Interaction of plasmenylcholine with free radicals in selected model systems

Interaction of plasmenylcholine with free radicals in selected model systems

Author’s Accepted Manuscript Interaction of plasmenylcholine with free radicals in selected model systems A. Broniec, A. Żądło, A. Pawlak, B. Fuchs, R...

2MB Sizes 0 Downloads 41 Views

Author’s Accepted Manuscript Interaction of plasmenylcholine with free radicals in selected model systems A. Broniec, A. Żądło, A. Pawlak, B. Fuchs, R. Kłosiński, D. Thompson, T. Sarna www.elsevier.com

PII: DOI: Reference:

S0891-5849(17)30090-4 http://dx.doi.org/10.1016/j.freeradbiomed.2017.02.029 FRB13222

To appear in: Free Radical Biology and Medicine Received date: 13 September 2016 Revised date: 10 January 2017 Accepted date: 13 February 2017 Cite this article as: A. Broniec, A. Żądło, A. Pawlak, B. Fuchs, R. Kłosiński, D. Thompson and T. Sarna, Interaction of plasmenylcholine with free radicals in selected model systems, Free Radical Biology and Medicine, http://dx.doi.org/10.1016/j.freeradbiomed.2017.02.029 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Interaction of plasmenylcholine with free radicals in selected model systems. A.Broniec1*, A.Żądło1, A.Pawlak1, B.Fuchs2, R.Kłosiński1, D.Thompson3, T.Sarna1 1

Biophysics Department, Biochemistry, Biophysics and Biotechnology Faculty, Jagiellonian University, Krakow, Poland 2

Institute of Medical Physics and Biophysics, Medical Faculty, University of Leipzig, Germany 3

Department of Chemistry, Purdue University, West Lafayette, IN, USA

*

Correspondence to: Biophysics Department, Biochemistry, Biophysics and Biotechnology Faculty,

Jagiellonian University, Cracow, Poland. Tel.: +4812 664 65 26. [email protected] Abstract: Plasmalogens (Plg) - naturally occurring glycerophospholipids with the vinyl-ether group in the sn-1 position are generally viewed as physiological antioxidants. Although there are numerous examples of antioxidant action of plasmalogen in cell cultures and in experimental animals, this hypothesis is far from being satisfactorily proven due to substantial limitations of such studies. Thus, plasmalogen reactivity in cells results in the accumulation of toxic byproducts and the experimental design is usually too complicated to evaluate the protective function of solely one type of lipid molecular species. In this study, experiments were performed in homogenous and heterogeneous model systems consisting of solutions in organic solvents as well as micelles and liposomes containing pure synthetic plasmenylcholines. Under the experimental conditions used, chemical reactivity of plasmalogens could be attributed to specific fatty acid esterification pattern. This is important because the chemical reactivity cannot be separated from physicochemical properties of the lipids. Time-dependent formation of phospholipid and cholesterol hydroperoxides were determined by iodometric assay and HPLC-EC. EPR oximetry and Clark electrode were employed to detect the accompanying changes in oxygen concentration. Oxidation of the studied lipids was monitored by standard colorimetric TBARS method as well as MALDI-TOF mass spectrometry. Our data indicate that the reactivity of sn-2 monounsaturated vinyl ether lipids in peroxyl radical-induced or iron-catalyzed peroxidation reactions is comparable with that of their diacyl analogs. In samples containing cholesterol and plasmalogens, oxidative processes lead to accumulation of the radical oxidation product of cholesterol. It can be concluded that the antioxidant action of plasmalogens takes place intramolecularly rather than intermolecularly and depends on the degree of unsaturation of esterified fatty acids. Thus, it is questionable if plasmalogens can really be viewed as “endogenous antioxidant”, even though they may exhibit, under special conditions, protective effect. Abbreviation AAPH,

2,2’-Azobis(2-methylpropionamidine)dihydrochloride;

AMVN,

2,2’-Azobis(2,4-

dimethylvaleronitrile); Asc, ascorbate/ascorbic acid; BHT – butylated hydroxytoluen; BME, (1-biphenyl-4yl-1-methyl-ethyl)-tert-butyl diazene; BME●, 1-biphenyl-4-yl-1-methyl-ethyl radical; ChOOH– cholesterol

hydroperoxide; DHA, docosahexaenoic acid; DHB–2,5-dihydroxybenzoic acid; DMPC, 1,2-dimyristoylsn-glycerol-3-phosphocholine; Fe(HQ)3, ferric-8-hydroxyquinoline; LA-linolenic acid (18:3); LOOH, lipid hydroperoxides; MALDI-TOF, matrix-assisted laser desorption ionization with time of flight analyzer; MDA, malondialdehyde; mHCTPO, 4-protio-3-carbamoyl-2,2,5,5-tetraperdeuteromethyl-3-pyrroline-1oxyl; MUV-multilamellar liposomes; MUFA, monounsaturated fatty acids; SUV, small unilamellar liposomes;

TX-100,

triton

X-100;

Plg,

plasmalogens;

PlgPC,

plasmenylcholine;

PlgPE,

plasmenylethanolamine; PlgPC 16:0/18:1, 1-O-1’-(Z)-hexadecenyl-2-oleoyl-sn-glycero-3-phosphocholine; PlgPC 18:0/18:1, 1-O-1’-(Z)-octadecenyl-2-oleoyl-sn-glycero-3-phosphocholine; PlgPC 18:0/20:4, 1-O-1’(Z)-octadecenyl-2-arachidonoyl-sn-glycero-3-phosphocholine; PlgPC 18:0/22:6 -1-O-1’-(Z)-octadecenyl2-docosahexaenoyl-sn-glycero-3-phosphocholine;

POPC-1-palmitoyl-2-oleoyl-sn-glycero-3-

phosphocholine (PC 16:0/18:1); PUFA, polyunsaturated fatty acids; SAPC, 1-stearoyl-2-arachidonoyl-snglycero-3-phosphocholine phosphocholine

(PC

(PC 18:0/22:6);

18:0/20:4); SOPC

SDPC-1-stearoyl-2-docosahexaenoyl-sn-glycero-3-

-1-stearoyl-2-oleoyl-sn-glycero-3-phosphocholine

(PC

18:0/18:1),TBA, thiobarbituric acid; TCA, trichloroacetic acid; VE, vinyl ether/vinyl ether bond Keywords:

plasmalogens,

plasmenylcholine,

radical-induced

oxidation

of

phospholipids,

lipid/phospholipid oxidation

1. Introduction Plasmalogens (Plg) are a class of phospholipid molecules with different head group but with conserved vinyl-ether bond in the sn-1 position of glycerol backbone. This chemical bond exerts some additional reactivity and structural diversity to Plg compared with their diacyl analogs [1]. Typical plasmalogen molecules contain saturated/monounsaturated fatty acid in the sn-1 position (i.e.16:0, 18:0, 18:1) while the fatty acid at the sn-2 position could be highly unsaturated such as 20:3, 20:4, 22:4, 22:5, 22:6 in some tissues [2],[3]. The hydrophobic-hydrophilic interface of Plg molecule includes two sp2 hybridized carbon atoms [4] which are additionally activated by electrodonating capability of the nearby oxygen atom [5]. The reactivity of enol-ethers is useful property in organic synthesis whereas in food industry and living organisms could pose serious problems with membrane stability [6]. From biological point of view it is rather intriguing why highly metabolically active tissues such as brain or retina are enriched in sn-2 polyunsaturated fatty acids (PUFA) plasmenylethanolamine (PlgPE) [7]. The high metabolic activity usually means high probability of generation of partially reduced oxygen species (PROS), which abstract PUFA bis-allylic hydrogens or attack vinyl-ether bond. This suggest that the degradation of Plg in such tissues is not detrimental but could be important for other unknown reasons like biochemical signaling or protein function maintenance. On the other hand, if we address any structural role to sn-2 monounsaturated fatty acids (MUFA) Plg, such as is supposed to be in myelin sheet, then the oxidative degradation should

