(Received for publication, November 22, 1996, and in revised form, March 21, 1997)
From the Entry of monocytes into the vessel wall is an
important event in atherogenesis. Previous studies from our laboratory
suggest that oxidized arachidonic acid-containing phospholipids present in mildly oxidized low density lipoproteins (MM-LDL) can activate endothelial cells to bind monocytes. In this study, biologically active
oxidized arachidonic acid-containing phospholipids were produced by
autoxidation of
1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine (Ox-PAPC) and analyzed by liquid chromatography and electrospray ionization mass spectrometry in conjuction with biochemical
derivatization techniques. We have now determined the molecular
structure of two of three molecules present in MM-LDL and Ox-PAPC that
induce monocyte-endothelial interactions. These lipids were identified as 1-palmitoyl-2-(5-oxovaleryl)-sn-glycero-3-phosphocholine
(m/z 594.3) and
1-palmitoyl-2-glutaryl-sn-glycero-3-phosphocholine (m/z 610.2). These two molecules were produced
by unambiguous total synthesis and found to be identical by analytical
techniques and bioactivity assays to those present in MM-LDL and
Ox-PAPC. Evidence for the importance of all three oxidized
phospholipids in vivo was suggested by their presence in
fatty streak lesions from cholesterol-fed rabbits and by their
immunoreactivity with natural antibodies present in ApoE null mice.
Overall, these studies suggest that specific oxidized derivatives of
arachidonic acid-containing phospholipids may be important initiators
of atherogenesis.
Atherosclerosis is a chronic inflammatory disease responsible for
profound human morbidity and mortality (1). The atherosclerotic lesion
typically develops from a clinically silent fatty streak lesion,
characterized by the presence of lipid-laden monocyte-macrophages in
the subendothelial space, to an advanced fibro-fatty lesion that can
rupture, resulting in acute thrombosis and vascular occlusion (2, 3).
There is a growing body of evidence suggesting that oxidative
modification of low density lipoprotein
(LDL)1 is involved in the development and
progression of atherosclerosis (4-9). The process by which LDL
undergoes free radical-induced oxidation in vitro can be
described as having three stages (10-13): the lag phase, during which
lipophilic antioxidants (vitamin E, Tissue culture media, serum, and supplements were
obtained from sources previously reported (14, 17, 20). Acetonitrile, chloroform, hexane, methanol, 2-propanol, triethylamine, and water (all
Optima or HPLC grade) were obtained from Fisher. Gelatin (endotoxin-free, tissue culture grade), methoxyamine hydrochloride, sodium borohydride, N,N-diisopropylethylamine,
pentafluorobenzyl bromide, and butylated hydroxytoluene (BHT) were
obtained from Sigma. Authentic
L- Human aortic endothelial cells at
passages 4-7 were cultured as described (14, 17, 20) in Medium 199 supplemented with fetal bovine serum.
Rabbits were fed for 24 weeks with a
semipurified atherogenic diet containing 20% protein (casein), 46%
carbohydrate (sucrose), and 11% fat (butter). The diets also contained
0.09% cholesterol, 1% corn oil to provide essential fatty acids, 14%
cellulose, and levels of vitamins and minerals similar to those used
previously in similar diets (21). For some rabbits the diet was
supplemented with the antioxidants vitamin E (147 IU LDL (1.019-1.069 g/ml)
was isolated from the sera of normal blood donors by density gradient
ultracentrifugation as described (22) and stored at 4 °C until use
within 1-2 weeks of isolation. Lipoprotein concentration is expressed
in terms of protein content throughout this report. MM-LDL was produced
by dialysis at 5 mg protein/ml in 5 µM
FeSO4·7H2O in phosphate-buffered saline
(diluted 1:9 with sterile water) + 0.15 M NaCl, pH 6.8, for
~48 h at room temperature. Oxidation was terminated by the addition
of EDTA and BHT (100 and 0.3 µM, respectively). The
concentration of bacterial endotoxin in medium containing each agonist
was <20 pg/ml (determined by chromogenic assay), which is
approximately 100-fold less than that required to induce monocyte
binding to endothelial cells (data not shown). PAPC was oxidized by
transferring 1 mg in 100 µl of chloroform to a clean 16 × 125 mm glass test tube and evaporating the solvent under a stream of
nitrogen. The lipid residue was allowed to autoxidize while being
exposed to air for 24-48 h. The extent of oxidation was monitored by
positive ion flow injection electrospray mass spectrometry.
Lipids were extracted from rabbit aortic
tissue with chloroform/methanol (2:2, v/v) (23, 24). An aliquot of this
extract was supplemented with 20 µmol BHT/ml, covered with argon,
sealed, and stored protected from light. These extracts were
supplemented with dimyristoyl phosphatidylcholine (50 ng/mg tissue) as
an internal standard for the quantitative analysis of phospholipids
(25). Phospholipids, free fatty acids, and neutral lipids were
separated by the method of Kaluzny using aminopropyl solid phase
extraction chromatography (26). Phospholipid recovery typically ranged between 95 and 98% as determined by including
1-palmitoyl-2-[1-14C]-arachidonoyl-sn-glycero-3-phosphocholine
as an internal standard in the lipid-purification procedure. Lipid
fractions not used immediately were dried under nitrogen, resuspended
in chloroform containing 0.01% BHT, covered with argon, and stored at
Oxidized phospholipids of interest
were semipurified by preparative normal phase HPLC and treated with
either methoxyamine hydrochloride, pentafluorobenzyl bromide, or sodium
borohydride, and then lipids were extracted with chloroform/methanol
(27). Methoxyamine derivatives were produced by resuspending 250 µg of Ox-PAPC in a solution of 0.921 mM methoxyamine
hydrochloride in phosphate-buffered saline (pH 4.0). The solution was
incubated at 37 °C for 45 min, after which 3.5 ml of
chloroform/methanol (2:1, v/v) and 0.5 ml of water was added. The
solution was mixed thoroughly and centrifuged at 2,000 × g for 15 min. The chloroform phase was transferred to a
clean glass test tube and 0.5 ml of methanol, 0.5 ml of water, and 40 µl of formic acid was added. The solution was again mixed thoroughly
and centrifuged at 2,000 × g for 15 min. The
chloroform phase was then collected for analysis. The second extraction
step, which included formic acid, was necessary to displace salts
(mostly sodium) that associated with phospholipids during incubation
with phosphate-buffered saline. Without this step both the free
phospholipids and phospholipid salts were detected by positive ion
electrospray mass spectrometry. Pentafluorobenzyl ester derivatives of
carboxylic acids were produced by resuspending 250 µg of Ox-PAPC in
50 µl of a 10% solution of pentafluorobenzyl bromide in acetonitrile
and 50 µl of a 20% solution of
N,N-diisopropylethylamine in acetonitrile. The
solution was mixed thoroughly and incubated for 45 min at room
temperature. The solvent was dried under argon, and the lipids were
re-extracted by addition of 3.5 ml of chloroform/methanol (2:1) and 0.5 ml of water. The solution was mixed thoroughly and centrifuged at
2,000 × g for 15 min. The chloroform phase was transferred to a clean glass test tube and dried under nitrogen prior
to lipid analysis. Ox-PAPC was reduced by resuspending 250 µg of
lipid residue in 0.5 ml of a 0.1 M borate buffer (pH 8.0) and adding ~10 µg of sodium borohydride. The lipids were reduced instantaneously and then extracted by addition of 3.5 ml of
chloroform/methanol (2:1, v/v). Salts were displaced by re-extraction
in the presence of formic acid as described above.
