PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Int J Mass Spectrom. Author manuscript; available in PMC 2011 December 1.
Published in final edited form as:
Int J Mass Spectrom. 2010 December; 298(1-3): 78–90.
doi:  10.1016/j.ijms.2010.02.007
PMCID: PMC2990699
NIHMSID: NIHMS182074

Metabolic Profiling of Human Blood by High Resolution Ion Mobility Mass Spectrometry (IM-MS)

Abstract

A high resolution ion mobility time-of-flight mass spectrometer with electrospray ionization source (ESI-IM-MS) was evaluated as an analytical method for rapid analysis of complex biological samples such as human blood metabolome was investigated. The hybrid instrument (IM-MS) provided an average ion mobility resolving power of ~90 and a mass resolution of ~1500 (at m/z 100). A few µL of whole blood was extracted with methanol, centrifuged and infused into the IM-MS via an electrospray ionization source. Upon IM-MS profiling of the human blood metabolome approximately 1,100 metabolite ions were detected and 300 isomeric metabolites separated in short analyses time (30 minutes). Estimated concentration of the metabolites ranged from the low micromolar to the low nanomolar level. Various classes of metabolites (amino acids, organic acids, fatty acids, carbohydrates, purines and pyrimidines etc) were found to form characteristic mobility-mass correlation curves (MMCC) that aided in metabolite identification. Peaks corresponding to various sterol derivatives, estrogen derivatives, phosphocholines, prostaglandins, and cholesterol derivatives detected in the blood extract were found to occupy characteristic two dimensional IM-MS space. Low abundance metabolite peaks that can be lost in MS random noise were resolved from noise peaks by differentiation in mobility space. In addition, the peak capacity of MS increased six fold by coupling IMS prior to MS analysis.

Keywords: Mobility-Mass Correlation Curves (MMCC), Trend lines, IMS, isomer separation, time-of-flight mass spectrometry, metabolomics, gas phase separation, blood metabolites, IM-MS

Introduction

1.1 Metabolomics

A parallel branch of biological science to proteomics, genomics, and transcriptomics that deals with the “omics” of metabolites and metabolism is called “metabolomics”. Metabolites are small molecular weight compounds (< ~1000 Da) that are employed as building blocks or produced as end products in various metabolic pathways and cellular regulatory processes in a biological system (1, 2). The entire collection of metabolites in a biological system, whether at the cellular, pathway or organism level, is known as a “metabolome”. Levels of these metabolites in a metabolome are either dictated by the genome, proteome, and/or transcriptome of the biological system or imposed by environmental perturbations and results in changes in phenotype (35). Thus metabolomics can be applied to map or identify the cause of alteration in phenotype and understand correlations between “omics”(6).

1.2 Blood analysis

Determination of metabolites present in blood samples provide insight into many human disease mechanisms and identify biomarkers for disease diagnosis (710). For example, metabolic profiling of amino acids and acylcarnitines is a powerful diagnostic tool in the diagnosis of inborn errors of metabolism (IEM’s) such as phenylketonuria PKU, medium-chain acyl-CoA dehydrogenase deficiency MCAD; glutaric acidemia GA, multiple carboxylase deficiency MCD, isovaleric acidemia IVA and other inherent metabolic disorders IMD’s (11, 12). Similarly, monitoring of various metabolic biomarkers present in blood is utilized for diagnosis for several diseases. Multiple sclerosis (13), HIV-1(14), rotavirus gastroenteritis patients (15), influenza-associated encephalopathy (16), cystic fibrosis (17, 18), congenital adrenal hyperplasia CAH (19), antley-bixler syndrome ABS and apparent pregnene hydroxylation deficiency APHD (20), gaucher type 3 disease (21), cancer (22), fabry's disease, glycolipid lipidosis (23), atherosclerosis (24), blood-circulation disorders (25), acute hepatitis (26), anemia and glucose-6-phosphatase deficiency (27) are some of the diseases where metabolite(s) monitoring is applied for diagnostic/screening purposes.

1.3 Current analytical methods

Different metabolite analysis strategies can be applied to characterize a metabolome, such as 1) Metabolite target analysis, 2) Metabolite profiling, 3) Metabolomics, 4) Metabolic fingerprinting, 5) Metabonomics, and 6) Metabolic footprinting (10, 2830). However, lack of efficient instrumental analytical techniques that can rapidly analyze complex biological samples and can provide sufficient visualization of the metabolome with diverse chemical and physical nature of metabolites require an array of analytical techniques for comprehensive analysis of a metabolome. Selection of the most suitable technique generally requires a compromise among speed, selectivity and sensitivity. Analytical separation methods such as gas chromatography (3134), high-performance liquid chromatography (3541), capillary electrophoresis (4245), and ion-exchange chromatography (46) in tandem with mass spectrometry have been investigated for their application to metabolomics. Spectroscopic methods such as NMR have also been extensively explored for metabolomics studies (4750). Examples include application of NMR to assess effects of chemical, physical, and biological stressors on organisms (51, 52). Currently, NMR, GC, HPLC, and CE are extensively used in metabolic profiling and targeted metabolite analysis for disease diagnosis and treatment, detection and identification of biomarkers and drug synthesis and metabolism investigations (5368).

In general, however, all of these methods require labor intensive and time consuming sample preparation (extraction, derivatization, concentration, etc.) and are mostly suited for analysis of a specific class of metabolites. Thus, there is a need for innovative approaches for measuring, monitoring and identifying the metabolome (69, 70). Techniques that would allow rapid and simultaneous determination of several classes of metabolites would not only increase sample throughput but would also make profiling of metabolome possible under different stress conditions, and understanding relative variation of metabolites or metabolite classes under a specific stress condition. Such an analytical tool would enable identification of more biomarker(s) for confirmation of diseases.

Ion mobility spectrometry coupled with mass spectrometry (IM-MS) is a rapid two dimensional analytical tool and has been applied to various biological (7173), pharmaceutical (7478), environmental (7982) and security related applications (8385). Feasibility of using IM-MS as a rapid analytical tool for metabolome profiling has been recently demonstrated where rapid separation and detection of hundreds of metabolites present in intracellular E-coli metabolome along with simultaneous separation of isomeric and isobaric metabolites was achieved (86). This manuscript evaluates the potential of IM-MS for application to blood metabolome profiling.

Experimental

2.1 Chemicals and Sample preparation

High performance liquid chromatography grade solvents (methanol, water and acetic acid) were purchased from J. T. Baker (Phillips burgh, NJ). Blood samples were collected anonymously from the finger tips of the donor. The finger tips were first cleaned with sterilizing alcohol and then pierced with sterilized Glucose meter lancets (OneTouch UltraSoft) purchased from local departmental store (Rite Aid, Pullman, WA). Blood drops from finger tips were directly collected into methanol as extraction solvent. Approximately 50 µL of blood was added to 950 µL of methanol and acetic acid (1%) solution maintained at 40°C over a heated water bath. The hot methanol-acetic acid-blood solution was kept in 40°C water bath for 30 minutes and then centrifuged for 1 hour at 1500 revolutions per minute. The supernatant was subjected to analysis by ESI-IM-MS without further sample preparation.

2.2 Instrumentation

The ESI-IM-MS instrument; schematic and picture shown in Figure 1; was comprised of an electrospray ionization source, an ion mobility spectrometer operated at ambient pressure and a time-of-flight mass spectrometer.

Figure 1
Schematic of the electrospray ionization atmospheric pressure ion mobility time of-flight mass spectrometer used for the analysis of human blood metabolome. This instrument is comprised of nine primary units: (1) electrospray ionization source; (2) heated ...

2.2.1 Electrospray source

For electrospray ionization, samples were infused into a 15 cm long, 50µm inner diameter silica capillary transfer line by a KD Scientific 210 syringe pump (New Hope, PA) at a flow rate of 4µl/min. The capillary transfer line was connected to a 10 cm long 50µm inner diameter silica capillary through a stainless steel junction. A positive voltage of 15.12 kV was applied at the stainless steel junction to generate the electrospray. This 10cm long capillary served as the electrospray needle and was centered ~0.5cm from a target screen in the IMS.

2.2.2 Ion mobility spectrometer

A stacked-ring design ion mobility spectrometer used in this study was constructed at Washington State University as described in previous publications(87, 88). The APIMS tube was divided by a Bradbury-Nielsen ion gate into a 7.5 cm long desolvation region and a 17.3 cm long drift region. Both regions consisted of alternating alumina spacers and stainless steel rings with high temperature resistors connecting the stainless steel rings (500 kΩ resisters for the desolvation region, 1MΩ resisters for the drift region). The ion gate was held at an electrical potential of 9934 volts; with last ring of IMS at 285 V. Thus, the electric field was ~558 V/cm throughout the drift region. Using the Bradbury-Nielsen gate, ions were pulsed into the drift region with a pulse width of 0.2 milliseconds. The temperature inside the IMS tube measured at the center of the desolvation region was 179°C. Nitrogen was used as the drift gas at a flow rate of 1100 ml/min. During these studies the atmospheric pressure in Pullman, WA, was 703–705 Torr. The ESI source, the ion mobility spectrometer, and supporting electronics were designed and assembled at Washington State University, Pullman.