be harmful to the tissue [8]. It is obvious that the major radical or hydrolysis degradation products of Plg such as lysolipids, free fatty acids, short chain aldehydes, are all surface active compounds that could change molecular organization of the membrane of their origin. All above suggest that the oxidative decomposition of Plg could positively or negatively affect cell functions in a way being highly dependent on the plasmalogen structure. Since sn-1 and sn-2 positions are both independent targets for free radicals attack the fatty acids esterification pattern determines the spectrum of observed products [8]. The interaction between sn-1/sn-2 position and the type of radical initiator used (hydrophilic/hydrophobic) also complicate the mechanism of Plg peroxidation. In this study, different synthetic plasmenylcholines both in organic solvents and liposomal membranes or micellar suspensions were examined. The effect of PlgPC on peroxidation of unsaturated lipids, induced by hydrophobic iron complex, Fenton system and azo-initiators (AAPH, AMVN) was analyzed. Although different oxidants, such as ROS, RNS, halogenating species have been used to induce oxidative stress, in experiments described in this paper, only simple iron-catalyzed oxidation of PC or peroxyl radical-mediated reactions with PC, were employed. The results clearly show that radical-induced peroxidation of PlgPC involves different chemistry compared to reactions with diacyl phospholipids. Importantly, the antioxidant action of plasmenlcholine depends on the composition of neighboring lipid molecules (intermolecular mechanism) and even more so on the PlgPC esterification pattern (intramolecular mechanism). It underlines the importance of intra- and inter-molecular interactions in the resulting oxidation pathway. While sn-2 monounsaturated PlgPC radical reactivity does not differ, to any significant extent, from their diacyl analogs, the sn-2 PUFA PlgPC supports the intramolecular protective function of PlgPC. The lack of antioxidant action of the mono-unsaturated PlgPC in multicomponent lipid systems is worth emphasizing, considering that the observed result is inconsistent with the postulated protective function of plasmalogens in vitro [9] . The simplest model, examined in this study consisted of one-component homogenous solution of lipids. In such systems, structural changes caused by oxidation and interaction between lipids, could be neglected. The analysis of the effects of plasmalogens on peroxidation of unsaturated lipids was then continued in micelles and liposomes, which can be viewed as very simple models of biomembrane organization. This is an important consideration because such model systems were expected to provide information about structure-dependent oxidation pathways. The additional complexity in lipid research arises from the fact that the headgroup reactivity is often neglected. By choosing PlgPC as the experimental model, the studied lipid interactions were substantially simplified. This is because of the absence of reactive amine group (compared to PlgPE), which enabled us to focus only on the fatty acyl chains impact on peroxidation of lipids. Such analysis was particularly informative in the case of the synthetic PlgPC 16:0/16:0, which allowed us to assess the impact of sn-1 vinyl-ether bond on the lipid peroxidation. As this molecule is a weak substrate of peroxidation monitored by oxygen uptake we systematically introduced other

neighboring unsaturated lipids in our research. The obtained results show that the presence of other unsaturated lipids in mixtures such as LA or PC 18:0/20:4 is not sufficient to unambiguously demonstrate antioxidant properties of PlgPC. On the other hand the protective action of PlgPC becomes apparent when the plasmalogen molecule contains in the sn-2 position highly unsaturated PUFA. At the end, we introduce cholesterol molecule to PlgPC16:0/16:0 liposomes as it was speculated in literature that Chol interact with plasmalogen both structurally [10] and mechanistically [11].

Materials and Methods 1.1 Chemicals Lipids: DMPC, DPPC, SAPC, SOPC, Chol and plasmalogens (PlgPC 16:0/18:1, PlgPC 18:0/18:1, PlgPC 18:0/20:4, PlgPC 18:0/22:6) were purchased from Avanti Polar Lipids, Inc., Alabaster, USA. The saturated plasmenylcholine, PlgPC 16:0/16:0 was synthesized inprofessor D.H.Thompson laboratory, Purdue

University,

West

Lafayette,

IN,

USA.

The

organic

solvents,

benzene,

carbon

tetrachloride,methanol,acetonitryl,isopropanol were from Merck, Darmstadt, Germany.The inorganic salts for PBS buffer preparation, KCl, NaCl, Na2HPO4*12H2O, KH2PO4 and TCA were from Polish Chemicals POCH, Gliwice, Poland. The NaClO4, a support electrolyte for HPLC-EC mobile phase, was from KochLight-Laboratories Ltd., England. Spin label, mHCTPO for EPR oximetry measurements was a gift from professor H.J.Halpern, Chicago University, Chicago, IL, USA. Other chemicals, LA, BHT and TBA were from Sigma-Aldrich Chemie, Steinheim, Germany and DHB was from Fluka, Taufkirchen, Germany. 1.2 Methods Lipid hydroperoxide determination The total amount of lipid hydroperoxides was determined by iodometric assay. The initial extraction step of sample lipids with 2 : 1 (v/v) chloroform: methanol with the addition of 0.4 mM BHT was performed. Next, potassium iodide (KI)in water (1.2 g KI in 1 ml of water) was added to lipids dissolved in a 3: 2 (v/v) mixture of acetic acid : chloroform. The reaction was stopped after 10 minutes by addition of 0.5 % (w/v) cadmium acetate and centrifuged. The absorption spectra of aqueous phase was measured at 353 nm. TBARS determination A simple thiobarbituric acid reactive substance test was use in 1- or 2- component lipid mixture to detect low molecular weight aldehydic compounds. Shortly, 150 μl of sample was mixed with 200 μl of 20% (w/w) TCA and 11 μl of 250 mM BHT in ethanol, than mixed and centrifuged to sediment precipitated

contaminations. In next step the 250 μl of supernatant was mixed with 250 μl of 0.5% (w/w) TBA and heated 15 minutes at 100oC. The absorbance was read at 535 nm with correction for 800 nm background. EPR oximetry: The oxygen consumption rate in a lipid mixtures was determined using X-band EPR oximetrywith Bruker spectrometer EMEX-AA ER 081(90/30), Bruker BioSpin, Germany. 100 μM mHCTPO spin probe and selected liposome suspension was placed in a quartz EPR flat cell. In case the reaction was initiated by decomposition of azo-compounds in elevated temperature (47oC) the relative oxygen uptake rate was calculated based on phenomenological R-parameter kinetic plots. The R-parameter is calculated based on the EPR spectra of mHCTPO spin label and is described in detail in [12].The detection parameters were set as follows: modulation amplitude 0.063 Gs, time constant 20.48 ms, conversion time 20.48 ms, microwave power 0.265 mW. Clark electrode experiments To register and equilibrate oxygen concentration in a liposome or micellar suspension before adding iron catalyst we used Clark electrode for oxygen uptake measurements. The thermostated micro-electrode vessel OX1-LP, Qubit Systems Inc., Ontario, Canada with a software Logger Pro 3.2 was used. HPLC-EC (Hg-CGMDE) The high performance liquid chromatography separation with electrochemical detection was used to evaluate oxidized cholesterol concentration, namely isomers of cholesterol hydroperoxides (ChOOH). We used 400 EG&G Princeton Applied Research EC detector, USA with reverse-phase column UltrasphereTM XL-ODS (3μm x 4.6 mm x 70mm), Beckman Coulter Inc., USA and HP ChemStation Rev.A.06.01software. As a mobile phase we used 72 % (v/v) methanol, 11% (v/v) acetonitrile, 8% (v/v) isopropanol and9% (v/v) water with 1 mMNaClO4 as a supporting electrolyte. The ChOOH determination was based on synthetic standards of 7-α/β-ChOOH epimers and 5α-ChOOH with the correction for cholesterol loading. MALDI-TOF-MS Chloroform solutions of plasmalogens and chloroform extracts of liposome suspensions were directly mixed (1:1, v:v) with the DHB matrix solution (0.5 M in methanol). The resulting solutions (about 1 µl per sample)were then spotted onto the MALDI target (MTP 384 massive gold target). After evaporation of the solvents, samples were directly analysed by MALDI-TOF MS. All MALDI-TOF mass spectra were acquired on a Bruker Autoflex mass spectrometer (Bruker, Germany). The system utilizes a pulsed nitrogen laser, emitting at 337 nm, the extraction voltage was 20 kV and gated matrix suppression was applied to prevent detector saturation. 200 single laser shots were used for each mass spectrum to average inhomogeneous sample deposition. The laser power was kept about 5% above

threshold to obtain an optimum signal-to-noise ratio. All spectra were acquired in the positive ion reflector mode using delayed pulsed extraction. Further analysis of the spectra was done by the instrument software “Flex Analysis 3.0” (Bruker Daltonics, Bremen).

