSB-2 was purchased from the American Type Culture Collection (ATCC) and cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS). The GS strain of HHV-6 variant A was propagated in cord blood mononuclear cells (CBMCs). Viral titers were determined as viral DNA equivalents by quantitative PCR. A multiplicity of infection (MOI) of 20 virus DNA copies per cell was used for all the experiments. The infection of HSB-2 cells by HHV-6A strain GS was performed as previously described [44].

Chemical reagents
Rapamycin and Torin1 was purchased from the Selleck and used at 100 nM, and 5 nM respectively. 2-deoxy-D-glucose (2-DG) was also purchased from the Selleck and used at a final concentration of 1mM.

Real-time quantitative RT-PCR analysis
Total RNA was extracted from mock and HHV-6A-infected HSB-2 cells using TRIzol reagent (Invitrogen, USA), cDNA was generated using a PrimeScript II RT kit (Takara) according to manufacturer's instructions. Quantitative RT-PCR was performed on a 7500 fast PCR System (Applied Biosystems) using the TB Green Premix Ex Taq (Takara). Data were normalized to the expression of β-actin. Additionally, viral DNA was extracted from HHV-6A-infected T cells and supernatants, respectively, and viral DNA was quantified using the U22 primer set. Quantitative PCR primer pairs are listed in S1 and S2 Tables.

Glucose uptake assay
HSB-2 cells were infected with or without HHV-6A for the indicated time points. The 1640 culture medium was replaced by glucose-free medium with 10% FBS. After incubating the cells for another 30 min, 2-NBDG (Cayman Chemical, USA) was added at a final concentration of 100 μM for 15 min and cells were collected and washed twice with PBS. The cells were ready to be analyzed by flow cytometry.

Measurements of glucose consumption and lactate production
Cells (5× 10 5 /well) were seeded in a 6-well plate, and infected with HHV-6A in the presence or absence of inhibitors (Rapamycin or 2-DG). Mock-infected and HHV-6A-infected HSB-2 cells were cultured for 24, 48 and 72 h. Glucose and lactate concentrations in the culture media were measured using the Glucose Oxidation Assay Kit (Applygen Technologies) and the Lactate Assay Kit (Jiancheng Technologies), respectively, according to the manufacturer's instructions. Glucose consumption and lactate measurements were normalized to cell numbers.

Metabolic assays
The measurement of extracellular acidification rate (ECAR) were determined with the Seahorse XF96 Extracellular Flux Analyzer (Seahorse Bioscience). Experiments were performed according to the manufacturer's protocols. Briefly, mock and HHV-6A-infected HSB-2 cells were cultured for 24 h, and the cells were then plated on cell-tak-coated Seahorse XF96 cellculture microplates at a density of 5×10 4 cells per well. The ECAR were measured under basal condition after sequentially adding 10mM glucose, 4μM oligomycin and 100mM 2-DG (XF Glycolysis Stress Test kit, Agilent, USA) at indicated time points. For some experiments, the metabolic profiles were evaluated in the presence or absence of rapamycin in cultured HSB-2 cells. The parameters of cellular glycolysis (including glycolysis, glycolytic capacity and glycolytic reserve) were calculated from ECAR profiles. Glycolysis was determined after the addition of glucose, and the glycolytic capacity was assessed after the addition of oligomycin. Glycolytic reserve was defined as the difference between glycolytic capacity and glycolysis.

RNA sequencing analysis
Mock-infected and HHV-6A-infected HSB-2 cells were cultured for 72 h. Total RNA was extracted using TRIzol reagent (Invitrogen, USA) following the manufacturer's instructions. A total amount of 3 μg of RNA per sample was used as input material for the RNA sample preparations. Sequencing libraries were generated using NEBNext Ultra Directional RNA Library Prep Kit for Illumina (NEB, USA) following the manufacturer's recommendations. After cluster generation, the library preparations were sequenced on an Illumina Hiseq 2500 platform by CapitalBio Technology. Differentially expressed genes were identified by fold-change screening. Gene ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed to systematically identify differentially expressed genes and the pathways associated with HHV-6A infection.

Viral genome and protein quantification
Cells (5 × 10 5 /well) were infected with HHV-6A and seeded in a 6-well plate with or without inhibitors (2-DG, Rapamycin or Torin). The medium was replaced with fresh medium at 48 h post-infection, and the cells were incubated for an additional 24 h. Total DNA was extracted from HHV-6A-infeted cells and supernatant respectively. Viral genomes were quantified by the real-time PCR using the primers for viral U22 gene detection. HHV-6 immediate early protein, IE1 and late protein, gB were detected by Western blot analysis.

Statistical analysis
Statistical analysis was performed using GraphPad Prism5 software. The data are presented as mean ± standard deviation (SD). Statistical significance was determined by a Student's t-test or one-way analysis of variance (ANOVA). P value of < 0.05 was considered significant.