Murine cytomegalovirus infection exacerbates complex IV deficiency in a model of mitochondrial disease

The influence of environmental insults on the onset and progression of mitochondrial diseases is unknown. To evaluate the effects of infection on mitochondrial disease we used a mouse model of Leigh Syndrome, where a missense mutation in the Taco1 gene results in the loss of the translation activator of cytochrome c oxidase subunit I (TACO1) protein. The mutation leads to an isolated complex IV deficiency that mimics the disease pathology observed in human patients with TACO1 mutations. We infected Taco1 mutant and wild-type mice with a murine cytomegalovirus and show that a common viral infection exacerbates the complex IV deficiency in a tissue-specific manner. We identified changes in neuromuscular morphology and tissue-specific regulation of the mammalian target of rapamycin pathway in response to viral infection. Taken together, we report for the first time that a common stress condition, such as viral infection, can exacerbate mitochondrial dysfunction in a genetic model of mitochondrial disease.


Introduction
Mitochondrial diseases (MDs) are a group of progressive tissue-specific or multi-systemic disorders that are caused by defects in energy production [1]. MDs are caused by mutations in mitochondrial genes, or nuclear genes that encode mitochondrial proteins, which together constitute the most commonly inherited metabolic disorders worldwide [2]. Although not reported in the literature, in the clinic, MD patients commonly experience progressive deterioration in their clinical symptoms and suffer from profound fatigue and lethargy during infections. This effect is hypothesized to occur due to the large metabolic demand and stress imposed on mitochondria required to provide the necessary energy to overcome infection. MD patients are expected to make a full recovery following infection, however, in the case of severe infections, prolonged or reoccurring infections, patients may experience permanent regression in their clinical symptoms. It remains unknown how environmental stress conditions such as infections could potentially trigger the onset or potentiate the severity of debilitating symptoms suffered by MD patients.
Recently we developed a model of mitochondrial disease that is caused by a mutation in a nuclear gene encoding the translational activator of cytochrome c oxidase subunit 1 (TACO1). To date, TACO1 is the only known translational activator in mammalian mitochondria [3,4]. A homozygous mutation in the TACO1 gene has been identified in patients suffering from Leigh Syndrome (LS) who have an isolated cytochrome c oxidase (or complex IV) deficiency that leads to progressive cognitive dysfunction, dystonia and visual impairment [3]. The identified TACO1 homozygous mutation in patients is a single-base-pair insertion in position 472 (472insC) that results in a frame shift and generates a premature stop codon, which consequently causes loss of the TACO1 protein in LS patients. The TACO1 protein is a member of a large family of poorly characterised proteins that contains a 297 amino acid DUF28 domain and is imported into the mitochondrial matrix. TACO1 is composed of three domains that form a hook-like structure, domain 1 has positively charged residues that associate with the mt-Co1 mRNA and domain 2 is required for TACO1 stability by linking domains 1 and 3 [4]. The Taco1 mutation in mice results in the loss of the TACO1 protein and causes similar pathologies to those observed in LS patients harbouring TACO1 mutations. Molecular defects are detectable in Taco1 mutant mice from 4 weeks of age, with MD symptoms evident from 20 weeks of age [4], analogous to the patients who show clinical symptoms from ages 7-11 [3].
A large proportion of the human population (50-80%) harbours a latent human CMV (HCMV) infection that is typically contracted in early childhood. The CMVs are a family of species-specific viruses with murine CMV (MCMV) widely utilized as a model for HCMV infection due to the similarities in structure and biology between the two viruses [5]. We used our Taco1 model of MD, to investigate the effects of a common viral infection on the onset and progression of MD. Here we show that infection with MCMV in a model of cytochrome oxidase c deficiency exacerbates the complex IV deficiency and mitochondrial dysfunction, causes alterations in mTOR signalling, and morphological changes in neuromuscular connections. These findings reveal that viral infection is a significant environmental modifier of genetic mitochondrial disorders and the avoidance or treatment of influencing environmental factors is vital in the management of MD.