Liposomes and micelles preparations Multilamellar liposomes (MUV) were prepared from chloroform solution, the lipid films were dried under vacuum overnight and than resuspended in a heated to 37oC PBS buffer (150 mM) pH 7.4. In case of small unilamellar liposomes (SUV), MUV were additionally freeze-thaw in liquid nitrogen and passed through 100 nm-diameter membrane with extruder. Micelles were dried from chloroform with the final 4 mM concentration of Triton-X 100 and than resuspended in PBS buffer. Statistical analysis The results are presented as a mean of repeated measurements ± standard error of the mean (SEM). As the independent samples were repeated two to three times the correction for SEM according to two-tailed Student t-distribution was done with the probability α=0.3174. Statistical significance was assessed by tStudent test with the probability α = 0.05. In case of kinetic plots the representative data were chosen. 2. Results and Discussion 2.1 Oxidation of plasmalogens in homogeneous phase The simplest model of plasmalogen oxidation is homogeneous solution of the lipid with strictly defined esterification pattern in organic solvent. We chose a synthetic plasmenylcholine with saturated fatty acids in sn-1/sn-2 position, namely PlgPC 16:0/16:0. In this simple model, only one bond, the vinylether functionality is susceptible to radical attack. Dissociation energy of this bond is calculated to be around 81 kcal/mol what enables a relatively easy abstraction of an allylic hydrogen atom from this position [13]. The oxidation reaction was carried out in solvents of different polarity and in the presence of different radical initiators. Fig.1. refers to such a reaction in carbon tetrachloride induced by the decomposition of thermo-labile hydrophobic azo-initiator AMVN. Under the conditions used, there was no measurable lipid hydroperoxides in the solution of PlgPC 16:0/16:0.

The observable results could be due to the following: First, in AMVN-derived peroxyl radical reactions no production of 1’-allylic hydroperoxide occurs, unlike in reaction between plasmalogens and singlet oxygen, as demonstrated before [14]. Second, the radical formed is unstable in CCl4 or low yield of the reaction prevents detection of the hydroperoxide by iodometric assay employed in this study. Similar results were obtained in benzene solution (Fig.2) in the

presence of 1-biphenyl-4-yl-1-methyl-ethyl radicals (BME●) produced by photo-induced breakdown of its precursor diazene. In order to study the allylic 1’-hydroperoxide products of plasmalogen oxidation MALDI-TOF MS analysis was performed on air-oxidized PlgPC 16:0/16:0 after complete evaporation of chloroform (dried film). By performing solid state oxidation, we avoided the possibility of selective loss of oxidation products during the extraction process. At the same time, the changes in the physicochemical properties of bilayers caused by oxidation products could be neglected. However, even under these conditions, we were unable to detect any molecule with a m/z 750.6 Da, that would be the expected hydroperoxide product of PlgPC 16:0/16:0 oxidation. Nevertheless, the lyso- (m/z 496.3 Da) and formyl- (m/z 524.3 Da) products were found in abundance (data not shown). It should be noted, however, that the stability of lipid hydroperoxide products could be reduced in MALDI-TOF MS analysis by UV irradiation or hydroxyl radical generation from the acidic DHB matrix. We repeated this experiments for plasmenylcholines with increasing double bonds in sn-2 position, namely PlgPC 18:0/18:1, PlgPC 18:0/20:4 (both data sets not shown) and PlgPC 18:0/22:6 (Fig.3A and B). For those lipids the corresponding lyso- and formylderivatives could be easily detected until 16 h of solid phase oxidation. However, in the case of PlgPC 18:0/22:6 at longer oxidation times such as 40 h and 64 h of solid phase oxidation only the lyso- (m/z 568.4) and their oxidation products and no more formyl-lipid (m/z 596.5) are detected (Fig.3B). Interestingly, the MALDI-MS spectra of PlgPC 18:0/22:6 extracted from oxidized liposomes were dominated by high intensity peaks of lyso- (m/z 568.4) and formyl- (m/z 596.5) products even after 64 h (data not shown). This observation suggests that either the mechanism or kinetics of oxidation differs in organic solution compared to liposomes suspension. It must be pointed out that plasmenylcholines underwent immediate decomposition into their corresponding lyso- and formyl-products upon handling (Fig.3A) if not stored under argon what does not hold true for their diacyl analogs (Fig.3C). After 64 h of solid phase oxidation of the diacyl analog PC 18:0/22:6 (SDPC) corresponding oxidation products of the sn-2 fatty acyl residue as well as the lyso-product (m/z 524.4) are detected (Fig.3D). The m/z values and the corresponding structural assignments of the PlgPC 18:0/22:6 and SDPC peaks and their autooxidation products are shown in Tab.2. and Tab.3. in Supplementary Material. Further structural explanation of the oxidation products is shown in Scheme 2. and Scheme 3. The lyso-lipid is the stable end product of singlet oxygen addition to vinyl-ether bond (both by ene- and dioxethane- pathway) or acid-catalyzed hydrolysis [15]. It is also possible to detect lyso-lipid after enzymatic cleavage of vinyl ether by the mechanism dependent on myeloperoxidase [16]. On the other hand, the transient formyl-product is formed in reactions with electrophilic singlet oxygen or ozone. It could be an intermediate in a complex Criegee rearrangement of ozonide-PL [17] and a thermal decomposition product of dioxethane-PL pathway. The formyl-PL product is also supposed to arise from

allylic plasmenylcholine hydroperoxide degradation by Hock cleavage. Although this pathway is chemically feasible it seems physiologically irrelevant [14], [18]. In summary, it was not possible to detect plasmenylcholine hydroperoxide in radical induced oxidation of PlgPC 16:0/16:0 in organic solvents. The solid phase oxidation of sn-2 saturated PlgPC produced degradation products of limited complexity, namely sn-1 formyl- and lyso- derivatives. The sn-2 position was stable until we introduced PUFA which increased the amount of products formed as it was in the case of SDPC oxidation. Murphy and his colleagues [13],[19],[20] demonstrated that there is a strict interaction between esterified fatty acyl chains in plasmalogen molecules and that both chains could generate the first initiating radical species. Then the radical can be transferred intramolecularly between the fatty acyl chains or intermolecularly between adjacent molecules. It could be assumed that molecular organization of lipids in liposomes or micelles enhances the intermolecular reaction that should not be the case in homogenous solutions.

