Metabonomic signature analysis in plasma samples of glioma patients based on 1H-nuclear magnetic resonance spectroscopy
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.177606
Source of Support: None, Conflict of Interest: None
Objective: The presence of a glioma is associated with increasing mortality. In this study, nuclear magnetic resonance (NMR) based metabonomics has been applied to investigate the metabolic signatures of a glioma in plasma. The purpose of this study was to assess the diagnostic potential of this approach and gain novel insights into the metabolism of glioma and its systemic effects.
Keywords: Glioma; metabonomics; nuclear magnetic resonance spectra
A glioma is the most common tumor from amongst primary central nervous system tumors, accounting for more than 70% of all primary brain tumors.,, The prognosis of a patient detected to be having a malignant glioma is very poor and the recurrence rate is still very high. The survival period in malignant lesions is generally not more than 2 years. Thus, an early diagnosis of these lesions is the key to success in their management.,
Metabonomics is a quantitative measurement of the dynamic and multivariate metabolic response of complex multicellular organisms of living systems to pathophysiologic stimuli or genetic modification and the consequent disruption of system regulation., It is a powerful tool for analyzing the chemical composition and in providing important information regarding the disease processes. It has, therefore, been widely used in the diagnosis of diseases , and in biomarker screening., Nuclear magnetic resonance (NMR) is a noninvasive technique that is widely used in metabonomic studies for acquiring the metabolic profiles of biological specimens without extensive sample preparation. NMR-based metabonomics is also used to find possible biomarkers of different cancers such as esophageal cancer,, colorectal cancer, prostate cancer, breast cancer, cervical cancer, and brain cancer. NMR based metabonomics coupled with data-reduction techniques offer a powerful approach to the metabolic data analysis of biofluids and tissues, thus, providing a “metabolic fingerprint” of the pathophysiological changes in the body.
In this study, the NMR profiling of human plasma was explored as a method to identify the glioma-related metabolic signature. This has the potential to provide valuable insights into the metabolism of gliomas and the systemic effects associated with its presence.
Collection of plasma samples
The Ethics Committee of Xinjiang Medical University approved the study protocol, and all subjects gave a written, informed consent. A total of 70 healthy volunteers were recruited following their extensive medical examination; and, 70 patients having a glioma (World Health Organisation grade III and grade IV) from the First Affiliated Hospital of Xinjiang Medical University were enrolled during the period from June 2012 to March 2013. All patients showed a tumor of greater than or equal to 3cm in size. The patient groups were not evenly matched based on their gender, age or stage of the disease, to maximize the patient diversity in the study. Patients with cardiovascular, hepatic, renal or inflammatory disease, and pregnant women were excluded from the study. The average age of participants was 64.2 years, ranging from 46 to 77 years.
Plasma samples were obtained by centrifugation of blood samples from every individual in the morning before breakfast and were immediately stored at −80°C until they were subjected to NMR spectroscopy.
1 H-nuclear magnetic resonance spectroscopy of blood plasma
The plasma samples were prepared for NMR analysis by mixing 200 μL of plasma with 400 μL of heavy water buffer solution (1.5 M NaH2 PO4 + 1.5 M K2 HPO4 in 20% v/v D2O and 80% v/v H2O, pH 7.4), and were kept at room temperature for 10 min. The mixture was subjected to centrifugation at 10,000 rpm for a further 10 min at 4°C. The clear supernatant (550 μL) was placed in a 5-mm NMR tube for spectroscopic analysis. The samples were analyzed by 1 H-NMR spectroscopy at 599.95 MHz using a Varian Unity Inova 600 (California, USA) spectrometer at 298 K. A combination of presaturation and the Carr–Purcell–Meiboom–Gill pulse sequence suppressed the water signals and broad protein resonance signals. For each sample, 128 scans were converted into 32,768 data points with a spectral width of 10,000 Hz, which resulted in an acquisition time of 1.64 s and a relaxation delay of 2 s. For assignment purposes, several two-dimensional (2D) NMR experiments including 1 H-1 H homonuclear correlation spectroscopy (COSY), total COSY (TOCSY), and J-resolved spectroscopy (J-Res) were also performed for the selected samples.,,