Taco1 mutant mice exhibit no overt immunological defects following MCMV infection
Taco1 mut/mut or Taco1 wt/wt mice were infected with MCMV and their body weights monitored. Over the course of the experiment weight gain by MCMV-infected Taco1 mut/mut mice was similar to uninfected Taco1 mut/mut mice ( Fig 1A). Furthermore, no difference in weight gain between wild-type mice and Taco1 mut/mut mice infected with MCMV was noted ( Fig 1A). Next lymphocyte populations in the spleen and liver, two organs targeted by MCMV, were assessed at 23 weeks post-infection (p.i.) by flow cytometry. The total number of lymphocytes in the spleen and liver of Taco1 mut/mut was similar to that of wild-type mice, with MCMV infection having no significant impact on lymphocyte numbers (Fig 1B and 1E). T cells and natural killer (NK) cells are critical for control of MCMV infection. We did not detect a difference in the total number of these two lymphocyte populations in the spleens of Taco1 mut/mut mice after MCMV infection compared to controls (Fig 1C and 1D). By contrast, MCMV-infected Taco1 mut/mut mice had increased numbers of T cells in the livers compared to controls while NK cell numbers were not significantly different (Fig 1F and 1G). Importantly however, the number of MCMV-specific m38 + CD8 + T cells in the spleen and liver of Taco1 mut/mut mice was equivalent to that of Taco1 wt/wt mice ( Fig 1H). Overall these findings indicate that Taco1 mut/mut mice are capable of generating an effective immune response to MCMV.

Murine CMV infection exacerbates isolated complex IV deficiency in the heart
We investigated the effects of MCMV infection on the abundance of TACO1 and mitochondrial proteins by immunoblotting in the heart and liver of 30 week old mice, and specifically the levels of cytochrome c oxidase subunit 1 (COXI) a key component of the respiratory chain that catalyses the reduction for oxygen to water, which is regulated by TACO1. The levels of TACO1 were dramatically lowered in the presence or absence of infection and there was a reduction in the abundance of COXI in the hearts and livers (Fig 2A and 2B, respectively) of Taco1 mut/mut mice from both groups compared to their respective control mice. This reduction of COXI was more severe in MCMV infected Taco1 mut/mut mice than non-infected Taco1 mut/mut mice in mitochondria isolated from hearts, but not in mitochondria isolated from livers. There were no differences in the abundance of protein subunits from the remaining complexes, further validating that this mutation only affects Complex IV (Fig 2A and 2B). Other mitochondrial and nuclear-encoded subunits of Complex IV were also decreased, suggesting that loss of COXI leads to decrease of other Complex IV subunits (Fig 2C and 2D) and the loss affected the mitochondrially encoded subunits in the heart more severely. Next, we analyzed the levels of the assembled respiratory complexes by blue native polyacrylamide gel electrophoresis (BN-PAGE) in heart and liver mitochondria (Fig 3A and 3B, respectively). There was a significant reduction of complex IV in the hearts of Taco1 mut/mut mice from both groups compared to their respective control mice ( Fig 3A). This reduction of complex IV was more severe in MCMV infected Taco1 mut/mut mice compared to the non-infected Taco1 mut/mut mice, and there were no major differences in the levels of the remaining respiratory complexes in the heart (Fig 3A). Complex IV levels were reduced in both infected and non-infected livers of Taco1 mut/mut mice compared to control mice, however the infection did not exacerbate the decrease further ( Fig 3B). The levels of the remaining respiratory complexes were not different in the livers of the infected and non-infected mutant and control mice (Fig 3B). These findings confirm the isolated complex IV deficiency previously reported [4] and show that chronic

