NMR Metabonomics in an in vitro Model of HIV-1 latency

Abstract

Background: Metabolic disorders have been identified in patients infected with the human immunodeficiency virus (HIV). These disorders include lipodystrophy, wasting syndrome, cardiovascular disease and glucose intolerance. Highly active antiretroviral treatment (HAART) administered to patients can successfully suppress the virus and decrease the prevalence of opportunistic infections associated with AIDS but increase incidence of metabolic disorders. Glucose tolerance test, dual-energy x-ray absorptiometry and CD4 counts are some of the conventional tests that are used to detect and monitor disease progression and metabolic disorders. These are single result tests that are time-consuming and provide limited information about the metabolic disorder. A metabonomic approach allows for the measurement of multiple metabolites simultaneously; which could lead to the identification of markers of disease progression. Most HIVmetabonomics studies to date used nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS) spectrometry to detect multiple metabolites from blood and urine samples simultaneously, fewer studies utilized cell culture supernatants as metabolite source. In the body HIV is able to seize control of cellular networks and exists as an active or latent virus. HIV latency is mainly responsible for survival of the virus through the production of reservoirs throughout the body. It is then of great interest to investigate the effects of active and latent virus on cell metabolic networks. In vitro cell models such as U1 cells, a promonocyte latently infected with HIV-1 are suitable for such investigations because the models allow for control over activation of the virus with stimulants. To date only one study has compared the metabolic profiles of the active virus against non-infected cells but no study has comparatively investigated actively and latently infected immune system U1 cells, which is what is presented here. In this study a metabonomic approach was used to investigate active and latently infected U1 promonocytic cells and the metabolites were detected by NMR spectroscopy. Methods: U1 and U937 cells were cultured and lysed by the freeze-thaw method to extract the supernatant, the extent of cell lysis was determined through flow cytometry. NMR spectroscopy was used to detect metabolic profiles of uninfected U937 cells (parent cell line from which U1 cells were derived) as well as actively and latently infected U1 cells. Phorbol myristate acetate (PMA) was used as a stimulant to activate the virus. NMR data was preprocessed for statistical analysis with Mestrenova 10.0 software, and the metabolites were putatively identified with the use of Chenomx software, literature searches, human metabolome database (HMDB) and Kyoto encyclopedia of genes and genomes (KEGG). For statistical data analysis SPSS 20.0 software was used to determine group separation and metabolic profile differences. Results: Glucose, lactate, glutamine/glutamate, leucine, alanine, choline, phosphocreatine and lipid are some of the metabolites that were detected by NMR spectroscopy and through ANOVA analysis the metabolites were determined to be significantly different (P-value<0.05) between the three groups. A multiple comparison table presented the group significant differences and LDA correctly classified the experimental groups with 100% accuracy. U937 and actively infected cells produced similar results to what was seen in other investigations where sera and plasma were used as metabolite source. Latently infected cells produced the more distinguishable separation among the experimental groups and the metabolites responsible for this separation were those mainly involved in glycolysis and lipid biosynthesis pathways. All the cell lines were treated with lactate to evaluate the influence of one prominent metabolite on the virus and cell metabolism. Lactate was selected because the metabolite was found to be significantly present in the initial experiments and for its role in glycolysis (indicates anaerobic respiration). Cysteine, an indicator of oxidative stress was produced and some of the metabolites such as alanine and taurine were no longer detectable. Conclusion: NMR spectroscopy successfully elucidated metabolic profiles of U937 cells, U1 cells (latent virus) and PMA induced U1 cells (active virus). The technique was highly reproducible with minimal sample preparation. Most metabolites that were detected are those primarily associated with metabolic disorders involving glycolytic energy metabolism. Through multiple comparisons it was determined that latent HIV-1 infection had a profound effect on cell glycolysis as seen by the significant alteration of lactate and the occurrence of aerobic glycolysis and mitochondrial disruption. In this study, it was observed that the virus supports biosynthetic pathways more than the production of energy through oxidative phosphorylation. Cells that were exposed to lactate produced a different metabolic profile from those that were not treated, this indicated that an increase or decrease in concentration of a particular metabolite can affect cell metabolism in HIV infected cells.