Novel Application of Reversed-Phase UPLC-oaTOF-MS for Lipid
Analysis in Complex Biological Mixtures: A New Tool for Lipidomics
Paul D. Rainville,*,† Chris L. Stumpf,† John P. Shockcor,† Robert S. Plumb,† and
Jeremy K. Nicholson‡
Waters Corporation, 34 Maple Street, Milford, Massachusetts 01757, and Biological Chemistry, Division of
Biomedical Sciences, Sir Alexander Fleming Building, Imperial College London SW7 2AZ, United Kingdom
Received November 18, 2006
Ultra-Performance LC (UPLC) utilizing sub-2-µm porous stationary phase particles operating with high
linear velocities at pressures >9000 psi was coupled with orthogonal acceleration time-of-flight (oaTOF)
mass spectrometry and successfully employed for the rapid separation of lipids from complex matrices.
The UPLC system produced information-rich chromatograms with typical measured peak widths of
3 s at peak base, generating peak capacities in excess of 200 in 10 min. Further UPLC coupled with
MSE technology provided parent and fragment mass information of lipids in one chromatographic run,
thus, providing an attractive alternative to current LC methods for targeted lipid analysis as well as
lipidomic studies.
Keywords: Lipidomics • Metabolic Profiling • Ultra Performance LC • Mass Spectrometry
Introduction
Lipids are an important class of biomolecule that exist in
great variety in higher organisms. The great diversity exhibited
by these molecules is most probably due to the many biochemical functions in which these molecules are involved.
Lipids are used as energy stores to fuel metabolism, as
structural components in cell membranes, and are involved
in many metabolic processes such as signal transduction,
morphogenesis, secretion, and vesicle trafficking.1,2 Metabolic
diseases such as Diabetes mellitus have been shown to have a
direct relationship with a disorder in the lipids and fatty acids
that make up phospholipids,3 a class of lipids that are defined
by various specific polar head groups. This class of lipid has
shown commercial use in biomembranes, skin care formulations, and in the making of liposomes, which are used in drug
delivery as well as in cosmetics and detergents.3 The analysis
and profiling of lipids has therefore become increasingly
important in the fields of food analysis, commercial applications, and metabolic profiling. The ability to profile the lipids
in biological fluids is also important, as it allows for the analysis
of the effects of many candidate pharmaceuticals of metabolic
pathways such as cholesterol synthesis. Lipidomics is a field
of study that has rapidly expanded the study of lipids based
on the advances made by electrospray ionization mass spectrometry4 that focuses on the study of lipid classes as they exist
in their natural environment.
The analysis of lipids has been performed by a diverse variety
of approaches reflecting the diverse chemical subclasses. Gas
chromatography (GC) and GC mass spectrometry (GC/MS)
* Corresponding author. E-mail: [email protected].
†
Waters Corporation.
‡
Imperial College London.
552
Journal of Proteome Research 2007, 6, 552-558
Published on Web 12/23/2006
approaches provide a rapid, sensitive, and relatively cheap
method of analysis, but require prior time-consuming chemical
derivatization which can leads to inter laboratory variations.5
1
H NMR provides a relatively fast method of profile analysis
that can generate information on a range of lipids including
lipoproteins, but the disadvantage of this approach is the
difficulty in separating multiple overlapped lipidic species with
the exception of lipoproteins. Lipids have been successfully
analyzed by normal-phase high-performance liquid chromatography (HPLC)6-8 and normal-phase HPLC coupled with
mass spectrometry;3,9-11 this method of analysis has high
sensitivity and does not require any sample derivatization.
However, the analysis times are typically long, 30-60 min, and
peak resolution is typically poor, with peak width in the order
of 20-30 s. Direct infusion of lipids into a mass spectrometer
has also been shown as a method for characterization of cellular
lipids,4 but this type of analysis of simultaneously introducing
multiple analytes into a mass spectrometer can affect the
detection of low-abundance species due to ion suppression.
