Stable isotope labeled 4-(dimethylamino)benzoic acid derivatives of glycerophosphoethanolamine lipids

Article abstract of DOI:10.1021/ac900583a

A set of four (D₀, D₄, D₆, and D ₁₀) deuterium enriched 4-(dimethylamino)benzoic acid (DMABA) N-hydroxysuccinimide (NHS) ester reagents was developed that react with the primary amine group of glycerophosphoetha

Full text of DOI:10.1021/ac900583a

Anal. Chem. 2009, 81, 6633–6640
Stable Isotope Labeled 4-(Dimethylamino)benzoic
Acid Derivatives of Glycerophosphoethanolamine
Lipids
Karin A. Zemski Berry,† William W. Turner,‡ Michael S. VanNieuwenhze,‡ and Robert C. Murphy*,†
Department of Pharmacology, University of Colorado Denver, Aurora, Colorado 80045, and Department of Chemistry,
Indiana University, Bloomington, Indiana 47401
A set of four (D0, D4, D6, and D10) deuterium enriched
4-(dimethylamino)benzoic acid (DMABA) N-hydrox-
ysuccinimide (NHS) ester reagents was developed that
react with the primary amine group of glycerophos-
phoethanolamine (PE) lipids to create derivatives
where all subclasses of DMABA labeled PE are de-
tected by a common precursor ion scan. The positive
ion collision induced dissociation data from (D0, D4,
D6, and D10)-DMABA labeled PE standards indicated
that a precursor ion scan of m/z 191.1, 195.1, 197.1,
and 201.1 could be used to selectively detect (D0, D4,
D6, and D10)-DMABA modified PE, respectively, in a
complex biological mixture. The PE lipids from a time
course (0, 30, 60, and 300 min) of 2,2′-azobis-(2-
amidinopropane) hydrochloride (AAPH) treatment of
liposomes made of RAW 264.7 cell phospholipids
were each labeled with the D0-, D4-, D10-, and D6-
DMABA NHS ester reagents, respectively. The DMABA
derivatives revealed loss of endogenous PE lipids and
an increase in oxidized PE lipid throughout the time
course of AAPH treatment. These DMABA NHS ester
reagents provide a universal scan for diacyl, ether, and
plasmalogen PE lipids that cannot be readily observed
otherwise, enable differential labeling, and provide an
internal standard for each PE lipid.
eases.4-7 These findings have demonstrated the importance of
obtaining a detailed analysis of phospholipid species present in
certain tissues and cells.
One very powerful technique used to analyze phospholipids
employs electrospray tandem mass spectrometry. Tandem mass
spectrometry of the [M + H]+ of phospholipids yields unique
product ions and neutral losses that are indicative of the polar
headgroup.8 For diacyl and ether glycerophosphoethanolamine
(PE) lipids, collision induced dissociation (CID) of the [M +
H]+ results in the neutral loss of phosphoethanolamine (141
amu).9 However, the collision induced dissociation behavior
of the [M + H]+ of plasmalogen PE lipids, which contain a
vinyl ether substituent at the sn-1 position of the glycerol
backbone, results in two prominent fragment ions; one that
was characteristic of the sn-1 position and one that was
characteristic of the sn-2 position with a very minor ion that
results from the neutral loss of 141 amu (NL141).10 Since all
subclasses of PE lipids do not fragment similarly in positive
ion mode, there is not a universal scan for PE lipids that allows
for selective detection of all PE subclasses in a biological
sample.
Few studies have used tandem mass spectrometry to detect
oxidized PE in samples. In most of these studies the oxidation
products were first predicted and then detected either in the
negative ion mode using a precursor ion scan of a predicted
oxidized fatty acid11 or in the positive ion mode using multiple
reaction monitoring (MRM) of the NL141 of predicted oxidized
PE species.12 The detection of oxidized PE in both of these
previous studies depended on correctly predicting the oxidized
PE products that would most likely be observed; however, it is
possible to inadvertently exclude many different oxidized species
in a complex biological sample. Additionally, the MRM method
Phospholipids are the main component of cellular membranes
that have important structural and functional properties. In
addition to forming the physical boundary of cells, phospholipids
are involved in key regulatory cellular functions. Phospholipids
are precursors for biologically active molecules, such as plate-
let activating factor (PAF)1 and arachidonate derived lipid media-
tors,2 and signaling lipids, such as diacylglycerol.3 Additionally,
free radical oxidation of polyunsaturated fatty acids (PUFA)
containing phospholipids has been implicated in numerous
diseases including atherosclerosis and neurodegenerative dis-
(4) Montine, T. J.; Morrow, J. D. Am. J. Pathol. 2005, 166, 1283–1289
(5) Morrow, J. D. Arterioscler., Thromb., Vasc. Biol. 2005, 25, 279–286
(6) Berliner, J. A.; Watson, A. D. New Engl. J. Med. 2005, 353, 9–11
(7) Fruhwirth, G. O.; Loidl, A.; Hermetter, A. Biochim. Biophys. Acta 2007,
1772, 718–736
(8) Pulfer, M.; Murphy, R. C. Mass Spectrom. Rev. 2003, 22, 332–364
.