Normal phase high
performance liquid chromatography (NP-HPLC) was performed by injecting
isolated phospholipids (resuspended in chloroform) onto a silica column
and eluting isocratically with a mobile solvent of
acetonitrile/methanol/water (77:8:15, v/v/v). NP-HPLC analysis was
performed using an analytical column (Spherisorb, 150 mm × 4.6 mm, 5 µm; Alltech Associates, Inc.) at a flow rate of 1.0 ml/min or a
preparative column (Adsorbosphere, 250 mm × 10 mm, 5 µm;
Alltech Associates, Inc.) at a flow rate of 5.0 ml/min. UV absorbance
was detected with a diode array detector (L-3000, Hitachi, Ltd., Tokyo,
Japan) scanning from 200 to 350 nm. In some cases, fractions were
collected from the column under sterile conditions, dried to a residue
under nitrogen at 37 °C, and either resuspended in tissue culture
medium for monocyte adhesion assays or used as a semipure preparation
for further modification and/or analysis by mass spectrometry. Reversed
phase HPLC (RP-HPLC) was performed using an ODS column (Adsorbosphere
HS, C18, 250 mm × 4.6 mm, 5 µm, Alltech Associates,
Inc.). Phospholipids were eluted with a mobile phase of 90% methanol,
which was changed linearly over a period of 30 min to 100% methanol,
at a flow rate of 1 ml/min. Fractions were collected at 1-min
intervals. For a near homogeneous preparation of an oxidized
phospholipid of interest, fractions obtained from preparative NP-HPLC
were further purified by analytical RP-HPLC.
Electrospray mass
spectrometry was performed with a API III triple-quadrupole
biomolecular mass analyzer (Perkin-Elmer) fitted with an articulated,
pneumatically assisted nebulization probe and an atmospheric pressure
ionization source. The tuning and calibration solution consisted of a
mixture of polypropylene glycol 425, 1000, and 2000 (3.3 × 10 The procedure of Huckstep (28) was
modified. Amberlyst-15 resin (4 g) was added to a magnetically stirred
solution of 5-pentanolide (2.16 g, 21.6 mmol) in dry methanol (40 ml).
The reaction mixture was boiled under reflux under dry nitrogen for
18 h, then filtered through florisil. The solvent was removed
under reduced pressure to produce methyl 5-hydroxypentanoate (2.8 g,
99%), which, without further purification, was oxidized with
pyridinium chlorochromate as described previously (28). Flash
chromatography on silica gel with ethyl acetate/hexane (1:3, v/v)
delivered 5-oxopentanoate (1.94 g, 69%). The 1H NMR
spectra of methyl 5-hydroxypentanoate and methyl 5-oxopentanoate were
identical with those reported previously (28). To a magnetically stirred solution containing ammonium nitrate (80 mg, 1 mmol) in dry
methanol (11 ml) was added freshly distilled trimethyl orthoformate (2.96 g, 28 mmol, 4 eq) and methyl 5-oxopentanoate (907 mg, 7 mmol).
The reaction mixture was stirred under nitrogen and monitored by TLC.
The reaction was stopped at 24 h, and solvent was removed under
reduced pressure. The residue was titrated with ether (30 ml), and the
ether solution was filtered to remove ammonium nitrate. Solvent was
removed under reduced pressure. Flash chromatography on a column (50 mm × 180 mm) of silica gel (40 µm, J. T. Baker Inc.) with
ethyl acetate/hexane (1:3, v/v) yielded methyl 5,5-dimethoxypentanoate (883 mg, 78.6%). The 1H NMR spectrum of methyl
5,5-dimethoxypentanoate was identical with that reported previously
(29). The methyl ester (142 mg, 0.81 mmol) was stirred with a solution
of sodium hydroxide (161.4 mg, 4 mmol, 5 eq) in 4 ml of
water/methanol/tetrahydrofuran (2:5:3, v/v/v) at room temperature.
After 1.5 h, the reaction mixture was acidified to pH 3 with
ice-cold 0.1 N HCl and then extracted with ethyl acetate
(3 × 15 ml). The combined organic extracts were dried over
MgSO4 and filtered, and the solvent was removed under
reduced pressure to provide 5-dimethoxypentanoic acid (129 mg, 99%).
The 1H NMR spectrum of 5-dimethoxypentanoic acid was
identical with that reported previously (29). Esterification of
lysophosphatidylcholine with 5-dimethoxypentanoic acid was accomplished
by the general method of Hebertt et al. (30). Traces of
moisture were removed from a mixture of acid (86.5 mg, 0.53 mmol, 3.1 eq) and 1-palmitoyl-2-hydroxy-sn-glycero-3-phosphocholine (82.2 mg, 0.17 mmol) by azeotropic distillation with toluene (3 × 5 ml) under reduced pressure at 20 °C. Freshly distilled dry CHCl3 (10 ml), dicyclohexylcarbodiimide (105 mg, 0.51 mmol,
3 eq), and
p-(N,N-dimethylamino)pyridine (62 mg,
0.51 mmol, 3 eq) were added, and then the flask was flushed with
nitrogen and protected from light. The reaction was monitored by TLC
and was stopped at 26 h, and the solvent was removed under reduced
pressure. The resulting solid was flash chromatographed on a column (20 mm × 180 mm) of silica gel (40 µm, J. T. Baker Inc.) with
CHCl3/MeOH/H2O (40:50:10, v/v/v) to produce
1-palmitoyl-2-(5-dimethoxypentanoyl)-sn-glycero-3-phosphocholine (199 mg, 93%). 1H NMR: d 5.21 (dd, 1H, J = 4 Hz, 4 Hz), 4.4-4.3 (4H), 4.1 (dd, 1H, J = 4 Hz, 7 Hz), 3.97 (t, 2H,
J = 7 Hz), 3.79 (m, 2H), 3.35 (s, 9H), 3.28 (s, 6H), 2.32 (t, 2H,
J = 8 Hz), 2.25 (t, 2H, J = 9 Hz), 1.62 (m, 2H), 1.24 (24H),
0.86 (t, 3H, J = 8 Hz). Acetal hydrolysis was catalyzed by an
acidic resin (31). Thus,
1-palmitoyl-2-(5-dimethoxypentanoyl)-sn-glycero-3-phosphocholine (30 mg) was dissolved in 6 ml of acetone/water (6:1, v/v) and stirred
magnetically with Amberlyst-15 resin (15 mg). The reaction was
monitored by TLC and was stopped at 6.5 h. The solvent was removed
under reduced pressure, and the crude product was flash chromatographed
on a column (10 mm × 150 mm) of silica gel (40 µm, J. T. Baker
Inc.) with CHCl3/MeOH/H2O (40:50:10, v/v/v) to yield
1-palmitoyl-2-(5-oxovaleryl)-sn-glycero-3-phosphocholine (POVPC, 21 mg, 76%). 1H NMR: d 9.78 (s, 1H), 5.23 (m, 1H),
3.85-4.45 (8H), 3.37 (s, 9H), 2.57 (t, 2H, J = 7 Hz), 2.39 (m,
2H), 2.28 (t, 2H, J = 7 Hz), 1.92 (t, 2H, J = 7 Hz), 1.58 (m,
2H), 1.26 (24H), 0.88(t, 3H, J = 7 Hz). This spectrum agrees
closely with that reported previously (32).