Reversed-phase HPLC has been the technique of choice for
both pharmaceutical and bioanalytical LC/MS/MS analysis.12
This is due to the high efficiency of the separations, the
compatibility of the mobile phase with biological and lipophilic
samples, and the easy interfacing with a variety of detectors
including UV, fluorescence, evaporative light scattering detection, radio-chemical detection, mass spectrometry, and NMR.
In fact, a majority of liquid chromatographic separations used
today are based upon reversed-phase LC. However, as previously stated, lipid analysis is often carried out using normalphase chromatography and becomes difficult if one is switching
between the two different modes of chromatography on a single
system. The process of changing from reversed-phase to
normal-phase chromatography entails replacing not only the
10.1021/pr060611b CCC: $37.00
 2007 American Chemical Society
A New Tool for Lipidomics: Reversed-Phase UPLC-oaTOF-MS
research articles
Figure 1. Reversed-phase UPLC separation of phospholipid standards shown with corresponding ESI negative MS specta. Shown are
1-oleoyl-2-hydroxy-sn-glycero-3-phosphocholine, 3.9 min; 1,2-dioleoyl-sn-glycero-3-[phospho-L-serine], 8.20 min; and 1,2-dioleoyl-snglycero-3-phosphoethanolamine, 8.63 min.
Table 1. Mass Accuracy of Phospholipids Standards Detected
with Reversed-Phase UPLC/MS Method
standard
1-oleoyl-2-hydroxysn-glycero-3phosphocholine
1,2-dioleoyl-snglycero-3[phospho-L-serine]
1,2-dioleoyl-snglycero-3phosphoethanolamine
retention theoretical
measured
ppm
time
mass (M - H) mass (M - H) error
3.90
566.3458
566.3463
-0.9
8.20
786.5285
786.5299
-1.8
8.63
742.5387
742.5399
-1.6
mobile phase, but also the wash and purge solvents. This task
is time-consuming, as it is essential to remove all traces of water
prior to the commencement of analysis, and can result in
several days of down time while the instrumentation become
“dry”. Methods have been developed utilizing both aqueous13,14
and nonaqueous15,16 reversed-phase chromatography for the
analysis of lipids, but to the authors’ knowledge, the analysis
of lipids has not been shown using the high-resolution capabilities generated by ultra-performance liquid chromatography.
Recent advances in porous particle manufacturing and
chromatographic hardware has led to the generation of highresolution “Ultra Performance” liquid chromatography (UPLC).
UPLC builds on the principles of HPLC but employs alkylbonded porous particles with a particle size less than 2 µm in
diameter operated with increased mobile phase linear velocities. These small particles can generate significantly higher
resolution separations than 5 or 3 µm stationary phases. This
increased resolution results in superior separations, increased
sensitivity, and faster analysis.12 Because of the increased
Journal of Proteome Research • Vol. 6, No. 2, 2007 553
research articles
Rainville et al.
Figure 2. Total ion current (TIC) chromatogram of UPLC/MS separation of analytes extracted from the supernatant of protein precipitated
rat plasma; (A) electospray negative and (B) electrospray positive.
sensitivity and resolution that UPLC offers, it only seems logical
to apply it to the field of lipid analysis.17 However, the smaller
particles used in UPLC generate greater back pressure, making
it necessary to have a chromatography system that can operate
at pressures in excess of 10 000 psi. The need to handle elevated
pressures in excess of 10 000 psi can be mitigated to some
extent by the use of increased column temperature. This is
possible because of the reduction in viscosity of the mobile
phase solvents with increases in temperature. However, increases in column temperature can and often do result in a
change in selectivity and peak elution order and as such is
nonpredictable. Raising the column temperature also increases
the optimal linear velocity required to run the column effectively, in turn, requiring the need to handle elevated
pressures and negating the benefits of raising the column
temperature. Here, we demonstrate the use of UPLC operated
in reversed-phase mode at moderate temperatures for the
analysis of lipids from rat plasma samples. We further show
the utility of MSE, a technique whereby both precursor and
fragment mass spectra are simultaneously acquired by alternating between high and low collision energy during a single
chromatographic run.18 The advances presented here may be
significant for lipidomic research, which has so far been
technically limited to relatively low-throughput applications.