2.2.3 Pressure interface

To couple an ambient pressure IMS system to a low pressure time-of-flight mass spectrometer, a interface was constructed at Ionwerks Inc., Houston, TX. The interface consisted of three ion lenses (nozzle, focusing, and skimmer) enclosed in a stainless steel chamber. The voltages applied on each element were: nozzle + 74 V, focusing lens +62V, skimmer + 35 V. These setting provided optimum ion transmission through the interface with minimum ion fragmentation at the interface. However, any variation in the position of the IMS and/or voltage on the last IMS electrode would require optimization of the lens voltages. Ions exiting the IMS entered a differentially pumped chamber through a 250 µM aperture in the nozzle and entered the ion beam region of the MS through a 250 µM aperture in the skimmer.

2.2.4 Time-of-flight mass spectrometer

The time-of-flight mass spectrometer and supporting electronics were designed and constructed at Ionwerks Inc., Houston, TX. Ions entering through aperture were guided into the extraction region of the time-of-flight analyzer by a series of ion lenses. The ions were extracted orthogonally and accelerated to 4.5 kV into the 40-cm-long reflectron drift tube and detected by a bipolar time-of-flight detector (Burle electro-optics Inc. Sturbridge, MA). The detector signal was processed by a time-to-digital converter (TDCX4, Ionwerks, Houston, TX). The ToF-MS operating voltages were as follows: lens1 +24.6 V, lens2 −22.1 V, deflector up −10.4 V, deflector down −14.0 V, lens3 +108 V, reflector back 915.7 V, reflector grid −389.9 V, and bias 4.5 kV. The detector voltages were: photomultiplier tube −0.45 kV, scintillator + 2.8 kV, and micro-channel plate (MCP) 0.9 kV.

2.2.5 Pressure considerations and Data acquisition

Pressure in the interface (1.5 torr), ion beam region (2×10−2torr) and reflectron region (1.4×10−6 torr) was maintained using two Varian rough pumps (100 L/min DS102 and 600 L/min DS602) and two Varian turbo pumps (68 L/s-V70 and 250 L/s-V250). The IMS and the MS were scanned at 25 Hz and 25 kHz respectively. Typical timing sequence and data acquisition has been described previously (89). The two dimensional IM-MS data was displayed as contour plots using IDL Virtual machine based software developed by Ionwerks Inc.

2.3 Basic IMS theory and calculations

Ion mobility spectrometry separates ions on the basis of the differences in their mobility K (cm2V−1 s−1) while the ions are drifting through a drift gas in a weak homogenous electric field gradient. The mobility of an ion through the drift region of the ion mobility tube is given as the ratio of the average ion velocity (vd = L/td) to the applied electric field (E = V/L)

equation M1
(1)

where, L is the length of the drift region in cm, td is the drift time in seconds (defined as the time an ion takes to travel through the drift region), and V is the voltage applied to the ion gate in volts. Experimental results are reported in terms of ion mobility reduced to standard temperature and pressure defined as: [Ko]

equation M2
(2)

Separation of ionic species in IMS is achieved due to the differences in the collision cross sections of the ions. Equation 3 defines the average ion-neutral collision cross section (Ω) measured in Å2 in terms of measured mobility of the ion: (90)

equation M3
(3)

where, NA is the number density of the drift gas in molecules per cm3, µ [=mM/(m+M)] is the reduced mass in kilograms of an ion of mass m g/mol and the neutral drift gas of M g/mol, k is Boltzmann’s constant in J/K, z is the number of the charge(s) on the ion, e is the charge of one proton in coulombs and K is the mobility of the ion in cm2 V−1 s−1. Number density NA is calculated as NA = (P/ k T) where P is the atmospheric pressure in atmospheres, k is the Boltzmann’s constant in L*atm. / K and T is the temperature in Kelvin.

Results and Discussion

3.1 Resolving power, sensitivity, reproducibility, and dynamic range

The mass range of the time-of flight mass spectrometer was calibrated through ESI-IM-MS analysis of 2,4-lutidine (m/z 107.2) and stachyose (m/z 666.6). Stachyose was detected as a sodium adduct (M + Na)+ and 2,4-lutidine as a protonated monomer (M + H)+. Average MS resolving power; measured using the equation R = m/Δm, where m is the mass of the peak and Δm is the full width at half maximum (FWHM), was 1500. Average resolving power of IMS defined as the ratio of the drift time of the ion to the peak width (in time) at half height (Rp = td/wh) was measured to be 90. Reduced mobility value measured for 2, 4-lutidine (m/z of 108) was 1.96 ± .01 cm2V−1 s−1 (literature value = 1.95). Average resolving power of first generation IM-MS constructed previously in our laboratory was 700 for the MS and 60 for the IMS(89).

To determine the reproducibility, detection limit, and dynamic range of IM-MS measurement, mixture solutions containing eleven metabolites (serine, nicotinamide, adenine, lysine, glutamine, ribose, phenylalanine, arginine, glucose, lactose, and maltose) each at 1 mM, 50 uM, and 500 nM concentrations prepared in 50:50 MeOH-H2O with 1% acetic acid were analyzed by ESI-IM-MS. Interface Definition Language (IDL) files developed by IonWerks Inc. Houston, and executed using IDL virtual machine platform (IDL VM 6.2, ITT Visual Information Solutions, Boulder, CO) were used for data processing and visualization. In an unprocessed IM-MS spectrum the whole 2D IM-MS space is filled with noise and analyte signals as shown in Figure 2S frame d (see supporting information). Adjustment of ion count threshold and averaging is employed to filter out noise peaks from the raw spectrum and the processed spectrum as shown in Figure 4a is obtained. Due to variation in instrumental and chemical noise signals from one data set to another the magnitude of threshold and averaging adapted to visualize a noise free spectrum varies from one data set to another. At present due to lack of an automatic noise recognition and removal tool “stare and compare strategy” was employed during noise removal. Theoretically, even one ion count above the noise level should be considered as an analyte signal. However, a simple curve (peak in this case) as defined by second degree polynomial equation (y = ax2 + bx + c) will need three points, peaks with less than three ion counts were discarded. For reference, zoomed in IM-MS spectra showing IM-MS signals with 3, 5 and 45 ions counts for known analytes are also shown in frames a, b, and c in Figure 2S (supporting information). During presentation of some examples further in the text, the threshold was increased higher than needed to exclude noise signals just to reduce clutter of peaks in the IM-MS spectrum and or to minimize labor intensive statistical data processing and metabolite identification. Low abundance metabolite peaks were thus not accounted for in the examples. An IM-MS spectrum for a 500 nM mixture solution of eleven metabolites is shown in Figure 2. On the x-axis is shown the mass-to-charge ratio of the ions in Da and the y-axis represents the IMS drift time of ions in microseconds. All of the metabolites in the mixture were detected and identified as one or more than one of the following adduct ions {M (H2O)n + Na}+, {M (H2O)n + H}+, and {M (CH3OH)n + H}+ (Table 1). Despite the fact that all metabolites were present in equimolar quantities in the mixture solution (500 nM), the number of ion counts for each metabolite peak ranged between 3 to ~100 counts. This variation in ion count with metabolite identity can be attributed to different ionization efficiencies of metabolites.

Figure 2
2D IMMS spectrum of an equimolar (500 n Molar) metabolite mixture solution containing serine, nicotinamide, adenine, lysine, glutamine, ribose, phenylalanine, arginine, glucose, lactose, and maltose in 50:50 MeOH and H2O with 1% acetic acid and 30 minute ...
Figure 4
Figure 4a: Two dimensional spectrum of metabolic features measured in methanol extract of human blood. Figure 4b: A zoomed in section of the IM-MS spectrum in the m/z range of 300–400 Da illustrating the peaks detected in the region with 5 ion ...
Table 1
List of ions detected in an equimolar (500 nM) metabolite mixture solution

Various isomeric/isobaric ions (see table; lysine and glutamine, lysine and ribose, lactose and maltose) were separated by IM-MS. Maximum standard deviations (triplicate measurements for 1mM, 50uM, and 500 nM mixture solutions) of 0.25 Da, 0.05 ms and 0.02 cm2 V−1s−1 in respectively m/z, drift time, and reduced mobility values were measured. Figure 3 illustrates the reproducibility of the technique where Figure 3 a is an IM-MS contour plot of 1µM solution of ribose analyzed in January 2006 and Figures 3b, 3c and 3d are 2D IM-MS plots of 1mM, 50 µM, and 500 nM solutions equimolar metabolite mixture analyzed in April 2006 respectively. The reduced mobility value for ribose (m/z 173 Da) measured from mixtures in April 2006 (Ko = 1.58 cm2 V−1s−1) matched with the reduced mobility value of standard ribose solution measured in January 2006 (Ko = 1.58 cm2 V−1s−1).

Figure 3
Demonstration of reproducibility of reduced mobility value (Ko) of ribose as (M + Na)+ ion measured for four different measurements at identical experimental conditions except for atmospheric pressure. Notice the reproducibility of Ko between measurements ...