For synthesis of
1-palmitoyl-2-glutaryl-sn-glycero-3-phosphocholine (PGPC),
1-palmitoyl-2-hydroxy-sn-glycero-3-phosphocholine (50 mg,
0.1 mmol) in a 15-ml flask was dried by azeotropic distillation with
toluene (3 × 5 ml) under reduced pressure and then was connected to high vacuum (0.25 mm Hg) for 3 h. After adding glutaric
anhydride (33.7 mg, 0.3 mmol),
p-(N,N-dimethylamino)pyridine (18.5 mg, 0.15 mmol), and 3.5 ml of freshly distilled chloroform, the above
flask was flushed with argon and the resulting solution was stirred at
room temperature. The progress of the reaction was monitored by TLC
(CHCl3/MeOH/H2O, 70:26:4, v/v/v). After 21 h the reaction mixture was acidified to pH 3.0 with cold 0.1 N HCl, extracted with CHCl3/MeOH (2:1, v/v,
3 × 30 ml), and washed with MeOH/H2O (pH 3.0, 1:1,
v/v, 30 ml). The combined organic phase was evaporated under reduced
pressure, and the resulting mixture was applied to a silica gel column
(15 mm inner diameter × 200 mm), which was eluted with
CHCl3/MeOH/H2O (40:50:10, v/v/v) to produce
PGPC. 1-Palmitoyl-2-glutaryl-sn-glycero-3-phosphocholine:1H
NMR (in CHCl3): These studies were performed
essentially as described previously (14). Briefly, for monocyte
adhesion assays human aortic endothelial cells were incubated with test
medium for 4 h at 37 °C. A suspension of human monocytes was
added for 12-15 min, and nonadherent monocytes were removed. Bound
monocytes were counted and expressed as monocytes/microscopic field.
Data were analyzed using model 1 analysis of variance.
Monoclonal autoantibodies that recognized
oxidation-specific epitopes of oxidized LDL were cloned from hybridomas
generated from apolipoprotein E-deficient (ApoEØ) mice that had not
been immunized exogenously as described (33). The supernatants from hybridomas produced by fusing B-lymphocytes obtained from the spleens
of two ApoEØ mice with a myeloma cell line were screened for binding
to either highly oxidized LDL, 4-hydroxynonenal-LDL, or
malondialdehyde-modified LDL. Fractions obtained from LC-MS analysis of
Ox-PAPC were dried under nitrogen at 37 °C and resuspended in 500 µl of absolute ethanol, and 25 µl of a 1:100 dilution in ethanol
was added to each well of a 96-well white round bottomed MicroFluor
(Dynatech) microtitration plate. Absolute ethanol was added to blank
wells. The ethanol was evaporated to dryness over 10-15 min under a
gentle stream of nitrogen. Purified murine monoclonal autoantibodies
were diluted into Tris-buffered saline buffer with 3% bovine serum
albumin, 0.27 mM EDTA, and 20 µM BHT
(dilution buffer). After adding 50 µl of solutions containing the
primary antibody per well (10 µg/ml), the plates were incubated at
room temperature for 1 h. The wells were washed four times with
washing buffer (Tris-buffered saline containing 0.27 mM
EDTA, 0.02% NaN3, and 0.001% aprotinin), and the amount
of antibody bound was measured with alkaline phosphatase-labeled goat
anti-mouse IgM (Sigma) by incubating for 1 h at room temperature.
After four more washes with washing buffer, plates were washed four
times with distilled water. To each well was then added 25 µl of a
50% solution of LumiPhos 530, and the plates were incubated for 1 h at room temperature in the dark. Luminescence was determined using a
Lucy 1 Luminometer and WINLCOM software (Anthos Labtec Instruments,
Salzburg, Austria). Binding to native LDL and modified LDL were
determined as described (34). Data are expressed as total flashes of
light measured over the indicated times.
Measurements of cell protein content and
those of lipoproteins were performed by microtiter plate assay based on
the method of Lowry et al. (35).
HPLC and Spectral Characteristics of Unoxidized and Oxidized
PAPC
PAPC was analyzed by normal phase HPLC before and after
autoxidation by exposure to air for 48 h. The eluant was passed
through the flow cell of a diode array detector that was scanning
between 200 and 360 nm. Unoxidized PAPC eluted at ~14 min and
possessed maximal absorbance in the 200-205-nm range typical of
polyunsaturated hydrocarbon chains (Fig. 1A).
Autoxidation of PAPC resulted in the formation of a complex array of
compounds with differing chromatographic and spectral characteristics
(Fig. 1B). A peak of absorbance at 200 nm near 14 min (Fig.
1B, asterisk) suggested the presence of residual
unoxidized PAPC. Several peaks were detected after autoxidation that
had different HPLC retention times and absorbed at wavelengths greater
than 215 nm (Fig. 1B).
Electrospray Ionization Mass Spectrometry of Oxidized PAPC
Analysis of unoxidized PAPC by FIA-ESI-MS in the positive ion mode
revealed a single predominant ion corresponding to the protonated
molecule (M+H+) at m/z 782.7 with few
impurities (Fig. 2A). Positive ion FIA-ESI-MS of Ox-PAPC confirmed the HPLC analysis showing that a variety of
oxidation products derived from PAPC were present in Ox-PAPC (Fig.
2B). These derivatives included molecules that had a larger molecular mass than PAPC (m/z > 782.7),
presumably due to the addition of oxygen atoms to the arachidonoyl
moiety of PAPC, and molecules that had a smaller molecular mass than
PAPC (m/z < 782.7), presumably due to
oxidative fragmentation of PAPC. To verify that the observed ions were
not fragmentation artifacts produced by the ionization process, Ox-PAPC
was analyzed while the orifice voltage was increased from 60 to 120 at
20-unit increments. As the orifice voltage increased there was a
uniform decrease in the abundance of all ions except for
1-palmitoyl-lysophosphatidylcholine (M+H+ = 496) and
phosphocholine (M+H+ = 183), which increased proportionally
to orifice voltage (data not shown). This was presumably due to
nonspecific fragmentation at the sn-2 and sn-3
positions, respectively, by the ionization energy. In addition, fast
atom bombardment mass spectrometry was performed and gave essentially
identical ions in terms of m/z and relative
abundance as those seen with positive ion FIA-ESI-MS (data not
shown).
