Article pubs.acs.org/jpr
Combination of LC−MS- and GC−MS-based Metabolomics to Study the Effect of Ozonated Autohemotherapy on Human Blood Michal Ciborowski,†,‡ Alina Lipska,§ Joanna Godzien,‡,∥ Alessia Ferrarini,‡ Jolanta Korsak,⊥ Piotr Radziwon,§,# Marian Tomasiak,† and Coral Barbas*,‡ †
Department of Physical Chemistry, Medical University of Bialystok, Kilinskiego 1, 15-089 Bialystok, Poland Regional Centre for Transfusion Medicine in Bialystok, Sklodowskiej-Curie 23, 15-950 Bialystok, Poland ‡ Center for Metabolomics and Bioanalysis (CEMBIO), Facultad de Farmacia, Universidad CEU San Pablo, Campus Monteprincipe, 28668 Boadilla del Monte, Madrid, Spain ∥ Department of Molecular Biology, Faculty of Mathematics and Natural Sciences, The John Paul II Catholic University of Lublin, Krasnicka 102, 20-718 Lublin, Poland ⊥ Department of Transfusiology, Military Institute of Health, Szaserow 128, 04-141 Warsaw 44, Poland # Department of Haematology, Medical University of Bialystok, Kilinskiego 1, 15-089 Bialystok, Poland §
S Supporting Information *
ABSTRACT: Ozonated autohemotherapy (O3-AHT) is a medical approach during which blood obtained from the patient is ozonated and injected back into the body. Despite an increasing number of evidence that O3-AHT is safe, this type of therapy remains controversial. To extend knowledge about the changes in blood evoked by O3-AHT, LC−MS- and GC−MS-based metabolic fingerprinting was used to compare plasma samples obtained from blood before and after the treatment with potentially therapeutic concentrations of ozone. The procedure was performed in PVC bags utilized for blood storage to study also possible interactions between ozone and plastic. By use of GC−MS, an increase in lactic acid and pyruvic acid was observed, which indicated an increased rate of glycolysis. With LC−MS, changes in plasma antioxidants were observed. Moreover, concentrations of lipid oxidation products (LOP) and lysophospholipids were increased after ozone treatment. This is the first report of increased LOPs metabolites after ozonation of blood. Seven metabolites detected by LC−QTOF-MS only in ozonated samples could be considered as novel biomarkers of oxidative stress. Several plasticizers have been detected by both techniques in blood stored in PVC bags. PVC is known to be an ozone resistant material, but ozonation of blood in PVC bags stimulates leaching of plasticizers into the blood. KEYWORDS: metabonomics, fingerprinting, biomarkers, blood ozonation
■
INTRODUCTION Ozone as a medicine can be used for the treatment of several diseases (e.g., peripheral obstructive arterial diseases, agerelated macular degeneration, chronic infectious diseases or pulmonary diseases). Its antibacterial properties can also be used for the treatment of bacterial, viral, and fungal infections, aphthous ulcers, burns, abscesses, and osteomyelitis. It can be administered intramuscularly; subcutaneously; rectally or vaginally via catheter; aurally via a stethoscope; and intravenously (mixed with the patient’s blood or injected directly). Medical ozone can also be administered orally in the form of ozonated drinking water or on the surface of wounds in the form of ozonated oils.1 So far, the most advanced and reliable approach is ozonated autohemotherapy (O3-AHT). It is a medical approach during which blood is collected from the patient, mixed with medical ozone (mixture of gaseous oxygene and ozone), and then injected back into the body through a vein or a muscle.1 O3-AHT has already been found effective in several pathologies, such as infections, vascular diseases, acute © 2012 American Chemical Society
and chronic viral diseases, orthopedic pathologies, and peripheral obstructive arterial disease.1,2 On the other hand, ozone therapy is considered controversial because of the phenomenon of oxidative stress, which is deeply involved in ozone’s mechanism of action,3,4 and is thought to contribute to the development of various diseases like cancer, atherosclerosis, Alzheimer’s or Parkinson’s diseases.2,4 The effects of ozone on human blood were the subject of several investigations and have been already reviewed.1,4,5 Ozone is 10-fold more water-soluble than oxygen, so when added to blood it quickly dissolves in plasmatic water, where it instantaneously reacts with antioxidants like uric acid, ascorbic acid, glutathione (GSH), vitamin E, as well as albumin-thiol groups.1,5−8 The other potential targets for ozone action are polyunsaturated fatty acids (PUFA), proteins, and carbohydrates.1 Received: September 23, 2012 Published: November 13, 2012 6231
dx.doi.org/10.1021/pr3008946 | J. Proteome Res. 2012, 11, 6231−6241
Journal of Proteome Research
Article
Figure 1. Schematic representation of blood and saline samples treatments performed. (A) Blood and (B) saline samples were ozonated or aerated. Composition of gas used for ozonation is described in Materials and Methods section. BO, before ozonation; AO, after ozonation; BA, before aeration; AA, after aeration; 1−10 mL dose; 2−20 mL dose.