554
Journal of Proteome Research • Vol. 6, No. 2, 2007
Experimental Procedures
Sample Preparation. Acetonitrile, chloroform, and methanol
were all obtained from Fisher Scientific (NH). Ammonium
acetate, acetic acid (spectroscopic grade), and leucine-enkephalin were purchased from Sigma/Aldrich (MO). Distilled
water was purified “in-house” using a MilliQ system Millipore
(MA). Phospholipid standards were obtained from Avanti Polar
Lipids (AL) as dry powder and were reconstituted in chloroform
and sonicated until dissolved. Rat plasma was obtained from
Equitech-Bio, Inc. (TX). Lipids were extracted from rat plasma
by precipitating rat plasma with 3:1 (acetonitrile/rat plasma)
and centrifuging at 14 700g for 5 min. The supernatant was then
removed and injected into the UPLC/MS system.
Chromatography. Reversed-phase analysis was performed
on a Waters ACQUITY Ultra Performance LC system using a
ACQUITY UPLC BEH C8 1.7 µm, 2.1 × 100 mm, analytical
column. The column was maintained at 60 °C and eluted using
a linear gradient of 35-95% B over 9 min where A consisted of
50 mM ammonium acetate, pH 5.0, and B consisted of
acetonitrile at a flow rate of 0.6 mL/min.
Mass Spectrometry. Mass spectrometry was performed on
a Waters Q-Tof Premier mass spectrometer operating in
negative and positive ion electrospray mode. The capillary and
cone voltages were set at 3.2 kV and 35 Vfor positive electrospray mode and 2.5 kV and 45 V for negative electrospray mode.
A New Tool for Lipidomics: Reversed-Phase UPLC-oaTOF-MS
research articles
Figure 3. (A) Fragmentation pattern of 1-oleoyl-2-hydroxy-sn-glycero-3-phosphocholine. (B, top panel) High collision energy mass
spectra of 1-oleoyl-2-hydroxy-sn-glycero-3-phosphocholine during MSE experiment. (B, bottom panel) Low collision energy mass spectra
of 1-oleoyl-2-hydroxy-sn-glycero-3-phosphocholine during MSE experiment.
The desolvation temperature was set to 350 °C and the source
temperature to 120 °C; the cone gas was set to a flow rate of
50 L/h, and the desolvation gas flow was maintained at 800
L/h. MSE analysis was performed on a Waters Q-Tof Premier
mass spectrometer set at 5 eV for low collision energy and 25
eV for high collision energy.
Data Analysis. Principal component analysis (PCA) was
carried out using MarkerLynx software from Waters Corporation (MA). Orthogonal partial least-squares (OPLS) analysis was
carried out using SIMCA-P software from Umetrics (Sweden).
Results and Discussion
To illustrate the performance of the UPLC separation, a
standard containing three phospholipids was injected into the
UPLC system. The peak shape for all three phospholipids
standards is excellent with little if any tailing, Figure 1. The
average peak width for this separation is 3 s at peak base, giving
a peak capacity of 200 for a 10 min separation. The mass
spectrum and corresponding chemical structure of the phospholipid peaks is also displayed in Figure 1. The accurate mass
measurement of these peaks is shown in Table 1. This data
shows that it is possible to obtain a sufficient number of points
across these narrow chromatographic peaks to obtain good ion
statistics.
