Background
Enteroviruses (EVs) are responsible for 30,000 to 50,000 hospitalizations for aseptic meningitis per year in the United States. Echovirus 30 (E30) (genus:
Enterovirus; family,
Picornaviridae) is one of the most frequently isolated EVs, causing extensive outbreaks of E30 in temperate climates in several countries [
1,
2].
Metabolomics is a new approach that facilitates multivariate profiling of the integrated metabolic responses of complex systems to pathophysiological stress and provides corresponding information to genomics and proteomics [
3]. The use of high-resolution nuclear magnetic resonance (NMR) spectroscopy to study metabolomics in biological fluids, tissues, and cells and to investigating the biochemical consequences of human disease is rapidly increasing [
4‐
6]. This approach is attractive in the study of cell metabolism owing to its unique non-invasive characteristics which can generate information on multiple pathways monitored simultaneously in a single step.
During viral infection in the host cell, delivery of viral nucleic acid involves a complex series of tightly regulated events where the invading virus utilizes the biosynthetic components of the host cell to reach its goal, altering cell metabolism. Depending on the host cell and virus type, different altering mechanisms have been proposed [
7]. Studies on virus-cell interactions have proven to be valuable in elucidating viral-cellular processes [
8‐
13].
Cell culture is widely used as an in vitro model due to increased experimental control of physiological parameters than tissue or in vivo studies. The widespread use of animal cells grown in tissue culture as models for investigation of properties of normal and malignant cells intensifies the importance of a clear understanding of the dynamics of cellular metabolism [
14]. The current understanding of the effect of viral infection on the metabolism of all metabolites in tissue culture is still relatively inadequate.
Thus far, there have been no reports on metabolic studies of E30 infection in human cell lines using NMR spectroscopy. In this study, we use high resolution 1H NMR spectroscopy to investigate the effect of E30 infection during acute, sub-acute and chronic stage infections in a rhabdomyosarcoma embryonic (RD) human cell line. NMR spectroscopy can play a strategic role in this scientific challenge as it can be used to study the interactions, both from thermodynamic and kinetic points of view, between and among biomolecules and small ligands, monitoring whole metabolic processes. In addition, NMR spectroscopy can be used to solve the structure of biomolecules and tell how the biomolecules interact. Theoretically, the combination of a metabolite and an analogue capable of antagonizing its biological action at very low concentrations could provide a sensitive system for determining slight amounts of metabolites. This study investigates the interaction between E30 and human cells using NMR spectroscopy to determine the mechanistic contribution of a viral infection.
Methods
Preparation of cell
RD cells were obtained from the Center for Disease Control and Prevention, Atlanta, GA, USA and were cultivated in the Eagle’s minimum essential medium (MEM) (Eagle 1959) supplemented with 10% fetal bovine serum (FBS), 100 U ml
−1 of penicillin, and 10 mg ml
−1 of streptomycin, essential amino acids, and L-glutamine (obtained from Sigma-Aldrich Co). The cells were cultured in 75 cm
2 flasks (Corning Inc. USA). The cell monolayer was washed twice with 1× phosphate buffered saline solution without calcium and magnesium ions (Sigma-Aldrich), then trypsinized (0.5 g trypsin; 0.2 g EDTA.4Na/l HBSS, Sigma-Aldrich) and kept at 36.5 °C in the incubator until the cells detached from the surface. Cells were counted using a hemocytometer by taking quadruplicate samples from each of three flasks using the procedure of Jenkin and Anderson [
15]. Fresh medium (30 ml) was then added into the flask. The cells were subcultured at 5 to 7 day intervals.
Virus
E30 used in this study was isolated from environmental specimens. The virus was serotyped according to World Health Organization’s protocol [
16] and confirmed by RT-PCR with specific primers [
17]. Virus stock was prepared by infecting RD cells in 25 cm
2 flasks (Corning Inc. USA). Virus infected and mock infected cells were incubated at 36.5 °C with 2% MEM. After 48 h of infection, the culture was aliquoted and kept at-80 °C for further use.