Murine CMV causes tissue-specific changes in mTOR signalling
Impaired OXPHOS function has been shown to stimulate the upregulation of the mechanistic target of rapamycin (mTOR) signalling pathway [6] and mTOR activation has been implicated in cytomegalovirus infections [7,8]. Therefore, we investigated the effects of Complex IV deficiency on mTOR signalling in the presence and absence of MCMV infection. We investigated the effects of the Taco1 mutation in response to MCMV infection on the mTOR signalling pathway by immunoblotting of heart and liver lysates isolated from uninfected and MCMV infected 30 week old mice. The levels of mTOR were not different in the hearts of uninfected week old Taco1 wt/wt mice and Taco1 mut/mut adult mice were resolved on 4-12% Bis-Tris gels and immunoblotted using antibodies to investigate the steady state levels of TACO1 and OXPHOS protein subunits in the heart (25 μg) (A) and in the liver (50 μg) (B), as well as mitochondrial and nuclear-encoded Complex IV protein subunits in the heart (C) and in the liver (D). CII (Succinate dehydrogenase complex subunit A (SDHA)) or VDAC1 were used as loading controls. Relative abundance of proteins was measured using Li-Cor Odyssey Classic software normalized to the loading control. All data are representative of results obtained from 4 mice of each strain and two independent biological experiments. Error bars indicate SEM; � p<0.05, �� p<0.01, compared with control treatments by a two-tailed paired Student's t-test. https://doi.org/10.1371/journal.pgen.1008604.g002

Fig 3. MCMV infection exacerbates mitochondrial dysfunction in the heart. (A)
Heart homogenates (80 μg) and (B) liver homogenates (160 μg) from uninfected and MCMV infected 30 week old Taco1 wt/wt mice and Taco1 mut/mut adult mice were resolved on a 4-16% BN-PAGE gels. Immunoblotting with the Blue Native OXPHOS antibody cocktail was used to visualize respiratory complexes. CII was used as a loading control. Relative abundance of proteins was measured using Li-Cor Odyssey Classic software normalized to the loading control. All data are representative of results obtained from 4 mice of each strain and two independent biological experiments. Error bars indicate SEM; � p<0.05, compared with control treatments by a two-tailed paired Student's t-test.
https://doi.org/10.1371/journal.pgen.1008604.g003 Taco1 mut/mut and Taco1 wt/wt mice (Fig 4), however the levels of mTOR and phosphorylated mTOR were increased in the hearts of infected Taco1 mut/mut mice compared to both infected and uninfected control mice (Fig 4). The steady-state levels of S6, a downstream target of mTOR, were not altered in the uninfected Taco1 mut/mut and Taco1 wt/wt mice, whereas infection reduced the phosphorylated levels of S6 in the hearts of both control and mutant mice (Fig 4). Interestingly, despite the reduction in the phosphorylation levels of S6 in the infected control and mutant mice relative to uninfected mice, the MCMV infection caused an increase in the phosphorylated S6 levels in the hearts of Taco1 mut/mut mice compared to control mice (Fig 4). This increase was consistent with the increase in both the steady-state and phosphorylated mTOR levels in the hearts of infected Taco1 mut/mut mice compared to Taco1 wt/wt mice (Fig 4), indicating that increased energy demand during an infection can result in upregulation of mTOR signalling via S6 in the heart specifically. There were no differences in the levels of another mTOR substrate, 4EBP1 and phosphorylated 4EBP1, in the hearts of uninfected and infected Taco1 mut/mut and Taco1 wt/wt mice (Fig 4). These results indicate that the isolated complex IV defect combined with an infection cause upregulation of mTOR signalling in the heart.
In contrast, in the livers of uninfected or MCMV infected mice, there were no differences in the steady state levels of mTOR between the Taco1 mut/mut and Taco1 wt/wt mice, whereas the levels of phosphorylated mTOR were decreased in the infected control and mutant mice compared to uninfected mice (Fig 5). The steady-state levels of S6 were not altered in the livers of uninfected Taco1 mut/mut and Taco1 wt/wt mice, however, the phosphorylated S6 levels were reduced in the uninfected mutant mice as well as the infected control and Taco1 mut/mut mice compared to uninfected mice (Fig 5). The increased phospho-S6 levels found in the hearts of the mutant mice were blunted in the liver and not different to the control mice ( Fig 5). Heart homogenates (25 μg) from uninfected and MCMV infected 30 week old Taco1 wt/wt mice and Taco1 mut/mut adult mice were resolved on 4-12% Bis-Tris gels and immunoblotted using antibodies to investigate the steady state levels of mTOR and its substrates. Glyceraldehyde-phosphate dehydrogenase (GAPDH) was used as a loading control. Relative abundance of proteins was measured using Li-Cor Odyssey Classic software normalized to the loading control. All data are representative of results obtained from 4 mice of each strain and two independent biological experiments. Error bars indicate SEM; � p<0.05; �� p<0.01, compared with control treatments by a two-tailed paired Student's t-test. https://doi.org/10.1371/journal.pgen.1008604.g004 Consistent with decreased levels of phosphorylated mTOR and S6, the infection caused reduction in the phosphorylated levels of 4EBP1 in the livers of the control mice relative to uninfected control mice as well as in infected mutant mice relative to uninfected control and mutant mice (Fig 5). These findings indicate that in the liver an isolated complex IV defect leads to downregulation of mTOR signalling and consequently reduced S6 and 4EBP1 phosphorylation.
Taken together, these results show tissue-specific regulation of the mTOR signalling pathway in mitochondrial disease that depends on the energy requirements of individual tissues. In the heart mTOR signalling is upregulated upon infection, indicating that additional energy impost on the mutant mice activates the mTOR stress response in the heart, while in the liver mTOR signalling is downregulated.

