Low-concentration ozone reacts with plasmalogen glycerophosphoethanolamine lipids in lung surfactant

Article abstract of DOI:10.1021/tx900306p

Ozone is a common environmental toxicant to which individuals are exposed to on a daily basis. While biochemical end points such as increased mortality, decrements in pulmonary function, and initiation of inflammatory processes are known, little is actually understood regarding the chemical mechanisms underlying changes in pulmonary health, especially for low concentrations of ozone. This study was undertaken to investigate ozone-induced oxidation of endogenous lipids that are potentially exposed to environmental ozone within lung, specifically focusing on plasmalogen glycerophospholipids present in pulmonary surfactant. Sensitive liquid chromatography - mass spectrometry methods were developed to follow oxidation of diacyl and plasmalogen phosphatidylethanolamine (PE) phospholipids and to identify and quantitate products generated by ozonolysis. Using a unilamellar vesicle system containing a 1:1 molar mixture of 1-O-octadec-1′-enyl-2-octadecenoyl-PE and 1,2-dihexadecanoyl-PC, these studies revealed that the vinyl ether bond of plasmalogens was oxidized preferentially at low concentrations of ozone (100 ppb), when compared to olefinic bond oxidation on ω-9 of the fatty acyl chain in the same phospholipids. Major phospholipid products generated were identified as 1-formyl-2-octadecenoyl-PE and 1-hydroxy-2-octadecenoyl-PE. Heptadecanal and heptadecanoic acid production was also quantitated using gas chromatography - mass spectrometry, and production was consistent with oxidation of the vinyl ether, at low concentrations of ozone. Analysis of murine lung surfactant from C57Bl/6 mice revealed several plasmalogen PE lipid species, encompassing ～38% of total PE species. Upon exposure of ozone (0 and 100 ppb) to murine surfactant, plasmalogen PE molecular species preferentially reacted, as compared to diacyl PE molecular species. Lysophospholipids, pentadecanal, and nonanal were found to be the primary products of surfactant ozone oxidation.

Full text of DOI:10.1021/tx900306p

1
08
Chem. Res. Toxicol. 2010, 23, 108–117
Low-Concentration Ozone Reacts with Plasmalogen
Glycerophosphoethanolamine Lipids in Lung Surfactant
Kelly M. Wynalda and Robert C. Murphy*
Department of Pharmacology, UniVersity of Colorado DenVer, MS 8303, 12801 E. 17th AVe.,
Aurora, Colorado 80045
ReceiVed August 26, 2009
Ozone is a common environmental toxicant to which individuals are exposed to on a daily basis.
While biochemical end points such as increased mortality, decrements in pulmonary function, and initiation
of inflammatory processes are known, little is actually understood regarding the chemical mechanisms
underlying changes in pulmonary health, especially for low concentrations of ozone. This study was
undertaken to investigate ozone-induced oxidation of endogenous lipids that are potentially exposed to
environmental ozone within lung, specifically focusing on plasmalogen glycerophospholipids present in
pulmonary surfactant. Sensitive liquid chromatography-mass spectrometry methods were developed to
follow oxidation of diacyl and plasmalogen phosphatidylethanolamine (PE) phospholipids and to identify
and quantitate products generated by ozonolysis. Using a unilamellar vesicle system containing a 1:1
molar mixture of 1-O-octadec-1′-enyl-2-octadecenoyl-PE and 1,2-dihexadecanoyl-PC, these studies
revealed that the vinyl ether bond of plasmalogens was oxidized preferentially at low concentrations of
ozone (100 ppb), when compared to olefinic bond oxidation on ω-9 of the fatty acyl chain in the same
phospholipids. Major phospholipid products generated were identified as 1-formyl-2-octadecenoyl-PE
and 1-hydroxy-2-octadecenoyl-PE. Heptadecanal and heptadecanoic acid production was also quantitated
using gas chromatography-mass spectrometry, and production was consistent with oxidation of the vinyl
ether, at low concentrations of ozone. Analysis of murine lung surfactant from C57Bl/6 mice revealed
several plasmalogen PE lipid species, encompassing ∼38% of total PE species. Upon exposure of ozone
(
0 and 100 ppb) to murine surfactant, plasmalogen PE molecular species preferentially reacted, as compared
to diacyl PE molecular species. Lysophospholipids, pentadecanal, and nonanal were found to be the
primary products of surfactant ozone oxidation.
Introduction
Ozone is a chemically reactive gas present in air pollution
and, as such, is one of the most common environmental toxicants
1). Ozone is generated through various photochemical-driven
nents of the surfactant layer covering the respiratory tract or
cells that are not always covered by the surfactant, such as
alveolar macrophages. Therefore, the damaging effects are not
directly due to ozone per se but, rather, are mediated by a
cascade of secondary ozonation products generated by reactions
with these targets (11). Recently, particular interest has been
focused on the oxidation of lipids within the surfactant, which
has been shown to generate secondary signaling cascades within
the lung (12, 13). For instance, ozone-induced oxidation of
cholesterol generates products such as the 5ꢀ,6ꢀ-epoxy cho-
lesterol, both in vitro and in vivo, which can cause necrosis in
bronchial epithelial cell lines (14). Another product of surfactant
oxidation, 1-hexadecanoyl-2-nonanal-PC, has been shown to
initiate apoptosis of peripheral monocytes at low nanomolar
concentrations (15).
(
reactions between oxygen and nitrogen oxides or volatile organic
compounds (2) and, therefore, more prominent during summer
months in most areas. Concentrations of ozone found in many
suburban areas often reach 100 ppb or higher, commonly
exceeding the current EPA limit of 75 ppb (1). Epidemiological
studies have linked increases in environmental ozone with
increased mortality, especially in susceptible populations such
as asthmatics or individuals with chronic obstructive pulmonary
disease (3). Other effects on pulmonary function, such as
neutrophilia, acute inflammatory responses, decrements in peak
expiratory flow rates, forced expiratory volumes and forced vital
capacities, eye and airway irritation, decreased phagocytosis by
macrophages, and increased incidence of asthma attacks in
children (4-10) have also been associated with concentrations
of ozone as low as 100 ppb. Despite the wealth of knowledge
to date regarding the biochemical and physiological end points
of ozone, little is currently known regarding the underlying
mechanisms that link the chemistry of ozone with subsequent
biological effects.
Another potential surfactant target of ozone oxidation could
be plasmalogen glycerophospholipids. Plasmalogens are a
chemically unique type of ether lipid that possess a vinyl ether
linkage at the sn-1 position of the glycerol backbone (Figure
1A, 18:0p/18:1-PE). The vinyl ether bond has been shown to
have a higher reactivity to most reactive oxygen species,
including superoxide, peroxyl radicals, and HOBr/HOCl, gener-
ated by myeloperoxidase (16, 17). The sn-2 position in
plasmalogen molecular species generally contains a highly
polyunsaturated fatty acid such as docosahexanoic or arachidonic
acid, and the sn-3 position is usually an ethanolamine or choline
headgroup. Plasmalogens are found in variable concentrations
in the various tissues; in human lung homogenates, plasmalogens
Because of its high reactivity, ozone is not thought to directly
act with pulmonary epithelium but instead react with compo-
*
To whom correspondence should be addressed. Tel: 303-724-3356.
E-mail: Robert.Murphy@ucdenver.edu.
1
0.1021/tx900306p  2010 American Chemical Society
Published on Web 11/16/2009

Ozonolysis of Plasmalogen Glycerophospholipids
Chem. Res. Toxicol., Vol. 23, No. 1, 2010 109
Figure 1. Structure of the ozonolysis products derived from 18:0p/18:1-PE (A), in a vesicular model system exposed to 100 ppb ozone (1 h),
yielded five primary products: 1-formyl-2-octadecenoyl-PE (B), 1-hydroxy-2-octadecenoyl-PE (C), nonanal (D), heptadecanal (E), and heptadecanoic
acid (F).
are reported to total ∼12% of all phospholipids, but the exact
location or cell type containing these unique lipids was not
investigated (18). While the exact function of plasmalogens has
not been completely elucidated, they have been suggested to
serve a role as endogenous antioxidants within cell membranes
and lipoproteins (19, 20).