.
.
.
.
* To whom correspondence should be addressed. Robert C. Murphy, Ph.D.,
Department of Pharmacology, University of Colorado Denver, Mail Stop 8303,
12801 E. 17th Ave., Aurora, CO 80045-0511. Phone: 303-724-3352. Fax: 303-724-
3357. E-mail: Robert.Murphy@ucdenver.edu.
(9) Murphy, R. C. Mass Spectrometry of Phospholipids: Tables of Molecular and
Product Ions; Illuminati Press: Denver, CO, 2002.
(10) Zemski Berry, K. A.; Murphy, R. C. J. Am. Soc. Mass Spectrom. 2004, 15,
1499–1508
.
† University of Colorado Denver.
(11) Maskrey, B. H.; Bermudez-Fajardo, A.; Morgan, A. H.; Stewart-Jones, E.;
Dioszeghy, V.; Taylor, G. W.; Baker, P. R. S.; Coles, B.; Coffey, M. J.; Kuhn,
‡ Indiana University.
(1) Honda, Z.; Ishii, S.; Shimizu, T. J. Biochem 2002, 131, 773–779
(2) Funk, C. D. Science 2001, 294, 1871–1875
(3) Taylor, C. W. Cell 2002, 111, 767–769
.
H.; O’Donnell, V. B. J. Biol. Chem. 2007, 282, 20151–20163
(12) Gugiu, B. G.; Mesaros, C. A.; Sun, M.; Gu, X.; Crabb, J. W.; Salomon, R. G.
Chem. Res. Toxicol. 2006, 19, 262–271
.
.
.
.
10.1021/ac900583a CCC: $40.75  2009 American Chemical Society
Published on Web 07/17/2009
Analytical Chemistry, Vol. 81, No. 16, August 15, 2009 6633

Scheme 1
nium bicarbonate buffer, and methoxyamine hydrochloride (MOX)
were purchased from Sigma Chemical Company (St. Louis, MO).
Sodium hydroxide (1 N) was purchased from J. T. Baker
(Phillipsburg, NJ). HPLC solvents were purchased from Fisher
Scientific (Fair Lawn, NJ) and used for HPLC and extraction.
Synthesis of DMABA-NHS Ester Reagents. The details of
the synthesis, purification, and characterization of D4-4-(dimethy-
lamino)benzoic acid, D6-4-(dimethylamino)benzoic acid, and
D10-4-(dimethylamino)benzoic acid are provided in the Sup-
porting Information. Additionally the synthesis, purification, and
characterization of the (D0, D4, D6, and D10)-DMABA NHS ester
reagents using 1-ethyl-3-[3-dimethylaminopropyl]carbodiimide
hydrochloride (EDC) coupling chemistry are also provided in
the Supporting Information
.
using the NL141 of the predicted oxidized PE species would
completely discriminate against any of the oxidized products that
are plasmalogen PE lipids. Typically, plasmalogen PE contain high
amounts of esterified PUFA.13,14 Since PUFA are highly suscep-
tible to oxidation, many oxidized products would be absent using
the above MRM method. Therefore, a selective precursor ion scan
that universally detects all subclasses of PE and potential oxidized
products would be of significant value.
Synthesis of DMABA Amide Tagged PE Lipids. A solution
of 18:1a/18:1-PE (20 µg) in chloroform was dried down under a
stream of nitrogen and resuspended in ethanol (65 µL) and 0.25
M triethylammonium bicarbonate buffer (15 µL). The (D0, D4,
D6, or D10)-DMABA NHS ester reagent in methylene chloride
(20 µL of a 10 mg/mL solution) was then added to 18:1a/18:
1-PE and allowed to incubate at 60 °C for 1 h. Water (0.4 mL)
was added to hydrolyze any unreacted DMABA NHS ester
reagent for 30 min. The phospholipids were then extracted by
addition of chloroform-methanol according to the method of
Bligh and Dyer.15 The extracted DMABA labeled PE was
resuspended in 60:20:20 (v/v/v) methanol/acetonitrile/water
with 1 mM ammonium acetate at a concentration of 1 ng/µL
for mass spectrometric analysis.
Electrospray Ionization Tandem Quadrupole Mass Spec-
trometry. The (D0, D4, D6, or D10)-DMABA labeled 18:1a/18:
1-PE lipids were infused into a Sciex API 2000 QTRAP mass
spectrometer (PE Sciex, Toronto, Canada) at a flow rate of 5
µL/min. Unless specifically indicated, enhanced product ion
spectra utilizing the terminal linear ion trap of this triple
quadruple are reported as product ion scans. In the positive
ion mode, an electrospray voltage of 5000 V, a declustering
potential of 50 V, and a collision energy of 30 V was used to
obtain the product ion spectra. In the negative ion mode, the
experimental parameters used to obtain the product ion spectra
were an electrospray voltage of -4500 V, a declustering
potential of -50 V, and a collision energy of -50 V. For some
experiments, the terminal quadrupole was not operated as an
ion trap and product ions were obtained under normal triple
quadrupole operation. A NL141 scan was used to detect
unlabeled PE lipids from RAW 264.7 cells in the positive ion
mode at a collision energy of 25 V.
