The Cytomegalovirus protein pUL37×1 targets mitochondria to mediate neuroprotection

There is substantial evidence that mitochondrial dysfunction plays a significant role in the pathogenesis of Parkinson disease (PD). This contribution probably encompasses defects of oxidative phosphorylation, mitochondrial turnover (mitophagy), mitochondrial derived oxidative stress, and apoptotic signalling. Human cytomegalovirus immediate-early protein pUL37 × 1 induces Bax mitochondrial translocation and inactivation to prevent apoptosis. Over-expressing pUL37 × 1 in neuronal cells protects against staurosporin and 6-hydroxydopamine induced apoptosis and cell death. Protection is not enhanced by bax silencing in pUL37 × 1 over-expressing cells, suggesting a bax-dependent mechanism of action. pUL37 × 1 increases glycolysis and induces mitochondrial hyperpolarization, a bax independent anti-apoptotic action. pUL37 × 1 increases glycolysis through activation of phosphofructokinase by a calcium-dependent pathway. The dual anti-apoptotic mechanism of pUL37 × 1 may be considered a novel neuroprotective strategy in diseases where mitochondrial dysfunction and apoptotic pathways are involved.

Mitochondrial dysfunction is thought to play an important role in the pathogenesis of Parkinson disease (PD) (review by Schapira) 1 . This includes the mitochondrion's role in oxidative phosphorylation, free radical generation, triggering apoptosis and alteration of mitochondrial turnover (mitophagy). The mitochondrion therefore presents multiple pathways for which to target interventions that could prevent or reverse the deficits and potentially favourably influence the course of PD.
Many viruses, including human cytomegalovirus (CMV), encode proteins that inhibit apoptosis, a powerful innate defence mechanism against viral infection 2,3 . UL37 exon 1 protein (pUL37 × 1), which is also known as a viral mitochondria-localized inhibitor of apoptosis, is encoded by the immediate early ul37 × 1 gene 4 . pUL37 × 1 inactivates Bax and inhibits apoptosis by causing the mitochondrial translocation and conformational change of Bax [5][6][7] . The Bax-binding function of pUL37 × 1 is essential for the apoptosis prevention, survival of the host cell and replication of the virus. Cells in which the bax gene has been silenced are not protected from staurosporin-induced apoptosis.
The anti-apoptotic action of pUL37 × 1 therefore offers a unique and novel mechanism for neuroprotection. We now demonstrate that pUL37 × 1 over-expression protected against toxin-induced cell death and apoptosis in PD cellular models. pUL37 × 1 over-expression protected against cell death and although Bax translocated to mitochondria, apoptosis was prevented. In addition, pUL37 × 1 over-expression also increased cellular glycolysis and hyperpolarized mitochondria, actions which contributed to the neuroprotective mechanism of pUL37 × 1 in these models.

pUL37 × 1 Over-expression Protected Against Toxin-Induced Apoptosis and Cell Death.
Cytochrome c, an electron carrier between complex III and IV, is also known to induce apoptosis when it is released into the cytoplasm. pUL37 × 1 over-expression significantly prevented the release of cytochrome c from mitochondria to cytoplasm. Staurosporine is a protein kinase C inhibitor and apoptosis inducer; 500 nM treatment for 3 hours caused increased cytochrome c release in the control which was blocked in pUL37 × 1 over-expressing cells ( Fig. 2A). In addition, pUL37 × 1 over-expression significantly reduced the cytosolic cytochrome c in the absence of toxin.