Plasmalogens have also been shown to be present within lung
surfactant; however, the exact content of various species has
not been previously examined. Choline and ethanolamine
plasmalogens have been reported to comprise ∼1% of total
surfactant phospholipids in rat (21). Plasmalogens were also
reported to be significantly lower in patients with respiratory
distress syndrome; however, the analysis used was unable to
distinguish specific lipid classes or fatty acyl chains (22). The
work described here investigates one of the potential contribu-
tions of plasmalogens within lung surfactant and studies their
susceptibility to oxidation by levels of ozone that typically occur
in environmental settings.
min to equilibrate ozone concentrations before any experiments
were conducted. The concentration of ozone in delivered gas was
confirmed in initial studies using a model 202 Ozone Monitor (2B
Technologies) to monitor ozone concentration exiting the chamber.
Vesicle Model System. Small unilamellar vesicles, comprised
of 18:0p/18:1-PE and 16:0a/16:0-PC (1:1 mol ratio), were prepared
using an extruder (Avanti Polar Lipids), per the manufacturer’s
instructions. The samples were then sonicated for 30 min. Vesicles
were made to a final volume of 1 mL/sample in phosphate-buffered
saline and contained no added oxidizing or reducing equivalents.
Prior to ozonolysis, vesicles were transferred into a 10 mL beaker,
which was placed inside the ozonolysis chamber for all ozone
exposures.
Isolation and Ozonolysis of Bronchoalveolar Lavage
(
BAL). C57Bl/6 mice (3-8 months) were sacrificed by CO
2
asphyxiation, followed by cervical dislocation. BAL was obtained
by tracheal instillation of 1 mL of PBS by means of a safelet
catheter (18 gauge needle) and was repeated four times. Lavages
were pooled and spun at 2000g to remove cells. The supernatant
was collected and spun at 10000g (24) to isolate endogenous
surfactant as a pellet. The surfactant pellets collected were pooled
and resuspended in PBS, and a 1,2-di-(3,7,11,15-tetramethylhexa-
decanoyl)-PE internal standard was added to ensure equivalent
splitting. The entire pool of surfactant was then split to yield three
mice per sample. All procedures were approved by the Animal Care
and Use Committee of the University of Colorado Denver.
Materials and Methods
1
-O-octadec-1′-enyl-2-octadecanoyl-sn-glycero-3-phosphatidyle-
thanolamine (18:0p/18:1-PE), 1,2-dihexadecanoyl-PC (16:0a/16:0-
PC), 1-tetradecanoyl-2-hydroxy-PE (14:0a/OH-PE), 1,2-tetradecanoyl-
PE (14:0a/14:0-PE), 1-octadecanoyl-2-hydroxy-PE (18:0a/OH-PE),
To ascertain the presence of PE and PC plasmalogens within
lung surfactant, the surfactant from two C57Bl/6 mice was isolated,
extracted, and separated on aminopropyl solid-phase extraction
column (Supelco), according to the method of Kim et al. (25). The
phospholipids were then separated using normal phase high-pressure
liquid chromatography, with mass spectrometry detection (NP-
HPLC/MS). Negative ion, full mass spectral scans were run, and 1
min fractions were collected. PE and PC lipid classes were pooled
separately and infused for positive and negative ion collision-
induced dissociation (CID) experiments. Identification of individual
PE phospholipids present in lavage was confirmed according the
method of Zemski Berry (26). Additional verification of plasmalo-
gen phospholipids was done by treating a portion of the sample
with methanolic HCl, which hydrolyzes the vinyl ether bond of all
plasmalogens within each sample, leaving only diacyl phospholipids
to be detected (27). Upon identification of all plasmalogen lipids
present within surfactant, multiple reaction monitoring (MRM)
methods were developed for quantitation of each molecular species.
Ozonolysis (0 and 100 ppb) of surfactant was conducted on
pooled surfactant extracts, equivalent to 3 mice per sample. Samples
(1 mL) were ozonized or exposed to air (scrubbed of ozone) in a
10 mL beaker that was placed inside of the ozone chamber for 1 h.
Samples were then extracted (see Lipid Extractions for the method)
and derivatized with either pentafluorbenzyl (PFB) hydroxylamine
1
,2-di-(3,7,11,15-tetramethylhexadecanoyl)-PE, 1-octadecanoyl-2-
octadecenoyl-PE (18:0a/18:1-PE), and (13,14,15,16-d )-hexadecanal
48.9 atom %, deuterated hexadecanal) were obtained from Avanti
9
(
Polar Lipids, Inc. (Alabaster, AL). HPLC solvents, formic acid,
methanolic HCl, dodecanal, pentafluorobenzyl bromide, and tri-
ethylammonium buffer (1 M) were purchased from Sigma Aldrich
(Milwaukee, WI). O-(2,3,4,5,6-Pentafluororbenzyl) hydroxylamine
hydrochloride (99+%) was purchased from Alfa Aesar (Ward Hill,
MA). The synthesis of N-hydroxy-succinimide-dimethylaminoben-
zoic acid (DMABA) reagents (deuterium labeled) has been previ-
ously described (23). Heptadecanoic acid was purchased from
Supelco (Bellefonte, PA) and (18-d
%
3
)-octadecanoic acid (99.5 atom
, deuterated stearic acid) was purchased from MSD Isotopes
(
Montreal, Canada).
Ozonolysis. Ozone, at precise concentrations, was generated
using a model 306 Ozone Calibration source (2B Technologies,
Boulder, CO), which delivered gas flow into a modified vacuum
desiccator, which served as the exposure chamber. Briefly, the
source was linked to a Teflon-coated tube, which transferred gas
from the generator to the ozonolysis chamber, at a flow rate of 2
L/min. A lab rotator was placed underneath the chamber to allow
shaking of samples during ozonolysis experiments (2 revolutions
per second). The generator and chamber were given at least 15

1
10 Chem. Res. Toxicol., Vol. 23, No. 1, 2010
Wynalda and Murphy
(
(
aldehydic compounds) or with dimethylaminobenzoic acid reagents
aminophospholipids).
Lipid Extractions. Following ozonolysis, internal standards (14:
Quantitation of all phospholipids was carried out using RP-LC/
MS/MS, on a 150 mm × 2.0 mm C18 (5 µm) column (Phenomenex)
and API QTRAP tandem quadruple 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 1 mM
aqueous ammonium acetate (6:2:2) (solvent A) and methanol made
to 1 mM ammonium acetate (solvent B). The gradient was run from
10 to 99% (solvent B) over 15 min, held at 99% for 15 min, and
then returned to 10% for re-equilibration. MRM was carried out in
negative ion mode with an electrospray voltage of -4000 V, a
declustering potential of -80 V, a collision energy of -40 V, a
collision gas of nitrogen, and a dwell time of 100 ms per mass.
Masses monitored included m/z 728.7f281.1 (18:0p/18:1-PE),
0
a/14:0-PE, 14:0a/OH-PE, and deuterated hexadecanal) were added
to each sample. Aldehydes were extracted by the addition of
acetonitrile and hexane (1:0.5:3 buffer/acetonitrile/hexane). The
sample was vortexed for 30 s, and the upper hexane layer was
removed. The extraction procedure was repeated three times. The
remaining aqueous layer, which contained the phospholipids and
polar products of ozonolysis, was extracted using a modified Bligh
and Dyer extraction, substituting methylene chloride for chloroform
(
28). Extraction was conducted three times per sample, and the
organic layer was pooled.
4
78.3f281.1 (1-hydroxy-2-octadecenoyl-PE), 506.4f281.1 (1-
Aldehyde Detection (GC/MS). The aldehydes removed in the
hexane extraction were taken to dryness using a stream of nitrogen
and then were derivatized to a PFB-oxime, using pentafluorbenzyl
hydroxylamine HCl, according to the method of Luo et al. (29).
The samples were analyzed by capillary column gas chromatog-
raphy-negative ion electron capture ionization mass spectrometry
formyl-2-octadecenoyl-PE), 634.3f227.2 (14:0a/14:0-PE), 424.3f
2
27.2 (14:0a/OH-PE), 618.4f171.1 [1-O′-octadec-1′-enyl-2-nona-
nal-PE (18:0p/9al-PE)], 368.2f171.1 (1-hydroxy-2-nonanal-PE),
and 396.2f171.1 (1-formyl-2-nonanal-PE).