Biological Activity of Lipids Isolated from Ox-PAPC
Our laboratory has previously shown that endothelial cells treated
for 4 h with Ox-PAPC bound significantly more human monocytes than
did endothelial cells treated with unoxidized PAPC (19). A LC-MS system
was assembled that provided simultaneous diode array detection, mass
spectrometry, and fraction collection of oxidized phospholipids
isolated by HPLC. Ox-PAPC (3 mg) was injected onto this system using a
preparative NP-HPLC column, and each 1-min fraction was collected and
dried under a stream of nitrogen at 37 °C. For biological activity
analysis, fractions were resuspended by vortexing for 1 min in tissue
culture medium that contained 5% serum and then added to endothelial
cells for 4 h at 37 °C. After incubation, the lipids were
removed from the endothelial cells and human monocytes were added for
12-15 min at 37 °C. Nonadherent monocytes were removed, and
adherent monocytes were quantitated by light microscopy. Endothelial
cells treated with oxidized phospholipids collected in fractions 16, 18, 19, 20, and 22 bound significantly more monocytes than did
sham-treated endothelial cells (Fig. 3A). Biological activity was closely associated with the elution of oxidized
phospholipids with m/z 594-595, 610-611, and
828-829 as determined by reconstructed selected ion chromatograms
(RSIC) for these ions (Fig. 3B).
We then analyzed phospholipids extracted from MM-LDL using LC-MS to
determine if the same biologically active phospholipids were present.
Phospholipids with similar mass and retention time were found in
phospholipids isolated from MM-LDL compared with those that possessed
biological activity in Ox-PAPC (Fig. 3C). In addition, the
molecule(s) with m/z 828-829 eluted as a bimodal peak at ~13.3 and 15 min from both Ox-PAPC and MM-LDL (Fig. 3, B and C). The molecules with
m/z 594.3, 610.2, and 828.6 were purified to near
homogeneity by preparative normal phase LC-MS followed by analytical
reversed phase LC-MS (RP-LC-MS) for use in monocyte binding assays
(Fig. 4A). Purification of each molecule was
based on ions observed in each fraction using positive ion FIA-ESI-MS.
Endothelial cells treated with purified m/z
594.3, 610.2, and 828.6 bound significantly more monocytes than did
cells treated with medium alone (Fig. 4B). Dose-response
curves for individual molecules were difficult to obtain because of the
variability in the response of endothelial cells and monocytes from
different donors. However, in most endothelial cell isolates all three
molecules induced maximal response at concentrations of about 20-30
µg/ml (data not shown). Dilution experiments showed that
m/z 594.3 and 610.3 were capable of inducing a
significant increase in monocyte adhesion to endothelial cells at a
concentration as low as 1-3 µg/ml. The molecule with
m/z 828.6 was capable of inducing monocyte binding as low as 400 ng/ml, if it is assumed that this molecule was
exclusively responsible for the biological activity of the fraction.
However, contaminating ions present in this fraction (Fig.
4A) were more abundant in fractions with less biological activity (data not shown).
Electrospray Mass Spectrometry of Derivatized Ox-PAPC
Ox-PAPC (250 µg) was added to each of four test tubes and the
solvent dried under nitrogen. The lipid residue in each tube was
treated with either buffer alone, sodium borohydride, methoxyamine hydrochloride, or a mixture of pentafluorobenzyl bromide and
diisopropylethylamine to derivatize particular functional groups such
as hydroperoxides, aldehydes, ketones, and carboxylic acids that may
have been present in our Ox-PAPC preparations. The lipids from each
sample were re-extracted with chloroform/methanol (2:1, v/v) as
described under "Experimental Procedures" and analyzed by positive
ion FIA-ESI-MS. It has been well documented that sodium borohydride can
reduce hydroperoxides, epoxides, and carbonyls to hydroxides. Reduction of hydroperoxides to hydroxides results in a net decrease of 16 Da due
to the loss of a single oxygen atom. Reduction of aldehydes and ketones
to hydroxides results in a net increase of 2 Da due to the addition of
two hydrogen atoms to the carbonyl-containing carbon.
The mass spectrum of
Ox-PAPC treated with buffer alone (Fig. 5A)
was similar to the mass spectrum of Ox-PAPC that had not been incubated
with buffer (see Fig. 2B). The mass spectrum of Ox-PAPC
treated with the reducing agent, sodium borohydride (Fig. 5B), was significantly different from that of Ox-PAPC
treated with buffer alone (Fig. 5A). The ions observed at
m/z 594.3, 620.4, 650.4, 828.6, 846.6, 862.5, and
878.7 were diminished, and the ions observed at
m/z 596.4, 622.5, 652.5, 832.5, 848.7, 864.6, and
880.5 became predominant (Fig. 5B). This result indicated that many components of Ox-PAPC contain reducible functional groups. We
suspect that the increase in 1-palmitoyl-lysophosphatidylcholine (m/z 496.2) was due to nonspecific hydrolysis of
the fatty acid at the sn-2 position during the workup
procedure. This may account for the decrease of some ions after
derivatization, such as m/z 732.6 (Fig. 5). With
regard to the molecules in Ox-PAPC that possessed biological activity
(m/z 594.3, 610.2, and 828.6), the ions at m/z 594.3 and 828.6 were diminished and the ion
at m/z 610.2 was not altered significantly (Fig.
5B). It was concluded that the former ions possessed sodium
borohydride-reducible groups and the latter ion did not. These data
suggested that the molecule with m/z 594.3 was
reduced to m/z 596.4 (presumably due to the reduction of an aldehyde or ketone) and the molecule with
m/z 828.6 was reduced to 832.5 (presumably due to
the reduction of two carbonyl groups). Alternatively, the
m/z 832.5 ion could have been produced by the
reduction of a molecule with m/z 846.6 that possessed a carbonyl group and a hydroperoxide. Reduction of a hydroperoxyl group (
Based on the molecular mass
of the m/z 610.2 ion we hypothesized that it was
an oxidatively fragmented phospholipid with a 5-carbon moiety at the
sn-2 position containing a terminal carboxylic acid
(glutaric acid) that would produce a M+H+ ion at
m/z 610.2 (Fig. 6A,
structure). To test this hypothesis, the molecule with
m/z 610.2 was partially purified by NP-LC-MS (Fig. 6A) and then treated with pentafluorobenzyl bromide
and N,N-diisopropylethylamine, which are commonly
used to derivatize carboxylic acids to pentafluorobenzyl esters. After
treatment, positive ion FIA-ESI-MS revealed that the ion at
m/z 610.2 had diminished and a novel ion at
m/z 790.2 had appeared (Fig. 6B). This
observation was consistent with the addition of a pentafluorobenzyl group (180 Da) to the ion at m/z 610.2 (Fig.
6B, structure). The attachment of the
pentafluorobenzyl group to the 5-carbon group at the sn-2
position was confirmed by FIA-ESI-MS/MS in the negative ion mode. The
ion (M+H+) observed at m/z 790.2 by
positive ion FIA-ESI-MS produced a pseudo-ion (M-15) at
m/z 774.2 by negative ion FIA-ESI-MS due to loss
of a methyl group from the choline moiety of the molecule (36).
Prominate daughter fragments produced by collision-induced dissociation
of this pseudo-parent ion (m/z 774.2) was the
M-15 ion of 1-palmitoyl-lysophosphatidylcholine
(m/z 480), the carboxylate anion of palmitic acid
(m/z 255), and a third fragment with
m/z 311 (Fig. 6C). The latter fragment
(m/z 311) corresponded to the calculated
molecular mass of the carboxylate anion of a 5-carbon pentafluorobenzyl
ester that was predicted to be in the sn-2 position of this
molecule. Therefore, the molecular structure of one biologically active
oxidized phospholipid in Ox-PAPC and MM-LDL was confirmed to be
1-palmitoyl-2-glutaryl-sn-glycero-3-phosphocholine (PGPC) (Fig. 6A, structure). This assignment was
confirmed by comparative analysis of PGPC prepared by unambiguous total
synthesis, which had properties identical to those of the identified
molecule when analyzed by positive ion FIA-ESI-MS, NP-LC-MS, and
RP-LC-MS.