chromatography−mass spectrometry (GC−MS) was used to investigate the effects of potentially therapeutic concentrations of ozone on human blood. Application of LC−MS or GC−MS technologies for biomarkers discovery is not only complementary to NMR15 but also demonstrate several advantages over it, including greater sensitivity and dynamic range.16 Application of these methodologies to study the metabolic profiles of plasma obtained from ozonated blood may extend our knowledge about the processes that take place in human blood after ozonation. Additionally, the possible interaction between ozone and components of the plastic bag in which the process was performed was also investigated.
Peroxidation of PUFAs leads to the formation of lipid oxidation products (LOP), aldehydes, and H2O2, which are active ozone messengers evoking several reactions when present in blood.1,4,5,9,10 Ozone is not a radical molecule, but it is far more reactive than oxygen and readily generates some of the reactive oxygene species (ROS) produced by oxygen (e.g., H2O2, lipoperoxyl radicals, hydroperoxides). Therefore, mixing of blood with ozone implies a calculated and precise oxidative stress that causes changes in homeostasis. The oxidative stress induces a biological response leading to an adaptive phenomenon. The initial disruption of homeostasis due to ozone oxidation is followed by the rapid reestablishment of homeostasis by triggering several biochemical reactions in blood cells and by the induction of an adaptive process due to the up-regulation of the antioxidant enzymes.5 All of those changes in ozonized blood induce a set of biological responses when this blood is infused into the donor patient. Further biological responses had already been described such as improved blood circulation and oxygen delivery to ischemic tissue, improved general metabolism by improved oxygen delivery, up-regulated antioxidant system, mild activation of the immune system, and enhanced release of growth factors from platelets; moreover, a majority of patients report a feeling of wellbeing.2,6 Despite the increasing number of reports that O3-AHT is safe when O3 is administrated in appropriate doses, which were experimentally defined by several authors as 10−80 μg of gaseous ozone per 1 mL of blood,11−13 this type of therapy remains partly prohibited in the U.S.A. and poorly regarded in other developed countries.1 However, private medical services are using this therapy worldwide;4 therefore, evaluation of comprehensive effects of ozone on human blood is of great importance. So far, in most of the investigations, the effect of ozone on human blood was studied by measuring a limited number of compounds like antioxidants, LOPs, hemoglobin, ROS, lipids, and pO2.8,9,11 Metabonomics, quantitative measurement of the dynamic multi parametric metabolic response of living systems to physiological stimuli or genetic modification,14 has already been applied to investigate the effects of ozone on the metabolite profile of human blood measured by use of nuclear magnetic resonance (NMR).6 In the study presented here, for the first time, metabolic fingerprinting of plasma by use of liquid chromatography−mass spectrometry (LC−MS) and gas
■
MATERIALS AND METHODS
Gas Mixtures
The oxygen−ozone gas mixtures were generated from medicalgrade oxygen by using ATO-3 MINI ozone generator (CryoFlex, Poland) attested by the Polish Health Ministry. This device is able to generate medical ozone, an oxygen− ozone gas mixture composed of 95−99% of oxygen and 1−5% of ozone with ozone concentration up to 70 μg/mL. The concentration of ozone in the gas mixture generated for this study was 30 μg/mL. The purpose of this study was to evaluate the effects of O3-AHT on human blood, with control sample blood being treated with appropriate doses of medical-grade air. All gases were kept at normal atmospheric pressure. Blood Collection