To further illustrate the capability of the reversed-phase
UPLC/MS system for the analysis of lipids, lipids were extracted
from rat plasma and analyzed with the same method parameters as the previous sample. Both positive and negative ion
electrospray were utilized because not all lipid species may
ionize effectively in one mode. The chromatograms produced
are shown in Figure 2; again we can see that the UPLC system
produced a high efficiency separation. We then coupled the
UPLC with a Q-tof Premier mass spectrometer operating in
electrospray positive MSE mode. Illustrated in Figure 3 are the
chemical structure and the electrospray positive MSE spectra
of 1-oleoyl-2-hydroxy-sn-glycero-3-phosphocholine. Observed
are fragmentation ions 184 and 104 m/z generated from the
choline head group under high-energy conditions. An extracted
ion chromatogram of the high collision energy scans shows that
a vast majority of analytes present in supernatant from protein
precipitated rat plasma contain ions 184 and/or 104 m/z, Figure
4. We then repeated this experiment using three different lots
of rat plasma. PCA carried out using the top 10 most intense
ions detected during the analysis showed separation of the
different lots of rat plasma tested, indicating that a statistical
difference exists between different lots of rat plasma based on
the top 10 most intense ions, Figure 5. OPLS S plot further
generated a list of ions (blue data points) responsible for the
inter-lot difference, Figure 6. Ions calculated from the OPLS S
plot as contributing the greatest to the inter-lot difference are
shown in Table 2. UPLC/MSE analysis carried out on these ions
listed in Table 2 generated high-energy spectra containing
Journal of Proteome Research • Vol. 6, No. 2, 2007 555
research articles
Rainville et al.
Figure 4. Electrospray positive extracted ion chromatogram (XIC) and TIC of UPLC/MS separation of choline-containing lipids extracted
from rat plasma. (A) XIC of m/z 184 and 104 and (B) TIC.
Figure 5. PCA scores plot showing variability between three different lots of rat plasma. Six repeat injections from each lot are also
shown.
major fragment ions of 184 and 104 m/z. An example of this is
shown for precursor mass 496.3397, Figure 7. A database search
of the monoisotopic precursor mass was done using the LIPID
556
Journal of Proteome Research • Vol. 6, No. 2, 2007
MAPS Web site and produced two possible matches within an
error of 1.3 ppm: 2-lyso glycerophospho-ethanolamine (2-lyso
GPEtn) and 2-lyso glycerophosphocholine (2-lyso GPCho).
research articles
A New Tool for Lipidomics: Reversed-Phase UPLC-oaTOF-MS
Figure 6. OPLS S-plot, data point in blue indicate ions responsible for variation between lots 1 and 2 of protein precipitated rat plasma.
Table 2. Top 10 Parent Mass Ions Contributing to Differences
between Rat Plasma Lots 1 and 2
m/z
R.T.
weight
confidence
496.33
524.36
520.33
522.35
521.34
544.33
496.61
802.60
774.57
722.60
3.88
4.78
3.65
4.2
3.657
3.72
3.88
7.96
7.36
10.1
0.350881
0.292548
0.227726
0.141613
0.124123
0.114852
0.0983533
0.0842536
0.079154
0.0763954
0.992827
0.965764
0.99346
0.980016
0.987388
0.960234
0.996497
0.97337
0.997217
0.982884
These two database matches produce different diagnostic
fragments associated with their headgroup. 2-Lyso GPCho is
expected to produce ions of 184 and 104 m/z, while 2-lyso
GPEtn is expected to produce ions of 141 m/z.19,20 The high
collision spectrum in Figure 7 shows the diagnostic fragments
identifying 2-lyso GPCho as the correct structural match from
the database. This information gathered by the UPLC/MSE
method coupled with the fact that lysophospholipids and
phospholipids are known to exist in the supernatant proteinprecipitated plasma19 indicate that a majority of the analytes
present in the supernatant of protein precipitated rat plasma
are lipids that contain the choline group.
Conclusion
Ultra-Performance Liquid Chromatography (UPLC) utilizing
sub-2-µm porous stationary phase particles operating with high
linear velocities at pressures in excess of 9000 psi was coupled
with oaTOF mass spectrometry and successfully employed in the
analysis of lipids from mammalian plasma. Chromatographic
peak widths of 3 s at peak base were observed, thus, generating
separations with peak capacities in excess of 200 in just 10 min.