Chronic murine CMV infection results in reduced Complex IV activity
To determine the effects of chronic MCMV infection in the hearts and livers of control and mutant mice on mitochondrial function, we measured Complex IV activity in uninfected and MCMV infected 30 week old mice. Complex IV activity was significantly reduced in both infected and uninfected mutant mice compared to control mice ( Fig 6A). In addition, the MCMV infection caused a greater decrease in Complex IV activity in the hearts of infected compared to uninfected Taco1 mut/mut mice (Fig 6A), indicating that MCMV infection can exacerbate mitochondrial dysfunction. In the liver complex IV activity was also significantly reduced in both infected and uninfected mutant mice compared to control mice (Fig 6B), however, infection did not exacerbate this dysfunction. We measured Complex IV activity in the hearts and livers of 9 week old mice both infected (day 4 p.i.) and uninfected Taco1 mut/mut mice and Taco1 wt/wt mice (S1A Fig). The Taco1 mutation reduced Complex IV activity and this reduction was not further exacerbated by the infection indicating that chronic viral infection can further promote the Complex IV deficiency. Oxygen consumption measurements in liver mitochondria of uninfected and MCMV infected 9 week old mice, validate these findings, where uninfected and infected Taco1 mut/mut mice have an overall reduced oxygen consumption rate with reduced basal respiration, ATP production, proton leak, maximal respiration, non-mitochondrial respiration and spare capacity compared to Taco1 wt/wt mice (S1B Fig). This shows that energy production is impaired as a consequence of the Taco1 mutation, as it has been observed previously [4]. Interestingly, in the infected mice, we also found that Taco1 mut/mut mice had increased proton leak compared to control mice and uninfected mutant mice (S1 Fig

Murine CMV infection causes motor axon thinning of neuromuscular junctions in skeletal muscle
Deterioration of neuromuscular junctions (NMJs) has been observed following mitochondrial dysfunction in a mouse model of amyotrophic lateral sclerosis [9]. To investigate if changes occur in the morphology of the neurons and their synaptic connections, we analyzed skeletal NMJs by immunofluorescent labelling, in uninfected and infected 30 week old Taco1 mut/mut and Taco1 wt/wt mice. Neuromuscular junctions were stained using antibodies against neurofilaments and synaptic vesicles to label the motor axons and their terminal endings (pre-synaptic components), together with labelled α-bungarotoxin to identify post-synaptic acetylcholine receptors. Confocal imaging of the gastrocnemius showed a significant reduction in motor axon thickness projecting to NMJs in Taco1 mut/mut mice compared to Taco1 wt/wt mice in both infected and uninfected groups (Fig 7A and 7C). Examination of the extensor digitorum longus, which is a predominantly fast twitch muscle, showed a similar significant reduction in NMJ axonal thickness in Taco1 mut/mut mice compared to Taco1 wt/wt mice for the uninfected cohort ( Fig 7B). While this reduction in axonal thickness was further reduced at NMJs from MCMV infected Taco1 wt/wt mice compared to uninfected Taco1 wt/wt and Taco1 mut/mut mice, there was no further decrease in NMJ axonal thickness at NMJs from infected Taco1 mut/mut mice (Fig 7B and 7C). These observations suggest that the mutation in Taco1 causes morphological changes to the NMJ axons in the gastrocnemius and extensor digitorum longus

PLOS GENETICS
Viral infection exacerbates mitochondrial disease muscles. While the presence of MCMV infection did cause further reductions in NMJ axonal thickness for NMJs of the gastrocnemius muscle, it did not exacerbate the reduction in NMJ axonal thickness of EDL NMJs from Taco mut/mut mice. Taken together these findings suggest that MCMV infection can cause alterations in neuromuscular axonal morphology, that only intensify the phenotype in a muscle-type specific manner.