In the current study, a set of four (D0, D4, D6, and D10)
deuterium enriched 4-(dimethylamino)benzoic acid (DMABA)
N-hydroxysuccinimide (NHS) ester reagents (Scheme 1) were
developed that react with the primary amine group of PE lipids.
These reagents were developed in order to create DMABA PE
derivatives where all subclasses of DMABA labeled PE could be
detected by a common precursor ion scan. In addition, this set of
reagents permitted differential labeling of PE and thereby generate
stable isotope labeled variants of all PE molecular species by
comparison of control lipids with treated lipids. The tandem mass
spectrometric behavior in both the positive and negative ion mode
of DMABA labeled PE lipids was determined using (D0, D4, D6,
and D10)-DMABA labeled 1,2-dioleoyl-sn-glycero-3-phosphoet-
hanolamine (18:1a/18:1-PE) standards. (Abbreviations for
individual PE molecular species used in this paper: n:jk/s:t-
PE, where n is the number of carbon atoms in the sn-1
substituent and j is the number of double bonds in the sn-1
hydrocarbon chain; k represents the type of sn-1 linkage, where
p refers to plasmalogen (1-O-alk-1′-enyl), e refers to ether (1-
O-alkyl), and a refers to acyl; s is the number of carbons and
t is the number of double bonds in the sn-2 substituent.)
Additionally, the DMABA NHS ester reagents were used to
assess the changes that occurred in the distribution of PE lipids
after exposure of liposomes made from phospholipids extracted
from RAW 264.7 cells to 2,2′-azobis-(2-amidinopropane) hydro-
chloride (AAPH).
RAW 264.7 Cells. The cells used in these experiments were
the RAW 264.7 macrophage cell line that was obtained from ATCC
laboratories (Manassas, VA). In brief, the cells were grown in an
incubator with a 5% CO2 humidified atmosphere maintained at
37 °C. The cells were grown until 80% confluent at which time
they were harvested and used in the experiments described
below.
Preparation of Liposomes from Lipids Extracted from
RAW 264.7 Cells and Oxidation Procedure. Lipids were
extracted from RAW 264.7 cells (40 × 106) by a Bligh-Dyer
extraction. The phospholipid extract was split equally into four
EXPERIMENTAL SECTION
Materials. 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (18:
1a/18:1-PE), 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphoet-
hanolamine (16:0a/20:4-PE) and 1,2-dimyristoyl-sn-glycero-3-
phosphoethanolamine (14:0a/14:0-PE) were purchased from Avanti
Polar Lipids, Inc. (Alabaster, AL). Hanks’ balanced salt solution
(1×) (HBSS) was obtained from Invitrogen (Carlsbad, CA). 2,2′-
Azobis-(2-amidinopropane) hydrochloride (AAPH), triethylammo-
(13) Chilton, F. H.; Murphy, R. C. J. Biol. Chem. 1986, 261, 7771–7777
(14) Gross, R. W. Biochemistry 1984, 23, 158–165
.
.
(15) Bligh, E. G.; Dyer, W. J. Can. J. Biochem. Physiol. 1959, 37, 911–917.
6634 Analytical Chemistry, Vol. 81, No. 16, August 15, 2009

different test tubes, taken to dryness under a stream of
nitrogen, and resuspended in HBSS (0.4 mL). Small, unilamel-
lar vesicles were prepared by vortexing the phospholipid
suspensions for 30 min followed by sonication for 10 min. The
liposomes were then subjected to AAPH treatment (10 mM
final concentration), and a time course was performed. The
liposomes were incubated with AAPH for 0, 30, 60, and 300
min at 37 °C and the phospholipids were extracted using a
Bligh-Dyer extraction after the internal standard (14:0a/14:
0-PE, 1 µg) was added. Methoxime derivatives of ketone or
aldehyde groups present on the phospholipids were prepared
using a gas phase MOX procedure.16
then subjected to AAPH treatment (10 mM final concentration)
for 1 h at 37 °C. The reaction was terminated by immersion in an
ice bath and the addition of chloroform-methanol according to
the method of Bligh and Dyer.15 At this point, half of the oxidized
16:0a/20:4-PE sample was saved and half was labeled with the
D4-DMABA NHS ester reagent according to the above protocol.
The oxidized 16:0a/20:4-PE and the D4-DMABA labeled 16:
0a/20:4-PE were resuspended in 60:20:20 (v/v/v) methanol/
acetonitrile/water with 1 mM ammonium acetate at a concen-
tration of 1 ng/µL for mass spectrometric analysis.