The 18:0p/18:1-PE and 16:0a/16:0-PC, using each as their own
reference standard and using 14:0a/14:0-PE as the internal standard,
were used to generate standard curves. Lysophospholipids were
quantitated using the standard curve generated from 18:0a/OH-PE
as a reference standard and 14:0a/OH-PE as an internal standard.
For quantitation of the formyl products, synthetic 1-formyl-2-
octadecanoyl-PE was used as a reference standard as described
above for 14:0a/OH-PE as an internal standard.
(GC-NI-ECMS). GC-MS experiments were carried out using a 30 m
(
30 m × 0.2 mm i.d. × 0.25 mm film thickness) ZB-1 polydim-
ethysiloxane capillary gas chromatograph column (Phenomenex,
Torrance, CA) attached to a ThermoFinnigan (San Jose, CA) Trace
DSQ mass spectrometer. Helium was used as the carrier gas, at a
flow rate of 1.5 mL/min, and methane was used as the collision
gas; samples were injected by splitless injection, with an injection
port at 230 °C. The column temperature was initially set at 70 °C
for 2 min, then increased by 10 °C/min until 310 °C, and then held
at 310 °C for 6 min. The source parameters were as follows:
temperature, 200 °C; and electron energy, 70 eV. A specific ion
for each derivative was selected, corresponding to the loss of [M
DMABA Derivatization. Deuterium-labeled DMABA reagents
were synthesized as previously described, and derivatization
methods were followed according to the method of Zemski Berry
(
4
23). A solution of d -DMABA reagent for control and d10-DMABA
for ozone samples was added to each sample for aminophospholipid
derivatization. After derivatization, re-extraction was conducted
using the modified Bligh and Dyer described above to remove
excess reagent. A detailed work flow diagram has been included
in the Supporting Information to schematically indicate steps used
to differentially label experimental samples. Quantitation of DMABA-
labeled phospholipids was achieved by creating standard curves of
-
-
-
HF] or [M - HFNO] for selected ion monitoring (SIM), with
a dwell time of 100 ms per mass. Standard curves were made using
dodecanal as a reference standard and an internal standard of
deuterium-labeled hexadecanal. The linear regression generated was
directly applied to the signal for heptadecanal and nonanal relative
to the added internal standard.
d
10-DMABA 18:0p/18:1-PE, d10-DMABA 18:0a/18:1-PE, and d10
-
Fatty Acid Extraction. Separate small unilamellar vesicles were
used for fatty acid analysis. Following vesicular air or ozone
exposure, deuterated stearic acid was added as an internal standard.
Fatty acids were extracted from buffer by the addition of ethyl
acetate (2:1 ethyl acetate/buffer). The sample was vortexed for 10 s,
and the upper ethyl acetate was removed. The extraction procedure
was repeated three times.
DMABA 18:0a/OH-PE as reference standards and d10-DMABA 14:
0
a/14:0a-PE and d10-DMABA 14:0a/OH-PE as internal standards.
Statistics. Data are expressed as means ( SEMs. Individual
comparisons between groups (ozone-treated versus control) were
confirmed by the two-tailed Student’s t test. Reported differences
were calculated using GraphPad Prism 4 (La Jolla, CA), and a p
value of less than 0.05 was considered to be statistically significant.
Fatty Acid Analysis. The fatty acids extracted were derivatized
to a PFB-ester using pentafluorbenzyl bromide (PFB-Br), according
to the method of Hadeley et al. (30). A specific ion for eac-h
derivative was selected, corresponding to the loss of [M - PFB]
for SIM, with a dwell time of 100 ms per mass. Standard curves
were made using heptadecanoic acid as a reference standards and
internal standard deuterated stearic acid. The linear regression
generated was directly applied to the signal for heptadecanoic acid
relative to the added internal standard.
Ozonolysis Phospholipid Product Analysis. Phospholipid samples
extracted in methylene chloride layer were taken to dryness under
a stream of nitrogen and were separated using NP-LC/MS (elec-
trospray ionization), according to the method of Zemski Berry (26),
on a 150 mm × 2.0 mm Ultremex 3 Silica column (Phenomenex).
Products of ozonolysis were identified by their molecular ions, CID
of both the positive and negative molecular ion, and HPLC retention
time.
Results
Ozone Oxidation of Plasmalogen PE. A unilamellar vesicle
system of defined phospholipid content was first examined to
evaluate plasmalogen oxidation using low concentrations of
ozone. The plasmalogen PE molecular species studied (18:0p/
18:1-PE, Figure 1A) was not the one found in murine surfactant
(see below) but did contain two different unsaturated, a vinyl
ether moiety connecting the sn-1 radyl chain to the glycero-
phospholipid backbone as well as an ω-9 olefinic bond of the
octadecenoyl chain. The progress of ozonolysis was followed
by product formation and loss of substrate. We hypothesized
that the vinyl ether bond of plasmalogens, in an ordered structure
such as a vesicle, would be preferentially oxidized at low
concentrations of ozone, when compared to fatty acyl chain
olefinic double bonds. Therefore, small unilamellar vesicles
comprised of 18:0p/18:1-PE and 16:0a/16:0-PC, in a mole ratio
of ∼1:1, were made in PBS and exposed for 1 h to ozone at 0,
The synthesis of 1-formyl-2-octadecenoyl-PE was carried out
according to previously published methods (31) with modifications
for the formylation of free hydroxyl, rather than acetylation. Briefly,
1
-hydroxy-2-octadecenoyl-PE (100 µg) was suspended in formic
5
0, 100, 300, or 1000 ppb.
acid (250 µL). The solution was stirred overnight, at room
temperature, and then extracted as detailed above (modified Bligh
and Dyer). The product was characterized by positive and negative
ion CID mass spectrometry (yielding CID spectra identical to Figure
Concentration-dependent oxidation was observed by monitor-
ing a decrease in 18:0p/18:1-PE content (Figure 2A). A loss of
approximately 19% 18:0p/18:1-PE was observed at 50 ppb, 31%
at 100 ppb, 36% at 300 ppb, and 52% at 1000 ppb ozone. The
3
C,D) experiments, with an approximate yield of 40%.

Ozonolysis of Plasmalogen Glycerophospholipids
Chem. Res. Toxicol., Vol. 23, No. 1, 2010 111
confirmation of identity was achieved by comparing retention
times of the commercially available standard, 1-hydroxy-2-
octadecenoyl-PE, as well as synthetic 1-formyl-2-octadecenoyl-
PE to the ozonolysis products (data not shown).
Two major phospholipid products were found to be produced
in the vesicular model system at 100 ppb, 1-formyl-2-octade-
cenoyl-PE (Figure 4A) and 1-hydroxy-2-octadecenoyl-PE (Fig-
ure 4B). This was consistent with the hypothesis of the vinyl
ether bond having a unique susceptibility to low concentrations
of ozone. At higher concentrations of ozone, these products were
still produced; however, further oxidation of the acyl olefinic
bond within these products was found, yielding the 1-hydroxy-
2-nonanal-PE and 1-formyl-2-nonanal-PE (Supporting Informa-
tion, Figure 1A,B). The 1-formyl-2-nonanal-PE did not display
significant production above control until 300 and 1000 ppb.
The 1-hydroxy-2-nonanal-PE increased at 100 and 300 ppb, with
the highest levels reached at 1000 ppb.
Fatty aldehydes measured by GC/MS (Figure 5A) included
the nonanal, which was likely produced by oxidation of the (ω-
9
) olefinic double bond in the oleoyl fatty acyl chain, and
heptadecanal, produced by oxidation of the vinyl ether moiety
of the plasmalogen. Aldehyde formation followed the same trend
as the phospholipid products recovered. The heptadecanal was
a prominent product at low concentrations of ozone when
compared to the nonanal, again consistent with preferential
oxidation of the vinyl ether moiety of the plasmalogen to low
concentration of ozone. Higher concentrations of ozone dis-
played oxidation of both olefinic double bond in the oleoyl fatty
acyl chain as well as the vinyl ether moiety.