Plaque assay
Plaque assays of E30 were performed on monolayers of RD cells in 60-mm six-well tissue culture plates. Tenfold serial dilutions of virus stock were added to 90% confluent monolayer. After 1 h incubation at 36.5 °C with 5% CO2, unabsorbed virus was removed by gentle washing with PBS and an overlay of fresh media (1% agarose in MEM with 1% FBS) was added. After 4 days incubation, the cells were fixed with 10% formaldehyde and stained with 0.5% crystal violet and plaques were counted.
Virus infection
After the cells were grown with 100% confluence, the growth medium (MEM with 10% FBS) was replaced with maintenance medium (MEM with 2% FBS) from all the flasks. Three flasks were inoculated with 200 μl of virus (1 × 10 [
5] Pfu/ml), while the other three flasks were inoculated with 200 μl of sterile PBS served as mock infected for each time point (12 h, 24 h, 48 h). After 12 h, one set of flasks both mock infected and virus infected was taken out from the incubator, photographs of the cells were taken under a Nikon inverted microscope at a magnification of X 400 and the maintenance medium was taken out and stored at −80 °C until 1H NMR experiments were performed for analyzing the extracellular metabolites from the medium. The cells were harvested by trypsinization, stained with trypan blue and counted, washed twice with ice-cold PBS 1X (4 °C, pH = 7. 4) and suspended in buffered D2O (50 ml PBS 10X and 450 ml D
2O) for 1H NMR experiments of the whole cells. Virus infection in the remaining two sets of flasks was continued for 24, 48 h period and, at the end of each period, photographs of cells were taken, the maintenance medium was taken out for 1H NMR experiments and the cells were harvested, counted and suspended in buffered D2O as described above for flasks with 12 h virus infected and mock infected cells.
Other components
Eagle’s minimum essential medium, HEPES buffer (N-2-hydroxyethylpiperazine-N0-2-ethanesulfonic acid), sodium bicarbonate, phenol red, deuterium oxide (D2O),trimethylsilylpropionic acid (TSP), sodium salt (all from Sigma Aldrich, USA), penicillin, streptomycin, FBS, L-Glutamine, PBS, fungizone, trypsin, trypan blue stain, (all from GIBCO, USA) and cell culture flask 75 cm2 (Corning, Inc. USA) were used for the study.
NMR experiments
One and two-dimensional NMR analyses were performed on a Bruker Biospin Avance 400 MHz NMR spectrometer using a 5 mm broadband inverse probe equipped with z-gradient. Analyses were performed according to the protocol described in a previously published article Tiwari et al. in 2012 [
16]. An experiment was performed in triplicates.
Collected extracellular media from virus infected cells and mock infected cell counterparts were thawed. An aliquot of 0.5 ml of each medium was drawn separately into 5 mm NMR tubes and 1H NMR spectra were recorded. After inserting a TSP control capillary, all spectra comparing the metabolite patterns excreted by virus infected and mock infected cells were recorded under identical experimental conditions.
All virus infected cells (12–48 h post infection) and mock infected cells (12–48 h post infection) were washed twice with 1× PBS and suspended in buffered D2O (50 μl PBS 10× and 450 μl D2O). Cells were then transferred to 5 mm NMR tubes and 1H NMR spectra were recorded under identical experimental conditions.
Water-soluble intracellular metabolites of virus infected and mock infected whole cells were extracted into PBS prepared in D2O (50 μl PBS 10× and 450 μl D2O) by sonicating the cells under ice-cold conditions. This was followed by centrifugation at 12,000 rpm at 4 °C for 10 min. The supernatants were stored for lipid extraction. The whole cell lysate was thoroughly mixed by vortexing the sample and brought up to a total volume of 0.5 ml. The whole cell lysates containing water-soluble cellular metabolites were then placed in 5 mm NMR tubes and one-dimensional 1H NMR spectra were recorded. Suppressing the residual water by pre-saturation allowed for spectra to be collected under identical conditions. For 48 h virus infected and mock infected cell extracts, a 1H–1H two-dimensional double-quantum filtered correlated spectroscopy (DQF-COSY) experiment was also performed by suppressing the residual water signal by pre-saturation. After recording the spectra, all samples were stored at-80 °C for lipid extraction.