Affects of charge competition and ion suppression was also evident upon analysis of equimolar mixtures of metabolites at varying concentrations. For example, peaks for lysine, adenine and nicotinamide were of highest intensity in the 500 nM mixture whereas in 1 mM and 50 µM metabolite mixture, the peak intensities for arginine and ribose were the strongest. Preferential ionization of an analyte was also evident. For example ribose as (M + H3O)+ ion at m/z 169 Da was observed at 5% intensity of the respective sodium adduct. 2D IM-MS plots for ribose detected with 5 and 45 ion counts as adducts of hydronium ion and sodium ion respectively are shown in frames b, and c in Figure 2S (supporting information).

Calibration curve for ribose as (M + Na)+ at m/z value of 173 Da showed a linear relationship between peak intensity and concentration with a slope of 0.2 in the concentration range studied (1 mM to 500 nM). Linear but functional, not proportional, dependence of peak intensity on metabolite concentration was measured.

3.2 Blood Analysis

3.2.1 IM-MS Spectrum

ESI-IM-MS analysis of ESI solvent (water, methanol and acetic acid) produced ions that were identified as {M (H2O)n + H}+, {M (H2O)n + Na}+, {M (CH3OH)n + H}+, {M (CH3COOH)n + H}+, {M (H2O)n + K}+, and {M (H2O)n + NH4}+ where M was either water, methanol, acetic acid. Figure 4a illustrates a 2D IM-MS spectrum of the metabolite ions detected in hot methanol blood extract. On the x-axis is the mass-to-charge ratio of the ions in Da and the y-axis reports the IMS drift times of the ions in microseconds. The IM-MS spectra showed several high intensity metabolite and low intensity metabolite regions with ion counts for the peaks ranging between 3 and ~80. Extent of metabolite recovery in methanol, metabolite abundance in blood, number of isomers and isobars for each metabolite and/or variation in ionization efficiency of metabolites contributes to variation in peak intensities of ions. Not all detected ions are visible in Figure 4a due to relative ion-intensity differences. To visualize low abundance ions, the data file was processed in increments of 100 Da mass range as shown in Figure 4b and signal threshold raised to exclude instrumental noise (dark counts) signals. A list of metabolites detected from the blood extract was generated with each peak characterized by its unique m/z value, drift time and intensity. Approximately ~1100 metabolite peaks with ion counts ≥ 3 and ~850 metabolite peaks with ion counts ≥ 5 were detected.

In an MS-only analysis, analyte peaks are characterized by their m/z values and their intensities. This is also true for the noise peaks. However in an IM-MS analysis the random noise peaks are not associated with unique mobility (drift time) values but instead are spread out in the mobility space in contrast to analyte peaks that are characterized by unique mobility values. Figure 5 illustrates three different conditions when mobility measurements provide added advantages to MS analysis. Figure 5(1) is a m/z spectrum where all detected peaks are shown and Figure 5(2) is a m/z spectrum where only those peaks that also had unique mobility values are shown. On the x-axis is shown the mass-to-charge ratio of peaks and on the y-axis is shown the intensity of peaks detected. Comparison of intensities of peaks (shown with solid line arrows) in Figures 5(1) and 5(2) illustrates an example where in the absence of mobility differentiation a noise peak can be considered to be an analyte peak. Similarly peaks shown with dotted line arrows illustrate an example when an analyte peak can be misinterpreted as noise peak. Contribution of noise peak intensity to the observed intensity of analyte peak is depicted through peaks shown with dashed line arrows in Figure 5. Approximately four fold decrease in peak intensity at m/z value of 237.4 relative to intensity of peak at m/z value of 265.4 was observed when contribution from noise peak intensity was considered.

Figure 5
One dimensional mass spectrum (1) compared to one dimensional mass spectrum with mobility differentiation (2). Three main features observed: a) noise peak accounted as real peak (bold line), b) contribution of random noise to peak intensity (dashed line), ...

3.2.2 Metabolite Identification

3.2.2.1 Database search

Peaks detected in the blood-extract were identified by comparing the m/z values of ions detected to the m/z values of metabolites listed in databases. The databases used were “METLIN-A metabolite mass spectral database”, “LMSD: LIPID MAPS structure database” and “Golm Metabolome Database” (9194). The peaks were identified as metabolite ions as {M (H2O)n + H}+, {M (H2O)n + Na}+, {M (H2O)n (CH3OH) + H}+, {M (H2O)n (CH3COOH) + H}+, {M (H2O)n + K}+, and/or {M (H2O)n + NH4}+ where M is a metabolite and n = 0 to 3. Peaks which were not identified could have been fragment ions of larger molecules or other adducts or clusters or unknown compounds. Similar to protein/peptide/metabolite identification via LC-MS methods that largely depend on spectral databases, development of IM-MS spectral databases becomes a necessity for identification of metabolites present in biological samples. In addition, as has been reported for proteins and peptides (9597), extension of two dimensional IM-MS technique to higher order of dimensionality in separation (Xm-IMn-MSo, where “X” is an added dimension in separation) such as IM-MS/MS or LC-IM-MS/MS would allow confirmed identification of known metabolites and discovery of unknown metabolites. Various amino acids (arginine, tryptophan etc), carbohydrates (ribose, glucose etc), sugar phosphates (glucose 6-phosphate, fructose 6-phosphate), steroids (estrogen derivatives), nucleotides, nucleosides organic acids, bile acids (deoxycholic acid, lithocholic acid etc), acyl fatty acids (prostaglandin D2, E2, I2 etc) glycerophospholipids (GPCho[17:0/14:1{9Z}], GPCho[16:0/18:1{9Z}], etc), phospholipids are some of the classes of metabolites that were detected in the extract. Selected classes of metabolites detected as protonated ions are highlighted in Figure 6S and Table 2S (see supporting information) lists the metabolites detected with ≥ 5 ion counts along with their tentative identification. Identity of amino acids, pentose sugars and hexose sugars detected in blood extract were validated by comparing the m/z values and reduced mobility values of peaks detected in the blood extract to that measured for standard solutions using ESI-IM-MS. Following sections discuss selected classes of metabolites detected (validated and tentative).

3.2.2.2 Mobility-Mass-Correlation Curves

Separation of various classes of compounds such as lipids and peptides based on characteristic mobility-mass-correlation curve (MMCC) formed in two dimensional IM-MS space has been shown(98). MMCC can provide a valuable tool in the identification of peaks detected in an IM-MS analysis of complex samples. Figure 7 shows the different MMCCs observed for the various classes of metabolites detected in the blood extract by IM-MS.

Figure 7
A: MMCC for amino acids detected in the blood extract. Peaks identified as amino acids in blood extract based on mass and reduced mobility data matched with that measured for standard solutions of amino acids. B: MMCC for various classes of metabolites ...

Figure 7A shows the 2D IM-MS plot in the m/z range of 50–250 and drift time range of 7–14 ms where peaks encircled correspond to the m/z values of protonated amino acids forming the MMCC. Identification of the peaks encircled was performed by comparing the reduced mobility values (Ko) measured using equation 2 for protonated ions of amino acids (50 µM standard solutions) to that detected in the blood sample. The insert in Figure 7A shows the MMCC obtained by plotting the 1/Ko values measured for the standard solutions of amino acid peaks and the peaks identified as amino acids in the blood extract against their respective m/z values (note the inverse relationship between drift time td and Ko in equation 2). Distribution of amino acid peaks found in the blood along the MMCC similar to the MMCC obtained for the standard amino acid solutions supported the identification of the peaks as that of amino acids especially in cases where multiple peaks were detected at same m/z values. For example, at m/z 205.2 two isomeric metabolite peaks were observed at Ko values of 1.39 and 1.48 cm2V−1s−1. The m/z values of the two peaks corresponded to m/z values of two endogenous metabolites, tryptophan (m/z 204.23) and oxalo-glutarate (m/z 204.14), detected as protonated ions. The peak at m/z 205.2 that had a Ko value of 1.39 cm2V−1s−1 was identified as that of tryptophan since both the Ko value and the m/z value of the detected peak was within the standard deviation of that measured for a protonated ion of tryptophan standard (1.41 ± 0.02 cm2V−1s−1). This peak also followed the MMCC for the amino acids which also supported the identification as tryptophan. Measured Ko values (blood and standard) and m/z values of detected amino acids are listed in Table 1. A maximum error of 2% was observed between the Ko values of amino acids detected in the blood extract and that of the standards. A deviation of 2% in Ko value is also generally observed in day-to-day IMS measurements.

Various isomers of monosaccharides were also detected and identification established. Sugars preferentially form sodium adducts upon ionization by ESI but are also detected in low abundance as adduct ions with proton, hydronium ion, protonated methanol, and potassium(99, 100). Pentose sugars were detected at m/z values of 151 (pentose + proton), 165 (methyl pentose + proton), 169 (pentose + hydronium ion), and 173 (pentose + sodim ion). Hexose sugars were detected at m/z values of 181 (hexose + proton), 195 (methylhexose + proton), 203 (hexose + sodium) and 219 (hexose + potassium). Similar to that of amino acids, isomeric sugars were distributed along a MMCC. Due to the isomeric nature of sugars, the sugar isomers were distributed in ~1 ms mobility space along the MMCC whereas the amino acids were confined in <0.5 ms mobility space. Identification of the detected peaks as sugar isomers were based on 1) comparison to Ko values of monosaccharides(99), 2) their respective m/z values, and 3) their position along the MMCC in the 2D IM-MS space.