The heptadecanoic acid was also measured by GC/MS as the
PFB-ester (Figure 5B). Heptadecanoic acid, produced by
oxidation of the heptadecanal, displayed a concentration-
dependent production at each concentration measured. Nonanoic
acid, a potential product of nonanal hydroxyhydroperoxide
oxidation, was observed at all four concentrations of ozone
exposure employed; however, it was also abundant in control
and procedural blanks at very high levels; therefore, absolute
quantitation was not carried out.
Presence of Plasmalogens in Murine Lung Surfactant. To
identify the plasmalogen lipid species present within lung
surfactant, BALs were taken, and surfactant was isolated
according to the method of Oulton et al. (24) as described in
the Materials and Methods. Because plasmalogen species are
most commonly found as PE or PC phospholipids, identification
of plasmalogens within lung surfactant focused on these two
lipid classes. The PE and PC lipids were separated from neutral
and anionic lipids using the aminopropyl solid-phase extraction
column (25). The collected eluant containing PE and PC was
then separated on NP-LC/MS/MS, and fractions were collected
for direct infusion and collisional activation of both positive
and negative molecular ions.
Figure 2. Ozone concentration effect: Small unilamellar vesicles
comprised of 18:0p/18:1-PE and 16:0a/16:0-PC were ozonized for 1 h,
at 0 (air scrubbed of ozone), 50, 100, 300, and 1000 ppb. (A) Loss of
1
8:0p/18:1-PE as measured by quantitative LC/MRM analysis. (B)
Quantity of 16:0a/16:0-PC in the same vesicles. Results are reported
as means ( SEMs (n ) g3).
quantity of 16:0a/16:0-PC present in these vesicles did not
change significantly (Figure 2B). To begin identification of
ozonolysis products, control and ozonized total ion chromato-
grams (of NP-LC/MS) were overlaid, allowing changes in ion
abundances to be compared (data not shown). Areas of increased
ion abundances were noted for identification of products during
infusion of fractions collected from NP-LC/MS. The products
generated were identified using positive and negative ion CID
experiments, and the alkyl aldehydic products and fatty acids
generated were derivatized and quantitated by GC/MS. The main
products formed included 1-formyl-2-octadecenoyl-PE (Figure 1B),
1-hydroxy-2-octadecenoyl-PE (Figure 1C), nonanal (Figure 1D),
heptadecanal (Figure 1E), and heptadecanoic acid (Figure 1F). The
formation of the 1-formyl-2-octadecenoyl-PE (Figure 1B) and
aldehydes (Figures 1C and 1D) follows typical Criegee ozonolysis
mechanisms. The production of the 1-hydroxy-2-octadecenoyl-PE
Infusion of the PE fraction (Figure 6A) showed numerous
-
(Figure 1C) is proposed to occur by loss of performic acid from
molecular species ([M - H] ). The most abundant lipid species
the hydroxyhydroperoxy intermediate, on carbon-1 of the vinyl
ether group.
Structural characterization of phospholipid products was based
upon product ions generated by CID experiments of observed
molecular ions, as well as HPLC retention time. Characterization
of the products in positive ion CID demonstrated that both the
(m/z 722) was identified as 1-O-hexadec-1′-enyl-2-eicosatet-
raenoyl-PE (16:0p/20:4-PE), as confirmed by CID experiments
(Figure 6B,C). As shown, in positive ion CID (Figure 6B),
abundant fragmentation to m/z 361 and 364, corresponds to the
sn-2 and sn-1 fragments. Fragmentation of PE plasmalogens to
these specific fragments has been studied in great detail by
Zemski Berry; the production of the sn-1 and sn-2 fragments
was proposed to occur by a complex series of molecular
rearrangements upon CID (26). In negative ion CID, the
arachidonate carboxylate anion (m/z 303) was the most promi-
nent fragment ion (Figure 6C), followed by loss of the
1
-formyl-2-octadecenoyl-PE and the 1-hydroxy-2-octadecenoyl-
PE fragmented to lose the phosphoethanolamine headgroup
Figure 3B,D, [M + H - 141]) (30). Negative ion CID yielded
(
m/z 281 as the primary fragment, indicative of an intact oleic
acid in the sn-2 position of the backbone (Figure 3A,C). Further

1
12 Chem. Res. Toxicol., Vol. 23, No. 1, 2010
Wynalda and Murphy
Figure 3. Characterization of phospholipid products produced by exposure of small unilamellar vesicles to 100 ppb ozone: Products of ozonolysis
were extracted and separated using NP-LC/MS. Fractions were infused for positive and negative ion CID mass spectrometric experiments. Major
products found are as follows: (A) 1-hydroxy-2-octadecenoyl-PE (negative ion mode), (B) 1-hydroxy-2-octadecenoyl-PE (positive ion mode), (C)
1
-formyl-2-octadecenoyl-PE (negative ion mode), and (D) 1-formyl-2-octadecenoyl-PE (positive ion mode).
Figure 4. Phospholipid products generated by ozone exposure (0, 50,
00, 300, and 1000 ppb) of 18:0p/18:1-PE. Products were measured
using LC/MRM analysis. (A) 1-Formyl-2-octadecenoyl-PE (blue, m/z
1
Figure 5. Concentration response of aldehydes produced by exposure
of small unilamellar vesicles to 0 (air control), 50, 100, 300, and 1000
ppb ozone for 1 h. (A) Aldehydes were extracted and derivatized to
the PFB-oxime; GC-MS was used for quantitation of aldehydes,
506f281) and (B) 1-hydroxy-2-octadecenoyl-PE (red, m/z 478f281).
Quantitation was based on standard curves. Results are expressed as
means ( SEMs (n ) g3).
-
monitoring the [M - HFNO] ion. Heptadecanal (red) was monitored
at m/z 399, and nonanal (blue) was monitored at m/z 287. (B)
Heptadecanoic acid produced by exposure of small unilamellar vesicles
to 0 (air control), 100, 300, and 1000 ppb ozone for 1 h. Fatty acids
were extracted and derivatized to the PFB-ester; GC-MS was used for
arachidonate as a neutral ketene (m/z 436). Other plasmalogens
species found included 1-O-hexadec-1′-enyl-2-docosahexaenoyl-
PE (16:0p/22:6-PE), 1-O-hexadec-1′-enyl-2-octadecenoyl-PE
quantitation of the heptadecanoic acid, monitoring the carboxylate anion
(
16:0p/18:1-PE), 1-O-hexadec-1′-enyl-2-hexadecenoyl-PE (16:
-
[
)
RCOO] at m/z 269. All results are expressed as means ( SEMs (n
0p/16:1-PE), 1-O-hexadec-1′-enyl-2-eicosapentaenoyl-PE (16:
0p/22:5-PE), 1-O-hexadec-1′-enyl-2-eicosatetraenoyl-PE (16:0p/
22:4-PE), and 1-O-octadec-1′-enyl-2-eicosatetraenoyl-PE (18:
0p/20:4-PE). Diacyl species were also identified and confirmed
3).
most abundant molecular species (accounting for >95% of all
PE phospholipids identified) is presented in Table 1. The 18:
0p/18:1-PE molecular species used in the vesicular model was
not found in the mouse lung surfactant. In total, plasmalogens
encompassed over ∼38% of the total PE phospholipids present;
using CID (Table 1), with the most abundant species confirmed
as 1-hexadecanoyl-2-octadecadienoyl-PE (16:0a/18:2-PE) and
1
-hexadecanoyl-2-octadecenoyl-PE (16:0a/18:1-PE). The 15

Ozonolysis of Plasmalogen Glycerophospholipids
Chem. Res. Toxicol., Vol. 23, No. 1, 2010 113
Figure 6. Identification of PE phospholipid molecular species by mass spectrometry. Isolated surfactant from two mice was pooled and separated
on an aminopropyl solid-phase extraction column (see the Materials and Methods for conditions). The collected eluant was additionally separated
using NP-LC/MS, fractions were collected, and all PE phospholipids were pooled for individual lipid species identification. (A) Pooled PE phospholipids
were infused. Shown is a negative ion full mass spectrum from m/z 620 to 800. Plasmalogen species are indicated in red text. (B) Positive ion CID
of m/z 724 (16:0p/20:4-PE) yielded abundant fragment ions at m/z 361 and m/z 364, corresponding to the sn-1 and sn-2 fragments indicative of
plasmalogens. A fragment corresponding to the NL 141 at m/z 583 was also observed. (C) Negative ion CID of m/z 722 (16:0p/20:4-PE) confirmed
sn-2 arachidonate carboxylate anion (m/z 303) as the most prominent fragment, as well as a neutral loss as a ketene (m/z 436).