Cell lipids
Cell lysates, stored at-80 °C after recording the proton NMR spectra, were subjected to lipid extraction following Folch’s extraction procedure [
18]. Supernatant and residue were lyophilized. Briefly, the residual cell pellet was mixed with 3.18 ml of chloroform, methanol and saline water in a ratio of 2.0:1.0:0.18. The resultant mixture was sonicated for 3 min under ice cold conditions and centrifuged at 2400 rpm for 10 min to allow a clear separation of the aqueous and non-aqueous phases. Subsequently, the aqueous (upper) phase was collected and subjected to re-extraction of the remaining lipids, if any, by treatment with 2.0 ml of chloroform and methanol in a ratio of 2:1, sonicated and centrifuged for 10 min. The non-aqueous (lower) phase from both extraction steps were pooled and dried using nitrogen gas. The residue obtained was dissolved in 0.5 ml CDCl
3:CD
3OD mixtures in a ratio of 2:1 and
1H NMR spectra were recorded.
Statistical analysis
The multiple comparisons of time intervals corresponding to various chemical components (dependent variable) and their significant value were calculated. The Confidence interval (CI) 95% was also estimated. The One-way ANOVA test was performed with all time interval groups were also calculated. A significant association was found between time interval (12 h, 24 h, and 48 h) and metabolites likewise Alanin, Lactate, Acetate, Glutamate, Tyrosine, Histidine, Phenylalanine, Creatine, Choline and Formate respectively released by cell during infection, which is highly significant (p < 0.01).
Discussion
A significant association was found between virus infected and mock infected cells at time intervals12, 24, and 48 h for metabolites Alanine, Lactate, Acetate, Glutamate, Tyrosine, Histidine, Phenylalanine, Creatine, Choline and Formate respectively released by cell during infection, which is highly significant (
p < 0.01). Some insignificant association was also found during the infection period that cells did not use some proportion of metabolites after 12, 24 and 48 h, namely Glutamine and Phosphocholine (Fig.
3).
Cell death through viruses is associated with inhibition of cellular macromolecular synthesis and changes in cellular morphology. The morphological changes are not direct consequences of viral inhibition of host macromolecular synthesis [
21]. All cellular components were utilized as time was increased, whereas cell lipid components like phosphatidylcholines, phosphatidylethanolamine, cholesterol, backbone of glycerol and some additional functional group lipid (−CH = CH-), lipid (=CH-CH
2-CH=), lipid (−CH
2COO-), lipid (=CH-CH
2-CH
2), lipid (−CH
2-CH
2-CH), and lipid (−CH
3) were increased. Thus, these results clearly indicating that EV30 virus does not utilize cell lipids during its viral particle synthesis, in spite of all other cellular components being utilized when cellular damage takes place, as observed by the optical microscopy (Fig.
1). Cell lipid components like, phophatidylethanolamine and phospholipids were also increases in
Adenoviruses infected on HEK cell along with uninfected cells [
22]. Fetal bovine serum in the medium helps the maintenance of viability of the virus to about the same level as observed in medium containing calf serum; therefore, bovine serum was used for this study. The stimulatory effect of sodium oleate on Japanese encephalitis virus growth is clearly shown, where the viral infectivity was almost 50-fold higher in the presence of 9-18:1 (20 μg ml
−1) than in control medium [
23]. It may be noted that E30 does not contain envelop in its outer structure so unable to utilize cell lipids during its viral particle synthesis, 1H NMR results indicate an increase in the cell lipids as time was increased. This hypothesis could account for the increase of enterovirus growth in the presence of fetal bovine serum in the medium, in addition to the virus protective effect of the FBS. There would be one possible reason for increasing concentration of cell lipid with infection time that is the cell death, which is normally an increase in a programmed cell death. A rounding of the cells when come in contact with an antigen and, when mixed in suspension, by the inability of cells to spread normally [
24], manifests this phenomenon. Presumably, these effects can be related to changes in cellular membranes and are reflected in the alteration in the lipid metabolism noted.
Interestingly, the intensity of the total lipid signals in the spectra of the cells was similar for the mock infected [
25] and 12 h virus infected cells, whereas they gradually increase with virus infection time (Fig.
