`‘1:’ 1991 hy The American Society for Biochemistry and Molecular Biology, Inc
`
`Vol. 266, No. 17, Issue of June 15, pp. 11095-11103,1991
`Printed in U. S. A.
`
`Human Plasma Platelet-activating Factor Acetylhydrolase
`OXIDATIVELY FRAGMENTED PHOSPHOLIPIDS
`AS SUBSTRATES*
`
`Kay E. Stremler, Diana M. Stafforini, Stephen M. Prescott, and Thomas M. McIntyreS
`From the Nora Eccles Harrison Cardiovascular Research and Training Institute and the Departments of Internal Medicine and
`Biochemistry. University of Utah, Salt Lake City, Utah 841 12
`
`(Received for publication, June 4, 1990)
`
`phosphocholine, PAF’) is a potent phospholipid autacoid
`Human plasma platelet-activating factor (PAF) ace-
`sn-2 acetyl residue of
`whose actions are mediated by receptors on the surface of
`tylhydrolase hydrolyzes the
`PAF, but not phospholipids with long chain sn-2 resi-
`target cells that specifically recognize PAF. At concentrations
`as low as 10-“’ M it activates platelets, neutrophils, and other
`low density lipoprotein
`dues. It is associated with
`(LDL) particles, and is the LDL-associated phospholi-
`leukocytes, and it
`induces hypotension, increased vascular
`pase A, activity that specifically degrades oxidatively
`permeability, and shock in animals (reviewed in Refs. 1 and
`damaged phospholipids (Stremler, K. E., Stafforini, D.
`2). The potency and nature
`of its effects suggest that the
`M., Prescott, S. M., Zimmerman, G . A., and McIntyre,
`presence of PAF should be strictly controlled and, in fact, the
`T. M. (1989) J. Biol. Chem. 264, 5331-5334). To
`enzymatic activities
`responsible for PAF biosynthesis are
`identify potential substrates, we synthesized phospha-
`tightly regulated (1-3). Additionally, accumulation of PAF in
`tidylcholines with sn-2 residues from two to nine car-
`some cells
`is controlled by its rate of degradation (4, 5 ) .
`bon atoms long, and found the V/k ratio decreased as
`Degradation also plays a major role in the potential for PAF
`the sn-2 residue was lengthened: the C5 homolog was
`or function as a locally acting
`to circulate as a hormone
`50%, the C6 207’0, while the C9 homolog was only 2%
`autacoid due to the presence of plasma PAF acetylhydrolase.
`as efficient as PAF. However, the presence of an W-oxo
`This enzyme inactivates PAF by hydrolyzing the sn-2 acetyl
`function radically affected hydrolysis: the half-life of
`residue, and is exclusively responsible for the degradation of
`the sn-2 9-aldehydic homolog was identical to that of
`PAF in whole blood (6).
`PAF.
`The majority (60-70%) of PAF acetylhydrolase in human
`We oxidized [2-arachidonoyl]phosphatidylcholine
`plasma is associated with low density lipoprotein; the rest is
`associated with a subpopulation of high density lipoprotein
`and isolated a number of more polar phosphatidylcho-
`lines. We treated these with phospholipase C, deriva-
`that contains apolipoprotein E (HDL-with apoE) (6). The
`tized the resulting diglycerides for gas chromato-
`enzyme associated with LDL is immunologically equivalent
`to that in HDL and is transferred between the two lipoprotein
`graphic/mass spectroscopic analysis, and found a num-
`ber of diglycerides where the m/z ratio was consistent
`particles in a pH-dependent fashion (6). There is a sharp
`with a series of short to medium length sn-2 residues.
`deviation from Michaelis-Menten kinetics at low substrate
`We treated the polar phosphatidylcholines with acetyl-
`concentrations for LDL-associated activity; hydrolysis of 1
`nM PAF proceeds at only 1.7% of the predicted rate (7). HDL-
`hydrolase and derivatized the products for analysis by
`associated activity is even less efficient than LDL-associated
`gas chromatography/mass spectroscopy. The liberated
`activity at low substrate concentrations,
`residues were more polar than straight chain standards
`so that in whole
`and had mlz ratios from 129 to 296, consistent with
`plasma only LDL-associated activity appears to catalyze PAF
`short to medium chain residues. Therefore, oxidation
`hydrolysis (7). Lipoprotein-associated acetylhydrolase activ-
`fragments the sn-2 residue of phospholipids, and the
`ity also rapidly degrades endothelial cell-associated PAF (8),
`acetylhydrolase specifically degrades such oxidatively
`indicating that stimulated endothelial cells express this PAF
`fragmented phospholipids.