Table 1. Listed Are the 15 Most Abundant Lipid Species,
Which Were Monitored by LC/MRM Analysis, in Negative
a
Ion Mode
PE lipid
species
MRM transistion
mol % total
PE phospholipids
used (m/z)
1
1
1
1
1
1
1
1
1
1
1
6:0p/16:1
6:0a/16:1
6:0p/18:1
6:0a/18:2
6:0a/18:1
6:0p/20:4
6:0a/20:4
8:0a/18:1
6:0p/22:6
6:0p/22:5
6:0e/22:6
672f253
688f253
700f281
714f279
716f281
722f303
740f303
744f281
746f327
748f329
748f327
750f331
750f303
762f327
766f303
1.9
5.9
2.5
20.4
12.4
23.8
1.8
1.4
1.7
2.3
1.3
1
1
1
1
6:0p/22:4
8:0p/20:4
6:0a/22:6
8:0a/20:4
2.0
4.7
10.5
2.6
Figure 7. PE molecular species in murine lung surfactant and changes
a
MRM transitions used for identification are listed in column two,
caused by exposure to 100 ppb ozone for 1 h. Control samples were
and percent contribution of each species to total PE phospholipids is
listed in column three. All identifications were confirmed using both
positive and negative ion CID experiments. Results are expressed as
means ( SEMs (n ) 3).
4
derivatized with d -dimethylaminobenzoic acid (DMABA), and ozo-
nized samples were derivatized with d10-DMABA. Samples were then
re-extracted, pooled together, separated from PC using NP-LC/MS,
collected, and quantitated using RP-LC/MRM. Shown in blue are
control data, and shown in red are ozone data. Plasmalogens are shown
on the left side, and diacyl phospholipids are shown on the right. Data
are expressed as means ( SEMs (n ) 3). *Significant oxidation as
compared to control values (p < 0.05).
on the basis of the published composition of each lipid species
contribution (24) to total surfactant composition, plasmalogens
comprise approximately 1-2% of all of the lipid species.
Analysis of the PC fractions revealed no plasmalogen species
to be present (data not shown). The surfactant from male and
female mice was also compared and revealed no significant
differences in PE species content (data not shown).
Ozonolysis of Surfactant. Finding that plasmalogens com-
prised ∼38% of all murine surfactant PE lipid species, it was
of interest to monitor the effect of low concentrations of ozone
on lipid species present in this more complex system. The
surfactant was extracted and pooled according to the methods
described in the Materials and Methods. Samples were individu-
ally exposed to air (scrubbed of ozone) or ozone (100 ppb) for
plasmalogen species (Figure 7). Plasmalogens were found to
decrease on average ∼53%, whereas diacyl molecular species
decreased, however, not significantly as compared to control
values. The oxidation of the 16:0p/20:4-PE, the most abundant
plasmalogen, was reduced ∼56% following exposure to 100
ppb ozone. In comparison, 16:0a/18:2-PE, the most abundant
diacyl species, did not significantly decrease even though it
contained two double bonds.
Lyso and formyl products were also measured in response
to 100 ppb ozone (Figure 8A). Interestingly, significant increases
in the 1-hydroxy-2-eicosatetraenoyl-PE (20:4a/OH-PE) and
1
h, and a differential display method was used to label control
1
-hexadecanoyl-2-hydroxy-PE (16:0a/OH-PE) were observed.
versus experimentally treated samples (see the Supporting
Information, Scheme 1). DMABA labeled with four deuterium
atoms was used for control samples, while a label with 10
deuterium atoms was used to tag ozonolysis samples. This
allowed for a combination of samples to decrease inter-run
variability and increase the mass of each phospholipid monitored
by 151 and 157 mass units, respectively. The conversion of the
ion abundance ratios of labeled PE species to the internal
standard was used with a standard curve for quantitation.
The 16:0a/OH-PE was not an expected product because no
plasmalogens were identified that contained a saturated sn-2 with
palmitate; therefore, further experimentation was conducted to
confirm the regiochemistry (sn-1 versus sn-2) of the lysophos-
pholipids formed. Chromatographically, lysophospholipids can
be separated by regiochemistry, using the reverse phase system
described in the Materials and Methods (32). The DMABA-
labeled lysophospholipids, however, will not separate in this
manner. Therefore, additional samples were exposed to control
air or 100 ppb of ozone, not labeled using DMABA, and run
by RP-LC/MRM to monitor regiochemistry of the increased
When the d
4
/d10 ratios were measured for 100 ppb ozone,
the only PE components that revealed significant oxidation were

1
14 Chem. Res. Toxicol., Vol. 23, No. 1, 2010
Wynalda and Murphy
Figure 9. Proposed mechanism of formation of plasmalogen PE
ozonolysis products.
the sn-2 fatty acyl group distant from the glycerol backbone.
Products of ozonolysis were characterized using liquid and gas
chromatography with sensitive mass spectrometric techniques.
When low concentrations of ozone were exposed to these
vesicles, the major products observed were 1-hydroxy-2-
octadecenoyl-PE, 1-formyl-2-octadecenoyl-PE, and heptadeca-
nal, consistent with preferential attack of the vinyl ether moiety
by ozone. Attack at the fatty acyl double bond yielded nonanal
as a product, which increased in abundance as the concentration
of ozone increased above 100 ppb.
Figure 8. Lysophospholipid molecular species and aldehydes generated
from murine surfactant exposure to 100 ppb ozone for 1 h. (A) Control
4
samples were derivatized with d -DMABA, and ozonized samples were
derivatized with d10-DMABA. Samples were then re-extracted, pooled
together, separated from PC using NP-LC/MS, collected, and quantitated
using RP-LC/MRM. The appearance of lysophospholipids was mea-
sured in control vs ozone samples. (B) Appearance of aldehydes
generated during ozonolysis. Three aldehydes were found as measured
by GC-MS as the PFB-oxime. Significant increases were found in two
of the three aldehydes measured. Data are expressed as means ( SEMs
(
n ) 3). *Significant oxidation (p < 0.05).
Previous studies have shown that ozonolysis of short chain
vinyl ethers, such as ethyl 1-propenyl ether, has higher reaction
rate constants for reaction with ozone, as compared to olefinic
double bonds in compounds such as methyl methacrylate
lysophospholipid products formed by ozone oxidation. As shown
in the Supporting Information, Figure 2, the increase in 20:4a/
OH-PE was predominantly sn-1 lyso, whereas the increase in
1
6:0a/OH-PE was sn-2 lyso (32). This supported increased
(
36, 37). Other published studies showed that ozonolysis of
arachidonyl lysophospholipid, resulting from oxidation of the
vinyl ether moiety, whereas the 16:0a/OH-PE was likely a result
from the removal of a sn-2 fatty acid by secretory phospholi-
alkenes, in which the double bond was connected to an electron-
donating group, reacted many times faster than typical alkenes
and those connected to electron-withdrawing groups (38).
Phospholipid products of ozonolysis identified within this
vesicular plasmalogen model exhibited the same unique sus-
ceptibility of the vinyl ether bond; additionally, this effect was
specific to low concentrations of ozone (e.g., 100 ppb),
suggesting that when plasmalogens were present in lipid
bilayers, they were very likely initial targets of reactions with
ozone at the lipid interface.
2 2
pases A (sPLA ) or a PAF-acetylhydrolase, known enzymatic
components of surfactant (33-35). Other lysophospholipid
compounds measured showed only negligible production. The
formation of formyl products was also measured; however, no
increase in the formation of product was observed for ozonized
surfactant, as compared to controls, and standard deviations of
data were high (data not shown).