3). Jenkin and Anderson showed that the sodium salt investigation effect of the fatty acids on viral growth without background interference from the cell [
15]. This correlates to the fact that the total quantity of lipid is the same at all time points of mock infected and gradually increased in virus infected cells, as determined from subsequent lipid extraction from whole cells in Fig.
5. It is clearly seen that cell lipid were not utilized by the E30, a non-enveloped virus (Figs.
3,
4 and
5). Unraveling the effects of biological perspective is important only if the intracellular environment can influence the properties of metabolites. The most apparent difference between in-vitro sample and in-vivo is the solute concentration. Macromolecular solutes reach concentrations of hundreds of grams per liter in cells and other biological fluids, but most in-vitro studies are performed in the buffered solution with <1% of the cellular macromolecule concentration. These conditions give optimal signals, but may lack biological relevance [
26].
Aspartate and glutamate participate within the neurotransmitter family of substances which play important role in cell. Whereas glutamate is one of the most important excitatory transmitters of central nervous system in lower animals and may also be important in humans. Figure
3 shows that the extracellular metabolites released by cells after infection of E30; At 24 h post virus infection concentration of aspartic acid was high, while it was not observed in mock infected cells and after 48 h virus infection cells were consumed the glutamate and aspartic acid. Glutamate and aspartate are also very important in the TCA cycle. Excitatory transmitters such as aspartate lead to depolarization of the nerves in humans; on the other hand, inhibitory transmitters cause hyperpolarization. When cells showed an apoptotic effect during infection and showed an adverse stress effect. Some of the vital sources like alanine for the production of protein, essential for proper functions of the central nervous system cells and help form neurotransmitters. In-vitro study of uninfected control and infected cell released the alanine in both conditions. Alanine increases during 12 to 24 h post virus infection (Fig.
3) but at 48 h post virus infection it consumed completely. The consumption of alanine is due to readily conversion into glucose when sugar levels fall and amino acids are liberated from cells to provide energy. Alanine may help keep sugar levels stable during virus infection cells. When the energy required for the cells, alanine is converted into the glucose, and thus it may help to maintain the sugar level. Alanine is one of the simplest amino acid and is involved in the energy-production through breakdown of 7glucose. The cells utilized glucose through glycolysis and TCA metabolic pathways to produce several of the metabolic end products such as lactate, acetate and formate respectively into the extracellular medium [
16].
In –vitro cells study, the creatine was present in the mock infected cells as well as virus infected cells. It was gradually decreased because of the large number of cell death after 48 h infection. The cells get creatine with supplements from minimum essential media (Eagle’s Media) for cell growth. It was interesting that a significant increase in the acetate incorporation into lipids could be obtained even after infection with the virus. The specific activity of the extracted lipid from virus infected cells and mock infected cells are compared. These results suggested that the increased lipid metabolism might have not been induced by the infective virus, but by a component of the virus capsid. It is not known whether this lipid stimulation reflects the formation of membranes necessary for these early events or an introduction to the later synthesis and assembly of particles. It is of interest that a similar increase in lipid synthesis has been described in cells infected with vaccinia [
27] and poliovirus [
28,
29].
Conclusion
The present study shows the metabolomics of E30 infected mammalian cells. The NMR analysis revealed no ambiguous signals using the RD cell line. The present proton NMR spectroscopic study, monitoring cellular metabolic changes during virus infection in cells, is useful in probing the molecular mechanism of virus-cell interactions. The fact that the effects of viral infection could be observed through metabolic changes at much earlier stages than previously reported indicates an application for NMR spectroscopy in studying cell-virus interactions. Regardless of the actual mechanism leading to cell shrinkage, it could have significance in late viral functions.
In addition to the consumption of metabolites, the virus-specific RNA and protein synthesis may also be affected by cell shrinkage, the cellular concentration of amino acids, nucleotide concentrations, and other requirements for macromolecular synthesis.
Acknowledgements
Author Sarika Tiwari is exceedingly grateful to the Indian Council of Medical Research, New Delhi, India for providing a Senior Research Fellowship, and both authors are thankful to the World Health Organization for partial support. Thanks to Prof. Raja Roy for cooperation during the NMR experiments and analysis. Both authors are also thankful to Mr. Rishi for his kind assistance during experiments and writing of the manuscript, unfortunately he is no more in this world to read this manuscript.