`on their cell surface. It also shows that LDL can modulate
`
`I The abbreviations used are: PAF, platelet-activating factor (1-0-
`alkyl-2-acetyl-sn-glycero-3-phosphocholine); GPC, sn-glycero-3-
`phosphocholine; HDL, high density lipoprotein; LDL,
`low density
`lipoprotein; t-BDMS,
`tert-hutyldimethylsilyl; CV3988, rac-3-(N-
`Platelet-activating factor (l-O-alkyl-2-acetyl-sn-glycero-3-
`n-octadecylcarbamoyloxy)-2-methoxypropyl-2-thiazolioethy1 phos-
`5-(prop-2-ene)-2(3,4-dimethoxyphenyl)-3a,a-
`phate; Kadsurenone,
`methoxy-3-methyl-2,3,3a,6-tetrahydro-6-oxohenzofuran; L652,731,
`trans-2,5-bis(3,4,5-trimethox~henyl)tetrahydrofuran; WEB2086, 3-
`[4-(chlorophenyl)-9-methyl-ll-thieno[3,2-~[l,2,4]triazolo-[4,3~a]
`[1,4]diazepin-2-yl](4-morpholinyl)-l-propanone; L659,989,
`( & ) -
`tran~s-Z-(3-methoxy-5-methylsulfonyl-4-propoxyphenyl)-5-(3,4,5-tri-
`methoxypheny1)tetrahydrofuran; U66985, (l-O-octadecyl-2-acetyl-
`sn-glycero-3-phosphoric acid-6’-trimethylammoniumhexyl ester);
`WEB2170, 6-(2-chlorophenyl-8,9-dihydro-l-methyl-8-(4-morpholi-
`nylcarbonyl)-4H,~H-cyclopenta[4,5]thieno[3,2-~[1,2,4]triaozo[4,3-
`a][l,4]diazepine; SRI63441, cis-(f)-l-12-[hydroxy({tetrahydro-5-
`[~octadecylaminocarbonoyl)oxy]methyl)furan-2-yl)methoxyphosphi-
`nyloxy]ethyl}quinolinium hydroxide (inner salt); HPLC, high
`performance liquid chromatography; PC, phosphatidylcholine; TLC,
`thin layer chromatography; GLC, gas liquid chromatography.
`
`* This work was supported in part by funds from the Nora Eccles
`Treadwell Foundation; an Established Investigator Award (to S. M.
`P.) from the American Heart Association; and Grants 5T32 CA09602,
`HL35828, and HL34127 from the National Institutes of Health. The
`mass spectrometry
`facility was supported by Grant CHE-8100424
`from the National Science Foundation and by funds from the Uni-
`versity of Utah Institutional Funds Committee. The costs of publi-
`cation of this article were defrayed in part by the payment of page
`charges. This article must therefore
`he hereby marked “aduertke-
`mcznt” in accordance with 18 U.S.C. Section 1734 solely to indicate
`this fact.
`To whom correspondence should he addressed: Cardiovascular
`Research and Training Institute, Building 500, University of Utah,
`Salt Lake City, UT 84112. Tel.: 801-581-8183; Fax: 801-581-3128.
`
`11095
`
`RIMFROST EXHIBIT 1095 page 0001
`
`
`
`11096
`
`Hydrolysis of Oxidatively Fragmented Phospholipids
`
`
`
`
`this surface expression and thereby alter cellular interactions,
`EXPERIMENTAL PROCEDURES~
`such as neutrophil adhesion and activation, that is mediated
`Materials-l-Palmitoyl-2-[l-'4C]arachidonoyl-sn-glycero-3-phos-
`by endothelial cell-associated PAF (8).
`phocholine was purchased from Du Pont-New England Nuclear and
`We have purified and characterized acetylhydrolase from
`l-palmitoyl-2-[l-"C]oleoyl-sn-glycero-3-phosphocholine from Amer-
`sham Corp. l-Palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocho-
`human plasma (9) and have identified and partially charac-
`
`line, l-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine, and l-palmi-
`terized a distinct activity from human erythrocytes (10).