RESULTS
Mass spectrometric analysis was used to elucidate the products
The PE lipids in the control sample (0 min) were labeled with
the D0-DMABA NHS ester reagent, while the PE lipids in the
AAPH treated samples (30, 60, and 300 min) were labeled with
the D4-, D10-, and D6-DMABA NHS ester reagents, respectively,
according to the above protocol that was followed for the 18:
1a/18:1-PE standard. After the reaction of the DMABA NHS
ester reagent with PE lipids was complete, the control and
oxidized samples were all combined together and extracted
by the method of Bligh and Dyer.15 The combined DMABA
labeled control and oxidized samples were then introduced onto
an aminopropyl SepPak column (Supelco, Bellefonte, PA) that
was conditioned with hexane in order to separate phospholipids
by classes.17,18 The neutral lipids were eluted using chloroform/
2-propanol (2:1) (4 mL), and methanol (4 mL) was then added to
the column to elute the glycerophosphocholine (PC). The DMABA
labeled PE lipids were eluted using methanol/chloroform/3.6 M
aqueous ammonium acetate (60:30:8) and collected. Chloroform
(3 mL) and water (1 mL) were added to these fractions so that a
Bligh-Dyer extraction could be performed to remove the large
amount of ammonium acetate present in this solution.
DMABA Labeled PE Separation by RP-HPLC and Analy-
sis by Electrospray Ionization Mass Spectrometry. The D0-,
D4-, D6- and D10-DMABA labeled PE species were separated
according to lipophilicity by reverse phase HPLC with a Gemini
5 µm C18 (2.0 mm × 150 mm) column (Phenomenex, Torrance,
CA) coupled to a Sciex API 2000 QTRAP mass spectrometer
(PE Sciex, Toronto, Canada). The HPLC was operated at a flow
rate of 0.2 mL/min with a mobile phase of methanol/
acetonitrile/water 60:20:20 (v/v/v) with 2 mM ammonium
acetate (solvent A) and 2 mM methanolic ammonium acetate
(solvent B). The gradient was 25% solvent B to 100% solvent B
in 20 min, followed by isocratic elution at 100% solvent B for
20 min. The D0-, D4-, D6-, and D10-DMABA labeled PE lipids
from 1 × 106 cells were detected during one chromatographic
run by alternating between precursors of m/z 191.1 (P191),
m/z 195.1 (P195), m/z 197.1 (P197), and m/z 201.1 (P201)
scans, respectively, every 1.5 s with a collision energy of 35 V.
Oxidation of 16:0a/20:4-PE Standard. Standard 16:0a/20:
4-PE (10 µg) in chloroform was taken to dryness under a stream
of nitrogen and resuspended in HBSS (0.4 mL). Small, unilamellar
vesicles were prepared by vortexing the phospholipid suspensions
for 30 min followed by sonication for 10 min. The liposomes were
formed after the reaction of 18:1a/18:1-PE standard with the (D0
,
D4, D6, and D10)-DMABA NHS ester reagents. Prior to labeling
18:1a/18:1-PE with the (D0, D4, D6, and D10)-DMABA NHS
ester reagents, the [M + H]+ ion observed in the positive ion
mass spectrum was at m/z 744.7. Upon reaction of the
D0-DMABA NHS ester reagent with 18:1a/18:1-PE, the [M +
H]+ ion shifted to m/z 891.7 (data not shown), which cor-
responded to the addition of 147 amu onto the [M + H]+ of
18:1a/18:1-PE. The [M + H]+ ions at m/z 895.7, 897.7, and
901.7 were observed upon reaction of 18:1a/18:1-PE with each
of the (D4, D6, and D10)-DMABA NHS ester reagents, respec-
tively, which corresponded to an addition of 151, 153, and 157
amu onto the [M + H]+ of 18:1a/18:1-PE. The yield of the
DMABA NHS ester reagents when labeling the 18:1a/18:1-PE
standard was determined using the [M + H]+ peak height of
18:1a/18:1-PE and the DMABA labeled 18:1a/18:1-PE found
in the positive ion mass spectrum. It was found for each of the
(D0, D4, D6, and D10)-DMABA NHS ester reagents that >92%
of the 18:1a/18:1-PE standard was converted to (D0, D4, D6,
and D10)-DMABA labeled 18:1a/18:1-PE using the described
labeling conditions.