The activation of caspase-3 is an important component of the apoptosis pathway. pUL37 × 1 over-expression was expected to prevent the activation of caspase-3 due to blocking of the upstream apoptosis process. In the absence of toxin, pUL37 × 1 over-expression did not affect the activity of caspase-3 compared to control. However, following 500 nM staurosporine treatment for 4 hours, pUL37 × 1 over-expressing SH-SY5Y cells exhibited significantly lower caspase-3 activity compared to control (Fig. 2B). Based on these findings, pUL37 × 1 over-expression is able to block the apoptosis cascade induced by staurosporine.
pUL37 × 1 Over-expression Protected against Apoptosis and Neuronal Death in Rat Primary Cortical Cultures. Primary rat cortical cultures were used to investigate further the protection conferred by pUL37 × 1 over-expression. Liposomal transfecting reagent mixed with red fluorescence protein (RFP) plasmid (control) or a mixture of pUL37 × 1 and RFP plasmids at 10 to 1 ratio (experiment) were co-incubated with the cultures for 48 hours. pUL37 × 1 over-expression was identified by the expression of RFP.
resulted in a significant increase in apoptosis in cortical neurons. pUL37 × 1 over-expression alone did not increase apoptosis compared to control. However, following 20 μM 6-OHDA treatment for 6 hours, pUL37 × 1 over-expression significantly reduced the percentage of apoptotic cells (Fig. 3A,B). Similar results were obtained using Fluoro-Jade C (FJ-C), as a specific marker for degenerative neurons. pUL37 × 1 over-expression significantly reduced the neuronal death labelled by FJ-C compared with control following 10 μM 6-OHDA treatment for 24 hours (Fig. 3C,D). The transfection efficiency was approximately 5%. The representative images of apoptotic cell counting by positive cleaved caspase-3 stain (green) on RFP and pUL37 × 1 transfecting rat primary cortical culture. Because the ratio of concentration between pUL37 × 1 and RFP is 10:1, expression of RFP (red) is assumed as pUL37 × 1 expressing neurons. The apoptotic neurons were identified based on the RFP-expressed cells with positive cleaved caspase-3 stain (arrowhead) whereas RFP-expressed cells without cleaved caspase-3 stain were recognized as survival neurons. Low magnification images (left) were used to evaluate incidents of apoptosis whereas high magnification images (right, corresponding to the boxed area in the low magnification images) show the representative immunoreactivity of cleaved caspase-3 on transfected neurons. For each rat cortical culture preparation, at least ten low magnification images were taken to quantitative analysis. (B) pUL37 × 1 expression significantly reduced the percentage of apoptotic cells induced by 20 μM 6-OHDA treatment for 6 hours compared with RFP expressing ones (21.8 ± 1.6% in RFP-control versus 13.3 ± 2.1% in pUL37 × 1 over-expression, p < 0.05, n = 6). (C) The representative images of fluoro-jade C (FJ-C) stain. Similar to the detection of apoptotic neurons, RFPexpressed neurons (red) with positive FJ-C stain (green, double arrowhead) were counted as death neurons whereas RFP-expression only neurons were survival neurons. Low magnification images (left) were used for counting dead neurons whereas high magnification images (right, corresponding to the boxed area in the low magnification images) show the representative fluorescent FJ-C staining. (D) pUL37 × 1 expression significantly reduced the percentage of FJ-C positive neurons induced by 10 μM 6-OHDA treatment for 24 hours compared with RFP expressing ones (25.3 ± 1.9% in RFP-control versus 17.4 ± 2.7% in pUL37 × 1 over-expression, p < 0.05, n = 6). Data were presented as mean ± S.E.M. Statistics was performed by two-tailed Student's t-test. (NS, non-significant, *p < 0.05 ).
Scientific RepoRts | 6:31373 | DOI: 10.1038/srep31373 The protection of pUL37 × 1 is Bax-dependent. The mitochondrial sequestration and inactivation of Bax has been postulated as an anti-apoptotic mechanism of pUL37 × 1 5-7 . In the present study, pUL37 × 1 over-expressing cells exhibited some co-localization of Bax with mitochondria, supporting a mitochondrial translocation of Bax (Supplement S3A).