`toyl-2-hydroxy-sn-glycero-3-phosphocholine were obtained from
`Examination of the substrates utilized by plasma acetylhy-
`Avanti Polar Lipids (Birmingham,
`AL). 4-(N,N-Dimethylamino)-
`drolase (9, 11-13) shows that it functions as a short chain
`pyridine, propionic anhydride, and succinic anhydride were obtained
`acyl hydrolase. It will use substrates with either an ether or
`from Aldrich. Butyric anhydride, valeric anhydride, and hexanoic
`NY).
`ester bond at the sn-1 position, and it demonstrates little
`anhydride were purchased from Eastman Kodak (Rochester,
`Nonanoic anhydride and all short chain fatty acids (Ca through C7
`specificity for the sn-3 headgroup (11). It does not, however,
`(C, represents a straight carbon chain n atoms long)) were obtained
`effectively utilize sn-2 fatty acyl residues that are longer than
`from Sigma. N-Methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide
`six carbon atoms in length. We fragmented l-palmitoyl-2-
`was purchased from Pierce Chemical Co. CV3988 was a gift from M.
`arachidonoyl-GPC by ozonolysis to generate l-palmitoyl-2-
`Nishikawa, Central Research Division, Takeda Chemical Industries
`was approxi-
`(5-oxovaleroyl)-GPC (11) and found that it
`Ltd. Kadsurenone, L652,731, and L659,989 were kindly provided by
`J. C. Chabala, Merck Sharp and Dohme Research Laboratories.
`mately half as effective a substrate as PAF for the acetylhy-
`WEB2086 and WEB2170 were the kind gifts of Peggy Ganong,
`drolase. This suggests that other oxidatively generated species
`Boehringer Ingelheim Pharmaceuticals, Inc. BN50726 was kindly
`might be substrates for the acetylhydrolase, and, indeed, we
`provided by P. Braquet, Institut Henri Beaufour. D. J. Hanahan
`found that uncontrolled oxidation of a phospholipid contain-
`kindly provided
`U66985, a product of The Upjohn Company.
`SRI63441 was a gift from D. A. Handley, Sandoz Research Institute.
`ing a polyunsaturated sn-2
`residue generates a myriad of
`The PAF acetylhydrolase was purified from human plasma as de-
`unidentified compounds, some of which are substrates for the
`scribed (9).
`acetylhydrolase (11). We found that the more polar of these
`Synthesis of Phospholipids-The majority
`of the sn-2 short chain
`species were better substrates than the
`less polar species,
`phospholipids were synthesized by the coupling method of Gupta et
`suggesting that the more polar species contained either
`al. (16) and purified using a modification (11) of the reversed-phase
`HPLC procedure described by Brash (17). The sn-2 aldehydic phos-
`shorter sn-2 residues, more oxygenated residues, or a combi-
`phatidylcholines were prepared by reductive ozonolysis
`(18). The
`nation of these two modifications. The restricted substrate
`general procedures utilized for reductive ozonolysis and the coupling
`specificity of the acetylhydrolase is unusual, but guarantees
`are presented below; spectral characterization and experimental de-
`that only fragmented phospholipids are attacked without af-
`tails for each compound may be found in the supplementary material.
`fecting the bulk of unoxidized parent phospholipids. This
`General Procedure for Preparation of Phospholipids Containing an
`sn-2Aldehydic Chain-Reactions were carried out by ozone treatment
`substrate specificity allows acetylhydrolase to circulate in a
`of the starting phospholipid, either 1-palmitoyl-2-arachidonoyl-sn-
`fully active state, and Stremler et al. (11) and Steinbrecher
`glycero-3-phosphocholine or l-palmitoyl-2-oleoyl-sn-glycero-3-phos-
`
`and Pritchard (13) postulate that this makes it likely that an
`phocholine (64-128 Fmol of the unlabeled or 10 FCi of the sn-2 1-
`additional physiological function of this enzyme is to degrade
`"C-labeled compounds) dissolved in anhydrous dichloromethane at
`phospholipids that have been oxidatively fragmented. This is
`-78 "C, until a blue color persisted. Excess ozone was removed by
`exposing the solution to room temperature until it became clear. The
`consistent with the observation (14) that oxidative modifica-
`solution was rechilled to -78 "C, and 50 ~1 of dimethyl sulfide was
`tion of LDL results in the hydrolysis of as much as 40% of its
`added. Solvent and excess dimethyl sulfide were removed under
`phosphatidylcholine by an LDL-associated phospholipase A2
`nitrogen flow. The resulting residue was purified by reversed-phase
`activity.
