Investigation of Lipoproteins Oxidation Mechanisms by the Analysis of Lipid Hydroperoxide Isomers

The continuous formation and accumulation of oxidized lipids (e.g., lipid hydroperoxides (LOOH)) which are present even in plasma lipoproteins of healthy subjects, are ultimately considered to be linked to various diseases. Because lipid peroxidation mechanisms (i.e., radical, singlet oxygen, and enzymatic oxidation) can be suppressed by certain proper antioxidants (e.g., radical oxidation is efficiently suppressed by tocopherol), in order to suppress lipid peroxidation successfully, the determination of the peroxidation mechanism involved in the formation of LOOH is deemed crucial. In this study, to determine the peroxidation mechanisms of plasma lipoproteins of healthy subjects, we develop novel analytical methods using liquid chromatography-tandem mass spectrometry (LC-MS/MS) for 1-palmitoyl-2-linoleoyl-sn-glycero-3-phosphocholine hydroperoxide (PC 16:0/18:2;OOH) and cholesteryl linoleate hydroperoxide (CE 18:2;OOH) isomers. Using the newly developed methods, these PC 16:0/18:2;OOH and CE 18:2;OOH isomers in the low-density lipoprotein (LDL) and high-density lipoprotein (HDL) of healthy subjects are analyzed. Consequently, it is found that predominant PC 16:0/18:2;OOH and CE 18:2;OOH isomers in LDL and HDL are PC 16:0/18:2;9OOH, PC 16:0/18:2;13OOH, CE 18:2;9OOH, and CE 18:2;13OOH, which means that PC and CE in LDL and HDL are mainly oxidized by radical and/or enzymatic oxidation. In conclusion, the insights about the oxidation mechanisms shown in this study would be useful for a more effective suppression of oxidative stress in the human organism.

From these circumstances, in this study, in order to prove the proposed hypothesis (i.e., LOOH present in lipoproteins of healthy subjects is mainly derived from radical and/or enzymatic oxidation), we develop novel analytical methods for all the possible PC 16 Figure 1). Authentic references of these LOOH are prepared and analyzed using liquid chromatography-tandem mass spectrometry (LC-MS/MS). The analyses are performed in each of the lipoprotein classes (chylomicron (CM), very-lowdensity lipoprotein (VLDL), low-density lipoprotein (LDL), and high-density lipoprotein (HDL)) to investigate whether there are differences between the oxidation mechanisms of individual lipoproteins. The insights obtained in this study will provide a better understanding of the choice of antioxidants to be incorporated in our daily life, based on the oxidation mechanism taking place in our body.