Detection and confirmed identification of amino acids and sugars in the blood extract, demonstrate the strength of IM-MS for rapid analysis of complex biological samples to monitor targeted or comprehensive metabolites in the metabolome. Detection of free amino acids present in 50 µL of blood sample using IM-MS without any pre-concentration step demonstrate the potential of IM-MS to detect metabolites such as amino acids that are present in sub-micromolar concentration levels and usually require pre-concentration prior to analysis by chromatographic methods (101).

Figure 7B show some selected MMCCs formed by other classes of metabolites. Tentative identification of the metabolites forming the MMCCs was exclusively based on the m/z values of ions and only peaks corresponding to protonated ions (M + H)+ of the metabolites were included.

Because drift times of ions increase with size of the ions, MMCC also provide information on the structure of ions in addition to serving as a visual aide for identification of metabolites in a complex spectrum. For example, in the 2D IM-MS space the amines occupied IM-MS space towards the longer drift times while the unsaturated fatty acids occupied the 2D space towards the shorter drift time relative to amino acid-carbohydrate MMCC (Figure 6S, A). This suggests that amines form less compact ions whereas the unsaturated fatty acids tend to form more compact ions upon ionization by ESI, relative to amino acids and monosaccharides. Similarly, in the m/z range of 380 – 600 Da (Figure 6S, B) sterols, diacylglycerophosphocholines, and sphingolipids were detected in the drift time range of 17 to 25 ms whereas the monoacyl glycerophosphocholines and isoprenoids occupied the 2D space in the drift time range of 15–18 ms. Sterols and their derivatives such as stigmasterol, dihydrocholesterol, dihydroxy lanosterol, cholestane-triol have structures that constitute multiple 5 or 6 membered rings with a short hydrocarbon chain, that could prevent folding of the molecule upon ionization. Diacyl glycerophosphocholines, with two hydrocarbon chains, formed less compact ions (drift time range of 17 to 25 ms) compared to the monoacylglycerophosphocholines and isoprenoids (drift time range of 15–18 ms). Review of the structures of sphingolipids and endogeneous blood isoprenoids such as carotenoids and retinyl oleate(22, 102), showed that the structures of sphingolipids are less favorable to folding due to two fatty acid tails and usually modified heads when compared to the structures of the isoprenoids that are in general single chain hydrocarbon derivatives. N-(hexadecanoyl)-sphinganine (ceramides), Oleoyl-sphingosine, (3'-sulfo)Galβ-N-(acyl)-sphing-4-enine, and N-(9Z-octadecenoyl)-sphing-4-enine, diacyl-glycerophospho-cholines, diacyl-glycerophospho-ethanolamines are few of the sphingolipids and Bisdehydro-β-carotene/ Tetradehydro-β-carotene (m/z 533), β-carotene (m/z 537), and 4-Keto-γ-carotene/ Keto-γ-carotene (m/z 551) are few of the isoprenoids detected.

MMCC for protonated glycerophospho-ethanolamines (for example at m/z 688, 698, 714, 736), and diacylglycerols (for example at m/z 605, 617, 618, 615, 642, 645, 660, 685, 701, 730) also show the effect of structural makeup of molecules on size of the ions. Note the high intensity peaks (Figure 6S, C) at m/z values of 672, 685, 688, and 701 Da corresponding to protonated glycosphingolipid [GlcCer(d18:1/14:0)], diacylglycerol [DG (19:0/22:5 (7Z,10Z,13Z,16Z,19Z)/0:0)], glycerophosphoethanolamine [GPEtn (16:1(9Z)/16:1(9Z))], and diacylglycerol [DG (20:1(11Z)/22:3(10Z,13Z,16Z)/0:0)]. Amongst these peaks, ions with higher drift times were that of glycosphingolipid at m/z 672 and glycerophosphoethanolamine at m/z 688 suggesting that these ions were relatively less dense than the diacylglycerols at m/z values of 685 and 701. This observation can be explained if the presence of comparatively larger head groups in the case of glycosphingolipid and glycerophosphoethanolamine compared to that of diacylglycerols resulted in formation of relatively less compact ions due to hindered folding of the aliphatic chains.

MMCC for estrogen derivatives and prostaglandins (lipid biosynthesis/metabolism mediators) are also shown in Figure 7B. Eicosanoids such as prostaglandins are biochemically synthesized from the arachidonic acid and are found in almost all tissues in the human body affecting many essential physiological functions such as contraction and relaxation of smooth muscle, the dilation and constriction of blood vessels, control of blood pressure, and modulation of inflammation (103). Relative distribution of these metabolites serve as indicators of affects of regulatory mechanisms on specific metabolic pathways and thus ability to separate isomeric forms, detect, and monitor them simultaneously becomes important (104). Examples of estrogen derivatives detected (circled in the Figure 6S, D) include hydroxyestrapentanone (m/z 267), estrone (m/z 217), estradiol (m/z 273), hydroxyestradiol (m/z 289), andosterone (m/z 291) and methoxyestrone (m/z 301). Different prostaglandins were detected and are encircled in Figure 6S, E. Along with separation of various isomeric prostaglandins different forms of prostaglandins (PGA, PGB, PGC, PGD, PGE, PGF, PGG and their dimethyl, deoxy, glyceryl, keto, hydroxyl, ethanolamine, and phenyl derivatives) were detected. For example, three isomeric peaks, mass identified as that of PGA3, PGB3, and PGJ3, were separated in the mobility space at m/z value of 333.3 Da. Similarly, isomeric peaks at m/z value of 429.3 Da, possibly that of 2-Glyceryl-PGF2α and 1(3)-Glyceryl-PGF2α were mobility separated.

Figure 6S, D also shows peaks at m/z values corresponding to that of octopine, arginosuccinic acid, deoxyinosine, guanosine, and xanthosine observed at m/z values of 265 (M+H3O+), 291 (M+H+), 253(M+H+), 284(M+H+) and 285(M+H+) respectively. These endogeneous metabolites belong to various metaboloic pathways such as arginine and proline metabolism pathway, alanine and aspartate metabolism pathway, and purine and caffeine metabolism pathway(94). Isomeric pentose phosphates, hexose phosphates, hexose alcohol phosphates, hexose phosphoacids, and hexose N-acetyl amines as protonated and/or sodiated adducts are also encircled in the Figure 6S, D. Peaks corresponding to the m/z values of protonated ions of purines and pyrimidines: adenine (m/z 137), guanine (m/z 152), thymine (m/z 127), cytosine (m/z 112), uracil (m/z 113), unsubstituted purine (m/z 121), xanthine (m/z 153), hypoxanthine (m/z 137) and uric acid (m/z 169) were detected.

Figure 6S, F (IM-MS contour plot in the m/z range of 600–750 Da., (see supporting information) illustrates another example where IMS when used as a pre-separation technique to MS measurements aids in the interpretation of peaks observed in a complex mass spectrum of biological samples. For example, isomeric pairs at m/z values of 619, 670, 696, and 712 Da (circled in Figure 6S, F). Interestingly, these isomers followed a MMCC suggesting that the ions observed at the above m/z values are structurally related (105107). Methylated heme (heme, M) has an m/z value of 618, thus the peaks at m/z values of 619, 670, 696 and 712 Da tentatively identified as (heme M+1)+, (heme M+52)+, (heme M+78)+, (heme M+94)+ ions respectively. Structure of heme M (shown as insert in Figure 6S, F) shows that there are two potential sites for methylation (A and B) which may result in two isomeric structures.

3.2.3 Isomer separation

Diversity in lipid and carbohydrate structures and existence of each in various isomeric forms render lipid and carbohydrate profiling a very complicated and difficult task (108). The most important contribution of ion mobility spectrometry in an IM-MS experiment is the separation of isomeric molecules simultaneous to detection by mass spectrometry. Figure 8a is 2D IM-MS spectra in the m/z range of 190–225 Da, with peaks less than 10 ions counts excluded by raising the data processing threshold, demonstrating the separation of isomers by IMS. For example at an m/z value of 219 Da, four mobility separated peaks possibly potassium adducts of hexose sugars or sodium adducts of hexose sugar acids were detected. Similarly, four mobility separated peaks at m/z 195 Da and two mobility separated peaks each at m/z values of 191, 197, 209, 217, and 203 Da were detected. Isomeric peaks at m/z 203 were identified as sodium adducts of methylated glucose and mannose by comparing the reduced mobility values and m/z values obtained for the peaks detected in blood to that of standards of methylated glucose and mannose. Figure 8b demonstrates the separation of over 200 isomeric/isobaric metabolites detected with ≥ 10 ion counts in blood extract by IM-MS analysis. On the x-axis is shown the mass-to-charge ratio of the ions in Da and the y-axis represents the drift time (IMS) of ions in milliseconds. Two metabolite peaks with m/z values within m/z standard deviation of ± 0.1 Da were considered isomers if their drift times were different by at least 0.2 ms. Isomeric ions corresponding to two glycerophosphoethanolamines GPEtn(18:1(9Z)/22:1(13Z); MW 799.6) and GPEtn(20:0/20:2(11Z,14Z); MW 799.6) detected at m/z 800.6 Da at drift times of 24.7 and 25.1 ms were separated with drift time difference of 0.4 ms. A drift time difference of 4.3 ms was observed between isomeric ions detected at m/z 801.7 and 801.8 Da corresponding to m/z values of protonated triacylglycerol, TG(16:1(9Z)/16:1(9Z)/16:1(9Z); m/z 800.7 Da) and that of dialkylglycerophosphocholine; m/z 800.6 Da) were observed at drift times of 27.6 and 23.1 ms. This experiment thus demonstrates that application of IM-MS to metabolomics would provide rapid detection of large number of metabolites present in complex biological matrix such as blood and simultaneous separation of isomeric species that cannot be easily differentiated by MS measurements alone.