`HPLC ( 5 - ~ m ODS Ultrasphere, column A = 0.46 X 25 cm or column
`PAF acetylhydrolase and the PAF receptor recognize the
`B = 1.0 X 25 cm; 905:70:25 methanol/water/acetonitrile containing
`20 mM choline chloride). Appropriate fractions based on the radio-
`same phospholipid, and share a requirement for a short sn-2
`activity profile of the eluant were combined, reduced in volume under
`residue. The PAF receptor also
`interacts with a variety of
`nitrogen flow, and extracted to remove the choline chloride using the
`compounds (15) that competitively block PAF binding to, and
`method of Bligh and Dyer (19). The resulting product was stored in
`activation of, this receptor. This raises the possibility that
`dichloromethane at -20 "C.
`receptor antagonists might also inhibit acetylhydrolase activ-
`General Procedure for Preparation of Phospholipids Containing an
`Residue-1-Palmitoyl-2-hydroxy-sn-glycero-3-
`sn-2 Short Chain
`ity. This could be an important
`issue from a physiological
`phosphocholine or the l-[l-'4C]palmitoyl-2-hydroxy-sn-glycero-3-
`perspective, since,
`if receptor antagonists prevented PAF
`phosphocholine (1 equivalent) was dissolved in anhydrous chloro-
`metabolism in plasma, the concentration of PAF would rise
`form/pyridine, 4:l (v/v). These solvents were dried over phosphorus
`and tend to counteract the competitive blockade of the recep-
`pentoxide and calcium hydride, respectively,
`and distilled immedi-
`ately prior to use. After addition of 4-(N,N-dimethylamino)pyridine
`tor. It would also suggest
`that receptor antagonists might
`(16) and the fatty acid anhydride (1.5-3 equivalents of each compound
`interfere with the metabolism of oxidatively fragmented phos-
`for unlabeled syntheses and 50-100 equivalents of each for the I4C-
`pholipids and possibly prolong the half-life of these reactive
`labeled compounds) the reaction flask was flushed with nitrogen and
`lipids.
`the resulting solution was allowed to stir at room temperature for 26-
`The goal of the experiments presented here was to define
`36 h. The solvent was then removed under nitrogen flow, and the
`the potential role of acetylhydrolase in the metabolism of
`resulting residue was purified by reverse-phase HPLC (ll), as above.
`Appropriate fractions based on the radioactivity profile of the eluant
`oxidatively fragmented phospholipids by examining the sub-
`were combined, reduced in volume under nitrogen flow, and extracted
`strate specificity of plasma PAF acetylhydrolase in detail.
`to remove the choline chloride using the method of Bligh and Dyer
`This included a definition of the sn-2 residues recognized by
`(19). The resulting product was stored in dichloromethane at -20 "C.
`the acetylhydrolase, identification of phospholipids derived
`For the ~-palmitoyl-2-succinyl-GPC
`([2-succinoyl]-PC) the reversed-
`from the oxidative fragmentation of arachidonoyl-containing
`phase HPLC solvent contained 25 mM ammonium formate instead
`of choline chloride. Appropriate fractions were combined, reduced in
`phosphatidylcholine, and characterization of the hydrolytic
`products obtained after acetylhydrolase treatment of these
`' Portions of this paper (including part of "Experimental Proce-
`oxidatively fragmented phospholipids. We also determined
`dures") are presented in miniprint at the end of this paper. Miniprint
`the effect of PAF receptor antagonists on the activity of PAF
`is easily read with the aid of a standard magnifying glass. Full size
`acetylhydrolase to determine
`if they might interfere with
`photocopies are included in the microfilm edition of the Journal that
`substrate degradation.
`is available from Waverly Press.
`
`RIMFROST EXHIBIT 1095 page 0002
`
`
`
`Hydrolysis of Oxidatively Fragmented Phospholipids
`11097
`homologs. Homologs of 1-[l-14C]palmitoyl-2-acyl-GPC con-
`volume under nitrogen flow, and the ammonium formate was removed
`by lyophilization or acidic extraction (19). The resulting product was
`taining sn-2 fatty
`acyl residues of increasing chain length
`stored in dichloromethane at -20 "C.