Accurate mass measurement of less than 2 ppm was observed
Figure 7. MSE spectra of 496.3397, the most statistically significant ion found by OPLS between protein precipitated rat plasma
lots 1 and 2.
in the mass spectra showing the capability of obtaining good
ion statistics for these narrow chromatographic peaks. UPLC
coupled with Q-Tof MS operating in MSE mode gave both fragment and parent information of lipids in a single chromatographic run. MSE experimental results further were used to
distinguish between two database matches for a single monoisotopic mass without further experimentation. Lastly, this method
coupled with PCA and OPLS indicated that choline-containing
lipids account for the variability observed between the supernatant of different lots of protein-precipitated rat plasma.
References
(1) Hermanson, M.; Uphoff, A.; Kakela, R.; Somerharju, P. Anal.
Chem. 2005, 77, 2166-2175.
Journal of Proteome Research • Vol. 6, No. 2, 2007 557
research articles
(2) Welti, R.; Wang, X. Curr. Opin. Plant Biol. 2004, 7, 1-8.
(3) Wang, C.; Kong, H.; Guan, Y.; Yang, J.; Gu, J.; Yang, S.; Xu, G.
Anal. Chem. 2005, 77, 4108-4116.
(4) Han, H.; Gross, R. W. J. Lipid Res. 2003, 44, 1071-1079.
(5) Sumner, L. W.; Mendes, P.; Dixon, R. A. Phytochemistry 2003, 62,
817-836.
(6) McNabb, T. J.; Cremesti, A. E.; Brown, P. R.; Fischl, A. S. Anal.
Biochem. 1999, 276, 242-250.
(7) Myher, J. J.; Kuksis, A. J. Chromatogr., B 1995, 671, 3.
(8) Christie, W. W. Adv. Lipid Methodol. 1996, 77.
(9) Uran, S.; Asmund, L.; Jacobeson, P. B.; Skotland, P. J. Chromatogr.,
B 2001, 758, 265-275.
(10) Wang, C.; Xie, S.; Yang, J.; Yang, Q.; Xu, G. Anal. Chim. Acta 2004,
525, 1-10.
(11) Hvattum, E.; Asmund, L.; Uran, S.; Michelsen, P.; Skotland, P. J.
Chromatogr., B 1998, 716, 47-56.
(12) Wilson, I. D.; Nicholson, J. K.; Castro-Perez, J.; Granger, J. H.;
Johnson, K. A.; Smith, B. W.; Plumb, R. S. J. Proteome Res. 2005,
4, 591-598.
558
Journal of Proteome Research • Vol. 6, No. 2, 2007
Rainville et al.
(13) Barrosa, B.; Bischoff, R. J. Chromatogr., B 2004, 814, 21-28.
(14) Reis, A.; Dominques, M. R. M.; Amado, F. M. L.; Ferrer-Correia,
A. J. V.; Domingues, P. Biomed. Chromatogr. 2005, 19, 129-137.
(15) Fauconnot, L.; Hau, J.; Aeschlimann, J.; Fay, L.; Dionisi, F. Rapid
Commun. Mass Spectrom. 2004, 18, 218-224.
(16) Byrdwell, W. C.; Neff, W. Rapid Commun. Mass Spectrom. 2002,
16, 300-319.
(17) Peterson, B. L.; Cummings, B. S. Biomed. Chromatogr. 2006, 20,
227-243.
(18) Bateman, R. H.; Carruthers, R.; Hoyes, J. B.; Jones, C.; Langridge,
J. I.; Millar, A.; Vissers, J. P. J. Am. Soc. Mass Spectrom. 2002, 13,
792-803.
(19) Little, J. L.; Wempe, M. F.; Buchanan, C. M. J. Chromatogr., B
2006, 833, 219-230.
(20) Pulfer, M.; Murphy, R. C. Mass Spectrom. Rev. 2003, 22, 332364.
PR060611B