Figure 8
a) Enlarged section of Figure 4a in the m/z range of 190–225 Da showing the separation of isomers/isobars by IMS. Four mobility separated peaks identified as potassium adducts of hexose sugars at m/z value of 219 Da along with four mobility separated ...

3.2.4 Peak capacity of IM-MS

With tandem separation techniques such as IM-MS, CE-MS, LC-MS, maximum number of peaks that can fit in a two dimensional space is defined as the peak capacity ([var phi]) of the tandem method (109). Thus peak capacity is a measure of the separation potential of the technique and depends on the resolution of the individual systems and difference in their separation mechanism (orthogonality). A 2D separation technique will have very high [var phi] value if the fundamental separation mechanisms of individual high resolution systems are completely different such as in LC-MS(110, 111). The peak capacity will also be defined by the complexity of the sample and the physical and chemical properties of the analytes in the sample. For example complex biological samples such as protein/peptide mixtures or metabolomes containing large number of analytes with varying chemical nature would provide larger peak capacities for the technique than that would be through analysis of a mixture of analytes of a specific class. For example the peak capacity of LC-MS, where LC separations occur due to differences in partition coefficient of analyte ions between mobile phase and stationary phase and MS separation is based on their m/z ratios, is quite high due to completely different mechanism of separations involved in the individual analytical method (112). With IM-MS as a tandem technique where separation occurs due to the differences in the Ω/z values (ion mobility) and m/z values (mass spectrometry), the plot of Ω/z versus m/z generally follow linear trend and thus exhibit smaller [var phi] than LC-MS. With MS resolution of 400, IMS resolution of 60, and a maximum deviation of 11% in drift time, 1 to 5 fold increases in the peak capacity of the IM-MS method has been reported for a mixture of tryptic peptides (m/z range 500–2500). Though peak capacities of upto 107 can be achieved for LC-MS upon analysis of protein/peptide mixtures (113, 114), number of metabolites detected in blood under single experimental condition has been reported to be between ~1000–3000 using LC-MS method (115117). However, even with limited orthogonality in separation, the high resolving power of IMS allowed separation of ~1100 metabolic features in a single experiment using the IM-MS method which is comparable to the number of metabolic features detected by LC-MS (~103). In addition, because ion mobility is a post-ionization gas phase separation method, temporal separation occurs in milliseconds resulting in very high peak production rate ([var phi]s−1). IM-MS as compared to LC-MS has at least three orders of magnitude greater [var phi]s−1 than LC-MS. An estimation of the peak capacity in IM-MS can be calculated as follows:

equation M4

Where, RIMS is the average resolving power of the ion mobility spectrometer, and RMS is the average resolution of the mass spectrometer in an m/z range. For example, in the mass range of 23–830 Da for the metabolites detected the minimum and maximum deviation in drift time along a vertical plane parallel to the y-axis defines the fraction of orthogonality as shown by double sided arrows in Figure 9. The theoretical MMCC shown as dotted line in the figure is the centre line of the shaded area along which a constant drift time deviation should be observed whereas the MMCC obtained for the experimental data is depicted as bold line in Figure 9. The length of the 2D space is defined by the m/z range and double sided arrow defines the drift time spread in the given m/z range. With a drift time spread of ~14.3 ms the total % deviation measures to be ~28 % [14.3 / (18.37/2 + 32.66/2)] of the total 2D space (Figure 9, shaded space in the X-Y quadrant). This gives an average deviation in drift time of ± 14% along the theoretical trend. With average IMS resolving power of 90, average MS resolution of 1500, and ± 14 % deviation, the estimated peak capacity for the instrument was ~18900. With MS resolution of 1500, peak width at m/z 404 is 0.27 Da (for the m/z range of 23–830 Da the average m/z value is equal to 404 Da) the estimated peak capacity of MS alone was 2989. By using IMS in tandem with MS, six fold increase in the peak capacity was observed (~19000) compared to peak capacity of the MS as a standalone method of separation (~3000).

Figure 9
Two dimensional space (shaded area in the X-Y quadrant) occupied by metabolites that were detected in human blood extract by ESI- IM-MS (Peak capacity) is illustrated. The theoretical MMCC (dotted line) is the centre line of the shaded area along which ...

Conclusions

The high resolution ion mobility mass spectrometer used in this study provided an average IMS resolving power of ~ 90 and MS resolution of ~1500. Reproducible reduced mobility values for metabolites in milli-molar to nano-molar concentration range can be achieved. Without any pre-concentration, over 1100 metabolites can be detected in methanol extracts of 50 µL of blood samples. Along with the detection of metabolites in the blood extract, simultaneous separation of over 300 isomeric/isobaric metabolites can also be accomplished. Various classes of metabolites including amino acids, carbohydrates, endogenous amines, purines and pyrimidines, organic acids, sterols, estrogens, prostaglandins, phosphocholines, mono- and di-acylglycerophosphoethanolamines, mono- and di-acylglycerols, sphingolipids, isoprenoids and various metabolic intermediates can be concurrently detected and separated in a 2D IM-MS analysis. Peak capacity of mass spectrometers can be increased by a factor of six by coupling of IMS as a pre-separation technique prior to MS analysis which can be further increased by increasing the separation power of IMS and MS.

An important advantage of IM-MS is separation of similar classes of metabolites along a unique MMC lines that can be used to complement and facilitate identification of metabolites detected by MS. Two dimensional IM-MS analysis facilitates separation of random noise peaks from low intensity metabolite peaks through differentiation in mobility space and thus make possible detection of low abundance metabolites.

When compared to LC-MS when applied to blood metabolite detection, IM-MS even with limited orthogonality in separation, demonstrated detection of comparable number of metabolic features in the blood metabolome due to high resolving power of the IMS. Ion mobility mass spectrometry with its ability to rapidly provide high resolution comprehensive complex sample analyses and its compatibility with various analytical methods such as chromatography appears to be a promising new analytical technique for assessment of metabolomes. Further improvement in IM-MS separation power and ion transmission efficiency would significantly enhance the potential of IM-MS as applied to metabolomics studies. Development of IM-MS spectral databases and multidimensional instrumentation and data processing software would allow rapid metabolome analysis and interpretation.

Supplementary Material

01

Table 2S (supporting information): Partial list of metabolites detected in human blood metabolome:

Table 2S lists the mass to charge ratio in Daltons (Da), drift time in milliseconds (ms), and tentative metabolic feature identity in columns 1, 2, and 3 respectively of ions detected in human blood metabolome with ≥ 5 ion counts. The ions were identified as (M+H)+, (M+H3O)+, (M+H5O2)+, (M+CH4OH)+, (M+CH4COOH)+, (M+Na)+, (M+K)+, (M+H2ONa)+, (M+H2OK)+, (M+NH4)+ and/or (M-H2O+H)+.

02

Figure 2S (supporting information)

Frames a), b), and c) illustrate peak shapes for IM-MS signals with 3, 5 and 45 ion counts respectively. Frame d) illustrates an unprocessed IM-MS contour plot illustrating random distribution of noise signals in the 2D IM-MS space along with analyte signals.

Figure 6S (supporting information): Zoomed in IM-MS frames highlighting various classes of metabolites detected in the blood extract.

Figure 6S, A: 2D IM-MS plot highlighting tentatively identified metabolites detected as protonated ions of various endogenous amines, methyl branched fatty acids, oxo-fatty acids, unsaturated fatty acids, purines, pyrimidines, and oxypurines.

Figure 6S, B: IM-MS contour plot in the m/z range of 380 – 600 Da highlighting peaks corresponding to m/z values of protonated sterols, diacylglycerophosphocholines, and sphingolipids in the drift time range of 17 to 25 ms and that of monoacyl glycerophosphocholines and isoprenoids in the drift time range of 15–18 ms.

Figure 6S, C: Peaks corresponding to mass identified metabolites glycerophospho-ethanolamines and diacylglycerols as protonated ions are highlighted.

Figure 6S, D: Peaks corresponding to protonated ions of various estrogen derivatives, metabolic pathway metabolites, and protonated and/or sodiated adduct of isomeric pentose phosphates, hexose phosphates, hexose alcohol phosphates, hexose phosphoacids, and hexose N-acetyl amines are highlighted.

Figure 6S, E: IM-MS spectrum in the m/z range of 300–450 Da highlighting peaks corresponding to various mass identified prostaglandins detected as protonated ions.

Figure 6S, F: IM-MS contour plot in the m/z range of 600–750 Da showing the MMCC formed by isomeric peaks at m/z values of 619, 670, 696 and 712 Da tentatively identified as (heme M+1)+, (heme M+52)+, (heme M+78)+, (heme M+94)+ ions respectively where heme M denotes methylated heme molecule. The insert shows Structure of heme shows that there are two potential sites for methylation (A and B) which may result in two isomeric structures.