The collision induced dissociation behavior of the [M + H]+
of the (D0, D4, D6, and D10)-DMABA labeled 18:1a/18:1-PE was
examined. One major fragmentation ion present in the positive
ion CID spectrum of the (D0, D4, D6, and D10)-DMABA labeled
18:1a/18:1-PE standards was at m/z 603.6 (Figure 1), which
was also observed upon CID of 18:1a/18:1-PE [M + H - 141]+
in the positive ion mode.9 The other major fragmentation ion
present in the positive ion CID spectra of (D0, D4, D6, and D10)-
DMABA labeled 18:1a/18:1-PE occurred at m/z 191.1 for D0-
DMABA labeled 18:1a/18:1-PE, m/z 195.1 for D4-DMABA
labeled 18:1a/18:1-PE, m/z 197.1 for D6-DMABA labeled 18:
1a/18:1-PE, and m/z 201.1 for D10-DMABA labeled 18:1a/18:
1-PE (Figure 1), which resulted from the cleavage at the
phosphate-ethanolamine bond. Equivalents of these fragmenta-
tion ions were not observed in the CID spectrum of 18:1a/18:1-
PE. There were also minor fragmentation ions in the positive ion
CID spectra of (D0, D4, D6, and D10)-DMABA-18:1a/18:1 labeled
PE at m/z 148.1, 152.1, 154.1, and 158.1, respectively, that were
due to cleavage of the amide bond that linked ethanolamine
to DMABA. This positive ion CID data suggested that the (D0,
D4, D6, and D10)-DMABA PE specific ions at m/z 191.1, 195.1,
197.1, and 201.1, respectively, could be used to selectively
detect (D0, D4, D6, and D10)-DMABA modified PE in a complex
biological mixture. The D0-DMABA labeled 18:1a/18:1-PE
standard was injected onto a reverse phase column and used
(16) Harrison, K. A.; Davies, S. S.; Marathe, G. K.; McIntyre, T.; Prescott, S.;
Reddy, K. M.; Falck, J. R.; Murphy, R. C. J. Mass Spectrom. 2000, 35, 224–
236
(17) Kim, H. Y.; Salem, N., Jr. J. Lipid Res. 1990, 31, 2285–2289
(18) Bodennec, J; Pelled, D.; Futerman, A. H. J. Lipid Res. 2003, 44, 218–226
.
.
.
Analytical Chemistry, Vol. 81, No. 16, August 15, 2009 6635

Figure 1. Positive ion CID spectra of the [M + H]+ of 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine labeled with (a) D0-DMABA NHS ester
reagent, (b) D4-DMABA NHS ester reagent, (c) D6-DMABA NHS ester reagent, and (d) D10-DMABA NHS ester reagent at a collision energy of
30 V. The origins of the ions that resulted from collisional activation are indicated in the inset structure.
to determine the limits of detection using a P191 scan. It was
found that 50 pg could be easily detected by a P191 scan and
additional sensitivity could be achieved using an MRM method.
One major product ion at m/z 281.1 was observed in the
negative ion CID spectrum of the [M - H ]- at m/z 889.7, 893.7,
895.7, and 899.7 of the (D0, D4, D6, and D10)-DMABA labeled
18:1a/18:1-PE (Figure 2). This ion was also present in the
negative ion CID spectrum of 18:1a/18:1-PE and corresponded
to the oleic acid carboxylate ion.9 There were also minor product
ions in the negative ion CID spectra of (D0, D4, D6, and D10)-
DMABA labeled 18:1a/18:1-PE at m/z 625.4, 629.4, 631.4, and
635.4, respectively, that corresponded to the loss of oleic acid
present at the sn-2 position as a neutral ketene. Both the fatty
acid ions and the neutral loss of the sn-2 position as a neutral
ketene have been observed in the CID of unlabeled 18:1a/18:
1-PE.9 The negative ion CID behavior of DMABA labeled PE
was largely unaltered in comparison to the unlabeled PE, which
is important because negative ion tandem mass spectrometry
can be used to characterize the fatty acyl substituents esterified
to the glycerol backbone of DMABA labeled PE lipids.
ion mass spectrum from m/z 690 to m/z 792 (Figure 3a). The
CID spectra in the negative ion mode of the [M - H]- of all PE
lipids in RAW 264.7 cells were obtained, and the fatty acids at
the sn-1 and sn-2 position of the glycerol backbone were
determined (Table S-1 in the Supporting Information). From
this data, the presence of abundant plasmalogen PE lipids as well
as acylated PUFAs in RAW 264.7 cells was established. A NL141
scan selectively detected diacyl and ether PE lipids, but plasmalo-
gen PE were very poorly detected using a NL141 scan (Figure
3b). Once the PE lipids from RAW 264.7 cells were labeled with
the D0-DMABA NHS ester reagent, a positive ion mass
spectrum scan revealed a distribution of PE lipids similar to
the positive ion mass spectrum of PE present in the RAW 264.7
cells (Figure 3a) but shifted in mass +147 amu (Figure 3c).