`
`were synthesized and tested as substrates for the purified PAF
`Assay of PAFAcetylhydrolase Activity-PAF acetylhydrolase activ-
`acetylhydrolase. Hydrolysis of each compound was examined
`ity was determined as previously described (6). Determination of half-
`at 12 concentrations, and the kinetic constants were deter-
`time (t7,J of PAF degradation at subsaturating concentrations of
`mined from a double-reciprocal plot. The results (Fig. 1) show
`substrate has been described in detail elsewhere (7). Kinetic param-
`that PAF acetylhydrolase was sensitive to the length of the
`eters for hydrolysis of sn-2 short chain phospholipids by purified
`plasma PAF acetylhydrolase were determined by varying the sub-
`sn-2 residue, but that it hydrolyzed sn-2 acyl groups of up to
`
`strate concentration between 0 and 40 p ~ . The phospholipids were
`five carbons in length without a dramatic reduction in cata-
`labeled either at the carboxyl group of the sn-2 residue or at the
`lytic efficiency, i.e. the V / k ratio was reduced only about 2-
`carboxyl group of the sn-1 palmitoyl residue. At the end
`of the
`fold for the valeroyl homolog. However, when the sn-2 residue
`incubation, the amount of radioactively labeled fatty acid released
`was extended to 9 carbon atoms the catalytic efficiency was
`from sn-2-labeled phospholipids was determined after separation
`reduced to 2% of
`that of PAF. This substrate
`specificity
`from remaining substrate by either reversed-phase Chromatography
`on C," Baker columns or by a Bligh-Dyer extraction (19). Both
`showed that PAF acetylhydrolase behaved
`as an efficient
`methods gave comparable results. For the s n - l labeled substrates, the
`short chain acyl hydrolase.
`amount of l-[l-'4C]palmitoyl-2-hydroxy-GPC produced was deter-
`We next determined the effect that an w-oxy function had
`mined after extraction, separation by TLC (20), and liquid scintilla-
`on catalytic efficiency by synthesizing phosphatidylcholine
`tion spectrometry.
`homologs with five-carbon sn-2 residues terminated with
`Oxidation of 1-Palmitoyl-2-arachidonoyl-sn-GPC and Analysis of
`either an aldehydic or a carboxylic function. We found that
`f'hospholipidProducts-l-Palmitoyl-2-[l-'"C]arachidonoyl~GPC (2.0
`pCi, 6.4 pmol) was incubated for 60 min at 25 "C with 7.5 pmol of
`the [2-(5-oxovaleroyl)]PC homolog was about two-thirds as
`sodium deoxycholate, 4.62 X lo5 unit,s of soybean lipoxygenase (Type
`efficient a substrate as PAF itself (Fig. l), consistent with our
`V, EC 1.13.11.12) in 1.5 ml of 0.2 M borate buffer (pH 9), as described
`previous results (11). We also found that the substrate with
`(17). The phospholipids were extracted, solvent was removed under
`an w-terminal carboxylic function was 56% as efficient as
`nitrogen flow, and the resulting thin
`film was stored under air
`at
`PAF, and that the valeroyl homolog was 51% as efficient a
`-20 "C for 7-10 days, allowing further oxidation to occur. Following
`substrate as PAF. In contrast to our prediction that
`acetyl-
`oxidation, the samples were examined by reversed-phase HPLC (11).
`Fractions were combined based on the radioactivity profile, extracted,
`hydrolase might preferentially recognize oxidized residues, it
`and then analyzed by positive ion
`fast atom bombardment mass
`appeared that it was insensitive to w-terminal oxy functions.