Discussion
We used a mouse model of late-onset Leigh Syndrome to investigate how a common viral infection can trigger the onset or progression of mitochondrial disease. We identified that infection with murine cytomegalovirus does not trigger an earlier onset of the disease but instead exacerbates the complex IV deficiency and causes alterations in mTOR signalling, energy metabolism and neuromuscular morphology. Cytomegalovirus is a common speciesspecific virus that belongs to the herpesvirus family. Infection with cytomegalovirus places a significant metabolic stress on a host cell due to the increased biosynthesis and energy utilization required for viral replication. Here we show that MCMV infection poses a significant stress on the heart in Taco1 mutant mice as mitochondrial biogenesis is severely impaired in the heart compared to the liver. It is possible that the infection imposes a greater cost on mitochondrial function in the heart because it predominantly relies on OXPHOS to maintain sufficient ATP levels [10]. Similar molecular defects were identified in the liver and viral infection increased the proton leak in the mutant mouse livers. However, the liver was able to cope with the additional energy demand, likely by using the spare respiratory capacity that may compensate to maintain mitochondrial function. We suggest that the ability of organs to use different metabolic and signalling pathways accounts for the tissue-specific molecular effects observed during viral infection. This differential ability of tissues to respond to mitochondrial stress has been observed recently in another model, where mitochondrial dysfunction caused by errorprone mitochondrial translation could recover due to liver-specific signalling that stimulated cellular proliferation and mitochondrial biogenesis [11]. It has also been demonstrated that human cytomegalovirus can itself disrupt cellular metabolism to cause metabolic reprogramming with an increase in biosynthetic reactions and glycolysis to maintain ATP levels for viral replication [12][13][14].
In the presence of infection, we identified downregulation of the mTOR signalling pathway via its downstream targets, S6 and 4EBP1, in the livers of Taco1 mutant mice. However, MCMV infection upregulated mTOR signalling through the phosphorylation of both mTOR and S6 in the hearts of Taco1 mutant mice. Upregulation in mTORC1 signalling has been observed previously in mitochondrial dysfunction that increases cytoplasmic protein synthesis in an effort to compensate for the severe OXPHOS defects [15,16]. This coincides with our findings where MCMV infection severely impairs mitochondrial biogenesis, particularly in heart mitochondria, that likely activates mTOR signalling in an attempt to overcome the OXPHOS defect, while the liver could maintain required ATP levels via glycolysis or gluconeogenesis and instead downregulates mTOR signalling, in an effort to preserve energy.
During infection cellular stress responses are activated that normally inhibit mTOR activation, which may act to preserve energy expenditure and prevent cytoplasmic translation of viral proteins [17,18]. However human cytomegalovirus has been shown to overcome these stress responses to activate mTOR signalling and maintain translation of viral proteins in in vitro studies [7,8,19]. These inconsistencies in mTOR signalling in response to viral infection may be attributed to the duration and stage of infection and species-specific effects. Furthermore, it still remains poorly understood how cytomegaloviruses can interact with a host cell to alter mTOR signalling and the consequences in different tissues.
Here we show for the first time that loss of TACO1 results in the development of significant neuromotor degeneration in skeletal muscle, exemplified by motor axon thinning including their terminal ends. This NMJ associated axonal thinning may contribute to the development of neurological symptoms that have been documented previously in this mutant mouse such as a loss of muscle grip strength and poor learning performance [4], which is commonly found in Leigh Syndrome patients [3]. In addition, we show that infection with cytomegalovirus exacerbates neurodegeneration. Little is known about the effects of human CMV on motor axons innervating skeletal muscle, however it is known that human CMV can alter the structure of CNS neurons including the retraction/alignment of their dendritic arbors [20], which along with effects on neural stem cells and progenitor cells in the brain, can result in brain disorders (reviewed in [21]). Pseudorabies virus and herpes simplex virus type 1 (HSV-1) infections have been shown to disrupt mitochondrial motility and morphology in the superior cervical ganglion (SCG) neurons of rodents by modulating mitochondrial proteins [22]. Our study extends this possible link where MCMV could also have the capacity to invade the nervous system and may have similar effects on mitochondrial function within axons, exacerbating the defects caused by reduced levels of Complex IV. We therefore provide a potential pathogenic link between the development of mitochondrial disease and the onset of neurological symptoms in Leigh Syndrome patients.
The characterization of this model of late-onset mitochondrial disease has provided insight into the contribution of acquired infections to the progression of mitochondrial disease. This study established that an environmental stress such as a viral infection can play a significant role in disease pathogenesis, neuromuscular morphology and the molecular defects caused by a genetic mutation. Infections can be potent modifiers of energy metabolism with the ability to regulate mTOR signalling. Future studies will focus on understanding the molecular and pathophysiological mechanisms in which environmental stresses impact mitochondrial disease and to establish treatments to alleviate clinical symptoms of both the infection and mitochondrial disease.