To investigate this protective mechanism further, both control and pUL37 × 1 over-expressing cells were treated with either scrambled siRNA or bax siRNA for 3 days. bax siRNA silencing substantially reduced the expression of Bax compared with scrambled siRNA (Supplement S3B). Following challenge with 30 nM staurosporin for 24 hours, bax siRNA silencing reduced the percentage of LDH release compared with scrambled siRNA silencing in control cells. In the presence of scrambled siRNA silencing, pUL37 × 1 over-expression remained protective against staurosporin. However, the combination of bax siRNA silencing with pUL37 × 1 did not modify LDH release (Fig. 4). These data indicate that the anti-apoptotic effect of pUL37 × 1 over-expression was dependent on Bax activity.

Glycolysis-induced Mitochondrial Hyperpolarization as a Protective
Mechanism of pUL37 × 1. Mitochondrial membrane depolarization is a common phenomenon during apoptosis 9 , and has been reported to take place before Bax translocation 10 . Hyperpolarization of mitochondria has been shown to be a protective strategy 11 . pUL37 × 1 over-expression induced hyperpolarization of mitochondria in SH-SY5Y cells (Fig. 5A,B). This hyperpolarization was reduced but was still significantly above control following exposure to 6-OHDA (Fig. 5C).
Mitochondrial hyperpolarization can result from complex V inhibition, insufficient ADP supply or increased glycolysis 12 . However, the two former result in energy deficit and lead to cell death, which did not occur at baseline in pUL37 × 1 over-expression (Fig. 5C). CMV infection has been found to increase glycolysis in host cells. pUL37 × 1 over-expression significantly increased glycolysis measured by the XF analyser compared with control in SH-SY5Y cells (Fig. 5D). This glycolysis-enhancing effect did not result from oxidative phosphorylation (OXPHOS) inhibition as there was neither significant difference in oxygen consumption rate (OCR) between control and pUL37 × 1 over-expressing SH-SY5Y cells (Fig. 5D), nor reduced ADP phosphorylation and enzymatic activity from each mitochondrial respiratory chain complex (Supplement S4).
To determine whether the glycolytic-dependent mitochondrial hyperpolarization contributes to the protective effect of pUL37 × 1 we assessed the effect of glycolysis inhibition with toxin exposure. Both 10 μM 2DG and 5 μM 3BP attenuated but did not prevent the protective effect of pUL37 × 1 over-expression against 6-OHDA (Fig. 5F).

Discussion
The present study demonstrated the neuroprotective effect of pUL37 × 1 over-expression against staurosporin and 6-OHDA-induced apoptosis and cell death in both SH-SY5Y and primary rat cortical cultured cells. This protective effect of pUL37 × 1 appears to be mediated through two separate mechanisms, both of which modulate mitochondria-dependent apoptosis. pUL37 × 1, located on the mitochondrial outer membrane, led to the mitochondrial translocation and inactivation of Bax. The second mechanism involved mitochondrial hyperpolarization with increased calcium release and enhanced PFK-mediated glycolysis and hyperpolarization. Cell death protection was attenuated with inhibitors of glycolysis and of calcium release (summary in Fig. 7). The strategy of using viral derived proteins as a means to target mitochondria for neuroprotection has recently been supported by using a protein derived from Borna disease virus 17 and the β-2.7 RNA from CMV 18 .
The advantages of using stable pUL37 × 1 over-expressing lines instead of transient transfection include long-term gene expression, consistency and the option of further genetic manipulation 19 . After antibiotic selection and pUL37 × 1-HA expression confirmation by western blot analysis, we selected three lines in order to eliminate the impact of possible cloning artefacts. All three lines presented a homogenous pUL37 × 1-HA expression by immunocytochemistry, which also confirmed the mitochondrial localization of pUL37 × 1 in SH-SY5Y cells. This has been reported previously in HeLa cells 20 and the first domain of pUL37 × 1 has been identified as responsible for mitochondria targeting and localization 21 .
in SH-SY5Y and primary rat cortical cells reduced the release of cytochrome c and activation of caspase-3 and decreased cell death upon staurosporin and 6-OHDA treatment. It is still important to measure cell death protection after confirmation of anti-apoptotic effect because in some occasions, down-regulation of apoptosis can promote necrosis 23,24 . In SH-SY5Y cells, the protection was present under several different conditions, including high dose, short-term toxin treatment and moderate dose, longer-term toxin treatment. These results indicate that pUL37 × 1 is anti-apoptotic in neuronal cells and neuroprotective in cellular models.