`spectrometry or derivatized for further separation and analysis. Fur-
`However, when we extended the length of the sn-2 residue to
`ther separation of the components of these HPLC fractions was
`nine carbon atoms, we found that the oxo function strongly
`obtained by treatment with diazomethane
`(21) or not, followed by
`affected catalysis. Although phosphatidylcholine with an sn-
`hydrolysis with phospholipase C, and conversion to tert-butyldimeth-
`2 nonanoyl residue was an extremely poor substrate for the
`ylsilyl derivatives (22) suitable
`for gas chromatography. The tert-
`butyldimethylsilyl standards were prepared from synthetic sn-2 short
`enzyme (its catalytic efficiency was only 2% of that of PAF),
`chain fatty acyl phosphatidylcholines using the same hydrolysis and
`its nine-aldehydic homolog had a catalytic efficiency 21-fold
`derivatization procedure. The tert-butyldimethylsilyl derivatives of
`higher (Fig. 1). This enhancement of hydrolysis by the w-oxo
`these standards and unknowns were analyzed using a mass spectrom-
`eter (VG Micromass 7050 double focusing high resolution mass spec-
`trometer with a VG data system) coupled to a gas chromatograph
`with a DB-5 capillary column (developed 210 "C for 10 min, 10 "C/
`min, 280 "C for 15 min). Spectral characterization of these compounds
`may be found in the Miniprint Supplement.
`Hydrolysis of Oxidized Phospholipids by PAF Acetylhydrolase-
`Following oxidation of 37-70 pmol of l-palmit0yl-2-[1-'~C]arachidon-
`oyl-GPC (0.03-0.05 pCi/pmol) and reversed-phase HPLC separation
`as described above, the oxidation products (fractions I through IV)
`were incubated with PAF acetylhydrolase (68 units) for 18 h at 37 "C.
`The solution was then acidified (pH 2) and extracted (19). Blanks
`were prepared by incubating oxidized products and PAF acetylhydro-
`lase separately under these same conditions. The apolar phases from
`the extraction were adjusted to pH 8 before the solvent was removed
`by evaporation under aspirator vacuum. The sn-2 acyl chains released
`by PAF acetylhydrolase, or present in the untreated control samples,
`were extracted from the residue with methanol. The extracts and
`short chain fatty acid standards were analyzed using a mass spec-
`trometer coupled to an OV-351 capillary column, developed at 70 "C
`for 1 min, 3 "C/min, 200 "C for 10 min. Spectral Characterization of
`these compounds may be found in the Miniprint Supplement.
`
`alkyl
`0 m-carboxylic
`0 o-aldehydic
`
`10
`
`Chain length
`
`2
`
`w-aldehydic
`
`RESULTS
`Effect of sn-2 Chain Length and Composition on the Sub-
`strate Specificity of PAF Acetylhydrolase-Oxidation
`of 1-
`palmitoyl-2-arachidonoyl-GPC generates a number of
`unidentified compounds that are hydrolyzed by the plasma
`acetylhydrolase (11). Since oxidation of unesterified polyun-
`saturated fatty acids generates fragments of varying size with
`varying oxygen functions (23), oxidation and fragmentation
`of phospholipids containing polyunsaturated fatty acyl resi-
`dues should generate phospholipids with varying sn-2
`resi-
`dues. To determine which of these compounds plasma PAF
`acetylhydrolase might attack, we determined the kinetic pa-
`rameters for the hydrolysis of a series of synthetic sn-2
`
`*carboxylic
`
`FIG. 1. Effect of sn-2 chain length and oxidation state on
`catalytic efficiency of 1-palmitoyl-2-acyl-GPC as a substrate
`for plasma acetylhydrolase. Purified plasma PAF acetylhydrolase
`(2.5 X lo-,'' units) was incubated for 30 min at 37 "C with a series of
`homologs of l-[l-'"C]palmitoyl-2-acyl-GPC, where the sn-2 residue
`was n carbon atoms in length, at concentrations ranging from 0 to 40
`The amount of [l-'4C]palmitoyl-GPC produced by this incubation
`p ~ .
`
`was determined as described under "Experimental Procedures." Al-
`ternatively, an equal amount of enzyme from the same preparation
`was incubated with [acetyl-"HIPAF, l-palmitoyl-2-([l-"C]5-oxoval-
`eroy1)-GPC, or l-palmitoyl-2-([l-'4C]9-oxononanoyl)-GPC, and the
`amount of radioactivity released from
`the phospholipid was deter-
`mined by phase/phase extraction (19).
`
`RIMFROST EXHIBIT 1095 page 0003
`
`
`
`11098
`
`
`
`Hydrolysis of Oxidatively Fragmented Phospholipids
`
`
`
`function overcomes much of the severe restriction on the
`acetylhydrolase could hydrolyze the sn-2 nine-carbonalde-
`length of the sn-2 residue and expands the number of potential
`hydic residue at low substrate concentrations. We found that
`substrates for acetylhydrolase.