Acknowledgements

This work was supported in part by a grant from National Institute of Health- Road Map Grant #R21-DK070274. We also would like to thank Mr. Roger H. Crawford at Washington State University in Pullman, WA and Drs. Agnes Tempez, and Thomas F. Egan at IonWerks in Houston, TX for their help throughout this project.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

1. David LN, Michael MC. In: Lehninger Principles of Biochemistry. Fourth Edition Freeman WH, editor. 2004.
2. Michal G. In: Biochemical pathways:An atlas of Biochemistry and molecular biology. Michal G, editor. New york: John Wiley & sons; 1999.
3. Goodacre R. Metabolomics-the way forward. Metabolomics. 2005;1(1):1–2.
4. Goodacre R, Vaidyanathan S, Dunn WB, Harrigan GG, Kell DB. Metabolomics by numbers: acquiring and understanding global metabolite data. Trends in Biotechnology. 2004;22(5):245–252. [PubMed]
5. Thomas GH. Metabolomics breaks the silence. Trends in Microbiology. 2001;9(4) 158-158. [PubMed]
6. Horning EC, Horning MG. Metabolic Profiles: Gas-Phase Methods for Analysis of Metabolites. Clin Chem. 1971;17(8):802–809. [PubMed]
7. Dunckley T, Coon KD, Stephan DA. Discovery and development of biomarkers of neurological disease. Drug Discovery Today. 2005;10(5):326–334. [PubMed]
8. Fell DA. Beyond genomics. Trends in Genetics. 2001;17(12):680–682. [PubMed]
9. Fernie AR, Trethewey RN, Krotzky AJ, Willmitzer L. Innovation - Metabolite profiling: from diagnostics to systems biology. Nature Reviews Molecular Cell Biology. 2004;5(9):763–769. [PubMed]
10. Fiehn O. Metabolomics - the link between genotypes and phenotypes. Plant Molecular Biology. 2002;48(1–2):155–171. [PubMed]
11. Kodo N, Millington DS, Norwood DL, Roe CR. Quantitative assay of free and total carnitine using tandem mass spectrometry. Clinica Chimica Acta. 1990;186(3):383–390. [PubMed]
12. Rashed MS, Bucknall MP, Little D, et al. Clinical Chemistry. 7. Vol. 43. Washington, D.C.: 1997. Screening blood spots for inborn errors of metabolism by electrospray tandem mass spectrometry with a microplate batch process and a computer algorithm for automated flagging of abnormal profiles; pp. 1129–1141. [PubMed]
13. Giovannoni G, Heales SJR, Silver NC, et al. Raised serum nitrate and nitrite levels in patients with multiple sclerosis. Journal of the Neurological Sciences. 1997;145(1):77–81. [PubMed]
14. Zangerle R, Fuchs D, Reibnegger G, et al. Serum nitrite plus nitrate in infection with human immunodeficiency virus type-1. Immunobiology. 1995;193(1):59–70. [PubMed]
15. Kawashima H, Inage Y, Ogihara M, et al. Serum and cerebrospinal fluid nitrite/nitrate levels in patients with rotavirus gastroenteritis induced convulsion. Life Sciences. 2004;74(11):1397–1405. [PubMed]
16. Kawashima H, Watanabe Y, Ichiyama T, et al. High concentration of serum nitrite/nitrate obtained from patients with influenza-associated encephalopathy. Pediatrics International. 2002;44(6):705–707. [PubMed]
17. Zarzycki PK, Kulhanek KM, Smith R, Clifton VL. Determination of steroids in human plasma using temperature-dependent inclusion chromatography for metabolomic investigations. Journal of Chromatography, A. 2006;1104(1–2):203–208. [PubMed]
18. Figueroa V, Milla C, Parks EJ, Schwarzenberg SJ, Moran A. Abnormal lipid concentrations in cystic fibrosis. American Journal of Clinical Nutrition. 2002;75(6):1005–1011. [PubMed]
19. Caulfield MP, Lynn T, Gottschalk ME, et al. The Diagnosis of Congenital Adrenal Hyperplasia in the Newborn by Gas Chromatography/Mass Spectrometry Analysis of Random Urine Specimens. J Clin Endocrinol Metab. 2002;87(8):3682–3690. [PubMed]
20. Shackleton C. Biochemical diagnosis of Antley-Bixler syndrome by steroid analysis. American Journal of Medical Genetics Part A. 2004;128A(3):223–231. [PubMed]
21. Gornati R, Bembi B, Tong X, Boscolo R, Bruno B. Total glycolipid and glucosylceramide content in serum and urine of patients with Gaucher's disease type 3 before and after enzyme replacement therapy. Clinica Chimica Acta. 1998;271(2):151–161. [PubMed]
22. Zhang X, Kiechle Frederick L. Review: Glycosphingolipids in health and disease. Annals of clinical and laboratory science. 2004;34(1):3–13. [PubMed]
23. Vance DE, Krivit W, Sweeley CC. Concentrations of glycosyl ceramides in plasma and red cells in Fabry's disease, a glycolipid lipidosis. J.Lipid Res. 1969;10(2):188–192. [PubMed]
24. Tall AR, Jiang XC. Detection and treatment of atherosclerosis based on plasma sphingomyelin concentration. New York, USA: Application: WO: Trustees of Columbia University in the City of; 2001. p. 76.
25. Mirzoyan SA, Mkheyan EE, Sekoyan ES, Grigoryan EK, Bakunts GO. Blood glycolipids in patients with brain blood-circulation disorders. Zhurnal Nevropatologiii Psikhiatrii imeni S.S.Korsakova. 1978;78(9):1317–1322. [PubMed]
26. Atzpodien W, Huewels G, Kremer GJ, Schnellbacher E. Study of plasma glycosphingolipids in acute hepatitis. Verhandlungen der Deutschen Gesellschaft fuer Innere Medizin. 1976;82(1):429–431. [PubMed]
27. Markuszewski MJ, Szczykowska M, Siluk D, Kaliszan R. Human red blood cells targeted metabolome analysis of glycolysis cycle metabolites by capillary electrophoresis using an indirect photometric detection method. Journal of Pharmaceutical and Biomedical Analysis. 2005;39(3–4):636–642. [PubMed]
28. Dettmer K, Hammock BD. Metabolomics - A new exciting field within the "omics" sciences. Environmental Health Perspectives. 2004;112(7):A396–a397. [PMC free article] [PubMed]
29. Dunn WB, Ellis DI. Metabolomics: Current analytical platforms and methodologies. Trac-Trends in Analytical Chemistry. 2005;24(4):285–294.
30. Fiehn O. Combining genomics, metabolome analysis, and biochemical modelling to understand metabolic networks. Comparative and Functional Genomics. 2001;2(3):155–168. [PMC free article] [PubMed]
31. Jiye A, Trygg J, Gullberg J, et al. Extraction and GC/MS analysis of the human blood plasma metabolome. Analytical Chemistry. 2005;77(24):8086–8094. [PubMed]
32. Hope JL, Prazen BJ, Nilsson EJ, Lidstrom ME, Synovec RE. Comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry detection: analysis of amino acid and organic acid trimethylsilyl derivatives, with application to the analysis of metabolites in rye grass samples. Talanta. 2005;65(2):380–388. [PubMed]
33. Katz JE, Dumlao DS, Clarke S, Hau J. A new technique (COMSPARI) to facilitate the identification of minor compounds in complex mixtures by GC/MS and LC/MS: Tools for the visualization of matched datasets. Journal of the American Society for Mass Spectrometry. 2004;15(4):580–584. [PubMed]
34. Taylor J, King Ross D, Altmann T, Fiehn O. Bioinformatics. Vol. 18. Oxford, England: 2002. Application of metabolomics to plant genotype discrimination using statistics and machine learning; pp. S241–S248. [PubMed]
35. Gamache PH, Meyer DF, Granger MC, Acworth IN. Metabolomic applications of electrochemistry/mass spectrometry. Journal of the American Society for Mass Spectrometry. 2004;15(12):1717–1726. [PubMed]
36. Lu W, Kimball E, Rabinowitz JD. A High-Performance Liquid Chromatography-Tandem Mass Spectrometry Method for Quantitation of Nitrogen-Containing Intracellular Metabolites. Journal of the American Society for Mass Spectrometry. 2006;17(1):37–50. [PubMed]
37. Pham-Tuan H, Kaskavelis L, Daykin Clare A, Janssen H-G. Method development in high-performance liquid chromatography for high-throughput profiling and metabonomic studies of biofluid samples. Journal of chromatography.B, Analytical technologies in the biomedical and life sciences. 2003;789(2):283–301. [PubMed]
38. Vigneau-Callahan KE, Shestopalov AI, Milbury PE, Matson WR, Kristal BS. Characterization of diet-dependent metabolic serotypes: Analytical and biological variability issues in rats. Journal of Nutrition. 2001;131(3) 924S–932S. [PubMed]