Because of the positive ion CID behavior of the D0-DMABA
labeled standards (Figure 1a), a P191 scan was used to selectively
detect the D0-DMABA labeled PE from RAW 264.7 cells (Figure
3d). The D0-DMABA labeled PE were easily detected using this
precursor ion scan, and the plasmalogen PE lipids were not
discriminated against with this scan. The PE from RAW 264.7
cells was also labeled with (D4, D6, or D10)-DMABA NHS ester
reagent, and the distribution of PE lipids similar to the mass
spectrum of nonlabeled PE from RAW 264.7 cells, but shifted
in mass 151, 153, or 157 amu, was observed using a P195, P197,
or P201 scan, respectively (data not shown). It was concluded
The PE lipids extracted from RAW 264.7 cells were labeled
with D0-DMABA NHS ester reagent in order to determine if
these reagents would work well for a complex mixture of PE
lipids from biological samples. The [M + H]+ for PE lipids
extracted from RAW 264.7 cells were observed in the positive
6636 Analytical Chemistry, Vol. 81, No. 16, August 15, 2009

Figure 2. Negative ion CID spectra of the [M - H]- of 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine labeled with (a) D0-DMABA NHS ester
reagent, (b) D4-DMABA NHS ester reagent, (c) D6-DMABA NHS ester reagent, and (d) D10-DMABA NHS ester reagent at a collision energy of
-50 V. The origins of the ions that resulted from collisional activation are indicated in the inset structure.
from this data that all subclasses of (D0, D4, D6, or D10)-DMABA
labeled PE species, including plasmalogens, could be detected
using a P191, P195, P197, or P201 scan, respectively.
The less lipophilic compounds (7-21 min) that eluted from
the column before the internal standard in the P191, P195, P201,
and P197 chromatograms (Figure 4) were the oxidized PE
products formed upon AAPH treatment. The negative ion CID
spectrum of one of the D6-DMABA labeled PE species with a
[M - H]- at m/z 760.6 that increased after 300 min of AAPH
treatment and eluted from the reversed phase column at 15.41
min was obtained (Figure 5). The CID spectrum revealed
fragmentation ions at m/z 633.5, 283.1, 144.1, and 112.1. The
fragmentation ion at m/z 283.1 corresponded to the carboxylate
ion of stearic acid, and the ion at m/z 633.5 corresponded to
the loss of the sn-2 substituent as a neutral ketene.9 Additionally,
the m/z 144.1 product ion corresponded to the MOX derivative
of 5-oxopentanoic acid and m/z 112.1 was consistent with the loss
of neutral methanol (32 amu) from the MOX derivative of
5-oxopentanoic acid.19 This data was consistent with proposed
identification of this product with an [M + H]+ at m/z 760.6 as
the MOX derivative of D6-DMABA labeled 1-stearoyl-2-(5-
oxovaleroyl)-sn-glycero-3-phosphoethanolamine (18:0a/5al-PE).
The negative ion CID spectra of other species in this chro-
matographic region of interest were also obtained (data not
shown) and permitted identification of several different types
of oxidized PE lipids, including lyso PE, chain shortened
ω-aldehydes at the sn-2 position, and direct oxidation of
polyunsaturated fatty acids at the sn-2 position, which have been
Control, 30 min AAPH treated, 60 min AAPH treated, and 300
min AAPH treated liposomes generated from RAW 264.7 phos-
pholipids were labeled with the D0-DMABA, D4-DMABA, D10-
DMABA, and D6-DMABA NHS ester reagents, respectively,
and then combined. The sample was then subjected to RP-
HPLC and the control D0-DMABA labeled PE, 30 min AAPH
treated D4-DMABA labeled PE, 60 min AAPH treated D10-
DMABA labeled PE, and 300 min AAPH treated D6-DMABA
labeled PE were detected by a P191, P195, P201, and P197 scan,
respectively (Figure 4). A scan rate of 1.5 s per precursor ion
scan was used which fit with our chromatographic separation;
however, faster scan rates are now available on newer instruments
and may remove this limitation. The area of the D0-, D4-, D10-,
and D6-DMABA internal standards were quite similar (Table
1
), and therefore the chromatograms could be visually compared
and changes in the PE distribution could be readily observed
(Figure 4). Lipophilic compounds (23-36 min) that eluted from
the column after the internal standard in the P191, P195, P201,
and P197 chromatograms (Figure 4) were the endogenous,
nonoxidized PE, which was confirmed by negative ion CID (data
not shown). The chromatograms indicate that the endogenous
PE decreased during the time course of AAPH treatment, which
can be explained in part by loss of endogenous PE species to
newly formed oxidation products during the time course of AAPH
exposure.
(19) Watson, A. D.; Leitinger, N.; Navab, M.; Faull, K. F.; Horkko, S.; Witztum,
J. L.; Palinski, W.; Schwenke, D.; Salomon, R. G.; Sha, W.; Subbanagounder,
G.; Fogelman, A. M.; Berliner, J. A. J. Biol. Chem. 1997, 272, 13597–13607
.