Aldehydic products were also measured in this experiment
of control (air) and ozone-treated (100 ppb) surfactant (Figure
The phospholipid products (1-formyl-2-octadecenoyl-PE and
1
-hydroxy-2-octadecenoyl-PE) could both arise from an initial
8
9
B). Aldehydes measured included nonanal (C ), a product of
ozonide formed at the vinyl ether moiety 18:0p/18:1-PE and
decomposition by the Criegee mechanism (39), giving rise to
an aldehydic product and an unstable hydroxyhydroperoxy
intermediate (Figure 9). When the aldehydic product is formed
at the 2′-carbon atom of the vinyl ether, the product would be
heptadecanal, while the hydroxyhydroperoxy product (shown
as an intermediate in brackets in Figure 9) would be a rather
unique ortho perracid intermediate, which would be expected
to be correspondingly unstable. The lyso PE product (A) would
directly result from decomposition of this intermediate through
the loss of performic acid. The loss of hydrogen peroxide from
this intermediate would lead to the formation of the 1-formyl-
PE product (B). This same product (B) would be formed if the
initial aldehydic product formed was at the 1′ carbon atom of
the vinyl ether in the initial ozonide. The nonanal would arise
from an ozonide formed at the ω-9 double bond of the fatty
oleoyl fatty acyl chain oxidation, pentadecanal (C15) and
heptadecanal (C17), products of oxidizing a 1-O′-hexadecyl-1′-
enyl or 1-O′-octadecyl-1′-enyl bond within plasmalogens. The
nonanal and pentadecanal both showed significant increases (25
and 7 nmol, respectively), in ozonized samples, as compared
to their control. The heptadecanal did not increase significantly,
an expected result considering there was only one 1-O-octadecy-
1
0
′-enyl containing plasmalogen found within the surfactant (18:
p/20:4-PE).
Discussion
In this study, a small unilamellar vesicle system (lipid bilayer
system) of known composition was used to evaluate the
reactivity of ozone. This vesicle system contained a plasmalogen
PE that had a vinyl ether bond, as well as an olefinic bond in

Ozonolysis of Plasmalogen Glycerophospholipids
Chem. Res. Toxicol., Vol. 23, No. 1, 2010 115
acyl group at sn-2, from starting material or subsequent
phospholipid products above.
development of bronchopulmonary dysplasia in preterm infants,
an effect attributed to protection against oxygen toxicity (45).
Therefore, it is possible that even small changes in the low
concentrations of plasmalogen PE by ozonolysis may have
implications on normal respiration.
The analysis of surfactant was then carried out because of it
being a likely target for ozonolysis reactions in the lung. Detailed
analysis of murine surfactant phospholipid molecular species
revealed a complex mixture of molecular species. The most
abundant phospholipids were the expected PC molecular species
Perhaps of greater importance is the potential pulmonary
toxicity of lysophospholipid products generated during ozone
exposure to surfactant. These molecules are known to have
biological activities. For example, lysophosphoethanolamine has
been shown to play a role in neutrophil calcium flux through
G2A receptor (46). It has also been reported to have growth
factor-regulating effects of growth plate chondrocytes through
activation of TGF-ꢀ 1 (47) and recently has been shown to have
antiapoptotic effects on PC-12 cells (48). Lysophospholipids
in general are also substrates for metabolism to other biologically
active lipids such as lysophosphatidic acid. In the lung,
lysophosphatidic acid has been shown to be involved in
numerous biological events such as pulmonary fibrosis, en-
hancement of epithelial cell barrier integrity, and stimulation
of cytokine secretion (49-51).
1-hexadecanoyl-2-hexadecenoyl-PC, 1-hexadecanoyl-2-octade-
cenoyl-PC, and 1-hexadecanoyl-2-octadecadienoyl-PC as previ-
ously reported (40). Major PE molecular species in mouse lung
surfactant contained unsaturated and polyunsaturated fatty acyl
groups esterified to the sn-2 glycerol backbone, as well as
plasmalogen PE species. Interestingly, no saturated PE molecular
species were detected in contrast to the major PC species
observed in lung surfactant. It was found that plasmalogen PE
accounted for 38% of all PE phospholipid species in the
surfactant isolated from C57/BL6 mice, with the major plas-
malogen molecular species containing a polyunsaturated fatty
acid at sn-2. Only two plasmalogen PE molecular species were
found that contained an acyl group with a single double bond.
The production of aldehydes may also have biochemical
implications both for lipids and proteins within the surfactant
as well as underlying cells, since these lipophilic molecules are
readily diffusible across lipid membranes. Adduction to proteins
can cause decreased activities and, in some cases, complete
alteration of protein properties. Nonanal, one of the aldehydes
detected in both surfactant and our small unilamellar vesicle
system, has been shown to adduct human plasma apolipoprotein
A1 (52) as well as adduct the heme group of cytochrome
P4502B4 (53). The reactivity of aldehydes with aminophos-
pholipids, such as phosphatidylethanolamine, has also been
Because surfactant was a considerably more complex mixture
of lipid molecules than the vesicle system, the strategy of using
a differential display of PE molecular species was used to
monitor the formation of new products, as well as the decrease
in the PE molecular species when murine surfactant was exposed
to ozone. Lysophospholipid molecular species were found to
be a primary ozonolysis product; however, the corresponding
formylated products observed in the small unilamellar vesicles
were not found as major components. An important difference
between the model vesicle system and the ozone exposed
surfactant system was likely the fact that the isolated lung
surfactant still could contain many active proteins and enzymes.
Any 1-formyl products could be a substrate for enzymes such
2
shown to lead to inhibition of secreted phospholipase A and
phospholipase D (54).
It is of interest to consider the low concentration of plas-
malogen PE (approximately 1% total lipid) that is present in
surfactant and the much higher concentration of other phos-
pholipids containing one or more double bonds in their fatty
acyl groups. How can, therefore, plasmalogen PE phospholipids
be targets of ozone since this reactive oxygen allotrope would
be consumed by olefinic bonds in the fatty acyl chains. However,
such considerations do not take into account the fact that in
ViVo, the epithelial lining fluid (surfactant) or even cell
membranes do not present molecules in a homogeneous solution
but rather in ordered systems into which ozone can diffuse or
even react with lipids at the air-liquid interface. These
experiments with surfactant preparations exposed to low con-
centrations of ozone suggest that even though plasmalogen PE
accounts for only a small percentage of total unsaturated lipids
in surfactant, they can be unique targets even when other fatty
acyl groups containing one or more double bonds may be more
abundant.
2
as PAF-acetylhydrolase or sPLA (Figure 9), enzymes known
to be present in lung surfactant and have broad specificity for
the hydrolysis of short chain acyl groups on both PE and PC
(
41-43).
Overall, the mass balance between oxidation of the most
abundant plasmalogen (16:0p/20:4-PE, 38 nmol oxidized) and
the production of PE lysophospholipids, predominately 1-lyso-
2
-arachidonoyl-PE (20 nmol), revealed that some products had
not been accounted for in this analysis. It was hypothesized
that the amount of 1-lyso-2-arachidonoyl-PE initially formed
was actually higher than what was eventually measured and
that this lysophospholipid was simply being further oxidized
by ozone at one or more of the double bonds of the polyun-
saturated fatty acyl group at sn-2. A shortened time course of
ozone exposure of 30 min was carried out and found that indeed
higher levels of this 1-lyso-2-arachidonoyl-PE were present (data
not shown).
While loss of either diacyl or plasmalogen phospholipids may
cause changes in the efficiency of the surfactant barrier in the
lung in terms of stability and protection of the underlying
epithelial layer from reactive oxygen species, the observed
susceptibility of plasmalogen PE to ozone corresponded to only
Conclusion
The mechanisms by which oxidants such as ozone initiate
and propagate cellular and tissue toxicity are a poorly understood
phenomenon. An alternate function of surfactant, other than
lowering surface tension, can be seen as a defensive shield for
the underlying epithelial cell layer. Reactions of ozone within
the surfactant layer are likely to be the source of many of the
biochemical products that mediate cellular responses to the
inhalation of ozone. The study was conducted to investigate
the reaction of ozone with plasmalogen glycerophospholipids,
lipids found to be a component of surfactant, and to discover
oxidation products that may be formed when low concentrations
1
-2% of total surfactant phospholipids, making such an effect
of loss of phospholipids minimal. Rudiger et al. (44) show that
as little as 2 mol % of added plasmalogens effectively reduced
the surface tension of surfactant like phospholipid mixtures.