`the purified acetylhydrolase not only hydrolyzed l-palmitoyl-
`Acetylhydrolase activity in plasma does not follow Michae-
`2-(9-oxononanoyl)-GPC at a concentration of 1 p ~ ,
`it did so
`
`lis-Menten kinetics at low substrate concentrations and de-
`at a rate that was undistinguishable from that of PAF (Fig.
`viation from calculated hydrolytic
`rates can be quite
`large:
`3). The very large effect of the w-oxo function is apparent as
`the rate of hydrolysis of 1 nM PAF is 60-fold less than that
`the 2-nonanoyl homolog was not hydrolyzed under these
`predicted from the kinetic constants (7). Since only subsat-
`conditions. Therefore, addition of an w-oxo function overcame
`urating concentrations of substrates are likely to be encoun-
`the effect of sn-2 chain elongation, even at low substrate
`tered in vivo, we determined whether hydrolysis of the hom-
`concentrations, but a plasma component(s) prevented access
`ologs with varying sn-2 residues also deviated from the pre-
`of the substrate to the enzyme. We do not know why the 5-
`dicted rate of hydrolysis under these conditions. We measured
`oxovaleroyl homolog did not demonstrate the same phenom-
`their half-life in whole plasma, and found that the relative
`enon, but it is clear that phospholipids with intermediate
`rate of hydrolysis differed somewhat from that catalyzed by
`length sn-2 residues are acetylhydrolase substrates, provided
`the purified acetylhydrolase (Fig. 2). For instance, the half-
`that they contain an a-oxo function.
`life of the sn-2 propionoyl homolog was shorter than that of
`Identification of Phospholipids derived from Oxidative Frag-
`PAF even though its V/k ratio was smaller than that of PAF.
`mentation of I -Palmitoyl-2-[1 -'4C]arachidon~yl-GPC-Our
`However, overall the half-lives of these homologs decreased
`next goal was to determine what types of phosphatidylcho-
`with increasing chain length, and the ratio
`of half-lives of
`lines were generated from an uncontrolled oxidative attack
`each compound relative
`to PAF was very similar to their
`on a phosphatidylcholine containing an sn-2 polyunsaturated
`ratios of V/k. For example, the half-life of the valeroyl hom-
`fatty acyl residue
`to define potential acetylhydrolase sub-
`olog was 39% of that of PAF, while its V/k ratio was 51% of
`strates. Phosphatidylcholine containing an sn-2 arachidonoyl
`that of PAF. We found that, as with the V/k ratio, addition
`residue was oxidized with soybean 15-lipoxygenase, to gener-
`of an w-terminal aldehydic
`or carboxylic function to the
`ate a few percent of the corresponding hydroperoxyphospho-
`valeroyl homolog did not affect the half-life in whole plasma.
`lipid, and then allowed to oxidize in air for several days as a
`These results showed that each of these compounds behaved
`thin film. When lipids from this oxidation were analyzed by
`like PAF, and therefore showed a large deviation from
`the
`reversed-phase HPLC, we found an unresolved mixture of
`predicted rate of hydrolysis at low concentrations in whole
`radiolabeled material that migrated in a position that showed
`plasma.
`they were more polar than the starting phospholipid (Fig. 4).
`l-palmitoyl-2-(9-oxonona-
`We unexpectedly found that
`Some of this material also absorbed light at 235 nm, indicative
`noy1)-GPC was not hydrolyzed when added at low concentra-
`of bond rearrangement to a conjugated system. We collected
`tions to whole plasma. Since purified acetylhydrolase did use
`the material in fractions 15 through 52 and chromatographed
`this phospholipid as a substrate (Fig. 11, this result either was
`it on straight phase thin layer plates, where it again migrated
`due to a failure of this phospholipid to localize to the appro-
`as material more polar than phosphatidylcholine (not shown).
`priate compartment in plasma, presumably LDL, or the result
`We found that these lipids could be visualized with molybdic
`of a very large deviation from the predicted rate of hydrolysis.
`acid reagent or Dragendorff reagent (24), indicating the pres-
`We examined the latter possibility by asking whether purified
`ence of phosphate and choline, respectively. The presence of
`14C radioactivity, derived from the carboxyl carbon of the sn-
`2 residue, showed that the sn-2 position remained esterified
`with at least a portion of the arachidonoyl residue.