Analytical Chemistry, Vol. 81, No. 16, August 15, 2009 6637

if all oxidized species present before DMABA labeling could be
detected after DMABA labeling. In order to examine this, 16:0a/
20:4-PE was oxidized with AAPH for 1 h and half of the sample
was derivatized with D4-DMABA NHS ester reagent. The [M
+ H]+ of the 16:0a/20:4-PE starting material is at m/z 740.7,
and the [M + H]+ of the D4-DMABA labeled 16:0a/20:4-PE
starting material is at m/z 891.7. The positive ion mass
spectrum of oxidized 16:0a/20:4-PE (Figure 6a) and D4-DMABA
labeled oxidized 16:0a/20:4-PE (Figure 6b) included molecules
that had molecular ions higher in mass than 16:0a/20:4-PE,
presumably due to the addition of oxygen atoms to the arachi-
donoyl moiety of 16:0a/20:4-PE, and molecular ions lower in mass
than 16:0a/20:4-PE, presumably due to oxidative fragmentation
of 16:0a/20:4-PE. The product distribution of oxidized 16:0a/20:
4-PE (Figure 6a) to D4-DMABA labeled oxidized 16:0a/20:4-PE
are very similar upon comparison, except that the D4-DMABA
labeled oxidized 16:0a/20:4-PE products are shifted in mass
by 151 amu. From this data it was concluded that the oxidized
PE species that are observable before labeling are still observ-
able after DMABA labeling and that PE oxidation can be
successfully monitored using the DMABA NHS ester reagents.
DISCUSSION
One of the most common ways to detect PE lipids using
tandem mass spectrometry is a neutral loss of 141 amu scan.
However, the disadvantage in using NL141 in order to determine
PE content in complex lipid mixtures is that only diacyl and ether
PE species undergo efficient neutral loss of phosphoethanolamine
and that this selective scan discriminates against the detection of
plasmalogen PE species. This difference in the CID behavior
displayed by the various subclasses of PE lipids has complicated
the detection of PE lipid because there is not a sensitive universal
scan that permits selective detection of all PE lipids. In this study
a set of four (D0, D4, D6, and D10) deuterium enriched DMABA
NHS ester reagents was developed that react with the primary
amine group of PE lipids in order to create derivatives where
all subclasses of DMABA labeled PE could be detected by a
common precursor ion scan. Indeed, the positive ion CID data
from the (D0, D4, D6, and D10)-DMABA labeled 18:1a/18:1-PE
suggested that a single, common precursor ion scan could be
used to selectively detect (D0, D4, D6, and D10)-DMABA
modified PE in a complex biological mixture. In fact, upon
reaction of PE species extracted from RAW 264.7 cells with
(D0, D4, D6, and D10)-DMABA NHS ester reagents, all sub-
classes of PE lipids were easily detected and the analysis of all
PE subclasses in a complex mixture became straightforward.
Additionally, the negative ion CID behavior of DMABA labeled
PE lipids was retained when compared to the unlabeled PE
species, facilitating the identification of the radyl groups present
at the sn-1 and sn-2 positions on the glycerol backbone, which
is particularly valuable in structural studies of lipids.
Figure 3. (a) Positive ion electrospray mass spectrum of the PE
lipids present in RAW 264.7 cells. The ions correspond to [M + H]+
of each PE present. (b) Neutral loss of 141 amu (NL141) scan to
analyze PE present in RAW 264.7 cells. (c) Positive ion electrospray
mass spectrum of the D0-DMABA labeled PE lipids present in RAW
264.7 cells. Note the shift in the [M + H]+ in the D0-DMABA labeled
PE lipids by 147 amu. The conversion of PE to DMABA labeled PE
in RAW 264.7 cells was determined to be 89% based on the ratio of
DMABA labeled ions (m/z 891) to the sum of DMABA labeled and
corresponding unlabeled species (m/z 744, not shown in the figure).
(d) Precursor ion scan of m/z 191.1 (P191) allows for specific
detection of D0-DMABA labeled PE lipids in RAW 264.7 cells.
previously observed.20,21 These chromatograms indicate that the
oxidized PE products increased during the time course of AAPH
treatment, which can be explained by formation of new oxidized
PE lipids.
The data above indicate PE oxidation products are detected
using the DMABA NHS ester reagents; however, it was not clear
While selective precursor ion scans allow for specific detection
of certain classes of phospholipids, a major challenge still
encountered has been the identification of those less abundant
phospholipid species formed by free radical reactions initiated in
certain disease states among an array of molecular species that
do not change. The use of mass spectrometry and the stable
isotope dilution strategy has been a powerful approach to address
changes in biomolecules, but there has been a lack of available
(20) Berliner, J. A.; Zimman, A. Chem. Res. Toxicol. 2007, 20, 849–853
(21) Khaselev, N.; Murphy, R. C. J. Lipid Res. 2000, 41, 564–572
.
.
6638 Analytical Chemistry, Vol. 81, No. 16, August 15, 2009

Figure 4. Reversed-phase HPLC separation and tandem mass spectrometry analysis (LC-MS/MS) of D0-DMABA labeled PE from the control
liposomes (blue), D4-DMABA labeled PE from 30 min AAPH treated liposomes (green), D10-DMABA labeled PE from 60 min AAPH treated
liposomes (black), and D6-DMABA labeled PE from 300 min AAPH treated liposomes (red). The elution of the D0-DMABA labeled PE from the
control liposomes from the reversed phase HPLC column was monitored by a precursor ion scan of m/z 191.1 (P191), while the elution of the
D4, D10, and D6-DMABA labeled PE from the AAPH treated liposomes from the reversed-phase HPLC column was monitored by a precursor ion
scan of m/z 195.1, 201.1, and 197.1 (P195, P201, and P197), respectively. Note that the internal standard (IS) is recovered equally in each of
the samples.