To determine the molecular
structure of the molecule with m/z 594.3, the
oxidized phospholipid was semipurified by NP-LC-MS (Fig.
7A) and treated with methoxyamine
hydrochloride. After treatment, analysis by positive ion FIA-ESI-MS
revealed that the ion at m/z 594.3 was diminished
and a novel ion at m/z 623.4 was observed (Fig.
7B). This result was consistent with the addition of a
methoxyamine group (29 Da) to the m/z 594.3 ion.
We hypothesized, based on molecular mass and presence of a carbonyl
group, that the molecule that produced the ion at
m/z 594.3 was an oxidatively fragmented phospholipid containing palmitic acid in the sn-1 position
and a 5-carbon chain with a terminal aldehyde (5-oxovaleric acid) in
the sn-2 position (Fig. 7A,
structure). To test this, negative ion FIA-ESI-MS/MS was
performed to determine daughter fragments of the
methoxyamine-derivatized molecule that produced the ion at
m/z 623.1 (M+H+) by positive ion
FIA-ESI-MS and the pseudo-ion at m/z 607 (M-15) by negative ion FIA-ESI-MS. Daughter fragments of the pseudo-parent ion
at m/z 607 included the M-15 ion for
1-palmitoyl-lysophosphatidylcholine (m/z 480),
the carboxylate anion from palmitic acid (m/z
255), and a fragment at m/z 144 that represented
the carboxylate anion of the methoxyamine-derivatized 5-carbon moiety
in the sn-2 position. The fragment at
m/z 549 represented the pseudo-parent ion after the loss of the methoxime group with the terminal carbon
(HC=N-O-CH3, 607-58 Da). The fragment at
m/z 113 represented the derivatized carboxylate
anion from the sn-2 position after the loss of the methoxy
group (O-CH3) (144-31 Da). Thus, we conclude that the molecular structure of a second molecule present in Ox-PAPC and MM-LDL
that induces endothelial cells to bind monocytes in vitro is
POVPC. This assignment was confirmed by comparative analysis to
1-palmitoyl-2-(5-oxovaleroyl)-sn-glycero-3-phosphocholine
(POVPC) prepared by unambiguous total synthesis, which had properties identical to those of the identified molecule when analyzed by positive
ion FIA-ESI-MS, NP-LC-MS, and RP-LC-MS.
Biological Activity of Synthesized Oxidized Phospholipids
Totally synthetic POVPC and PGPC were dried, resuspended in tissue
culture medium, and analyzed for their ability to induce endothelial
cells to bind monocytes. As did the same molecules produced by
oxidation of PAPC, the synthesized phospholipids caused a significant
increase in monocyte binding (Fig. 8). The dose-response curves for both synthetic and HPLC-isolated compounds were similar when
identical endothelial cells and monocytes were used (data not
shown).
Reactivity of Oxidized Phospholipids with Monoclonal
Antibodies
LDL, MM-LDL, PAPC, Ox-PAPC, and HPLC-isolated fractions from
Ox-PAPC were tested for reactivity to antibodies produced from hybridomas generated from apoE null mice. Of the nine autoantibodies tested, four (EØ-1, EØ-3, EØ-7, and EØ-14) displayed specific reactivity to Ox-PAPC compared with unoxidized PAPC and bovine serum
albumin (Fig. 9). Antibody EØ-6, which has been shown
to recognize an oxidized phospholipid epitope (34), also had increased binding to Ox-PAPC as well as to unoxidized PAPC and bovine serum albumin (data not shown), presumably due to epitopes already present. In general, the same antibodies that reacted specifically with Ox-PAPC
also recognized epitopes on MM-LDL and not the native LDL from which it
was derived (Fig. 9, inset). In control experiments, reactivity of these antibodies with bovine serum albumin was
essentially the same as reactivity with unoxidized LDL (data not
shown). Two of these antibodies (EØ-1 and EØ-3) were used in further
studies to determine immunoreactivity of HPLC-isolated fractions
collected from Ox-PAPC. Fractions were collected from Ox-PAPC (3 mg)
during LC-MS analysis on a preparative normal phase HPLC column. After collection the fractions were dried under nitrogen at 37 °C and resuspended in chloroform prior to analysis. Immunoreactivity of each
fraction that contained lipid (fractions 9-30) was determined and
expressed in terms of flashes/100 ms (Fig. 10). After
the level of reactivity of each fraction was determined the RSIC of the molecules with m/z 594-595, 610-611, and
828-829 were superimposed on the graph. In a separate experiment,
EØ-6 and EØ-7 were shown to react with the fractions containing ions
at m/z 828.6 and 610.2 (equivalent to fractions
13-19 in Fig. 3) but not fractions containing the ion at
m/z 594.3.
Presence of Biologically Active Oxidized Phospholipids in
Atherosclerotic Lesions
Phospholipids extracted from the aortae of rabbits on atherogenic
diets with or without supplementation with antioxidants were analyzed
by positive ion FIA-ESI-MS. The aortae from rabbits that were not
supplemented with antioxidants had a higher concentration of
phospholipids with characteristics identical to those of the biologically active oxidized phospholipids found in Ox-PAPC and MM-LDL
compared with aortae from rabbits that were on a chow diet (control) or
rabbits that were fed an atherosclerotic diet that was supplemented
with antioxidants (Fig. 11).
Our previous studies have suggested that the ability of MM-LDL to
induced monocyte-endothelial interactions is due to the oxidation of
arachidonic acid-containing phospholipids in LDL. The focus of these
studies was to identify the molecular structure of these lipids using
oxidized PAPC as a surrogate lipid. The products derived from the
autoxidation of PAPC were remarkably reproducible (Fig. 2). Only slight
variations were observed between preparations with regard to the time
necessary for oxidation, the retention time by normal and reversed
phase HPLC, and the relative abundance of each ion by positive ion
FIA-ESI-MS.
Although numerous oxidation products of PAPC were present in
preparations of Ox-PAPC (Fig. 2B) we concentrated on three
ions, m/z 594.3, 610.2, and 828.6, based on the
observation that NP-HPLC fractions that contained these species
possessed biological activity in monocyte binding assays. All three of
these species were present in MM-LDL. RSIC produced from normal phase
LC-MS analysis of phospholipids extracted from MM-LDL showed a similar
pattern and retention time to the oxidized phospholipids found in
Ox-PAPC (Fig. 4, B and C). These lipids were also
found in preparations of MM-LDL produced by incubation of LDL with
cocultures of human endothelial cells and smooth muscle cells (data not
shown).
We have observed that the presence of these particular species of
oxidized phospholipids appear transiently during the oxidation of LDL.
They appear to be most abundant during the late antioxidant-depletion phase and the early propagation phase of LDL oxidation (15, 19). This
may help to explain the transient ability of oxidized LDL to induce
monocyte-endothelial interactions, since the appearance of these
oxidized phospholipids and the presence of biological activity occur in
the same "window" of oxidation (9). We hypothesize that during the
late propagation and decomposition phase of LDL oxidation the
accumulation of end products of lipid oxidation such as
malondialdehyde, 4-hydroxynonenal, cholesterol oxides, and
lysophosphatidylcholine become the principle "biological
activators." It is these latter products that appear to be
responsible for protein modification (37-39), cell toxicity (40-42),
and intimal proliferation (43, 44) associated with highly oxidized
LDL.