Before Fourier transformation, free induction decays were multiplied by an exponential function equivalent to a 0.3 Hz line broadening factor. Fourier-transformed 1 H-NMR spectra were manually phased and baseline corrected, and chemical shifts were referenced to the anomeric proton signal of α-glucose at a chemical shift (δ) of 5.233 ppm. NMR spectra over the range of δ9.0–0.5 ppm were segmented into integral regions of 0.003 ppm, for each spectrum. All regions were normalized by the total integrated area of each spectrum. Due to the high variability in the intensity of water, the regions with δH = 5.22–4.68 ppm were excluded from the analysis. The pattern recognition analysis was carried out on the normalized NMR data sets using the SIMCA-P + software (Version 11.0, Umetrics Inc., Umeå, Sweden). We used the orthogonal partial least square discriminant analysis (OPLS-DA) method with unit variance scaling.,, OPLS-DA, a new method of data analysis, is a combination of orthogonal signal correction (OSC) with partial least squares discriminant analysis (PLS-DA). OPLS-DA, including an OSC in the PLS-DA, was used for the extraction of cancer-related biomarkers by removing the influence of systematic variations unrelated to the cancer pathology. OSC was capable of eliminating the influence of diet, age, sex, and environmental factors, and of decreasing sample heterogeneity (which is the common source of error in clinical investigations).,,
In this study, OPLS-DA comparisons between data of NMR spectra were obtained from the healthy controls and the patients with a glioma. The OPLS-DA model was constructed using the NMR data as the X matrix and the class information identifier for the different groups as the Y variable, using one PLS and one orthogonal component. The quality of the OPLS-DA model was described by the parameters R 2 X and Q 2; R 2 X represented the total explained variation for the X matrix, and Q 2 indicated the predictability of the model related to its statistical validity. The discriminative significance of metabolites between the different groups was determined by the Pearson's product–moment correlation coefficient.
Typical examples of the plasma samples obtained from the 1 H-NMR spectra of a healthy individual and a patient suffering from a glioma are shown in [Figure 1]. Resonance assignments were made by http://www.metabolomics.ca and confirmed by two dimensional (2D) NMR methods such as COSY, TOCSY, and J-Res spectra.
To optimize the separation of the two groups, we then utilized OPLS-DA to visualize the metabolic differences between the plasma samples of patients having a glioma and healthy controls. [Figure 2] shows that the two groups achieved a distinct separation in the score plot of the principal component (PC) 1 and PC2 of the OPLS-DA. In this study, R 2 X = 0.40 and Q 2 = 0.88, were significantly high, indicating that it is an excellent model suitable for data analysis. Based on the number of samples in each group, a correlation coefficient (Pearson's product–moment correlation coefficient) of 0.232 was used as the cutoff value for the statistical significance based on the discrimination significance at the level of P = 0.05. A positive value indicated a relatively lower concentration, and a negative value indicated a relatively greater concentration of metabolites in the plasma of patients with a glioma. The values of the correlation coefficients indicating the significance of the metabolites contributing to the separation between the healthy patients and the patient suffering from a glioma are summarized in [Table 1].
We applied 1 H-NMR to study the metabonomic profiling of human glioma and identified a total of 20 distinguishing metabolites. These metabolites are involved in the key metabolic pathways including glycolysis, amino acids metabolism, tricarboxylic acid (TCA) cycle, and fatty acid metabolism.
These correlation coefficients show that compared with the control group, the patients with a glioma had lower concentrations of isoleucine (δ0.93, δ1.00, δ1.96), leucine (δ0.95, δ0.97, δ1.72), valine (δ0.98, δ1.03, δ3.60), alanine (δ1.47, δ3.76), tyrosine (δ3.94, δ6.88, δ7.18), phenylalanine (δ7.32, δ7.37, δ7.42), 1-methylhistidine (δ7.06, δ7.79), glycoprotein (δ2.03), glutamate (δ2.13, δ2.36, δ3.75), myo-inositol (δ3.27, δ3.65), citrate (δ2.52, δ2.67), creatinine (δ3.06, δ4.05), choline (δ3.66, δ4.30), α-glucose (δ3.53, δ3.72, δ3.76, δ5.23), β-glucose (δ3.24, δ3.40, δ3.49, δ3.90, δ4.64), and lactate (δ1.33, δ4.11) (r > 0.232, P < 0.05), but significantly higher concentrations of very low density lipoprotein [VLDL] (δ0.85, δ0.88, δ1.26, δ2.22), low density lipoprotein [LDL] (δ1.26, δ4.24), unsaturated lipid (δ5.28, δ5.30) and pyruvate (δ2.37) [r>−0.232, P < 0.05].
The prognosis for patients with a glioma remains unclear despite significant progress in clinical therapies and related technologies. This is largely due to the inability of the current treatment strategies to address the highly invasive nature of this type of tumor. Malignant glial cells often spread throughout the brain, making it exceedingly difficult to target and treat all intracranial neoplastic foci. As a result, tumor recurrence is inevitable despite aggressive surgery as well as adjuvant radiotherapy, and/or chemotherapy.