3.2. Oxidation of plasmalogens in liposomal and micellar suspensions

There is a growing body of evidence that supramolecular organization of lipid molecules [21], the regioizomeric effect [22] and the influence of proteins in LDL or HDL [23] bioagregates have a dramatic effect on lipid oxidation pathways. In view of this complexity, we tried to compare the rate of oxygen uptake as a measure of lipid peroxidation in multilamellar liposomes (MUV) enriched with different molecular species of PlgPC. In this experiment (Fig.4) we used 1 mM PC 18:0/20:4 as a control oxidizing substrate. All other liposomes were composed of 1 mM PC 18:0/20:4 with increasing molar ratio of PlgPC, namely 0.5 mM, 1mM and 1.5 mM. The lipids used in this set of experiments were as follows; PlgPC 16:0/16:0, PlgPC 16:0/18:1, PlgPC 18:0/18:1 and PlgPC 18:0/20:4. The initiation of reaction was achieved by addition of ascorbate to a 2-minutes equilibrated mixture of liposomes, hydrogen peroxide and hydrophobic iron complex Fe(HQ)3 into the Clark-type electrode microcuvette. The plasmalogens were differentially effective in inhibiting the oxygen uptake. The most effective, PlgPC 16:0/18:1, reduced oxygen uptake by a factor of 8 in liposomes composed of 1mM PC 18:0/20:4 : 1.5 mM PlgPC 16:0/18:1 compared to control 1mM PC 18:0/20:4 (Fig.4). Such measurements were repeated in small unilamellar vesicles (SUV) composed of analogous lipids, in which even higher, 25-fold decrease of oxygen uptake was observed (data not shown). This indicates that the size and the amount of liposome’s layers affect the oxidation kinetics. To analyze the influence of structural parameters of the lipid molecule on oxidation mechanism, an additional control was employed with the diacylphosphatidylcholine, POPC (PC 16:0/18:1), that contained

the same acyl chain length but was devoid of vinyl-ether bond in sn-1. The presence of 1.5 mM POPC in 1mM PC 18:0/20:4 SUV diminished the rate of oxygen uptake two-fold compared with 1mM pure PC 18:0/20:4 (data not shown). Of particular interest is the fact that 0.5 mM plasmenylcholines (in 1:2 ratio with highly oxidizing control phospholipid) in MUV, irrespective of its sn-2 unsaturation did not cause any statistically significant reduction of oxygen consumption. However, if PlgPC was used in 1:1 (or higher) molar ratio to oxidizable PC18:0/20:4, its antioxidant action was evident. In view of these results, it is questionable to consider PlgPC as a classic antioxidant. Although PlgPC may inhibit peroxidation of other unsaturated lipids, the required concentration of the plasmalogen is comparable or even higher than that of the oxidizing substrate. The next step of our research was quantification of major degradation products of lipid peroxidation, in the presence of PlgPC. The simple experimental design enabled us to trace small molecular weight aldehydic compounds (TBARS) and lipid hydroperoxides (LOOH) formation from linolenic acid in TX100 micelles (Fig.5).

After preparing a micelle solution an aliquot was taken to determine the initial concentration of TBARS and LOOH. The rest of the solution was placed in 37oC thermostated microelectrode chamber. As an initiator of radical process we used argon-saturated FeSO4 solution, that was injected in two consecutive volumes into the micellar suspension. The oxygen uptake was monitored in real time during 13 minutes then the reaction was stopped by immersing the solution in ice. The second aliquots of micelles were taken for the final determination of the accumulated TBARS and LOOH. Given that peroxidation induced by transition metals critically depends on preformed lipid hydroperoxides (e.g. during lipid storage or handling), we compared the initial concentration of LOOH. It can be assumed that in the limit of statistical significance all the micelles shows comparable LOOH levels. Since iron-mediated lipid peroxidation could be a multi-reaction pathway with the formation of hypervalent iron and with the requisite of equilibrium between ferric and ferrous iron [24] we decided to stabilize micelles after first injection of FeSO4. Rather surprisingly, even though in the presence of PlgPC 18:0/18:1 the amount of TBARS and LOOH derived from LA was slightly diminished, the effect was not selective for PlgPC. Both structural analogs, namely diacyl SOPC (PC 18:0/18:1) and alkenyl-acyl (PlgPC 18:0/18:1) lowered the formation of these products to the same extent. It must be stressed that until now only few studies reported a clear antioxidant action of plasmalogens, in systems in which peroxidation of lipids was induced by cooper and iron ions [25],[26]. However, in the cited studies these authors used PlgPE isolated from brain, which contained a whole profile of tissue-specific acylation pattern. In our case we used pure synthetic PlgPC. The other noteworthy difference is that, Zommara et al. [26] started the reaction by addition of cumene hydroperoxide to generate peroxyl and alkoxyl radicals in the initiation phase of lipid peroxidation. The time-scale of performed experiments was also different.

A question arises why an acceleration of oxygen uptake in samples containing PlgPC 18:0/18:1 (Fig.5A) is evident. As the pure plasmenylcholine micelles shows a negligible oxygen consumption, it is not a cumulative effect of VE oxidation and PUFA oxidation in LA. We speculate that increased rate of oxygen uptake could either be an effect of structural modulation of LA by PlgPC or the intermolecular interaction between LA hydroperoxyradical and VE bond. The later hypothesis may be supported by the findings of plasmalogen epoxidation [27],[28] and by mass spectrometric analysis of products derived from free-radical induced plasmalogen oxidation reported by Murphy R.C. [13]. Although our experiments showed that the presence of plasmenylcholines at certain concentration reduce the rate of oxygen uptake in liposomes containing PC18:0/20:4 (Fig.4), this effect could not be attributed to only one type of chemical reaction. On the contrary, in LA micelles (Fig.5) enriched with PlgPC18:0/18:1, the enhancement of oxygen consumption was observed. These results strongly suggest that other products should be characterized to gain insight into the plasmalogen peroxidation process. Another possibility is the influence of plasmalogen, plasmalogen degradation products or linolenic acid degradation products on the susceptibility of LA to water-phase induced peroxidation. This observation indicates that the complexity of lipid mixtures, the molar ratio of their components, the head group as well as the nature of oxidant species and sampling time could dramatically modify the conclusion about the lipid reactivity. The protective effect of plasmalogens on PUFA-PL and PUFA-esters oxidation was unambiguously documented in AAPH-mediated reaction or autooxidation condition, in the presence of brain- and erythrocyte- isolated PlgPE, in micelles and solid phase oxidation [29],[30],[31]. The diversity of possible mechanisms responsible for degradation of plasmalogen molecule and suggestion that there is a strict interaction between VE bond and sn-2 acyl chain and vice versa [32] inspired us to come up with hypothesis that the antioxidant action of plasmalogens occurs only intramolecularly. Experiments described below were performed to provide further evidence supporting this statement. First, the results from EPR-oximetry documented the slower rate of oxygen consumption in SUV composed of DMPC-PlgPC 18:0/20:4 compared to DMPC-SAPC (Fig.6A,B). This is particularly interesting since we did not observe slower oxygen consumption in the presence of PlgPC with sn-2 monounsaturated phospholipid analogs, namely PlgPC 16:0/18:1 vs. POPC and PlgPC 18:0/18:1 vs. SOPC.

Second, in micelles as well as in liposomes composed of either 1 mM SAPC or 1 mM PlgPC 18:0/20:4 we noticed a reduced oxygen uptake in the presence of vinyl-ether bond (Fig.7A and 8A) if the peroxidation was induced by FeSO4. The reaction in micelles was very rapid and more effective (higher oxygen uptake rate) compared to that in liposomes. The lower oxygen uptake was accompanied by a lower accumulation of TBARS and LOOH in samples composed of PlgPC 18:0/20:4 (Fig.7B,C and 8B,C). Since these products originate mainly from esterified arachidonic acid, it seems that VE affects the reaction in the neighboring position of its own