Bax, a pro-apoptotic bcl-2 family protein, is thought to play an important role in apoptotic neuronal death in PD 25,26 . Bax ablation has been demonstrated to have an anti-apoptotic and neuroprotective effect. Vila et al. show that Bax deficient mutant mice were resistant to MPTP neurotoxicity 27 . In the present study, bax siRNA silencing reduced Bax protein expression levels and resulted in cell death protection against staurosporin in control cells. However, silencing of bax in pUL37 × 1 over-expressing cells did not provide further cell death protection against staurosporin compared with pUL37 × 1 expressing cells. Failure of cell death protection by bax siRNA silencing in pUL37 × 1 over-expressing cells indicated a Bax-dependent neuroprotective mechanism of pUL37 × 1, which support the Bax inactivation effect of pUL37 × 1 from previous reports [5][6][7] .
Mitochondria can become hyperpolarized due to the reverse of proton trafficking through complex V. When the intracellular ATP is sufficient, complex V consumes ATP to pump protons out rather than in. The main source of extra-mitochondrial ATP generation is glycolysis. Glycolysis is an alternative energy resource to oxidative phosphorylation, and though it might be seen as bioenergetically inefficient, due to low ATP and NADP yield, because of its fast reaction rate, glycolysis is able to provide sufficient energy to meet cellular demand 28 . Cancer cells are thought to generate ATP predominantly from glycolysis even in oxygen-rich conditions, the Warburg effect 29 . Owing to this alteration of glucose metabolism, cancer cells usually contain hyperpolarized mitochondria and are less sensitive to apoptosis [30][31][32] . Cells with higher glycolytic capacity, such as astrocytes and myocytes, also utilize glycolysis to protect cell death by enhancing mitochondrial membrane potential 33,34 . In in vitro PD models, higher glycolysis activity contributed to neuroprotection against MPTP [35][36][37] . We have demonstrated a novel protective effect of pUL37 × 1 that is mediated via the enhancement of glycolysis and leads to the hyperpolarization of the mitochondrial membrane potential. Glycolytic inhibitors attenuated the hyperpolarization and protective effects of pUL37 × 1 over-expression. Mitochondrial hyperpolarization also resulted in reduced cytosolic cytochrome c level in the absence of toxin. In physiological status, cytochrome c is loosely attached to the mitochondrial inner membrane due to the presence of mitochondrial membrane potential. Hyperpolarization of mitochondrial membrane potential enhances the binding affinity and reduce the leakage of cytochrome c to the cytoplasm 38,39 .
Glycolysis is tightly regulated at several stages and PFK is the most crucial control point in the glycolytic pathway. The enzymatic activity is tightly regulated by the ATP/AMP ratio, citrate, and fructose-2,6-bisphosphate. Intracellular calcium also plays an important role in the activity of PFK: high intracellular calcium leads to the binding of calmodulin to the high-affinity site of PFK monomer and stabilizes the PFK dimeric form with full activity 15 . MPTP-treated monkeys demonstrated a decrease of the enzyme activities related to the glycolytic pathway, especially the decrease in PFK 40 . It has been postulated that CMV infection increases glycolysis by activation of the calmodulin-dependent pathway 41 and pUL37 × 1 is essential for this biogenergetic effect because it triggers the release of calcium storage from the ER 16 . In the present study, pUL37 × 1 over-expression did not alter the expression level of PFK and other glycolytic enzymes, but did increase PFK activity. Moreover, calcium caging by BAPTA-AM attenuated the increased glycolytic capacity of pUL37 × 1, which also supports the role of calcium-dependent increased glycolysis. It is known that alteration of intracellular Ca 2+ is related to apoptosis. Also known is that Ca 2+ content in ER determines the sensitivity to apoptotic stimuli [42][43] . Dysregulated Ca 2+ signalling is considered as a contributing factor in the pathogenesis of PD 44 , in which the specific involvement In the present study, we demonstrated that pUL37 × 1 over-expression protected apoptotic death through two pathways: one protective mechanism of pUL37 × 1 depended on Bax inactivation and the other was due to mitochondrial hyperpolarization. pUL37 × 1 over-expression induced the increased intracellular calcium, which activated phosphofructokinase (PFK) and up-regulated glycolysis, followed by mitochondrial hyperpolarization.