Previous studies have suggested that PE contributes only 3%
to total lung surfactant phospholipids at most (24). It has been
found that even minor increases in the percentage of plasmalo-
gen phospholipids and polyunsaturated fatty acyl groups in
pulmonary surfactant systems were protective against the

1
16 Chem. Res. Toxicol., Vol. 23, No. 1, 2010
Wynalda and Murphy
of ozone typically encountered in the environment are present.
Plasmalogens were found to encompass ∼38% of total PE lipids
within surfactant, and the vinyl ether bond was shown to react
with ozone. Analysis of PE phospholipid products generated in
this experiment revealed a significant increase in lysophospho-
lipids, which are known to induce an array of biological
activities (46-48). The oxidation of plasmalogens may therefore
be a potential source of secondary bioactive lipids, produced
by low concentrations of ozone found in the environment.
(16) Stadelmann-Ingrand, S., Pontcharraud, R., and Fauconneau, B. (2004)
Evidence for the reactivity of fatty aldehydes released from oxidized
plasmalogens with phosphatidylethanolamine to form Schiff base
adducts in rat brain homogenates. Chem. Phys. Lipids 131, 93–105.
(17) Murphy, R. C. (2001) Free-radical-induced oxidation of arachidonoyl
plasmalogen phospholipids: Antioxidant mechanism and precursor
pathway for bioactive eicosanoids. Chem. Res. Toxicol. 14, 463–472.
(
18) Diagne, A., Fauvel, J., Record, M., Chap, H., and Douste-Blazy, L.
(
1984) Studies on ether phospholipids. II. Comparative composition
of various tissues from human, rat and guinea pig. Biochim. Biophys.
Acta 793, 221–231.
19) Zoeller, R. A., Lake, A. C., Nagan, N., Gaposchkin, D. P., Legner,
M. A., and Lieberthal, W. (1999) Plasmalogens as endogenous
antioxidants: somatic cell mutants reveal the importance of the vinyl
ether. Biochem. J. 338 (Part 3), 769–776.
(
Acknowledgment. This work was supported by a grant from
the National Institutes of Health (HL034303). Special thanks
to Dr. Lynn Heasely, Dr. Christine Wu, and PharmOptima for
their kind donation of mice to the surfactant studies and Dr.
Lori Nield, Brad Barrett, and Jason Fritz for their assistance
with BAL techniques and advice.
(
(
20) Engelmann, B. (2004) Plasmalogens: targets for oxidants and major
lipophilic antioxidants. Biochem. Soc. Trans. 32, 147–150.
21) Rustow, B., Kolleck, I., Guthmann, F., Haupt, R., Kunze, D., and
Stevens, P. (1994) Sythesis and secretion of plasmalogens by type II
pneumocytes. Biochem. J. 302, 665–668.
(
(
(
22) Schmidt, R., Luboeinski, T., Markart, P., Ruppert, C., Daum, C.,
Grimminger, F., Seeger, W., and Gunther, A. (2004) Alveolar
antioxidant status in patients with acute respiratory distress syndrome.
Eur. Respir. J. 24, 994–999.
23) Zemski Berry, K. A., Turner, W. W., VanNieuwenhze, M. S., and
Murphy, R. C. (2009) Stable isotope labeled 4-(dimethylamino)benzoic
acid derivatives of glycerophosphoethanolamine lipids. Anal. Chem.
Supporting Information Available: Figure of ozone con-
centrations of 1-hydroxy-2-nonanal-PE and 1-formyl-2-nonanal-
PE, figure of chromatographic separation of increasing lyso-
phospholipids found after ozonolysis of lung surfactant, and
work flow diagram to schematically indicate steps used to
differentially label experimental samples. This material is
available free of charge via the Internet at http://pubs.acs.org.
8
1, 6633–6640.
24) Notter, R. (2000) Lung biology in health and disease. In Lung
Surfactants: Basic Science and Clinical Applications (Lenfant, C., Ed.)
Marcel Dekker, New York.
(
25) Kim, H. Y., and Salem, N., Jr. (1990) Separation of lipid classes by
solid phase extraction. J. Lipid Res. 31, 2285–2289.
References
(
26) Zemski Berry, K. A., and Murphy, R. C. (2004) Electrospray ionization
tandem mass spectrometry of glycerophosphoethanolamine plasmalo-
gen phospholipids. J. Am. Soc. Mass Spectrom. 15, 1499–1508.
(
1) Ciencewicki, J., Trivedi, S., and Kleeberger, S. R. (2008) Oxidants
and the pathogenesis of lung diseases. J. Allergy Clin. Immunol. 122,
4
56–468.
(27) Snyder, F., Blank, M. L., and Wykle, R. L. (1971) The enzymic
synthesis of ethanolamine plasmalogens. J. Biol. Chem. 246, 3639–
3645.
(
2) Finlayson-Pitts, B. J., and Pitts, J. N. (2000) Chemistry of the Upper
and Lower Atmosphere, Academic Press, San Diego, CA.
(
3) Tong, D. Q., Yu, S., and Kan, H. (2009) Ozone exposure and mortality.
(28) Bligh, E. G., and Dyer, W. J. (1959) A rapid method of total lipid
extraction and purification. Can. J. Biochem. Physiol. 37, 911–917.
(29) Luo, X. P., Yazdanpanah, M., Bhooi, N., and Lehotay, D. C. (1995)
Determination of aldehydes and other lipid peroxidation products in
biological samples by gas chromatography-mass spectrometry. Anal.
Biochem. 228, 294–298.
(30) Hadley, J. S., Fradin, A., and Murphy, R. C. (1988) Electron capture
negative ion chemical ionization analysis of arachidonic acid. Biomed.
EnViron. Mass Spectrom. 15, 175–178.
(31) Clay, K. (1990) Arachidonate related lipid mediators. In Methods in
Enzymology (Murphy, R. C., Fitzpatrick, F. A., Ed.) Academic Press,
San Diego.
(32) Creer, M. H., and Gross, R. W. (1985) Separation of isomeric
lysophospholipids by reverse phase HPLC. Lipids 20, 922–928.
(33) Hurly, B. P., and McCormick, B. A. (2008) Multiple roles of
N. Engl. J. Med. 360, 2788–2789.
(
4) Michalec, L., Choudhury, B. K., Postlethwait, E., Wild, J. S., Alam,
R., Lett-Brown, M., and Sur, S. (2002) CCL7 and CXCL10 orchestrate
oxidative stress-induced neutrophilic lung inflammation. J. Immunol.
1
68, 846–852.
5) Bernard, A., Carbonnelle, S., Nickmilder, M., and de Burbure, C.
2005) Non-invasive biomarkers of pulmonary damage and inflam-
(
(
mation: Application to children exposed to ozone and trichloramine.
Toxicol. Appl. Pharmacol. 206, 185–190.
(
6) Brown, J. S., Bateson, T. F., and McDonnell, W. F. (2008) Effects of
exposure to 0.06 ppm ozone on FEV1 in humans: A secondary analysis
of existing data. EnViron. Health Perspect. 116, 1023–1026.
7) Schelegle, E. S., Morales, C. A., Walby, W. F., Marion, S., and Allen,
R. P. (2009) 6.6-h inhalation of ozone concentrations from 60 to 87
parts per billion in healthy humans. Am. J. Respir. Crit. Care Med.
(
2
phospholipase A during lung infection and inflammation. Infect.
Immun. 76 (6), 2259–2272.
1
80, 265–272.
(
8) Hoppe, P., Peters, A., Rabe, G., Praml, G., Lindner, J., Jakobi, G.,
Fruhmann, G., and Nowak, D. (2003) Environmental ozone effects in
different population subgroups. Int. J. Hyg. EnViron. Health 206, 505–
(34) De Luca, D., Silvia, B., Vento, G., Piastra, M, Pietrini, D., Romitelli,
F., Capoluongo, E., Romagnoli, C., Conti, G., and Zecca, E. (2008)
Secretory phospholipases A2 and neonatal respiratory distress: Pilot
study on broncho-alveolar lavage. IntensiVe Care Med. 34, 1858–1864.