molecule. Our study also demonstrate that plasmalogen molecule experiences an elongation of the lag time with decreasing concentration of catalytic iron (Fig.7A,8A). This could, in part, be explained by the fact that there must be an interfacial interaction of iron with VE before it reaches the highly oxidizable PUFA in sn-2. Other explanation for a lower oxygen uptake, assuming that PlgPC are more ordered in vesicles than diacyl PC [1] and posses a distinct conformational states of glycerol backbone [33] could be a lower oxygen diffusion into hydrophobic milieu of alkenyl-acyl bilayer. However, our preliminary data from EPR saturation recovery experiments with 5-doxylstearate labeled liposomes exclude this possibility. In all studied membranes, the 40 mol% enrichment of PC18:0/20:4 MUV in PlgPC 16:0/16:0, PlgPC 16:0/18:1, PlgPC 18:0/18:1 or PlgPC 18:0/20:4 did not change the oxygen transport parameter compared to pure PC 18:0/20:4 (data not shown). It is an intriguing question whether this behavior is similar in pure PlgPC liposomes. The strongest indication for a role of the surrounding lipids in the resultant reactivity could be inferred from data reported by Murphy et al. [34] where the liberation of oxidized arachidonate from the sn-2 position was observed only in pure PlgPC micelles. In mixed-micelles the oxidized arachidonate remained esterified in a phospholipid molecule [34]. Still unanswered issue is the relationship between reactivity and structural changes accompanying lipid peroxidation. Since the oxidation of plasmalogen vesicles creates a plethora of products in both sn-1 and sn-2 positions with the predominance of sn-1 lyso-PlgPC and formyl-PlgPC we have to take into account their detergent-like activity that could modulate lipid peroxidation process. It was documented in literature that plasmenylcholine bilayers are highly sensitive to perturbation caused by amphiphilic molecules [35]. Additionally, the recent molecular dynamics simulation revealed a great mobility of lipid peroxyl radicals with a preference of intramembrane location, which is against the popular hypothesis of their migration into polar interface [36]. If the mobility of lipid peroxyl radicals is a factor contributing to the rate of hydrogen abstraction from neighboring carbon atom, it could be subjected to modification by membrane packing density and chain geometry. In case of PlgPE, Rog et al. concluded that it formed more rigid and compressed membranes which were also thicker by extended acyl chain conformation [37]. For PlgPC the same conclusion could be drawn from monolayers study [38]. The other factor influencing lipid oxidation is the location of lipid hydroperoxides which are supposed to form transient hydrogen bonds with water, carbonyl and phosphate groups [36]. In plasmalogen molecules the substitution of one ester bond with vinyl ether bond could diminish sn-2 lipid hydroperoxide migration to the polar interface. This behavior would lower the rate of aqueous phase - derived oxidants attack on lipid hydroperoxides located at the ω-end of sn-2 acyl chain and the subsequent radical reaction with VE bond. Experiments carried out by Murphy et al. suggest that the close special proximity of hydroperoxy radicals both from sn-1 or sn-2 facilitates the selective formation of certain isomers of oxidized arachidonate [13]. Definitely the PUFA in sn-2 acyl chain has a tendency to form bended conformation thus positioning only certain carbon atoms (carbon-7 in arachidonic acid, carbon-6 in docosahexaneoic

acid) to be a primary targets for hydrogen abstraction upon interaction with VE bond. Unfortunately this hypothesis remains without direct evidence for DHA [20]. Since the other possibility to propagate the radical reaction between lipid molecules is the attack of hemiacetal hydroperoxy radical formed at 1’– carbon of VE bond on sn-2 acyl chain [13] we made an attempt to detect 1’-hydroperoxide of plasmenylcholine. Our experimental efforts were unsuccessful. There was no plasmenylcholine hydroperoxide generated in saturated PlgPC 16:0/16:0 either in homogeneous media (section 3.1) or in AAPH treated liposomes enriched with cholesterol (data not shown). The question arises whether 1’hydroperoxide from saturated PlgPC is not formed in our experiments or it is inherently unstable. We speculate that the PUFA in sn-2 position could be a necessary requirement for such a hydroperoxide to be formed in the radical-induced pathway. The main process contributing to the delayed oxidative degradation in PlgPC 18:0/20:4 liposomes or micelles remains yet to be established.

3.3 Plasmalogen influence on cholesterol oxidation The colocalization of cholesterol (Chol) and plasmalogens in lipid rafts [39], the role of plasmalogens in vesicular Chol transport from cell membrane surface as well as its HDL-mediated Chol efflux [40] are just a few examples of the interaction of these two molecules in vivo. Several lines of evidence suggest that there is a strict interaction between plasmalogen and cholesterol in lowering the oxidizability of lipid membranes [11],[41]. Additionally, Maeba et al proposed that cholesterol molecule relocates due to hexagonal phase formation facilitated by PlgPE and this translocation exerts antioxidative effect. However, other data published by Wang et al [42] suggests that the hexagonal phase is more prone to peroxidation compared to the lamellar phase. Given our insufficient knowledge of the influence of PlgPC on radical cholesterol oxidation, we analyzed the effect of PlgPC 16:0/16:0 on the accumulation of toxic oxysterols in model membranes. The kinetics of cholesterol hydroperoxides generation was monitored in AAPH-mediated peroxidation in multilamellar liposomes. The total lipid concentration was 2 mM and Chol content was 50 mol%. The ratio of DMPC and PlgPC was subjected to changes in that the PlgPC reaches 10, 20 and 30 mol%. At different time points the reaction was stopped by immersing samples in ice and the cholesterol hydroperoxides were detected by means of HPLC-EC(Hg). Contrary to the proposed protective effect of PlgPE, the saturated plasmenylcholine promotes the formation of 7-ChOOH which is an indicator of the free radical-induced process (Fig.9). We noticed the 3.7-time increase of 7ChOOH/Chol accumulation rate upon addition of 30 mol % of PlgPC 16:0/16:0. This result demonstrates that the presence of vinyl-ether bond is not sufficient to protect Chol against oxidation. Thus the question arises if other structural interaction between cholesterol and plasmenylcholine facilitates Chol oxidation or the effect results from efficient

peroxidation propagation phase and secondary peroxidation effect in the presence of plasmalogens. Consistently with the experiment described above, we noticed the PlgPC 16:0/16:0 concentrationdependent accumulation of 7-ChOOH in homogenous media of CS2 as a minor byproduct of photooxidation in the presence of photosensitizer (data not published).

3. Conclusion: Although peroxidation of lipids has been studied for years, these processes became recently of special interest because they could play an important role in signaling pathways, in which products of lipid breakdown [43] are involved. In literature, plasmalogens are usually considered as an uniform lipid group; however, substantial differences in the headgroup structure and chemical reactivity of these glycerophospholipids suggest that their role in lipid peroxidation should be analyzed separately. Although numerous studies seem to support the postulate about antioxidant action of plasmalogens [9] some researchers reported that they were unable to demonstrate clear antioxidant action of these phospholipids. Thus lack of the protective function of plasmalogens was found in bulk soybean oil [42], in the presence of superoxide anion radical [44] as well as in astrocytes subjected to lactic acid-induced cellular stress [45]. In our experiments we focused on synthetic plasmenylcholines with various unsaturation of fatty acids esterified in sn-2 position. The experiments conducted on such pure molecules enabled us to establish the role of the vinyl ether bond as well as methylene interrupted double bonds in overall reactivity of PlgPC. Among other contributing factors, the likely involvement of lipid molecular structure and its environment in peroxidation mechanism, has received much attention in recent years [22], [46]. The following general conclusions could be drawn based on our study: (1) The antioxidant effect of PlgPC is not only determined by the presence of the vinyl ether bond but it also depends on unsaturation of the sn-1/sn-2 chain, its length and conformation in membranes (2) The efficiency of PlgPC to interact with oxidizing radicals depends on the depth of the membrane penetration by the radicals, the reaction initiation site and structure of the surrounding lipids (e.g. PC 18:0/20:4 vs. LA). (3) Experiments, showing prooxidant effects of PlgPC 16:0/16:0 on cholesterol oxidation induced by radicals, are not consistent with the postulated protective role of plasmalogens. It must be emphasized here that cholesterol - phospholipids interactions are very complicated because they often lead to rearrangement of membrane domains, changes in membrane fluidity and creation of hydrophobic barrier for oxygen diffusion [47]. All these effects could dramatically affect peroxidation reactions.