The usage of two toxins, staurosporine and 6-OHDA in the present study, serves different purposes. Staurosporine is protein kinase c inhibitor and a strong apoptosis inducer as well. It triggers remarkable mitochondrial depolarization and Bax translocalization 48 . The protection of pUL37 × 1 against staurosporine-induced neuronal apoptotic death confirmed that pUL37 × 1 is able to be neuroprotective through anti-apoptosis. 6-OHDA is served to provide a PD in vitro model, which induced both apoptosis and necrosis though excessive oxidative stress [49][50][51] . pUL37 × 1 reduced not only caspase-3 activation but also total cell death, which reflect the summation of apoptosis and necrosis.
In conclusion, the mitochondrial mechanisms by which pUL37 × 1 inhibits apoptosis and prevents cell death may serve as a model for potential neuroprotective therapies that involve mitochondrial pathways.

Material and Methods
Cells and plasmids. SH-SY5Y cells were maintained as previous descriptions 52 with reagents supplied from Life Technologies (Paisley, UK). The method for setting up E18 primary rat cortical culture was adapted from previous well-established protocols 53 , and they were used 7 days post-culture. pul37 × 1 open reading frame (coding sequence complement joining 50262..51197,51302..51344,52573..53060; Genbank accession NC_006273) added with a 3′ haemagluttinin (HA) epitope (total reading frame length 1494 nucleotides) was cloned into pcDNA3.1 plasmid. Stable ectopic expression was established in SH-SY5Y cells.
Protein assay. Whole cell lysates for Western blot analysis were prepared in 10 mM Tris-HCl pH 7.5, 0.1% of SDS and in the presence of protease inhibitors (Fisher Scientific, Loughborough, UK), followed by DNAseI digestion (Promega, Southampton, UK). Protein samples were separated under reducing conditions by SDS-PAGE using the Novex system (Life Technologies), transferred to PVDF (Millipore, Watford, UK ) and were analyzed by a standard Western blot protocol using Amersham ECL Western Blotting Detection Reagent and Amersham Hyperfilm (GE Healthcare, Little Chalfont, UK). The following antibodies were used: anti-HA epitope antibody (clone 16B12, Life Technologies), anti-β-actin antibody (Abcam, Cambridge, UK), anti-Bax (NT) antibody (Millipore), anti-phosphofructokinase antibody, anti-pyruvate kinase antibody, and hexokinase 1 and 2 antibody (New England Biolabs, Hitchin, UK ). Protein concentration was decided by bicinchoninic acid assay (BCA) according to manufacturer's instructions (Fisher Scientific).
Immunocytochemistry and confocal microscopy. For immunocytochemistry, cells were fixed with 4% paraformaldehyde in phosphate buffered solution (PBS), permeabilized by adding cold methanol, and then processed in succession with a primary antibody, either anti-HA epitope antibody (Cambridge Bioscience, Cambridge,UK), anti-TOM20 antibody (Santa Cruz) or anti-Bax (NT) antibody (Millipore), Alexa Fluor secondary antibody (Life Technologies), and mounted with Citifluor (London, UK) in the presence of DAPI. Fluorescent images were obtained by epifluorescent microscope. In order to obtain image slices <1.5 μm, a Zeiss 510 laser scanning microscope was used.