39. Wagner S, Scholz K, Donegan M, Burton L, Wingate J, Voelkel W. Metabonomics and Biomarker Discovery: LC-MS Metabolic Profiling and Constant Neutral Loss Scanning Combined with Multivariate Data Analysis for Mercapturic Acid Analysis. Analytical Chemistry. 2006;78(4):1296–1305. [PubMed]
40. Wilson Ian D, Plumb R, Granger J, Major H, Williams R, Lenz Eva M. MHPLC-MS-based methods for the study of metabonomics. Journal of chromatography. B, Analytical technologies in the biomedical and life sciences. 2005;817(1):67–76. [PubMed]
41. Yang J, Zhao X, Liu X, et al. High Performance Liquid Chromatography-Mass Spectrometry for Metabonomics: Potential Biomarkers for Acute Deterioration of Liver Function in Chronic Hepatitis B. Journal of Proteome Research. 2006;5(3):554–561. [PubMed]
42. Anon. Research Profile; CE/MS for quantitative metabolome analysis. Journal of Proteome Research. 2003;2(5):463. [PubMed]
43. Sato S, Soga T, Nishioka T, Tomita M. Simultaneous determination of the main metabolites in rice leaves using capillary electrophoresis mass spectrometry and capillary electrophoresis diode array detection. Plant Journal. 2004;40(1):151–163. [PubMed]
44. Soga T, Ohashi Y, Ueno Y, Naraoka H, Tomita M, Nishioka T. Quantitative metabolome analysis using capillary electrophoresis mass spectrometry. Journal of Proteome Research. 2003;2(5):488–494. [PubMed]
45. Suresh BCV, Song EJ, Babar SME, Wi MH, Yoo YS. Capillary electrophoresis at the omics level: towards systems biology. Electrophoresis. 2006;27(1):97–110. [PubMed]
46. Chen X, Kong L, Su X, Pan C, Ye M, Zou H. Integration of ion-exchange chromatography fractionation with reversed-phase liquid chromatography-atmospheric pressure chemical ionization mass spectrometer and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry for isolation and identificationofcompounds in Psoralea corylifolia. Journal of chromatography.A. 2005;1089(1–2):87–100. [PubMed]
47. Lenz EM, Bright J, Wilson ID, Morgan SR, Nash AFP. A 1H NMR-based metabonomic study of urine and plasma samples obtained from healthy human subjects. Journal of Pharmaceutical and Biomedical Analysis. 2003;33(5):1103–1115. [PubMed]
48. Bhalla R, Narasimhan K, Swarup S. Metabolomics and its role in understanding cellular responses in plants. Plant Cell Reports. 2005;24(10):562–571. [PubMed]
49. Nicholls AW, Holmes E, Lindon JC, et al. Metabonomic investigations into hydrazine toxicity in the rat. Chemical research in toxicology. 2001;14(8):975–987. [PubMed]
50. Holmes E, Nicholls AW, Lindon JC, et al. Chemometric models for toxicity classification based on NMR spectra of biofluids. Chemical research in toxicology. 2000;13(6):471–478. [PubMed]
51. Wu H, Zhang X, Liao P, et al. NMR spectroscopic-based metabonomic investigation on the acute biochemical effects induced by Ce(NO3)3 in rats. J Inorg Biochem. 2005;99(11):2151–2160. [PubMed]
52. Viant MR, Rosenblum ES, Tieerdema RS. NMR-based metabolomics: a powerful approach for characterizing the effects of environmental stressors on organism health. Environ Sci Technol. 2003;37(21):4982–4989. [PubMed]
53. Pendyala G, Want EJ, Webb W, Siuzdak G, Fox HS. Biomarkers for neuroAIDS: The widening scope of metabolomics. Journal of Neuroimmune Pharmacology. 2007;2(1):72–80. [PubMed]
54. Dunn WB. Current trends and future requirements for the mass spectrometric investigation of microbial, mammalian and plant metabolomes. Phys Biol. 2008;5(1):11001. [PubMed]
55. Godin JP, Fay LB, Hopfgartner G. Liquid chromatography combined with mass spectrometry for 13C isotopic analysis in life science research. Mass Spectrom Rev. 2007;26(6):751–774. [PubMed]
56. Metz TO, Zhang Q, Page JS, et al. The future of liquid chromatography-mass spectrometry (LC-MS) in metabolic profiling and metabolomic studies for biomarker discovery. Biomark Med. 2007;1(1):159–185. [PMC free article] [PubMed]
57. Witkamp RF. Genomics and systems biology--how relevant are the developments to veterinary pharmacology, toxicology and therapeutics? J Vet Pharmacol Ther. 2005;28(3):235–245. [PubMed]
58. Adde Jong GJ. Foreword: new developments in CE, CIEF and CE for metabolomics. J Chromatogr A. 2008;1204(2):129. [PubMed]
59. Barbas C, Vallejo M, Garcia A, Barlow D, Hanna-Brown M. Capillary electrophoresis as a metabolomic tool in antioxidant therapy studies. J Pharm Biomed Anal. 2008;47(2):388–398. [PubMed]
60. Monton MR, Soga T. Metabolome analysis by capillary electrophoresis-mass spectrometry. J Chromatogr A. 2007;1168(1–2):237–246. discussion 236. [PubMed]
61. Ramautar R, Somsen GW, de Jong GJ. CE-MS in metabolomics. Electrophoresis. 2009;30(1):276–291. [PubMed]
62. Soga T. Capillary electrophoresis-mass spectrometry for metabolomics. Methods Mol Biol. 2007;358:129–137. [PubMed]
63. van der Kooy F, Maltese F, Choi YH, Kim HK, Verpoorte R. Quality control of herbal material and phytopharmaceuticals with MS and NMR based metabolic fingerprinting. Planta Med. 2009;75(7):763–775. [PubMed]
64. Queiroz EF, Wolfender JL, Hostettmann K. Modern approaches in the search for new lead antiparasitic compounds from higher plants. Curr Drug Targets. 2009;10(3):202–211. [PubMed]
65. Powers R. NMR metabolomics and drug discovery. Magn Reson Chem. 2009 [PubMed]
66. Vangala S, Tonelli A. Biomarkers, metabonomics, and drug development: can inborn errors of metabolism help in understanding drug toxicity? AAPS J. 2007;9(3):E284–E297. [PMC free article] [PubMed]
67. Orhan H. Analyses of representative biomarkers of exposure and effect by chromatographic, mass spectrometric, and nuclear magnetic resonance techniques: method development and application in life sciences. J Sep Sci. 2007;30(2):149–174. [PubMed]
68. Sem DS, Pellecchia M. NMR in the acceleration of drug discovery. Curr Opin Drug Discov Devel. 2001;4(4):479–492. [PubMed]
69. DeAngelis CD, Zerhouni EA. NIH Director Elias A. Zerhouni, MD, reflects on agency's challenges, priorities. Jama-Journal of the American Medical Association. 2003;289(12):1492–1493. [PubMed]
70. Zerhouni E. The NIH roadmap. Science. 2003;302(5642) [PubMed]
71. Valentine SJ, Plasencia MD, Liu XY, et al. Toward plasma proteome profiling with ion mobility-mass spectrometry. Journal of Proteome Research. 2006;5(11):2977–2984. [PubMed]
72. Fenn LS, Kliman M, Mahsut A, Zhao SR, McLean JA. Characterizing ion mobility-mass spectrometry conformation space for the analysis of complex biological samples. Analytical and Bioanalytical Chemistry. 2009;394(1):235–244. [PMC free article] [PubMed]
73. Kanu AB, Dwivedi P, Tam M, Matz L, Hill HH. Ion mobility-mass spectrometry. Journal of Mass Spectrometry. 2008;43(1):1–22. [PubMed]
74. Hyung SJ, Robinson CV, Ruotolo BT. Gas-phase unfolding and disassembly reveals stability differences in ligand-bound multiprotein complexes. Chem Biol. 2009;16(4):382–390. [PubMed]
75. Chan EC, New LS, Yap CW, Goh LT. Pharmaceutical metabolite profiling using quadrupole/ion mobility spectrometry/time-of-flight mass spectrometry. Rapid Commun Mass Spectrom. 2009;23(3):384–394. [PubMed]
76. Mie A, Ray A, Axelsson BO, Jornten-Karlsson M, Reimann CT. Terbutaline enantiomer separation and quantification by complexation and field asymmetric ion mobility spectrometry-tandem mass spectrometry. Anal Chem. 2008;80(11):4133–4140. [PubMed]
77. Dwivedi P, Hill HH., Jr A rapid analytical method for hair analysis using ambient pressure ion mobility mass spectrometry with electrospray ionization (ESI-IMMS) International Journal for Ion Mobility Spectrometry. 2008;11(1–4):61–69.
78. Eiceman GA, Blyth DA, Shoff DB, Snyder AP. Screening of solid commercial pharmaceuticals using ion mobility spectrometry. Anal Chem. 1990;62(14):1374–1379. [PubMed]
79. Borsdorf H, Nazarov EG, Eiceman GA. Atmospheric pressure chemical ionization studies of non-polar isomeric hydrocarbons using ion mobility spectrometry and mass spectrometry with different ionization techniques. J Am Soc Mass Spectrom. 2002;13(9):1078–1087. [PubMed]