Due to sex differences in oxidative stress,17 only male individuals were included in this study. Fifty healthy donors, aged 21−40 years, who had not taken any drug for at least 2 weeks prior to blood sampling, took part in this study. From each participant 25 mL of blood was collected and placed in polyvinyl chloride (PVC) blood transfer bags (JMS, Singapore). Blood was anticoagulated with citrate phosphate dextrose (CPD) solution (0.12% final concentration). Blood bags were placed on a Laboratory Shaker 358S (ELPAN, Poland) in order to mix blood with the anticoagulant. Blood and Saline Samples Ozonation
Different treatments of blood and saline samples performed in this study are shown in Figure 1. Experiments on blood (Figure 1A) were performed immediately after donation. From each bag, 5 mL of blood was collected and centrifuged at 2000× g 6232
dx.doi.org/10.1021/pr3008946 | J. Proteome Res. 2012, 11, 6231−6241
Journal of Proteome Research
Article
reversed-phase column (Discovery HS C18 15 cm × 2.1 mm, 3 μm; Supelco) with a guard column (Discovery HS C18 2 cm ×2.1 mm, 3 μm; Supelco). Chromatographic conditions were the same as described previously.18 Data were collected in positive and negative ESI ion modes in separate runs on a QTOF operated in full scan mode from 50 to 1000 m/z with a scan rate of 1 scan per second. Accurate mass measurements were obtained by means of an automated calibrant delivery system using a dual-nebulizer ESI source that continuously introduces a calibrating solution, which contains reference masses at m/z 121.0509 (protonated purine) and m/z 922.0098 [protonated hexakis (1H, 1H, 3H-tetrafluoropropoxy) phosphazine or HP-921] in positive ion mode; and m/z 119.0363 (proton abstracted purine) and m/z 966.0007 (formate adduct of HP-921) in negative ion mode. The capillary voltage was set to 3000 V for positive and 4000 V for negative ionization mode; the nebulizer gas flow rate was 10.5 L/min. Samples were analyzed in randomized order in separate runs (first for positive and second for negative ion mode). Saline solution samples were analyzed independently.
for 10 min to obtain nontreated samples (samples from 40 blood bags to be analyzed by LC−MS and from 10 by GC− MS). Blood remaining in bags (20 mL) was treated with a single dose of 10 mL (10 blood bags to be analyzed by LC− MS) or 20 mL (20 blood bags, 10 to be analyzed by LC−MS and 10 by GC−MS) of oxygen−ozone gas mixture or the same volume of air as a control (20 blood bags to be analyzed by LC−MS). Final concentrations of gaseous ozone added to blood were 15 or 30 μg/mL of ozone per 1 mL of blood, respectively. Gas mixtures were transferred into the blood by use of throwaway syringes made of ozone resistant material containing an antibacterial filter. Syringes were operated by use of an infusion pump set to a constant flow of 1 mL/min. During ozonation/aeration, blood bags were placed on laboratory shaker to ensure mixing of gas with blood. Immediately after the treatment blood samples were centrifuged at 2000× g for 10 min to obtain plasma samples. Additionally, 15 blood transfer bags (JMS, Singapore) were filled with 20 mL of sterile normal saline solution in each, and divided into three groups (5 bags in each). One group was left without any treatment, and two others were treated with either 10 or 20 mL of medical ozone (Figure 1B). Final concentrations of gaseous ozone added to saline were 15 or 30 μg/mL of ozone per 1 mL of saline solution, respectively. All samples (plasma and saline before and after the treatment) were stored in aliquots at −80 °C until the day of analysis. Investigations were performed with approval of the Committee on the Ethics of Research in Human Experimentation at Medical University of Bialystok, Poland and under the guidelines of the Helsinki Declaration for human research.