Preparation of Healthy Human Plasma and Lipoproteins
Blood samples were collected from 9 healthy subjects (4 men and 5 women, mean age 25 ± 4 years). The subjects were non-obese and had no diagnosed abnormality of the liver by health check. Additionally, they did not have any dietary restrictions. Participants ate bread (100 g) with butter (8 g) after 12 h of fasting. Three hours after the meal, blood samples were intravenously collected with ethylenediaminetetraacetic acid (EDTA). Obtained blood was centrifuged (1000× g for 10 min at 4 • C) to prepare plasma samples. Immediately, using the plasma (5.4 mL), the preparation of lipoproteins (CM, VLDL, LDL, and HDL) was performed as in a previous report with little modification [16]. PBS (including 250 mM EDTA, pH 7.4 (5.4 µL)) was mixed with the plasma. On top of the mixture, PBS (including 250 µM EDTA, pH 7.4 (1.8 mL)) was layered and the sample was centrifuged at 36,700× g for 30 min at 4 • C (Optima L -100 XP, Beckman Coulter, Inc., Brea, CA, USA) with a rotor (Type70.1Ti, Beckman Coulter, Inc., Brea, CA, USA). Upper layer (900 µL) was corrected as a CM fraction. Top 900 µL of the remainder was discarded to remove residual CM. Subsequently, PBS (including 250 µM EDTA, pH 7.4 (1.8 mL)) was layered and the sample was centrifuged at 370,500× g for 4.05 h at 4 • C. Upper layer (1080 µL) was corrected as a VLDL fraction. Top 720 µL of the remainder was discarded to remove residual VLDL. The remainder was mixed with potassium bromide aq. (50% (w/v), 1080 µL) to adjust the density to 1.063 g/mL. After centrifugation (370,500× g for 4.05 h at 4 • C), upper layer (1080 µL) was corrected as a LDL fraction. Top 1980 µL of the solution was discarded to remove residual LDL. Finally, the remainder was mixed with potassium bromide aq. (50% (w/v), 3574 µL) to adjust the density to 1.210 g/mL. After centrifugation (370,500× g for 6.46 h at 4 • C), upper layer (900 µL) was corrected as a HDL fraction. LDL and HDL fractions were dialyzed against PBS containing 250 µM EDTA. The purity of each lipoprotein was checked by agarose gel electrophoresis. Each lipoprotein sample was stored at −80 • C until the extraction.
With regard to the analysis of lipid oxidation products, it is generally stated that the analysis advances along with the evolution of mass spectrometry [20]. However, even in the use of latest instruments, the analysis of LOOH isomers is a great challenge [21][22][23]. Traditionally, the LOOH isomeric structure is analyzed in the form of fatty acid derivatives after some derivatizations (e.g., reduction, trimethylsilylation, and methanolysis) [24]. To overcome such a complication, Ag + coordination mass spectrometry without derivatization based on Hock fragmentation was previously developed [25][26][27]. This method, however, cannot distinguish half of the hydroperoxide isomers because Hock fragmentation provides the same product ions from the isomers originating from the same double bond (e.g., FA 18:2;9OOH and FA 18:2;10OOH). This means that it is impossible to distinguish radical oxidation from 1 O 2 oxidation (Figure 1). Under these circumstances, we discovered that collision-induced dissociation (CID) of sodiated PC 16:0/18:2;OOH provided hydroperoxyl group position specific product ions based on α-cleavage [5]. Using this method, we analyzed PC 16:0/18:2;9OOH and PC 16:0/18:2;13OOH in human plasma and hypothesized the contribution of radical and/or enzymatic oxidation to PC 16:0/18:2;OOH formation in human plasma lipoproteins [5]. Recently, we further showed the evidence that CID of sodiated LOOH is available to various lipid classes, including cholesterol ester [13]. Therefore, in this study, to ensure the above hypothesis (i.e., LOOH present in plasma lipoproteins are mainly derived from radical and/or enzymatic oxidation), we firstly developed novel analytical methods for PC 16:0/18:2;OOH and CE 18:2;OOH isomers.
Using the newly developed methods, PC 16:0/18:2;OOH and CE 18:2;OOH were analyzed in CM, VLDL, LDL, and HDL. In this study, PC 16:0/18:2;OOH and CE 18:2;OOH were mostly under LOQ for CM and VLDL. As a reason, it was considered that PC 16:0/18:2 and CE 18:2 concentrations in CM and VLDL solutions prepared in this study were very low (Table 1). On the other hand, PC 16:0/18:2;OOH and CE 18:2;OOH were clearly detected in LDL and HDL (Figure 4). Regarding the PC 16:0/18:2;OOH isomer in LDL and HDL, in agreement with our previous study [5] Regarding the prooxidants which associate with lipoproteins oxidation, some candidates were reported. For instance, in the subendothelial space, NADPH oxidase, xanthine oxidase, myeloperoxidase, and transition metals would contribute to the LDL oxidation [29]. Interestingly, most of them were radical generators, which was consistent with our results stating that LOOH in lipoproteins are mostly radical oxidation products. For reference, we previously reported that PC 16:0/18:2;OOH accumulated in human erythrocyte [6,30]. Considering the fact that LOOH generated in erythrocyte spontaneously migrates to LDL [31], erythrocyte oxidation may also contribute to the LOOH accumulation in lipoproteins. Additionally, it is interesting to note that photo-therapy also induces plasma lipid oxidation [32]. In human plasma, several photo sensitizers are present, and some of these induce type I photo oxidation (i.e., radical oxidation) [33]. These facts may suggest that the contribution of photooxidation to lipoproteins oxidation is quite possible. So far, it is not certain to what extent each prooxidant contributes to lipoproteins oxidation. However, this study showed that the oxidation mechanism of plasma lipoproteins before illness is comprehensively radical oxidation.
Furthermore, this study showed a higher oxidation ratio of HDL than that of LDL (especially for PC 16:0/18:2;OOH) ( Table 1). A lot of studies reported roughly two antioxidative systems of HDL. One is the detoxification of LOOH, for instance, based on the reduction in LOOH to LOH by Met attributed to apoAI, and the hydrolysis of oxidized fatty acid by PON 1 [34,35]. Another one is the enzymatic correction of LO(O)H in LDL by HDL [36]. These corrected LO(O)H are transported to the liver and metabolized. Summarizing these insights, HDL is considered to play a role for the vehicle of LO(O)H, which might result in the higher oxidation ratio of HDL. This study discussed the oxidation mechanisms focusing on the lipoproteins of healthy humans. Most oxidized lipoproteins would be metabolized in the liver [35,37]. However, the oxidized lipoproteins at a level exceeding the capacity of the metabolism might link to various diseases.

Conclusions
In summary, to determine the oxidation mechanisms of human plasma lipoproteins before illness, analytical methods for PC 16:0/18:2;OOH and CE 18:2;OOH isomers were developed. To the best of our knowledge, this was the first study which analyzed PC 16:0/18:2;OOH and CE 18:2;OOH isomers (partially including the cis-trans isomer) in lipoproteins without any derivatizations. As we hypothesized, it was found that PC and CE in the lipoproteins were mainly oxidized by radical and/or enzymatic oxidation, rather than 1 O 2 oxidation. In addition, from the analysis of cis-trans isomeric structures, it was considered that CE peroxidation was suppressed more effectively than PC. On the other hand, in this study, the concentrations of PC 16:0/18:2;OOH and CE 18:2;OOH in CM and VLDL were mostly under LOQ. Therefore, it is expected that the highly sensitive MS will provide further information about lipoproteins oxidation in future studies. In conclusion, the insights about the oxidation mechanisms shown in this study would be useful for a more effective suppression of oxidative stress.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
All of the data is contained within the article and the supplementary materials.

Acknowledgments:
We thank Junya Ito (Food and Biodynamic Chemistry Laboratory, Graduate School of Agricultural Science, Tohoku University, Sendai, Miyagi, Japan) for the technical assistance.

Conflicts of Interest:
The authors declare no conflict of interest.