The identification of the molecular structures of two of the oxidized
phospholipids present in Ox-PAPC (m/z 594.3 and
610.2) was relatively straightforward. The molecule with
m/z 594.3 was reduced by sodium borohydride (Fig.
6) and was derivatized by methoxyamine (Fig. 8). Based on the molecular
mass of this ion and the fragmentation pattern by negative ion
FIA-ESI-MS/MS the molecule was identified as POVPC. The molecule with
m/z 610.2 was not reduced by sodium borohydride
(Fig. 5) but was derivatized by pentafluorobenzyl bromide (Fig. 6).
Based on the molecular mass of this ion and the fragmentation pattern
by negative ion FIA-ESI-MS/MS the molecule was identified as PGPC. Both
of these assignments were confirmed by total unambiguous synthesis of
POVPC and PGPC. The synthetic molecules had identical analytical and biological properties compared with the molecules isolated from Ox-PAPC
(Fig. 8).
Tokumura et al. have reported that PGPC, as part of a lipid
extract from bovine brain (45) and as a pure compound (46), possessed
hypotensive properties. Others have reported the presence of
5-oxovaleric acid esterified to phospholipids and cholesteryl esters in
LDL (47, 48). In addition, POVPC has been shown to induce smooth muscle
cell proliferation and neutrophil activation in vitro (49,
50). This molecule may contribute to the platelet aggregation effects
associated with oxidized LDL, since it has structural similarity to
platelet-activating factor (51). Interestingly, the biological activity
of Ox-PAPC was decreased by 60-70% when treated with sodium
borohydride (data not shown). This suggests that particular functional
groups were necessary for the biological activity of Ox-PAPC. The
structural identification of the molecule with
m/z 828.6 was found to be more difficult. The
molecular composition of this complex oxygenated phospholipid is
currently under investigation.
There are several lines of evidence suggesting that the oxidized
phospholipids with m/z 594.3, 610.2, and 828.6 play a role in atherosclerosis in vivo. First, there is an
association between the extent of atherosclerotic lesions and the
abundance of these lipids in an animal model of atherosclerosis (Fig.
11). Second, the same lipids that induced endothelial cells to bind
monocytes are recognized by naturally occurring autoantibodies
immunoreactive with oxidized LDL epitopes. In addition, the same murine
autoantibodies that recognized several of the oxidation products of
PAPC also immunostain atherosclerotic lesions of rabbits and humans
(33). Oxidized LDL is immunogenic, and autoantibodies recognizing
epitopes of highly oxidized LDL have been described in plasma and in
atherosclerotic lesions of several species (52, 53). Moreover, higher
titers of these autoantibodies are present in patients with carotid
atherosclerosis, coronary artery disease, diabetes, peripheral vascular
disease, hypertension, and pre-eclampsia (54-61). These observations
suggest that under conditions of oxidative stress epitopes are produced that induce an immune response. It is not clear, however, whether the
epitopes that are recognized by the autoantibodies are intrinsic to the
oxidized phospholipids alone or are formed by oxidized lipid-protein
adducts. The significance of these antibodies in limiting the
development of atherosclerosis is shown by the observation that
injection of malondialdehyde-modified LDL (62) or oxidized LDL (63)
into rabbits before the initiation of cholesterol feeding caused a
reduction in fatty streak lesion formation. Thus, the results of our
studies together with previous observations cited above suggest that
the active derivatives of PAPC that we have identified may play an
important role in vivo in fatty streak formation.
The mechanism by which these oxidized phospholipids induced monocyte
binding is not known. We have hypothesized that all three molecules may
produce a common motif as part of a lipid monolayer or bilayer.
Thermodynamic considerations suggest that polar functional groups such
as hydroperoxides, hydroxides, and carbonyls associated with the fatty
acid moieties of phospholipids may engage in hydrogen bonding with the
aqueous medium in which the membrane is suspended. The
oxygen-containing acyl group at the sn-2 position of these molecules has the potential to protrude into the aqueous medium. This
has been suggested to be the mechanism by which phospholipase A2 preferentially hydrolyzes oxidized phospholipids (64,
65). We are investigating the possibility that this motif is also
responsible for interaction with particular receptors on the
endothelial cell surface.
The use of Ox-PAPC as a surrogate for MM-LDL greatly facilitates the
study of monocyte-endothelial interactions induced by oxidized
phospholipids. The active derivatives that we have identified probably
do not represent all of the active phospholipid oxidation products
present in MM-LDL or in the vessel wall. The synthesis of biologically
active oxidized phospholipids in conjunction with the powerful
analytical techniques of chromatography combined with mass spectrometry
should provide methods by which the metabolism of oxidized
phospholipids may be carefully dissected. Oxidized phospholipids have
been implicated in the pathogenesis of several chronic inflammatory
diseases such as rheumatoid arthritis (66, 67), inflammatory bowel
disease (68), antiphospholipid antibody syndrome (34), and multiple
sclerosis (69, 70). We are investigating the possibility that the
biologically active phospholipids reported in this paper may play a
fundamental role in the pathogenesis of these diseases.
The authors thank Drs. George Popják,
Linda L. Demer, A. Jake Lusis, Farhad Parhami, Peter A. Edwards, Jason
D. Morrow, Sampath Parthasarathy, Tom McIntyre, Christopher Fielding,
Richard Havel, John Taylor, and Michael Reidy for valuable discussions
and suggestions and Alan C. Wagner and Susan Hama for discussions and
expert technical assistance.
Department of Medicine,
Department
of Medicine, University of California, San Diego, California
92093-0682, the ** Department of Pathology, Bowman Gray University
School of Medicine at Wake Forest, Winston-Salem, North Carolina
27157-1072, and the
Department of
Chemistry, Case Western Reserve University, Cleveland, Ohio
44106-7027
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
-carotene, ubiquinol) are
depleted; the propagation phase, during which polyunsaturated fatty
acids and cholesterol become oxidized to become lipid hydroperoxides
and oxycholesterol, respectively; and the decomposition phase, during
which oxidized lipids undergo fragmentation to form aldehydes capable
of modifying amino groups on proteins and phospholipids. Minimally
modified LDL (MM-LDL) is LDL that has undergone antioxidant depletion
and oxidation of arachidonic acid-containing phospholipids, much less
linoleic acid oxidation, with little or no protein modification (14, 15). MM-LDL has unique biological properties that are not associated with native LDL or highly oxidized LDL, which may be involved in the
recruitment and entry of monocytes into the vessel wall and their
subsequent maturation into macrophages (14, 16-18). We have previously
shown that oxidized phospholipids isolated from MM-LDL, as well as a
mixture of products formed from the autooxidation of
1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine (Ox-PAPC), mimic the biological activity of MM-LDL (19). There is also
evidence that high density lipoproteins may inhibit
monocyte-endothelial interactions induced by MM-LDL by altering the
molecular structure of these biologically active oxidized phospholipids
(15, 19). In this report we combine liquid chromatography with
electrospray tandem mass spectrometry (MS/MS) to elucidate the
structural identification of biologically active phospholipids in
Ox-PAPC and confirm their presence in MM-LDL. In addition, we present
evidence for the presence of these oxidized phospholipids in animal
models susceptible to fatty streak formation and atherosclerosis.