LC−MS Data Treatment and Identification
The raw data collected by the analytical instrumentation was cleaned of background noise, and unrelated ions by the Molecular Feature Extraction (MFE) tool in the MassHunter Qualitative Analysis Software (B.04.00, Agilent Technologies). The MFE algorithm uses the accuracy of the mass measurements to group ions related by charge-state envelope, isotopic distribution, and/or the presence of adducts and dimers. The MFE then creates a listing of all possible components as represented by the full TOF mass spectral data. Each compound is described by mass, retention time and abundance. Parameters selected for data extraction by the MFE were the same as described previously.19 Briefly, the limit for the background noise was set to 200 counts, and to find coeluting adducts of the same feature following adduct settings were applied: +H, +Na, +K in positive ionization, and: −H, +HCOO for negative ionization. Dehydratation neutral losses were also allowed. Identification of compounds detected by LC−MS was performed as follows. Accurate masses of features were searched against the METLIN (www.metlin.scripps.edu), KEGG (www.genome.jp/kegg), LIPIDMAPS (www. lipidmaps.org/), and HMDB (www.hmdb.ca) databases. The identity of compounds was confirmed by LC−MS/MS by using a QTOF (model 6520, Agilent). Experiments were repeated with identical chromatographic conditions to the primary analysis. Ions were targeted for collision-induced dissociation (CID) fragmentation on the fly based on the previously determined accurate mass and retention time. Accurate mass data and isotopic distributions for the precursor and product ions can be studied and compared to spectral data of reference compounds, if available, obtained under identical conditions for final confirmation (HMDB, METLIN, LIPIDMAPS). In addition, confirmation with standards was performed by comparison of retention time, isotopic distribution, and fragments of commercially (Sigma) available reagents with those obtain in real samples. Phospholipids (Table 3) were confirmed with characteristic fragments of 184.07 m/z for phosphatidylcholines (PCs), of 184.07, 104.11, and 86.1 m/z for lysophosphatidylcholines (lyso PCs), and of (M + H − 141.02 m/z) for lysophosphatidylethanolamines (lyso PEs).20 Lyso PC (16:0), known as lyso platelet activating factor (lyso
Measurement of Complete Blood Cell Count and Level of Hemolysis
Automated hematological analyzer Sysmex XT-1800i (Sysmex, Kobe, Japan) was used to measure complete blood cell count (CBC), hematocrit (HCT), and total hemoglobin (Hbc). Plasma/LowHb Photometer (Hemocue AB, Angelholm, Sweden) was used to measure plasma level of hemoglobin (Hba). The percentage of hemolysis was calculated from the Hba, Hbc, and HCT values using the following equation. hemolysis (%) =
Hba × (1 − HCT) × 100% Hbc
Metabolic Fingerprinting with LC−MS
Eighty plasma samples obtained from blood donated by 40 men were fingerprinted by use of LC−MS. Samples were classified into four groups: (a) plasma obtained from blood treated with 10 mL of medical ozone (10 samples before and 10 samples after the treatment), (b) plasma obtained from blood treated with 20 mL of medical ozone (10 samples before and 10 samples after the treatment), (c) plasma obtained from blood treated with 10 mL of air (10 samples before and 10 samples after the treatment), and (d) plasma obtained from blood treated with 20 mL of air (10 samples before and 10 samples after treatment). Preparation of plasma samples for LC−MS analysis was performed as described previously.18 Samples of saline solution were just filtered through a 0.22 μm nylon filter before the analysis. Samples were analyzed by the HPLC system consisting of a degasser, two binary pumps, and thermostatted autosampler (1200 series, Agilent Technologies, Waldbronn, Germany) connected to an Agilent QTOF (6520) mass spectrometry detector. Samples (10 μL) were applied to a 6233
dx.doi.org/10.1021/pr3008946 | J. Proteome Res. 2012, 11, 6231−6241
Journal of Proteome Research
Article
respectively. The electron ionization source was operated at −70 eV. The mass spectrometer was operated in scan mode over a mass range 50−600 m/z at a rate of 1 spectra per second. A mixture of n-alkanes (C8−C28) dissolved in nhexane was run prior to the experimental samples for retention index determination.