In the recent years, evidence has accumulated in the reported literature that blood metabolic profiles may also exquisitely reflect different types of central nervous system cancer pathologies in humans, which also includes the presence of a glioma. In this work, the NMR profile of plasma has shown, for the first time to the best of our knowledge, that subsequent analysis of metabolite profiles of plasma samples can distinguish between patients with a glioma from healthy normal controls.
In this study, the tricarboxylic acid cycle and energy metabolism are dominantly altered in patients having a glioma. There were some metabolites that were shown to increase in the patients having a glioma (such as VLDL, LDL, unsaturated lipid, and pyruvate). The altered values of these metabolites may probably reflect an altered energy metabolism or a disregulated metabolism of the corresponding metabolites to compensate for the excessive energy consumption by the cancer cells. Several metabolites such as isoleucine, leucine, valine, lactate, alanine, glycoprotein, glutamate, citrate, creatinine myo-inositol, choline, tyrosine, phenylalanine, 1-methylhistidine, α-glucose, and β-glucose were less concentrated in patients with a glioma. This represents a typical signature in cancer patients. It has been previously confirmed that tumors rely on glycolysis as the main source of energy even in the presence of oxygen., Furthermore, the precursors of glucose in gluconeogenesis, such as lactate and alanine were found in lower concentrations in patients with a glioma, which clearly points towards an altered energy metabolism. The decrease of glucose, which is common in many cancers, indicates the increasing demand for higher energy in malignant tumors. Many blood amino acids were down-regulated in glioma patients compared with healthy controls, which indicate an increased demand for and overutilization of amino acids in the tumor tissue, as has been observed in other reports on varied malignant tumors., Fatty acid metabolism is also altered in the plasma of cancer patients, as evidenced by the increased levels of a number of unsaturated lipids, VLDL and LDL. This finding is also in accordance with the findings observed in the blood of patients suffering from other cancers.
Concentrations of the metabolites VLDL, LDL, unsaturated lipid, and pyruvate increased in the plasma of patients with a glioma as compared with healthy persons. Ketone bodies are intermediate products of fatty acid metabolism and enter the blood stream after their production in the liver. Usually, they are present in a low concentration in the blood, but this level increases with fat metabolism. Organs outside the liver produce a large quantity of acetyl coenzyme A, which uses the ketone bodies and inhibits pyruvate dehydrogenase so that pyruvate cannot enter the Kreb's cycle and its plasma concentration increases. Lower levels of creatine and myo-inositol are also related to fat mobilization. Carnitine is the main carrier of acetyl coenzyme A during fat metabolism and is involved in the biosynthesis and oxidation of fatty acids. A decrease in carnitine concentration also indicates an increased fat metabolism. α-Glucose, β-glucose, and their non-oxidative product, lactate, were decreased significantly in patients with a glioma compared with healthy controls. It is assumed that patients with a glioma rely on fat metabolism as their main source of energy.,
We observed that the plasma concentration of various amino acids, including leucine, alanine, citrulline, tyrosine, histidine, isoleucine and valine decreased significantly in patients with a glioma. The metabolites related to cell membrane protection and immune function  such as glutamine, myo-inositol, scyllo-inositol, and creatine also decreased in the plasma. The immune system is the main antitumor defense system in the body. Glutamine maintains the function of the immune system and increases the total lymphocyte count in combination with glycoprotein and alanine. Glutamine also protects cells, tissues, and organs from damage by free radicals. A decreased plasma glutamine level indicates a decrease in cell antioxidant activity, cell membrane damage, immune dysfunction, free radical injury, and an intensified oxidative damage. Some studies have demonstrated that a glioma produces many oxygen free radicals, which is the main cause of oxidant–antioxidant imbalance. Therefore, it is thought that a series of abnormalities such as immune dysfunction and oxidant-antioxidant imbalance occur in patients suffering from a glioma.
We have shown that the metabolic profiling of plasma, using a combination 1 H-NMR with multivariate statistical methods, revealed a detailed picture of metabolic changes in patients with a glioma compared with healthy controls. We discovered that a series of metabolites in glucose metabolism, fatty acid metabolism, and TCA cycle showed an altered expression in the plasma of patients with a glioma. These findings will not only enable an early diagnosis of this cancer at the molecular level but also improve our understanding of the initiation and development of a glioma. The plasma metabolite measurements have the capacity to revolutionize an early cancer detection and treatment.
Financial support and sponsorship
Conflicts of interest
The authors have declared no conflicts of interest for this article.
[Figure 1], [Figure 2]