The present knowledge on antioxidative function of plasmalogen was mainly built on experiments on highly complicated cell systems under photooxidative stress conditions [18]. As the photogenerated singlet oxygen oxidation of plasmenylcholine is a well-documented mechanism with a calculated high interaction rate constant in the range of 107 M-1s-1 [14], [48] only the free radical pathway still needs to be clarified. We were not able to detect 1’-hydroperoxide of plasmenylcholine both in organic solvent nor in liposomes. The challenges for the future will be a precise analysis of radicals that could react with plasmenylcholines. As some radicals are physiologically not relevant (AAPH derived peroxyl radical), unspecific due to high oxidation potential (●OH) or unreactive in lipid peroxidation reaction (O2-●) there must be probably a certain local condition achieved to observe radical oxidation of PlgPC in vivo. One of these requirements is the presence of PUFA in sn-2 position. We have demonstrated the protective intramolecular effect of PlgPC 18:0/20:4 on the oxidation of esterified arachidonate. The data reported by us were consistent in liposomes and micelles. It is worth mentioning, that the high level of unsaturation causes a particular instability in PlgPC 18:0/22:6 molecule compared to PC 18:0/22:6 also during storage. Findings of this study strengthen the concept that plasmalogens reveal its antioxidant function only under specific set of conditions. Lipid peroxidation is a complex process in which the generated products could modulate the overall outcome of the subsequent reactions. This modulation could be achieved both by chemical reactivity and by structural interaction which can destabilize bilayers or cause the detergent-like solubilization. Acknowledgments: The performed experiments were partially supported by a grant Sonata-Bis from the Polish National Science Center (2012/05/E/N23/00473). MALDI-TOF-MS experiments were supported by SET Project from Jagiellonian University (to AB) and German Research Council DFG FU 771/1-3 (to BF).We thank M. Dutka PhD for her help with statistical analysis.

References [1] J.H. Pak, V.P. Bork, R.E. Norberg, M.H. Creer, R.A. Wolf, R.W. Gross, Disparate molecular dynamics of plasmenylcholine and phosphatidylcholine bilayers, Biochemistry (Mosc.). 26 (1987) 4824–4830. [2] H.W. Mueller, J.T. O’Flaherty, D.G. Greene, M.P. Samuel, R.L. Wykle, 1-O-alkyl-linked glycerophospholipids of human neutrophils: distribution of arachidonate and other acyl residues in the ether-linked and diacyl species, J. Lipid Res. 25 (1984) 383–388. [3] R.W. Gross, High plasmalogen and arachidonic acid content of canine myocardial sarcolemma: a fast atom bombardment mass spectroscopic and gas chromatography-mass spectroscopic characterization, Biochemistry (Mosc.). 23 (1984) 158–165. [4] X. Chen, R.W. Gross, Potassium flux through gramicidin ion channels is augmented in vesicles comprised of plasmenylcholine: correlations between gramicidin conformation and function in chemically distinct host bilayer matrices, Biochemistry (Mosc.). 34 (1995) 7356–7364.

[5] A. Mellouki, Atmospheric Fate of Unsaturated Ethers, in: I. Barnes, K.J. Rudzinski (Eds.), Environ. Simul. Chamb. Appl. Atmospheric Chem. Process., Springer Netherlands, 2006: pp. 163–169. http://link.springer.com/chapter/10.1007/1-4020-4232-9_13 (accessed April 26, 2016). [6] G. Wang, T. Wang, The role of plasmalogen in the oxidative stability of neutral lipids and phospholipids, J. Agric. Food Chem. 58 (2010) 2554–2561. doi:10.1021/jf903906e. [7] S. Saab, J. Mazzocco, C.P. Creuzot-Garcher, A.M. Bron, L. Bretillon, N. Acar, Plasmalogens in the retina: from occurrence in retinal cell membranes to potential involvement in pathophysiology of retinal diseases, Biochimie. 107 Pt A (2014) 58–65. doi:10.1016/j.biochi.2014.07.023. [8] K. Gorgas, A. Teigler, D. Komljenovic, W.W. Just, The ether lipid-deficient mouse: tracking down plasmalogen functions, Biochim. Biophys. Acta. 1763 (2006) 1511–1526. doi:10.1016/j.bbamcr.2006.08.038. [9] D. Reiss, K. Beyer, B. Engelmann, Delayed oxidative degradation of polyunsaturated diacyl phospholipids in the presence of plasmalogen phospholipids in vitro, Biochem. J. 323 ( Pt 3) (1997) 807–814. [10] A. Hermetter, K. Lohner, G. Degovics, P. Laggner, F. Paltauf, Effect of cholesterol on vesicle bilayer geometry of choline plasmalogen and comparison with dialkyl-, alkylacyl- and diacylglycerophosphocholines, Chem. Phys. Lipids. 38 (1985) 353–364. [11] R. Maeba, N. Ueta, Ethanolamine plasmalogen and cholesterol reduce the total membrane oxidizability measured by the oxygen uptake method, Biochem. Biophys. Res. Commun. 302 (2003) 265–270. [12] H.J. Halpern, M. Peril, T.-D. Nguyen, D.P. Spencer, B.A. Teicher, Y.J. Lin, M.K. Bowman, Selective isotopic labeling of a nitroxide spin label to enhance sensitivity for T2 oxymetry, J. Magn. Reson. 1969. 90 (1990) 40–51. doi:10.1016/0022-2364(90)90364-F. [13] R.C. Murphy, Free-Radical-Induced Oxidation of Arachidonoyl Plasmalogen Phospholipids:  Antioxidant Mechanism and Precursor Pathway for Bioactive Eicosanoids, Chem. Res. Toxicol. 14 (2001) 463–472. doi:10.1021/tx000250t. [14] D.H. Thompson, H.D. Inerowicz, J. Grove, T. Sarna, Structural characterization of plasmenylcholine photooxidation products, Photochem. Photobiol. 78 (2003) 323–330. [15] O.V. Gerasimov, A. Schwan, D.H. Thompson, Acid-catalyzed plasmenylcholine hydrolysis and its effect on bilayer permeability: a quantitative study, Biochim. Biophys. Acta. 1324 (1997) 200–214. [16] K.R. Wildsmith, C.J. Albert, F.-F. Hsu, J.L.-F. Kao, D.A. Ford, Myeloperoxidase-derived 2chlorohexadecanal forms Schiff bases with primary amines of ethanolamine glycerophospholipids and lysine, Chem. Phys. Lipids. 139 (2006) 157–170. doi:10.1016/j.chemphyslip.2005.12.003. [17] K.M. Wynalda, R.C. Murphy, Low-concentration ozone reacts with plasmalogen glycerophosphoethanolamine lipids in lung surfactant, Chem. Res. Toxicol. 23 (2010) 108–117. doi:10.1021/tx900306p. [18] O.H. Morand, R.A. Zoeller, C.R. Raetz, Disappearance of plasmalogens from membranes of animal cells subjected to photosensitized oxidation, J. Biol. Chem. 263 (1988) 11597–11606. [19] N. Khaselev, R.C. Murphy, Susceptibility of plasmenyl glycerophosphoethanolamine lipids containing arachidonate to oxidative degradation, Free Radic. Biol. Med. 26 (1999) 275–284. [20] K.A. Zemski Berry, R.C. Murphy, Free radical oxidation of plasmalogen glycerophosphocholine containing esterified docosahexaenoic acid: structure determination by mass spectrometry, Antioxid. Redox Signal. 7 (2005) 157–169. doi:10.1089/ars.2005.7.157. [21] E.S. Budilarto, A. Kamal-Eldin, The supramolecular chemistry of lipid oxidation and antioxidation in bulk oils, Eur. J. Lipid Sci. Technol. EJLST. 117 (2015) 1095–1137. doi:10.1002/ejlt.201400200. [22] Z. Shen, C. Wijesundera, Effects of Docosahexaenoic Acid Positional Distribution on the Oxidative Stability of Model Triacylglycerol in Water Emulsion, J. Food Lipids. 16 (2009) 62–71. doi:10.1111/j.1745-4522.2009.01132.x. [23] A. Sevanian, F. Ursini, Lipid peroxidation in membranes and low-density lipoproteins: similarities and differences, Free Radic. Biol. Med. 29 (2000) 306–311. [24] Z. Cheng, Y. Li, What Is Responsible for the Initiating Chemistry of Iron-Mediated Lipid Peroxidation:  An Update, Chem. Rev. 107 (2007) 748–766. doi:10.1021/cr040077w. [25] P.J. Sindelar, Z. Guan, G. Dallner, L. Ernster, The protective role of plasmalogens in iron-induced lipid peroxidation, Free Radic. Biol. Med. 26 (1999) 318–324. doi:10.1016/S0891-5849(98)00221-4.