`
`loo
`
`1
`
`- -
`
`[P-nonanoyl]PC
`
`[2-(9-oxononanoyl)]PC
`
`1
`
`2
`
`6
`
`7
`
`\
`YPAF
`
`3
`5
`4
`sn-2 Chain Length
`FIG. 2. Effect of sn-2 chain length and oxidation state on
`1 0
`Tlme
`(mln)
`the half-life of 1-palmitoyl-2-acyl-GPC in plasma. Reaction
`FIG. 3. Effect of an W-oxo function on the half-life of 1-
`mixtures contained 250 p1 of plasma and 10-6-10-9 M [~cetyl-~H]PAF
`or l-[l-14C]palmitoyl-2-acyl-GPC
`palmitoyl-2-nonanoyl-GPC. [~cetyl-~H]PAF or l-palmitoyl-2-(9-
`(O), l-[l-'4C]palmitoyl-2-gluta-
`roy1)-GPC (O), or l-palmitoyl-2-(5-[l-'4C]oxovaleroyl-GPC (0) in a
`[l-14C]oxononanoyl)-GPC at a concentration of 1 PM was incubated
`total volume of 300 p1. The half-life was independent of substrate
`with 0.18 unit of purified plasma acetylhydrolase. At the stated times,
`aliquots were removed to quantitate the amount of substrate hydro-
`concentration over these ranges. Aliquots
`(50 pl) were removed at
`lyzed as described under Fig. 2. When 1-palmitoyl-2-nonanoyl-GPC
`various times (7), and the lipids were extracted and separated by thin
`was the substrate, the sensitivity
`of the assay was increased by
`layer chromatography. The loss of substrate was then quantitated by
`decreasing the substrate concentration to 1 nM and increasing the
`liquid scintillation counting by scraping the appropriate area of the
`amount of enzyme to 0.75 unit.
`plate.
`
`0 4
`0
`
`RIMFROST EXHIBIT 1095 page 0004
`
`
`
`
`
`
`
`
`
`11099
`
`15.48
`1
`
`16.92
`
`
`Hydrolysis of Oxidatively Fragmented Phospholipids
`I FRACTION I
`
`A
`
`v)
`0
`
`N a
`
`5 0
`
`6 0
`
`0
`
`1 0
`
`2 0
`
`3 0
`4 0
`Fraction
`FIG. 4. Reversed-phase HPLC chromatogram of products
`obtained from uncontrolled oxidation of phosphatidylcholine.
`l-Palmitoyl-2-[l-'4C]arachidonoyl-GPC was oxidized by 15-lipoxy-
`[l-'4C]15-hydroperoxyeicosatetraenoyl species,
`genase to the sn-2
`which comprises the largest radioactive and UV-absorbtive peak in
`the chromatogram. This product was allowed to oxidize by exposure
`to air as a thin film at -20 "C before lipid products were recovered
`by extraction (19). These lipids were separated by chromatography
`over an octadecylsilica column developed isocratically at 0.8 ml per
`min with methanol/water/acetonitrile (930:70:50) containing 20 mM
`choline chloride. The effluent was monitored for UV absorption at
`235 nm (0) with a flow-through detector before fractions (1 ml) were
`collected, and radioactivity (0) estimated by removing aliquots for
`liquid scintillation spectroscopy. Subsequently, fractions were com-
`bined as follows: I, HPLC fractions 15-21; 11, HPLC fractions 22-28;
`111, HPLC fractions 29-34; IV, HPLC fractions 35-52.
`
`We analyzed these polar phospholipids by mass spectrom-
`etry. Four fractions corresponding to fractions 15-21, 22-28,
`the HPLC chromatogram (Fig. 4) were
`29-34, and 35-52 of
`combined, extracted by an acidic Bligh and Dyer procedure,
`and subjected to fast atom bombardment mass spectrometry.
`The resulting spectra were complex, but were consistent with
`the presence of a mixture of phospholipids with shortened sn-
`2 residues. The four combined fractions were methylated with
`diazomethane, or not, and then treated with phospholipase C.
`The resulting diglycerides were then converted to their t-
`butyldimethylsilyl derivatives. These were separated by gas
`chromatography on a DB-5 capi