Animals and housing
Female wild-type (Taco1 wt/wt ) and homozygous mutant mice (Taco1 mut/mut ) carrying an ENUinduced point mutation, T491A in the Taco1 gene, were bred onto a C57Bl/6J background for 10-12 generations. Animals were housed in standard cages (45 cm Å~29 cm Å~12 cm) under a 12-h light/dark schedule (lights on 7 a.m. to 7 p.m.) in controlled environmental conditions of 22 + 2˚C and 50 + 10% relative humidity. Normal chow diet with physiological levels of thiamine (Rat & Mouse Chow, Specialty Foods, Glen Forrest, Western Australia) and water were provided ad libitum.

Ethics statement
The study was approved by the Animal Ethics Committee of the UWA and the Harry Perkins Institute of Medical Research Animal Ethics Committee (approved protocol numbers: RA/3/ 100/1256, RA/3/300/98, AE031/2015 and RA/3/300/125) and was performed in accordance with Principles of Laboratory Care (NHMRC Australian code for the care and use of animals for scientific purposes, 8th Edition 2013).

Flow cytometry
Spleens were passed through a steel mesh in order to generate a single cell-suspension. Singlecell suspensions from the liver were prepared by perfusing the liver via the portal vein with phosphate buffered saline. The liver was then passed through a steel mesh. The resulting preparation was resuspended in a 37.5% isotonic Percoll solution (Pharmacia) and centrifuged at 690 g for 12 min to separate lymphocytes from hepatocytes. Red blood cells were osmotically lysed using NH 4 Cl and cells washed in FACS buffer. The resulting single cell preparations were stained with antibodies specific for CD11b (M1/70), CD3 (145-2C11), CD4 (RM4-5), CD8 (53-6.7), NK1.1 (PK136), CD11c (HL3), CD19 (6D5). Virus-specific CD8 + T cells were identified using m38 tetramer (H-2K b -SSPPMFRV) from Immuno ID Tetramers (Melbourne, Victoria). Antibodies were obtained from BD Biosciences, BioLegend, or eBioscience. Fixable viability stain 620 (BD Biosciences) was used for live/dead discrimination. Samples were run on a LSRFortessa X-20 instrument (BD Biosciences) and results were analyzed using FlowJo software (TreeStar).