Apoptosis and cell death assay. Cytochrome c ELISA was performed according to manufacturer's instructions ( Life Technologies). Cells were lysed by a buffer containing 10 mM Tris, pH 7.4, 100 mM NaCl, 1 mM EDTA, 1 mM EGTA, and 10% glycerol incubated for 10 minutes. Cytosolic fraction was separated by a centrifugation at 12,000× g at 4 °C for 10 minutes. Caspase-3 activity was measured from cell lysates according to the manufacturer's instruction of the EnzChek ® Caspase-3 Assay Kit #2 (Life Technologies) which has been described previously 54 . Apoptotic neurons in primary rat cortical culture were identified by anti-cleaved caspase-3 antibody (New England Biolabs). Apoptosis was induced by 20 μM 6-OHDA treatment for 6 hours and followed by the aforementioned immunocytochemistry protocols. LDH release assay (Roche, Burgess Hill, UK) was used to quantify the cell death and the protocols had been described 55 . PI staining was used to distinguish between live and dead cells. Both cell death measurements are based on similar principle: the ratio of LDH release/PI fluorescence between experiments to Triton X-100 treated control would be recognized as the percentage of cell death. For the detection of neuronal death in primary culture, cells were stained with 0.001% Fluoro-Jade C (Millipore) after fixed with 4% paraformaldehyde. Positive stained cells were counted under by epifluorescent microscope. The count of the death/apoptosis neurons is made under blinded condition. For each cover slips, 8-10 fields were selected. The number of RFP expressed cells with positive FJ-C or cleaved caspase-3 stain divided by the total number of RFP expressed cells was the percentage of death/apoptosis. siRNA silencing. 10 5 SH-SY5Y cells suspended in 500 μl growth medium were incubated with 6 nM siRNA (Dharmacon, GE Health) and 4.5 μl of HiPerFect transfection reagent (Qiagen, Manchester, UK) at 37 °C/5% CO 2 for 24 hours. The efficiency of silencing was confirmed by Western blot analysis.

Mitochondrial membrane potential measured by Tetramethylrhodamine, methyl ester(TMRM).
Steady stable mitochondrial TMRM (Life Technologies) fluorescence was measured from confocal images as previously described 56,57 . Images were analyzed by ImageJ software (National Institutes of Health, Maryland, U.S.) following protocols of analysis that have been previously described 58 . OCR and extracellular acidification rate (ECAR). OCR and ECAR were measured by the XF extracellular flux analyzer following manufacturer's instructions (Seahorse Bioscience, MA, US). For measuring OCR, 25 mM glucose, 1 μg/ml oligomycin and 2 μM rotenone were loading sequentially whereas for ECAR, 25 mM glucose, 1 μg/ml oligomycin and 100 μM 2DG were applied. 30 nM BAPTA-AM (Life Technologies) which was used to cage intracellular calcium was applied in ECAR measurement after glucose addition.
SCIeNTIfIC REPORTS | 6:31373 | DOI: 10.1038/srep31373 PFK assay. PFK enzymatic activity was measured according to manufacturer's protocols (Sigma-Aldrich, Dorset, UK). ADP phosphorylation ability and mitochondrial respiratory chain complex enzymatic activity measurement. The measurement of ADP phosphorylation ability and respiratory chain complex enzymatic activity had been described by Gegg et al. 59 . Enzymatic activity was measured by spectrophotometric methods and ADP phosphorylation was measured by ATP Bioluminesence Assay kit HSII. Intracellular and mitochondrial calcium measurement. Cells were incubated with 5 ng/ml of Fluo-4 FF, AM and Rhod-2, AM (Life Technologies) in standard HBSS in the presence of 0.0025% Pluronic for 30 minutes at room temperature, followed by another incubation in HBSS alone for another 30 minutes. Steady-state confocal images were obtained using a Zeiss 510 laser scanning microscope, to calculate the average net fluorescence after background subtraction. Fluo-4 and Rhod-2 were excited at 48 nm and 543 nm respectively.
Statistics. All data were presented as mean ± standard error of mean (S.E.M.). Statistics was performed by either two-tailed Student's t-test or one-way ANOVA with post-hoc analysis. p value less than 0.05 is recognized as significant.