Materials
-1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine (PAPC) was obtained from Avanti Polar Lipids, Inc. (Alabaster, AL) or
Sigma, and
L-
-1-palmitoyl-2-[1-14C]-arachidonoyl-sn-glycero-3-phosphocholine
was purchased from DuPont NEN. Aminopropyl solid phase extraction
columns were obtained from J. T. Baker Inc.
-tocopherol
acetate/day/rabbit) or probucol (0.41 g/day/rabbit). After completion
of the dietary regimen the rabbits were euthanized and the aortae were
removed for the assessment of the extent of atherosclerosis.
Immediately after sacrifice, the entire aorta from the aortic arch to
just distal of the iliac bifurcation was removed and washed free of blood with 0.15 M phosphate buffer (pH 7.4). The aortae
were freed from adventitial debris, opened longitudinally, pinned flat,
and photographed. The aortic arch was separated from the descending thoracic aorta 1-2 mm distal to the ductus scar, and the descending thoracic aorta was separated from the abdominal aorta 1-2 mm proximal to the celiac artery. Aortic specimens were weighed, placed in a test
tube, covered with argon or nitrogen, and frozen at
70 °C protected from light prior to analysis.
44 °C.
5, 1 × 10
4, and 2 × 10
4 M, respectively) in
H2O/methanol (50:50, v/v) containing 2 mM ammonium formate and 0.1% acetonitrile. Calibration across the m/z range 10-2400 was effected by multiple ion
monitoring of eight polypropylene glycol positive ion solution signals
(typically the singly charged ions at m/z 58.99, 326.25, 334.30, 906.67, 1254.92, 1545.13, 1863.34, 2010.47, and the
doubly charged ion at m/z 520.4). The ion spray
voltage was typically operated at 4.5 kV in positive mode and 3.5 kV in
negative mode with 40 p.s.i. "zero"-grade air produced by a
Zero Air Generator (Peak Scientific, Chicago, IL) at 0.6 liter/min. A
curtain gas of N2 produced from the vapors of liquid
N2 was used at a flow rate of 0.8 liter/min. Normal and
precursor ion mass spectra were acquired by scanning quadrupole Q1.
Daughter ion spectra were obtained in the negative ion mode by
colliding the quadrupole Q1-selected pseudo-parent ion of interest
produced at an orifice voltage of
105 with a mixture of 5.2%
nitrogen in argon in quadrupole Q2 and scanning quadrupole Q3 to
analyze the fragment ion products. Normal scan spectra were recorded at
instrument conditions sufficient to resolve the isotopes of the
polypropylene glycol/NH4+ singly charged ion at
m/z 906 (40% valley definition). For precursor ion tandem MS/MS the analyzing quadrupole (Q3) and the mass selecting quadrupole (Q1) were operated at a resolution of about unit
m/z (using a 35% peak valley definition at low
m/z and a 60% peak valley definition at high
m/z). Reconstructed selected ion chromatograms were produced by software supplied by Perkin-Elmer. Phospholipids were
introduced into the electrospray mass spectrometer by flow injection
analysis or liquid chromatography (LC). For flow injection analysis
(FIA), 20 µg of phospholipid was resuspended in 20 µl of FIA
solvent consisting of acetonitrile/water/formic acid (50:50:0.1, v/v/v)
and injected into FIA solvent flowing at a rate of 25 µl/min into the
ion source of the mass spectrometer. The mass spectrometer was set to
scan from m/z 450-950 with an orifice voltage of
+65, a step size of 0.3, a dwell time of 3 ms, and a scan speed of ~4
s. For negative ion electrospray tandem mass spectrometry (step size = 1.0) a solvent of 100% methanol with 1 mM
ammonium acetate was used. Quantitative analysis of phospholipids was
performed essentially as described by Han et al. (25) using
dimyristoylphosphatidylcholine as an internal standard. Phospholipids
were quantitated based on their ion intensity relative to the internal
standard using positive ion FIA-ESI-MS. The equimolar ion intensities
of different species of phosphatidylcholines typically varied by less
than 5% (25).
5.25 (m, 1H), 3.75-4.35 (m, 8H), 3.32 (s, 9H), 2.38 (m, 2H), 2.27 (t, 4H, J = 8 Hz), 1.87 (m, 2H), 1.58 (m, 2H), 1.26 (m, 24H), 0.87 (t, 3H) J = 8 Hz).
Fig. 1.
HPLC chromatogram of unoxidized PAPC
(A) and Ox-PAPC (B). After autoxidation 3 mg of Ox-PAPC was injected onto a preparative normal phase HPLC column
and eluted isocratically with acetonitrile/methanol/water (77:8:15,
v/v/v) containing 1 mM ammonium acetate (pH 5.0) at 5 ml/min. UV absorbance was monitored using a diode array detector between 200 and 360 nm. Absorbance of Ox-PAPC is shown at 200 nm (),
235 nm (
-
), 250 nm (·····), and 270 nm (
- -
). Full scale absorbance (100% relative absorbance) was 0.900 absorbance unit, 0.370 absorbance unit, 0.370 absorbance unit, and 0.370 absorbance unit, respectively.
[View Larger Version of this Image (21K GIF file)]
Fig. 2.
Positive ion FIA-ESI-MS of unoxidized PAPC
(A) and Ox-PAPC (B). Unoxidized PAPC (250 µg/ml) was introduced into the mass spectrometry ion source dissolved
in flow injection solvent composed of acetonitrile/water/formic acid
(50:50:0.1, v/v/v) at 20 µl/min. Ox-PAPC, produced by autoxidation in
air for 24-72 h, was resuspended in flow injection solvent (1 mg/ml),
and 20 µl (20 µg) was analyzed as described for unoxidized
PAPC.
[View Larger Version of this Image (20K GIF file)]
Fig. 3.
Normal phase liquid chromatography/mass
spectrometry of biologically active oxidized phospholipids in Ox-PAPC
and MM-LDL. Ox-PAPC (3 mg) was analyzed by preparative LC-MS, and
the fractions eluting between 12 and 24 min were collected, dried under
argon, resuspended in tissue culture medium, and tested for their
ability to induce human aortic endothelial cells to bind monocytes
(A). p values were obtained by comparison to
control wells treated without exogenous lipids (n = 9).
Biological activity was expressed as monocytes bound to endothelial
cells above control as an average of three separate experiments, and
p values are shown compared with the number of monocytes
bound to untreated endothelial cells. A RSIC was prepared showing the
ions that colocalized with biological activity
(m/z 828.6, 610.2, and 594.3) (B). The
100% ion intensity for each RSIC is indicated in parentheses under
each ion label. Phospholipids isolated from MM-LDL by solid phase
extraction columns were analyzed by LC-MS in the same manner as above,
and RSIC of m/z 828.6, 610.2, and 594.3 are shown
(C).
[View Larger Version of this Image (24K GIF file)]
Fig. 4.