Table 1. Chromatographic Peak Area of the 14:0a/14:0-PE Internal Standard Used for the Time Course of
Liposomes Exposed to AAPH
[M + H]+ of DMABA labeled
14:0a/14:0-PE (m/z)
area of extracted
chromatographic peak
sample
retention time (min)
control (D0-DMABA)
783.7
787.7
793.7
789.7
21.73
21.63
21.49
21.57
5.58 × 107
5.45 × 107
5.39 × 107
5.62 × 107
30 min AAPH (D4-DMABA)
60 min AAPH (D10-DMABA)
300 min AAPH (D6-DMABA)
phospholipid stable isotope internal standards to facilitate such
studies. These factors have combined to make it problematic to
monitor minor phospholipid molecular species changes that arise
in the phospholipid molecular species mixture after cell stimulation
or in certain disease states. These deuterium enriched DMABA
NHS reagents address these issues because they enable dif-
ferential labeling of PE lipids and thereby generate an internal
standard for every PE lipid of interest. The deuterium enriched
DMABA reagents allow for a direct comparison of control PE
lipids to treated PE lipids and in addition direct comparison of
the changes in abundance of PE lipids.
in a complex mixture can be selectively detected using a P191,
P195, P197, and P201 scan, respectively. Additionally, modifica-
tion of the primary amine group of the PE lipid upon reaction
with (D0, D4, D6, and D10)-DMABA NHS ester reagents,
permitted the (D0, D4, D6, and D10)-DMABA labeled PE to be
separated from PC lipids in complex biological samples with a
facile aminopropyl solid phase extraction (SPE) procedure. This
aminopropyl separation of (D0, D4, D6, and D10)-DMABA labeled
PE lipids from PC lipids has the benefit of removing PC from
the sample, and therefore the detection of (D0, D4, D6, and D10)-
While labeling phospholipid extracts from cells or tissues does
add another step to sample preparation, the synthesis of the
DMABA NHS ester reagents and PE labeling procedures are
straightforward. In addition, there are specific advantages to
labeling PE lipids with DMABA NHS ester reagents. First, all
subclasses of (D0, D4, D6, and D10)-DMABA labeled PE lipids
Figure 5. Negative ion CID of the [M - H]- at m/z 760.6 using a
triple quadrupole mass spectrometer at a collision energy of -50 V
with a retention time of 15.41 min. The origins of the ions that resulted
from collisional activation are indicated in the structures of these
molecules. This data indicates that the [M + H]+ that eluted from the
column at 15.41 min is MOX derivatized D6-DMABA labeled 18:0a/
5al-PE.
Figure 6. Positive ion mass spectrum of (a) oxidized 16:0a/20:4-
PE and (b) D4-DMABA labeled oxidized 16:0a/20:4-PE. Note the shift
in the [M + H]+ in the D4-DMABA labeled oxidized PE lipids by 151
amu.
Analytical Chemistry, Vol. 81, No. 16, August 15, 2009 6639

DMABA labeled PE is not suppressed in positive ion mode.
Finally only one chromatographic run would be needed to
analyze up to four different experimental conditions because
of the differential display capability afforded by four different
DMABA NHS ester reagents. Once labeled with the (D0, D4,
D6, and D10)-DMABA NHS ester reagents, the samples from
four experimental variants could be mixed together and the
products could be detected in one single chromatographic run
by cycling through the P191, P195, P197, and P201 scans.
oxidation and apoptosis). This DMABA labeling method can
also compensate for the lack of PE internal standards currently
available.
ACKNOWLEDGMENT
This work was supported in part by a Lipid Maps Large Scale
Collaborative grant from General Medical Sciences (Grant
GM069338) and a grant from the Heart, Lung, and Blood Institute
(Grant HL34303) of the National Institutes of Health.
SUPPORTING INFORMATION AVAILABLE
Synthesis of DMABA NHS ester reagents and a table showing
the PE present in RAW 264.7 cells. This material is available free
of charge via the Internet at http://pubs.acs.org.
CONCLUSION
It has been found that the (D0, D4, D6, or D10)-DMABA NHS
ester reagents provide a novel and facile way to observe all
subclasses of PE lipids by tandem mass spectrometry. Ad-
ditionally, DMABA labeling is a novel and facile means to
observe changes in the distribution of PE lipids and appearance
of novel PE lipid products in a myriad of biological events (e.g.,
Received for review March 20, 2009. Accepted July 7,
2009.
AC900583A
6640 Analytical Chemistry, Vol. 81, No. 16, August 15, 2009