Two separate mechanisms are involved in membrane permeabilization during lipid oxidation

Abstract

Lipid oxidation is a universal degradative process of cell membrane lipids that is induced by oxidative stress and reactive oxygen and nitrogen species (RONS) in multiple pathophysiological situations. It has been shown that certain oxidized lipids alter membrane properties, leading to a loss of membrane function. Alteration of membrane properties is thought to depend on the initial membrane lipid composition, such as the number of acyl chain unsaturations. However, it is unclear how oxidative damage is related to biophysical properties of membranes. We therefore set out to quantify lipid oxidation through various analytical methods and determine key biophysical membrane parameters using model membranes containing lipids with different degrees of lipid unsaturation. As source for RONS, we used cold plasma, which is currently developed as treatment for infections and cancer. Our data revealed complex lipid oxidation that can lead to two main permeabilization mechanisms. The first one appears upon direct contact of membranes with RONS and depends on the formation of truncated oxidized phospholipids. These lipids seem to be partly released from the bilayer, implying that they are likely to interact with other membranes and potentially act as signaling molecules. This mechanism is independent of lipid unsaturation, does not rely on large variations in lipid packing, and is most probably mediated via short-living RONS. The second mechanism takes over after longer incubation periods and probably depends on the continued formation of lipid oxygen adducts such as lipid hydroperoxides or ketones. This mechanism depends on lipid unsaturation and involves large variations in lipid packing. This study indicates that polyunsaturated lipids, which are present in mammalian membranes rather than in bacteria, do not sensitize membranes to instant permeabilization by RONS but could promote long-term damage.

Figure 1

Plasma setup and lipids used in this study. (A) The plasma source of kINPen IND. One milliliter of LUV suspension is treated with plasma at a distance of 2 cm to prevent the plume touching the suspension. The vector gas stream agitates the suspension evenly during treatment. Milli-Q-H₂O is added into the suspension via a water pump during treatment to compensate for volume loss due to water evaporation. (B) Chemical structure of phosphocholine lipids. 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (PC(16:0/18:1)), 1,2-dioleoyl-sn-glycero-3-phosphocholine (PC(18:1/18:1)), 1-palmitoyl-2-linoleoyl-sn-glycero-3-phosphocholine (PC(16:0/18:2)), and 1,2-dilinoleoyl-sn-glycero-3-phosphocholine (PC(18:2/18:2)).

Figure 2

Permeabilization kinetics, characterized via K⁺ release, are drastically altered upon post-incubation. K⁺ release in percentage of LUVs made with PC(16:0/18:1) (red circles), PC(18:1/18:1) (green squares), PC(16:0/18:2) (purple diamonds), and PC(18:2/18:2) (blue triangles) induced by CAP after (A) direct treatment and (B) 20-h post-incubation. (C) Rate constant (k) of membrane permeabilization according to an exponential plateau model described in section “materials and methods.” One-way ANOVA analysis was used to compare k values between direct effects and 20-h post-effects for all LUVs, respectively. The curves represent an average from at least two independent experiments done in triplicate (n ≥ 6, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001). Error bars correspond to standard deviation.

Figure 3

Laurdan (A) and di-4-ANEPPDHQ GP (B) are altered differently upon CAP treatment, whereas the largest changes are observed in polyunsaturated lipids upon 20-h post-incubation. Effect on GP upon direct CAP treatment (dotted darker line) and 20-h post-incubation (dashed lighter line) in different LUV suspensions. Two-way ANOVA analysis was used to compare the differences in GP after treatment toward the corresponding negative controls. The GP values were from the average of at least two independent experiments, done in triplicate (n ≥ 6, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001). Error bars correspond to standard deviation.

Figure 4

Production of oxidized products in LUVs containing lipids with different degrees of unsaturations, induced by CAP directly after treatment (left panels) and 20-h post incubation (right panels) expressed as percentage of total phospholipids. (A) Total amount of hydroperoxide functional groups, (B) total amount of aldehyde and ketone functional groups, and (C) total amount of carboxylic acid functional groups as a percentage of total phospholipids. The total amount of hydroperoxide functional groups were determined by FOX assay (see section “materials and methods”). Two-way ANOVA analysis was used to compare the difference between CAP-treated samples and corresponding negative control for both direct treatment and 20-h post-incubation respectively. At least two independent experiments in duplicate were performed for each assay (n ≥ 4, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001). Error bars correspond to standard deviation.

Figure 5

Production and diversity of individual lipid hydroperoxides and lipid aldehydes and/or ketones induced by CAP. Production and diversity of lipid hydroperoxides in (A) PC(16:0/18:2) and (B) PC(18:2/18:2) LUVs induced by CAP after direct treatment and 20-h post-incubation. Distribution of oxidatively truncated lipid aldehydes and ketones (ALDO + ketone oxbreaks), ketone oxygen addition products (ketone adducts), or DNPH signal originating from peaks that contain both types of products (unseparated) out of total DNPH signal in (C) PC(16:0/18:1), (D) PC(18:1/18:1), (E) PC(16:0/18:2), and (F) PC (18:2/18:2) directly after CAP treatment and 20-h post-incubation. The proposed structures of lipid aldehydes and ketones are presented in supplementary files (Supplemental data file S1. Lipid oxides PC(16:0/18:1), Supplemental data file S2. Lipid oxides PC(18:1/18:1), Supplemental data file S3. Lipid oxides PC(16:0/18:2), Supplemental data file S4. Lipid oxides PC(18:2/18:2)). Two-way ANOVA analysis was used to compare the difference between CAP-treated samples and corresponding negative control for both direct treatment and 20-h post-incubation, respectively. At least two independent experiments were performed in duplicate for the quantification hydroperoxides, aldehydes, and/or ketones (n ≥ 4, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001). Error bars correspond to standard deviation.

Figure 6

Interfacial tension determination with pendant-drop tensiometry. (A) Interfacial tension of PBS buffer containing LUVs either directly after CAP treatment or after 20-h post-incubation; 0.1% Triton X-100 was used as positive controls. (B) Proposed process of interfacial tension reduction induced by OxPCs in all above LUVs. Since PC lipids tend to form lipid bilayers, in negative control (LUV Ctrl), there are no free lipids in solution or micelle formation on the timescale of the experiment. When a certain amount of OxPCs is formed, some OxPCs are released from lipid bilayers due to the increased hydrophilicity. Afterward, those released OxPCs adsorb at the buffer-air interface, resulting in a reduction of surface tension. Two independent experiments were performed for interfacial tension measurements and averages are displayed in the curves.

Figure 7

PCA displaying the development of average membrane properties and surface tension in parallel to the formation of specific OxLipids (brown arrows). (A) Untreated (red group) vs. 40-min CAP treatment (blue group), and (B) 40-min CAP treatment without (red group) vs. with 20-h post-incubation (blue group). Modifications of properties from individual LUV samples are highlighted by green (monounsaturated lipids) and purple arrows (polyunsaturated lipids).