Wholemount immunohistochemistry
Mice were anaesthetized with carbon dioxide and then euthanized by cervical dislocation. A minimum of five animals was used for each genetic and MCMV infection group. The gastrocnemius and extensor digitorium muscles were then dissected from the lower legs of these mice, pinned out at resting muscle length and fixed in 4% paraformaldehyde (PFA) in PBS for approximately 60 min at room temperature. They were then placed into 0.1M glycine in PBS for several hours, followed by washing in PBS. Small muscle bundles were dissected from these muscles and processed for whole mount immunostaining. In brief, muscle were incubated with 4% bovine serum albumin (BSA) 0.4% triton-X100 (TX-100) in PBS (blocking buffer) for 1 h at room temperature. For primary antibody staining, muscles were incubated in rabbit anti-neurofilament 1:500 and mouse anti-synaptic vesicle in 2% BSA 0.4% TX-100 in PBS overnight at 4˚C. The muscles were then washed three times with PBS for 10 min prior to incubation with secondary antibody 1:1000 goat anti-rabbit Alexa 488, 1:1000 goat anti-mouse Alexa 488 and co-stained for acetylcholine receptors at the NMJ with 1:1000 Alexa 555-αBTX in PBS. Muscles were then washed with PBS and mounted in Prolong Gold antifade reagent (Molecular Probes, Invitrogen). The motor axons and their NMJs were imaged using a Leica SP8 confocal. Images were taken at a resolution of 1024 by 1024 pixels in xy, using a 63x Oil objective with a z-step size of 0.3 μm, providing a voxel size of 0.124 x 0.124 x 0.3 μm. Images were captured using identical laser power levels, photomultiplier gain levels, scanning speed and pinhole size. Z-stacked images were projected into a 2D rendered image to show the top down view of NMJs and their innervating motor axons. The diameter of preterminal axons before the entry of motor axon into the nerve terminal region at each NMJ was determined using Imaris 8.1.2 software (Bitplane, South Windsor, CT, USA). Collected images were assembled in Adobe Photoshop and Adobe Illustrator (Adobe Inc. USA).

Mitochondrial isolation
Mitochondria were collected from homogenized livers and isolated by differential centrifugation as described previously [10], with some modifications. Livers were homogenized in buffer containing 250 mM sucrose, 5 mM Tris, 1mM EGTA, pH 7.4 with EDTA-free Complete protease inhibitor cocktail (Roche) before differential centrifugation. The mitochondrial protein concentration was quantified using the BCA assay using BSA as a standard.

BN-PAGE
Heart tissue was homogenized in 100 μl of 0.75 M aminocaproic acid, 50 mM BisTris-HCl, pH 7.0, and 1.2% n-dodecyl -D-maltoside (DDM) containing EDTA-free Complete protease inhibitor cocktail (Roche). The homogenate was centrifuged at 10,000 g for 5 min at 4˚C. The previous steps were repeated until a clear tissue homogenate was produced. The tissue homogenate protein concentration was quantified using the BCA assay using BSA as a standard. BN-PAGE was performed in 4-16% gradient gels according to recommendation of the Novex NativePAGE Bis-Tris gel System. PVDF membranes were used to transfer proteins from native gels followed by immunoblotting.

Complex IV measurements
Enzyme assays were carried out in a 1-ml cuvette at 30˚C using a PerkinElmer lambda 35 dual beam spectrophotometer. Complex IV was measured as the cyanide-sensitive oxidation of ferrocytochrome c as described before [4].

Statistical analysis
For statistical analysis a two-tailed Student's t-test with replicates was performed using statistical software (Excel, 2016). Numerical data for specific figures are in S1 Table. Supporting information

S1 Fig. Oxygen consumption in mitochondria isolated from hearts and livers of Taco wt/wt
and Taco mut/mut mice. (A) Complex IV activity was measured in heart and liver mitochondria from 9 week old uninfected and MCMV infected Taco1 wt/wt and Taco1 mut/mut mice. Complex IV activity was measured spectrophotometrically as nmol/min/mg of protein and enzyme activity is shown as percent of activity in control mice. (B) Oxygen consumption was measured in liver mitochondria from 9 week old uninfected Taco1 wt/wt mice and Taco1 mut/mut and MCMV infected 9 week old Taco1 wt/wt mice and Taco1 mut/mut using an Oroboros oxygen electrode using glutamate, malate and pyruvate as substrates under basal conditions followed by the sequential addition of rotenone, antimycin, oligomycin and FCCP. All data are representative of results obtained from 5 mice of each strain. R = routine/basal respiration, L = proton leak, ET-capacity = maximal respiration, ROX = residual oxygen. Error bars indicate SEM of 5 mice per genotype and treatment; � p<0.05; �� p<0.01; ��� p<0.001, compared with controls by a two-tailed paired Student's t-test. (TIF) S1