Induction of monocyte binding by purified
oxidized phospholipids. Fractions collected after sequential
normal and reversed phase HPLC were dried under nitrogen, resuspended
in tissue culture medium, and incubated with human aortic endothelial
cells. After 4 h of incubation the medium was removed and human
monocytes were added for 15 min. Unbound monocytes were removed after
incubation, and adherent monocytes were quantitated by light
microscopy. The mean number of monocytes bound in control wells
(n = 9) was subtracted from all values, which are
plotted as mean ± S.E.
[View Larger Version of this Image (16K GIF file)]
16 Da) and a carbonyl group (+2 Da) would also
produce an ion near m/z 832.5 (846.6
16.0 + 2.0 = 832.6).
Fig. 5.
Reduction of Ox-PAPC. Ox-PAPC was
incubated with sodium borohydride (20 µg/ml) in a 0.1 M
borate buffer (pH 8.0) or borate buffer alone for 15 min at room
temperature. After treatment the lipids were extracted and sham-treated
(A) and reduced (B). Ox-PAPC was analyzed by
positive ion flow injection electrospray mass spectrometry.
[View Larger Version of this Image (28K GIF file)]
Fig. 6.
Pentafluorobenzyl derivatization of
phospholipid with m/z 610.2. Ox-PAPC (3 mg) was analyzed by preparative normal phase LC-MS as described under
"Experimental Procedures." The fraction enriched in the
phospholipids with m/z 610.2 was dried and
treated with buffer alone (A) or pentafluorobenzyl bromide and diisopropylethylamine (B) and analyzed by positive ion
electrospray MS. Negative ion electrospray collision-induced
dissociation-MS/MS was also performed on the pseudo-ion at
m/z 774 (M-15) produced from the derivatized
molecule with M+H+ = 790.2 (C). Proposed
molecular structures and fragmentation patterns are inset in
each panel.
[View Larger Version of this Image (20K GIF file)]
Fig. 7.
Methoxyamine derivatization of phospholipids
with m/z 594.3. Ox-PAPC (3 mg) was
analyzed by preparative normal phase LC-MS as described under
"Experimental Procedures." The fraction enriched in the
phospholipids with m/z 594.3 was dried and
treated with buffer alone (A) or methoxyamine hydrochloride
(B) and analyzed by positive ion electrospray MS. Negative
ion electrospray collision-induced dissociation-MS/MS was also
performed on the pseudo-parent ion at m/z 607 (M-15) produced from the derivatized molecule with M+H+ = 623.1 (C). Proposed molecular structures and fragmentation patterns are inset in each panel.
[View Larger Version of this Image (20K GIF file)]
Fig. 8.
Biological activity of synthetically
manufactured oxidized phospholipids. The phospholipids
1-palmitoyl-2-(5-oxovaleryl)-sn-glycero-3-phosphocholine (synthetic POVPC) and
1-palmitoyl-2-glutaroyl-sn-glycero-3-phosphocholine (synthetic PGPC) were synthesized and tested for the ability
to induce endothelial cells to bind monocytes.
[View Larger Version of this Image (11K GIF file)]
Fig. 9.
Reactivity of LDL, MM-LDL, and Ox-PAPC with
ApoE monoclonal autoantibodies. Nine monoclonal antibodies
obtained from hybridoma lines obtained from ApoE null mice were tested
for reactivity with oxidized PAPC, Ox-PAPC, or bovine serum albumin
(control), which were adhered to the bottom of 96-well microtiter
dishes. Inset, antibodies EØ-1, EØ-3, and EØ-7 were
tested for reactivity with MM-LDL (black bars) and the LDL
preparations from which the MM-LDL was derived by mild oxidation
(white bars). Values are plotted as mean flashes of
light/100 ms ± S.D.
[View Larger Version of this Image (28K GIF file)]
Fig. 10.
Reactivity of isolated oxidized
phospholipids from Ox-PAPC with ApoE autoantibodies. Ox-PAPC (3 mg) was analyzed by normal phase LC-MS as described under
"Experimental Procedures." Fractions containing lipids (fractions
9-30) were dried under argon, resuspended in ethanol, and added to the
bottom of a 96-well microtiter plate. The solvent was evaporated, and
the lipids were treated with autoantibodies (EØ-1 and EØ-3) produced
by immortalized lymphocyte obtained from ApoE null mice that reacted
most specifically with Ox-PAPC in Fig. 11. A goat anti-mouse secondary
antibody with a fluorescent tag was used to quantitate the amount of
primary antibody attached to the lipids. Values are plotted as mean
fluorescent flashes/ms ± S.D. RSIC were produced for
m/z 828.6 (·····),
m/z 610.2 ( -
), and
m/z 594.3 (
) from the same LC-MS run.
[View Larger Version of this Image (23K GIF file)]
Fig. 11.
Presence of biologically active oxidized
phospholipids in atherosclerotic lesions. Aortae were obtained
from rabbits fed a normal chow diet (Control) or an
atherogenic diet with or without supplementation with the antioxidants
vitamin E or probucol. Lipids were extracted with chloroform/methanol
and phospholipids purified by solid phase extraction chromatography
after addition of the internal standard,
dimyristoylphosphatidylcholine. The phospholipid fraction was analyzed
by positive ion electrospray mass spectrometry, and the abundance of
the ions with m/z 594.3, 610.2, and 828.6 in
aortae from rabbits fed the control diet (white bars,
n = 10), rabbits feed an atherogenic diet (black
bars, n = 10), and rabbits fed an atherogenic diet
supplemented with antioxidants (hatched bars,
n = 14) were plotted in terms of ng/mg tissue. * = p < 0.01 compared with white bars; = p < 0.05 compared with black bars.
[View Larger Version of this Image (30K GIF file)]
*
This work was supported by U.S. Public Health Services
Grants HL 30568, HL 07386, and RR 865 and the Laubisch, Rachel Israel Berro, and M. K. Gray Funds (to A. D. W., N. L., M. N.,
A. M. F., and J. A. B.); a grant from the H. M. Keck Foundation
(to K. F. F.); U.S. Public Health Services Grant HL 14197 (to
J. L. W., S. H., W. P.); and the Academy of Finland (to S. H.).The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
§
To whom correspondence should be addressed: Dept. of
Medicine/Cardiology, Center for the Health Sciences, Rm. 47-123, UCLA Medical Center, Los Angeles, CA 90095-1679. Tel: 310-825-2436; Fax:
310-206-9133; E-mail: adwatson{at}ucla.edu.
1
The abbreviations used are: LDL, low density
lipoprotein; MM-LDL, mildly oxidized low density lipoprotein; Ox-PAPC,
1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine; MS/MS, tandem mass spectrometry; NP, normal phase; RP, reversed phase;
LC, liquid chromatography; FIA, flow injection analysis; ESI,
electrospray ionization; MeOH, methanol; ApoEØ, apolipoprotein E-deficient; BHT, butylated hydroxytoluene; HPLC, high performance liquid chromatography; PAPC,
1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine; PGPC, 1-palmitoyl-2-glutaryl-sn-glycero-3-phosphocholine;
POVPC, 1-palmitoyl-2-(5-oxovaleryl)-sn-glycero-3-phosphocholine;
RSIC, reconstructed selected ion chromatogram.
©1997 by The American Society for Biochemistry and Molecular Biology, Inc.