[26] M. Zommara, N. Tachibana, K. Mitsui, N. Nakatani, M. Sakono, I. Ikeda, K. Imaizumi, Inhibitory effect of ethanolamine plasmalogen on iron- and copper-dependent lipid peroxidation, Free Radic. Biol. Med. 18 (1995) 599–602. doi:10.1016/0891-5849(94)00155-D. [27] R. Felde, G. Spiteller, Plasmalogen oxidation in human serum lipoproteins, Chem. Phys. Lipids. 76 (1995) 259–267. [28] S. Stadelmann-Ingrand, S. Favreliere, B. Fauconneau, G. Mauco, C. Tallineau, Plasmalogen degradation by oxidative stress: production and disappearance of specific fatty aldehydes and fatty alpha-hydroxyaldehydes, Free Radic. Biol. Med. 31 (2001) 1263–1271. [29] W.N. Marmer, E. Nungesser, T.A. Foglia, Oxidation of ethyl hexadec-1-enyl ether, a plasmalogen model, in the presence of unsaturated esters, Lipids. 21 (1986) 648–651. doi:10.1007/BF02537215. [30] C. Leray, J.-P. Cazenave, C. Gachet, Platelet phospholipids are differentially protected against oxidative degradation by plasmalogens, Lipids. 37 (2002) 285–290. [31] N. Khaselev, R.C. Murphy, Structural characterization of oxidized phospholipid products derived from arachidonate-containing plasmenyl glycerophosphocholine, J. Lipid Res. 41 (2000) 564–572. [32] X. Han, X. Chen, R.W. Gross, Chemical and magnetic inequivalence of glycerol protons in individual subclasses of choline glycerophospholipids: implications for subclass-specific changes in membrane conformational states, J. Am. Chem. Soc. 113 (1991) 7104–7109. doi:10.1021/ja00019a003. [33] N. Khaselev, R.C. Murphy, Peroxidation of arachidonate containing plasmenyl glycerophosphocholine: facile oxidation of esterified arachidonate at carbon-5, Free Radic. Biol. Med. 29 (2000) 620–632. [34] X.L. Han, R.W. Gross, Alterations in membrane dynamics elicited by amphiphilic compounds are augmented in plasmenylcholine bilayers, Biochim. Biophys. Acta. 1069 (1991) 37–45. [35] J. Garrec, A. Monari, X. Assfeld, L.M. Mir, M. Tarek, Lipid Peroxidation in Membranes: The Peroxyl Radical Does Not “Float,” J. Phys. Chem. Lett. 5 (2014) 1653–1658. doi:10.1021/jz500502q. [36] T. Rog, A. Koivuniemi, The biophysical properties of ethanolamine plasmalogens revealed by atomistic molecular dynamics simulations, Biochim. Biophys. Acta. 1858 (2016) 97–103. doi:10.1016/j.bbamem.2015.10.023. [37] J.M. Smaby, A. Hermetter, P.C. Schmid, F. Paltauf, H.L. Brockman, Packing of ether and ester phospholipids in monolayers. Evidence for hydrogen-bonded water at the sn-1 acyl group of phosphatidylcholines, Biochemistry (Mosc.). 22 (1983) 5808–5813. doi:10.1021/bi00294a019. [38] L.J. Pike, X. Han, K.-N. Chung, R.W. Gross, Lipid rafts are enriched in arachidonic acid and plasmenylethanolamine and their composition is independent of caveolin-1 expression: a quantitative electrospray ionization/mass spectrometric analysis, Biochemistry (Mosc.). 41 (2002) 2075–2088. [39] N.J. Munn, E. Arnio, D. Liu, R.A. Zoeller, L. Liscum, Deficiency in ethanolamine plasmalogen leads to altered cholesterol transport, J. Lipid Res. 44 (2003) 182–192. [40] R. Maeba, N. Ueta, Ethanolamine plasmalogens prevent the oxidation of cholesterol by reducing the oxidizability of cholesterol in phospholipid bilayers, J. Lipid Res. 44 (2003) 164–171. [41] J. Wang, T. Miyazawa, K. Fujimoto, Z. Wang, T. Nozawa, The inverted hexagonal phase is more sensitive to hydroperoxidation than the multilamellar phase in phosphatidylcholine and phosphatidylethanolamine aqueous dispersions, FEBS Lett. 310 (1992) 106–110. [42] K.M. Eyster, The membrane and lipids as integral participants in signal transduction: lipid signal transduction for the non-lipid biochemist, Adv. Physiol. Educ. 31 (2007) 5–16. doi:10.1152/advan.00088.2006. [43] S. Wallner, G. Schmitz, Plasmalogens the neglected regulatory and scavenging lipid species, Chem. Phys. Lipids. 164 (2011) 573–589. doi:10.1016/j.chemphyslip.2011.06.008. [44] G.A. Jansen, R.J. Wanders, Plasmalogens and oxidative stress: evidence against a major role of plasmalogens in protection against the superoxide anion radical, J. Inherit. Metab. Dis. 20 (1997) 85– 94. [45] B. Fauconneau, S. Stadelmann-Ingrand, S. Favrelière, J. Baudouin, L. Renaud, A. Piriou, C. Tallineau, Evidence against a major role of plasmalogens in the resistance of astrocytes in lactic acidinduced oxidative stress in vitro, Arch. Toxicol. 74 (2001) 695–701. [46] A. Wolnicka-Glubisz, M. Lukasik, A. Pawlak, A. Wielgus, M. Niziolek-Kierecka, T. Sarna, Peroxidation of lipids in liposomal membranes of different composition photosensitized by

chlorpromazine, Photochem. Photobiol. Sci. Off. J. Eur. Photochem. Assoc. Eur. Soc. Photobiol. 8 (2009) 241–247. doi:10.1039/b809887e. [47] W.K. Subczynski, A. Wisniewska, J.S. Hyde, A. Kusumi, Three-dimensional dynamic structure of the liquid-ordered domain in lipid membranes as examined by pulse-EPR oxygen probing, Biophys. J. 92 (2007) 1573–1584. doi:10.1529/biophysj.106.097568. [48] A. Broniec, R. Klosinski, A. Pawlak, M. Wrona-Krol, D. Thompson, T. Sarna, Interactions of plasmalogens and their diacyl analogs with singlet oxygen in selected model systems, Free Radic. Biol. Med. 50 (2011) 892–898. doi:10.1016/j.freeradbiomed.2011.01.002.

Fig.1.The kinetics of lipid hydroperoxide accumulation in reaction with radicals generated from thermal decomposition of 10 mM AMVN at 40oC in carbon tetrachloride (CCl4). Circles – 1 mM PlgPC 16:0/16:0, triangles – 2mM POPC, diamonds – 10 mM Chol.