80. Zhao YY, Liu X, Boyd JM, Qin F, Li J, Li XF. Identification of N-nitrosamines in treated drinking water using nanoelectrospray ionization high-field asymmetric waveform ion mobility spectrometry with quadrupole time-of-flight mass spectrometry. J Chromatogr Sci. 2009;47(1):92–96. [PubMed]
81. Clowers BH, Steiner WE, Dion HM, et al. Evaluation of sulfonylurea herbicides using high resolution electrospray ionization ion mobility quadrupole mass spectrometry. Field Analytical Chemistry and Technology. 2001;5(6):302–312.
82. Mie A, Sandulescu M, Mathiasson L, Emneus J, Reimann Curt T. Analysis of triazines and associated metabolites with electrospray ionization field-asymmetric ion mobility spectrometry/mass spectrometry. Analytical sciences : the international journal of the Japan Society for Analytical Chemistry. 2008;24(8):973–978. [PubMed]
83. Steiner WE, Klopsch SJ, English WA, Clowers BH, Hill HH. Detection of a chemical warfare agent simulant in various aerosol matrixes by ion mobility time-of-flight mass spectrometry. Anal Chem. 2005;77(15):4792–4799. [PubMed]
84. Hill HH., Jr First Indo-US Workshop on Spectroscopy. Vol. 14. B.H.U Varanasi, India: Laser and Spectroscopy Society of India; 2006. Dwivedi P, Kanu AB. Reduction in false positive responses for explosives detection using ion mobility-mass spectrometry (IMMS) pp. 92–103.
85. Waltman MJ, Dwivedi P, Hill HH, Jr, Blanchard WC, Ewing RG. Characterization of a distributed plasma ionization source (DPIS) for ion mobility spectrometry and mass spectrometry. Talanta. 2008;77(1):249–255. [PubMed]
86. Dwivedi P, Wu P, Klopsch S, Puzon G, Xun L, Hill H. Metabolic profiling by ion mobility mass spectrometry (IMMS) Metabolomics (Published online: 11 December 2007) 2007
87. Asbury GR, Hill HH. Evaluation of ultrahigh resolution ion mobility spectrometry as an analytical separation device in chromatographic terms. Journal of Microcolumn Separations. 2000;12(3):172–178.
88. Wu C, Siems WF, Asbury GR, Hill HH., Jr Electrospray Ionization High-Resolution Ion Mobility Spectrometry-Mass Spectrometry. Analytical chemistry. 1998;70(23):10. [PubMed]
89. Steiner WE, Clowers BH, Fuhrer K, et al. Electrospray ionization with ambient pressure ion mobility separation and mass analysis by orthogonal time-of-flight mass spectrometry. Rapid Communications in Mass Spectrometry. 2001;15(23):2221–2226. [PubMed]
90. Revercomb HE, Mason EA. Theory of Plasma Chromatography/Gaseous Electrophoresis-A Review. Analytical Chemistry. 1975;47(7):970.
91. Sud M, Fahy E, Cotter D, et al. LMSD: LIPID MAPS structure database. Nucleic Acids Res. 2007;35:D527–D532. (Database issue) [PubMed]
92. Fahy E, Subramaniam S, Brown HA, et al. A comprehensive classification system for lipids. J.Lipid Res. 2005;46(5):839–862. [PubMed]
93. Smith CA, O'Maille G, Want EJ, et al. METLIN - A metabolite mass spectral database. Therapeutic Drug Monitoring. 2005;27(6):747–751. [PubMed]
94. Kopka J, Schauer N, Krueger S, et al. GMD@CSB.DB: the Golm Metabolome Database. Bioinformatics. 2005;21(8):1635–1638. [PubMed]
95. Koeniger SL, Merenbloom SI, Valentine SJ, et al. An IMS-IMS analogue of MS-MS. Analytical Chemistry. 2006;78(12):4161–4174. [PubMed]
96. Merenbloom SI, Koeniger SL, Valentine SJ, Plasencia MD, Clemmer DE. IMS-IMS and IMS-IMS-IMS/MS for separating peptide and protein fragment ions. Analytical Chemistry. 2006;78(8):2802–2809. [PubMed]
97. Valentine SJ, Kurulugama RT, Bohrer BC, et al. Developing IMS-IMS-MS for rapid characterization of abundant proteins in human plasma. International Journal of Mass Spectrometry. 2009;283(1–3):149–160.
98. Woods AS, Ugarov M, Egan T, et al. Lipid/peptide/nucleotide separation with MALDI-ion mobility-TOF MS. Analytical Chemistry. 2004;76(8):2187–2195. [PubMed]
99. Clowers BH, Dwivedi P, Steiner WE, Hill HH, Jr, Bendiak B. Separation of sodiated isobaric disaccharides and trisaccharides using electrospray ionization-atmospheric pressure ion mobility-time of flight mass spectrometry. Journal of American Society of Mass Spectrometry. 2005;16(5):660–669. [PubMed]
100. Dwivedi P, Bendiak B, Clowers BH, Hill HH., Jr Rapid Resolution of Carbohydrate Isomers by Electrospray Ionization Ambient Pressure Ion Mobility Spectrometry-Time-of-Flight Mass Spectrometry (ESI-APIMS-TOFMS) Journal of the American Society for Mass Spectrometry. 2007;18(7):1163–1175. [PubMed]
101. Britz-McKibbin P, Terabe S. On-line preconcentration strategies for trace analysis of metabolites by capillary electrophoresis. Journal of Chromatography A. 2003;1000(1–2):917–934. [PubMed]
102. Lidebjer C, Leanderson P, Ernerudh J, Jonasson L. Low plasma levels of oxygenated carotenoids in patients with coronary artery disease. Nutr Metab Cardiovasc Dis. 2007;17(6):448–456. [PubMed]
103. Kita Y, Takahashi T, Uozumi N, Nallan L, Gelb MH, Shimizu T. Pathway-oriented profiling of lipid mediators in macrophages. Biochemical and Biophysical Research Communications. 2005;330(3):898–906. [PMC free article] [PubMed]
104. Serhan CN. Mediator lipidomics. Prostaglandins & Other Lipid Mediators. 2005;77(1–4):4–14. [PubMed]
105. Jackson SN, Wang HYJ, Woods AS. Direct tissue analysis of phospholipids in rat brain using MALDI-TOFMS and MALDI-ion mobility-TOFMS. Journal of the American Society for Mass Spectrometry. 2005;16(2):133–138. [PubMed]
106. Ruotolo BT, Gillig KJ, Stone EG, Russell DH. Peak capacity of ion mobility mass spectrometry: separation of peptides in helium buffer gas. J Chromatogr B Analyt Technol Biomed Life Sci. 2002;782(1–2):385–392. [PubMed]
107. Ruotolo BT, Verbeck GF, Thomson LM, Woods AS, Gillig KJ, Russell DH. Distinguishing between phosphorylated and nonphosphorylated peptides with ion mobility-mass spectrometry. Journal of Proteome Research. 2002;1(4):303–306. [PubMed]
108. Morris M, Watkins SM. Focused metabolomic profiling in the drug development process: advances from lipid profiling. Current Opinion in Chemical Biology. 2005;9(4):407–412. [PubMed]
109. Ruotolo BT, Gillig KJ, Stone EG, Russell DH. Peakcapacity of ion mobility mass spectrometry: - Separation of peptides in helium buffer gas. Journal of chromatography. B, Analytical technologies in the biomedical and life sciences. 2003;782(1):8. [PubMed]
110. Frahm JL, Howard BE, Heber S, Muddiman DC. Accessible proteomics space and its implications for peak capacity for zero-, one- and two-dimensional separations coupled with FT-ICR and TOF mass spectrometry. J Mass Spectrom. 2006;41(3):281–288. [PubMed]
111. Rose DJ, Opiteck GJ. Two-dimensional gel electrophoresis/liquid chromatography for the micropreparative isolation of proteins. Anal Chem. 1994;66(15):2529–2536. [PubMed]
112. McLean JA, Ruotolo BT, Gillig KJ, Russell DH. Ion mobility-mass spectrometry: a new paradigm for proteomics. International Journal of Mass Spectrometry. 2005;240(3):301–315.
113. Liu XY, Valentine SJ, Plasencia MD, Trimpin S, Naylor S, Clemmer DE. Mapping the human plasma proteome by SCX-LC-IMS-MS. Journal of the American Society for Mass Spectrometry. 2007;18(7):1249–1264. [PMC free article] [PubMed]
114. Merenbloom SI, Bohrer BC, Koeniger SL, Clemmer DE. Assessing the peak capacity of IMS-IMS separations of tryptic peptide ions in He at 300 K. Analytical Chemistry. 2007;79(2):515–522. [PMC free article] [PubMed]
115. Want EJ, O'Maille G, Smith CA, et al. Solvent-Dependent Metabolite Distribution, Clustering, and Protein Extraction for Serum Profiling with Mass Spectrometry. Analytical Chemistry. 2006;78(3):743–752. [PubMed]
116. Roy SM, Anderle M, Lin H, Becker CH. Differential expression profiling of serum proteins and metabolites for biomarker discovery. International Journal of Mass Spectrometry. 2004;238(2):163–171.
117. Nordstroem A, O'Maille G, Qin C, Siuzdak G. Nonlinear Data Alignment for UPLC-MS, HPLC-MS Based Metabolomics: Quantitative Analysis of Endogenous and Exogenous Metabolites in Human Serum. Analytical Chemistry. ACS ASAP. [PubMed]