(35) Arbibe, L., Koumanov, K., Vial, D., Rougeot, C., Faure, G., Havet,
N., Longacre, S., Vargaftig, B. B., Bereziat, G., Voelker, D. R., Wolf,
C., and Touqui, L. (1998) Generation of lysophospholipids from
surfactant in acute lung injury is mediated by type-II phospholipase
A2 and inhibited by a direct surfactant protein A-phospholipase protein
interaction. J. Clin. InVest. 102, 1152–1160.
(36) Grosjean, E, and Grosjean, D. (1999) The reaction of unsaturated
aliphatic oxygenates with ozone. J. Atmos. Chem. 32 (1), 205–232.
(37) Grosjean, E, and Grosjean, D. (1998) Rate constants for the gas phase
reaction of ozone with unsaturated oxygenates. Int. J. Chem. Kinet.
30 (1), 21–29.
5
16.
(
9) Devlin, R. B., McDonnell, W. F., Mann, R., Becker, S., House, D. E.,
Schreinemachers, D., and Koren, H. S. (1991) Exposure of humans
to ambient levels of ozone for 6.6 h causes cellular and biochemical
changes in the lung. Am. J. Respir. Cell Mol. Biol. 4, 72–81.
(
(
(
10) Thurston, G. D., Lippmann, M., Scott, M. B., and Fine, J. M. (1997)
Summertime haze air pollution and children with asthma. Am. J.
Respir. Crit. Care Med. 155, 654–660.
11) Pryor, W. A. (1992) How far does ozone penetrate into the pulmonary
air/tissue boundary before it reacts? Free Radical Biol. Med. 12, 83–
8
8.
12) Dahl, M., Bauer, A. K., Arredouani, M., Soininen, R., Tryggvason,
K., Kleeberger, S. R., and Kobzik, L. (2007) Protection against inhaled
oxidants through scavenging of oxidized lipids by macrophage
receptors MARCO and SR-AI/II. J. Clin. InVest. 117, 757–764.
13) Postlethwait, E. M. (2007) Scavenger receptors clear the air. J. Clin.
InVest. 117, 601–604.
14) Pulfer, M. K., and Murphy, R. C. (2004) Formation of biologically
active oxysterols during ozonolysis of cholesterol present in lung
surfactant. J. Biol. Chem. 279, 26331–26338.
15) Uhlson, C., Harrison, K., Allen, C. B., Ahmad, S., White, C. W., and
Murphy, R. C. (2002) Oxidized phospholipids derived from ozone-
treated lung surfactant extract reduce macrophage and epithelial cell
viability. Chem. Res. Toxicol. 15, 896–906.
(38) Pryor, W. A., Giamalva, D., and Church, D. F. (1985) Kinetics of
ozonation. 3. Substituent effects on the rates of reaction of alkenes.
J. Am. Chem. Soc. 107, 2793–2797.
(
(
(39) Criegee, R. (1975) Mechanism of ozonolysis. Angew Chem. Int. Ed.
14 (11), 745–752.
(40) Bernhard, W., Wand, J., Tschernig, T., Tummler, B., Hedrich, H. J.,
and von der Hardt, H. (1997) Lung surfactant in a cystic fibrosis animal
model: Increased alveolar phospholipid pool size without altered
2
HGU/m2HGU
(
composition and surface tension function in cftrm
mice.
Thorax 52, 723–730.
(41) Nakos, G., Kitsiouli, E., Hatzidaki, E., Koulouras, V., Touqui, L., and
Lekka, M. (2005) Phospholipases A and platelet-activating-factor
2

Ozonolysis of Plasmalogen Glycerophospholipids
Chem. Res. Toxicol., Vol. 23, No. 1, 2010 117
acetylhydrolase in patients with acute respiratory distress syndrome.
Crit. Care Med. 33, 772–779.
42) Kudo, I., and Murakami, M. (2002) Phospholipase A enzymes.
2
(49) Tager, A. M., LaCamera, P., Shea, B. S., Campanella, G. S., Selman,
M., Zhao, Z., Polosukhin, V., Wain, J., Karimi-Shah, B. A., Kim,
N. D., Hart, W. K., Pardo, A., Blackwell, T. S., Xu, Y., Chun, J., and
Luster, A. D. (2008) The lysophosphatidic acid receptor LPA1 links
pulmonary fibrosis to lung injury by mediating fibroblast recruitment
and vascular leak. Nat. Med. 14, 45–54.
(50) He, D., Su, Y., Usatyuk, P. V., Spannhake, E. W., Kogut, P., Solway,
J., Natarajan, V., and Zhao, Y. (2009) Lysophosphatidic acid enhances
pulmonary epithelial barrier integrity and protects endotoxin-induced
epithelial barrier disruption and lung injury. J. Biol. Chem. 284, 24123–
24132.
(51) Medoff, B. D., Landry, A. L., Wittbold, K. A., Sandall, B. P., Derby,
M. C., Cao, Z., Adams, J. C., and Xavier, R. J. (2009) CARMA3
mediates lysophosphatidic acid-stimulated cytokine secretion by
bronchial epithelial cells. Am. J. Respir. Cell Mol. Biol. 40, 286–294.
(52) Szapacs, M. E., Kim, H. Y., Porter, N. A., and Liebler, D. C. (2008)
Identification of proteins adducted by lipid peroxidation products in
plasma and modifications of apolipoprotein A1 with a novel biotiny-
lated phospholipid probe. J. Proteome Res. 7, 4237–4246.
(
(
Prostaglandins Other Lipid Mediators 68-69, 3–58.
43) Min, J., Wilder, C., Aoki, J., Arai, H., Inoue, K., Paul, L., and Gelb,
M. (2001) Platelet activating factor acetylhydrolases: Broad substrate
specificity and lipoprotein binding does not modulate the catalytic
properties of the plasma enzyme. Biochemistry 40, 4539–4549.
44) Rudiger, M., Kolleck, I., Putz, G., Wauer, R. R., Stevens, P., and
Rustow, B. (1998) Plasmalogens effectively reduce the surface tension
of surfactant-like phospholipid mixtures. Am. J. Physiol. 274, L143–
L148.
(
(
(
(
(
45) Rudiger, M., von Baehr, A., Haupt, R., Wauer, R. R., and Rustow, B.
(
2000) Preterm infants with high polyunsaturated fatty acid and
plasmalogen content in tracheal aspirates develop bronchopulmonay
dysplasia less ofter. Crit. Care Med. 28, 1572–1577.
46) Frasch, S. C., Zemski-Berry, K., Murphy, R. C., Borregaard, N.,
Henson, P. M., and Bratton, D. L. (2007) Lysophospholipids of
different classes mobilize neutrophil secretory vesicles and induce
redundant signaling through G2A. J. Immunol. 178, 6540–6548.
47) Gay, I., Schwartz, Z., Sylvia, V. L., and Boyan, B. D. (2004)
Lysophospholipid regulates release and activation of latent TGF-beta1
from chondrocyte extracellular matrix. Biochim. Biophys. Acta 1684,
(
53) Kuo, C. L., Raner, G. M., Vaz, A. D., and Coon, M. J. (1999) Discrete
species of activated oxygen yield different cytochrome P450 heme
adducts from aldehydes. Biochemistry 38, 10511–10518.
54) Guichardant, M., Bernoud-Hubac, N., Chantegrel, B., Deshayes, C.,
and Lagarde, M. (2002) Aldehydes from n-6 fatty acid peroxidation.
Effects on aminophospholipids. Prostaglandins, Leukotrienes Essent.
Fatty Acids 67, 147–149.
(
1
8–28.
48) Nishina, A., Kimura, H., Sekiguchi, A., Fukumoto, R. H., Nakajima,
S., and Furukawa, S. (2006) Lysophosphatidylethanolamine in Grifola
frondosa as a neurotrophic activator via activation of MAPK. J. Lipid
Res. 47, 1434–1443.
TX900306P