Fig.2. The kinetics of lipid hydroperoxide accumulation in reaction with radicals generated during irradiation of 2 mM BME in benzene. Triangles – 2 mM PlgPC 16:0/16:0, circles – 10 mM Chol

Fig.3. Positive ion mode MALDI-TOF mass spectra of unsaturated plasmenylcholine and its diacyl analog prior and after 64 hours of solid phase autooxidation in the darkness: PlgPC 18:0/22:6 before autooxidation (3A) and after 64 hours of autooxidation (3B) as well as SDPC before autooxidation (3C) and after 64 hours of autooxidation (3D).

Fig.4. Electrochemical measurements of oxygen uptake in MUV composed of 1mM PC 18:0/20:4 with the addition of increasing molar ratio of different PlgPC. The reaction was initiated by Fenton chemistry with 10 μM Fe(HQ)3, 100 μM Asc, 5 μM H2O2 at 370C. Green bar represents a control of pure 1mM PC 18:0/20:4 liposomes and following bars are grouped with increasing molar ratio of 0.5 mM/ 1mM/ 1.5 mM of PlgPC molecular species. The inset shows the kinetics of oxygen uptake for control 1mM PC 18:0/20:4 liposomes measured by Clark-type microelectrode. The dashed line in the inset represents the time point of ascorbate addition. Statistical significance is calculated in comparison to control PC 18:0/22:4.

Fig.5. The set of experiments on the 200 uM FeSO4 oxidation of 1mM LA micelles enriched with 1mM DMPC or 1 mM SOPC or 1 mM PlgPC 18:0/18:1. The control one-component micelles were as follows: 1 mM LA, 1 mM SOPC, 1 mM PlgPC with 4mM TX-100. 5A – the oxygen uptake in micelles measured by Clark electrode. 5B – the accumulation of MDA in micelles, black bars represent TBARS concentration before reaction begins, striped bars represents TBARS concentration in the solution taken from the microelectrode chamber after reaction was stopped. 5C - the accumulation of LOOH in micelles, black bars represent LOOH concentration before reaction begins, striped bars represents LOOH concentration in the solution taken from the microelectrode chamber after reaction was stopped.

Fig.6. EPR- oximetry measurements of 5 mM SUV oxidation initiated by 10 mM AAPH at 470C. 6A circles – 5 mM DMPC, squares - 3 : 2 mM DMPC : PlgPC 18:0/20:4, triangles – 3 : 2 mM DMPC : PC 18:0/20:4. 6B- rate of parameter R increase as a measure of oxygen uptake calculated according to kinetic data obtained in 6A. Second, in

micelles as well as in liposomes composed of either 1 mM SAPC or 1 mM PlgPC 18:0/20:4 we noticed a reduced oxygen uptake in the presence of vinyl-ether bond (Fig.7A and 8A) if the peroxidation was induced by FeSO4.

Fig.7. The set of experiments on the FeSO4 oxidation of either 1 mM SAPC or 1 mM PlgPC 18:0/20:4 micelles. 7A – the kinetics of oxygen uptake measured by Clark microelectrode in samples of PlgPC 18:0/20:4 upon addition of different concentration of Fe(II) at 2 minutes after oxygen concentration was equilibrated, 100 μM Fe(II) – solid line, 50 μM Fe(II) – dotted line, 25 μM Fe(II) – dashed line. 7B – rate of oxygen uptake in SAPC micelles (empty bars) or PlgPC 18:0/20:4 micelles (green bars) oxidized by addition of different concentration of Fe(II). 7C – TBARS concentration determined after terminating the electrode reaction, SAPC (empty bars) and PlgPC 18:0/20:4 (checkered bars). 7D – LOOH concentration determined after terminating the electrode reaction, SAPC (empty bars) and PlgPC 18:0/20:4 (checkered bars). Blue bar represent the control oxidation of DMPC.

Fig.8. The set of experiments on the FeSO4 oxidation of either 1 mM SAPC or 1 mM PlgPC 18:0/20:4 multilamellar liposomes. 8A – the kinetics of oxygen uptake measured by Clark microelectrode in samples of PlgPC 18:0/20:4 upon addition of different concentration of Fe(II) at 2 minutes after oxygen concentration was equilibrated, 100 μM Fe(II) – solid line, 50 μM Fe(II) – dotted line, 25 μM Fe(II) – dashed line. 8B – rate of oxygen uptake in SAPC liposomes (empty bars) or PlgPC 18:0/20:4 liposomes (green bars) oxidized by addition of different concentration of Fe(II). 8C – TBARS concentration determined after terminating the electrode reaction, SAPC (empty bars) and PlgPC 18:0/20:4 (checkered bars). 8D – LOOH concentration determined after terminating the electrode reaction, SAPC (empty bars) and PlgPC 18:0/20:4 (checkered bars). Blue bars represent the control oxidation of DMPC.

Fig.9. Set of experiments on free radical-induced oxidation of Chol in liposomes. 9A – HPLC-EC(Hg) chromatogram of cholesterol oxidation products extracted from 2 mM MUV (DMPC-Chol-PlgPC 16:0/16:0 at 20 mol%-50mol%-30 mol% ratio) oxidized with 10 mM AAPH. In the order of decreasing signal intensity are: 7ChOOh standard (dashed red line), the products derived from MUV after 50-, 40-, 20-, 0- minutes incubation with AAPH. 9B – the kinetics of 7-ChOOH accumulation in 2 mM MUV subjected to 10 mM AAPH-mediated oxidation. Circles – DMPC-Chol (50 mol%- 50 mol%) MUV, squares – DMPC-Chol-DPPC (20 mol% - 50 mol% - 30mol%) MUV and triangles - DMPC-Chol-PlgPC (20 mol% - 50 mol% - 30 mol%) MUV. 9C – the HPLCseparation based calculated ratio of 7ChOOh/Chol in MUV enriched with different concentration of PlgPC16:0/16:0. Bar description: control - DMPC-Chol (50 mol%- 50mol%), 30mol% DPPC - DMPC-Chol-DPPC (20 mol% - 50 mol% - 30 mol%), 10 mol% PlgPC- DMPC-Chol-PlgPC (40 mol% - 50 mol% - 10 mol%), 20 mol% PlgPC - DMPC-CholPlgPC (30 mol% - 50 mol% - 20 mol%) and 30 mol% PlgPC - DMPC-Chol-PlgPC (20 mol% - 50 mol% - 30 mol%)

Scheme.1.The schematic representation of important factors affecting lipid peroxidation in liposomes. The two major products of sn-1 PlgPC oxidation, that is a formyl-PL and lyso-PL are a fingerprint of plasmalogen appearance in membranes. The increasing content of lipid degradation products with changed hydrophobic-hydrophilic balance can affect further peroxidation and also stability of lipid bilayers.

Scheme 2. Main oxidation products of PlgPC 18:0/22:6 from cleavage of the alkenyl ether in the sn-1 position: a mixture of 1-formyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine (formyl PC, formyl LPC 22:6) and 1- hydroxy2-docosahexaenoyl-sn-glycero-3-phosphocholine (lyso PC, LPC 22:6).

Scheme 3. Scheme of the cleavage of unsaturated fatty acyl residues at a defined double bond. The reaction is initiated by the addition of oxygen to the double bond and the subsequent cleavage of this product under generation of the corresponding aldehydes. The provided mass differences indicate the mass differences related to the lipid species containing a docosahexaenoyl residue in sn-2 position (PlgPC 18:0/22:6 and PC 18:0/22:6) found in the MALDI-TOF spectra in Fig. 3. Aldehydes are easily oxidized to the corresponding organic acids which lead to an additional peak with a mass difference of +16 Da in comparison to the corresponding aldehyde.

Highlights:    

Antioxidant function of plasmalogens is strongly dependent on its esterification pattern. The radical reactivity of sn-2 monoenoic plasmalogens is comparable to its diacyl analogs. The plasmalogen 1’-hydroperoxide was not detected in examined free radical-induced oxidation pathway. Cholesterol is not protected by the presence of PlgPC in liposomes oxidized by AAPH.