Table 3. Identification of Phospholipids Significantly Changing in Blood after Ozonationa change (%)
compound
ozone 10 mL
ozone 20 mL
CV in QCs (%)
RT (min)
monoisotopic mass (Da)
18.1
477.2855
+26NS
−40d
12
18.2
541.3168
+9
NS
−20b
12
20.7
481.3532
+17c
+27c
8
+41c
8
Lyso PE (18:2) Lyso PC (20:5) Lyso PC (16:0 − lyso-PAF) Lyso PC (18:1) PC#
25.0
507.3688
+12NS
24.8
635.4168
PC#
25.9
649.4323
PC#
29.9
677.4628
Appears Appears after O3e after O3e Appears Appears after O3e after O3e Appears Appears after O3e after O3e
GC−MS Data Treatment and Identification
Data was acquired using the Agilent MSD ChemStation Software. Data represented by Total Ion Current Chromatogram (TIC) was carefully examined by visual inspection of quality of chromatograms and internal standard signal. Peak detection and deconvolution were performed automatically with Automated Mass Spectrometry Deconvolution and Identification System (AMDIS). Metabolites were identified by comparing their mass fragmentation patterns with those available in the NIST mass spectral library and Fiehn RTL library. Significantly changing metabolites are summarized in Tables 1 and 4 including retention time, monoisotopic mass, percentage of change after the treatment, and CV for abundances in QCs. Additionally target and qualified ions, which were used for identification, are also presented in the tables.
a Metabolites appearing only after O3-AHT were not present in pooled plasma used for the preparation of QC samples; therefore, CV in QCs was not available for those compounds. #, for those metabolites no hits were found in searched databases; however, they were identified as PCs based on characteristic fragment 184.07 m/z observed in MS/MS spectra; NS, non-significant. bp ≤ 0.05. cp ≤ 0.01. dp ≤ 0.001. ep ≤ 0.0001.
Quality Control and Statistical Analysis
In order to check the stability of the LC−MS and GC−MS systems22 while running plasma samples, and to test the reproducibility of the plasma sample treatment procedures, quality control (QC) samples were prepared. For LC−MS analysis QC samples (n = 19) were prepared from a pool of human plasma obtained from healthy volunteers that was available in the laboratory. Each QC sample was prepared independently from this pooled plasma following the same procedure as for the rest of samples, and analyzed only once throughout the run. QCs were injected at the beginning of the run and after every 4−5 real samples. For GC−MS analysis QC samples were prepared by pooling equal volumes of plasma from each of the 20 samples that were analyzed with this technique. Those samples (n = 4) were independently prepared from this pooled plasma following the same procedure as for the rest of samples, and analyzed only once throughout the run. QCs were injected at the beginning of the run and after every 6−7 real samples. For statistically significant metabolites detected in QC samples CV for abundances in QCs was calculated and presented in Tables 1−4. Before statistical analysis, due to subtle retention time shifts during LC−MS analyses, samples need proper alignment to ensure the same metabolite is listed as the same feature within each sample analysis. Therefore, samples were multialigned using MassProfiler Professional (B.02.01, Agilent Technologies). Parameters applied for the alignment were 1% for retention time correction and 20 ppm for correction of the mass. Statistical analysis of LC−MS data was performed for four different comparisons: plasma before vs plasma after aeration (for both doses), and plasma before vs plasma after ozonation (for both doses). For GC−MS data profiles before and after the treatment with 20 mL of medical ozone were compared. For each comparison data were filtered by choosing only the features that were present in all samples in one of the compared groups (i.e., in all samples before or in all samples after a dose of ozone/air). Abundances of the peaks obtained with GC−MS were normalized against the IS before further data processing.
PAF), was confirmed with additional characteristic fragments of 166.06, 124.999, and 60.082 m/z. For all identified metabolites calculated mass error of measured mass in comparison to monoisotopic mass from database was
