TRAFFICKING OF OXIDATIVE STRESS-GENERATED LIPID HYDROPEROXIDES: PATHOPHYSIOLOGICAL IMPLICATIONS

Abstract

Lipid hydroperoxides (LOOHs) are reactive intermediates that arise during peroxidation of unsaturated phospholipids, glycolipids and cholesterol in biological membranes and lipoproteins. Non-physiological lipid peroxidation (LPO) typically occurs under oxidative stress conditions associated with pathologies such as atherogenesis, neurodegeneration, and carcinogenesis. As key intermediates in the LPO process, LOOHs are susceptible to one-electron versus two-electron reductive turnover, the former exacerbating membrane or lipoprotein damage/dysfunction and the latter diminishing it. A third possible LOOH fate is translocation to an acceptor membrane/lipoprotein, where one- or two-electron reduction may then ensue. In the case of cholesterol (Ch)-derived hydroperoxides (ChOOHs), translocation can be specifically stimulated by StAR family trafficking proteins, which are normally involved in Ch homeostasis and Ch-mediated steroidogenesis. In this review, we discuss how these processes can be impaired by StAR-mediated ChOOH and Ch co-trafficking to mitochondria of vascular macrophages and steroidogenic cells, respectively. Protective effects of endogenous selenoperoxidase, GPx4, are also discussed. This is the first known example of detrimental ChOOH transfer via a natural Ch trafficking pathway and inhibition thereof by GPx4.

1. Introduction

Unsaturated lipids, including phospholipids, glycolipids and cholesterol in lipoproteins and cellular membranes are susceptible to non-enzymatic oxidation under conditions of persistent oxidative stress. The membrane/lipoprotein damage caused by this oxidation, which is generally referred to as lipid peroxidation (LPO), can result in loss of normal physiological function [1–4]. These effects may occur in conjunction with (i) natural metabolic processes such as mitochondrial electron transport and NADPH oxidase activation, or (ii) exposure to exogenous oxidative pressure, e.g. from ultraviolet or ionizing radiation [1,2,5]. Free radical-initiated/mediated LPO is a degenerative process that is linked to a variety of pathological conditions, including neurodegeneration, atherogenesis, and carcinogenesis [6–8]. Non-enzymatic LPO can be initiated by oxidants such as hydroxyl radical (HO^●), hydroperoxyl radical (HOO^●), and nitrogen dioxide (NO₂^●), or non-radicals such singlet molecular oxygen (¹O₂), ozone (O₃), and peroxynitrous acid (ONOOH) (Fig. 1). The initiating step for radicals could be abstraction of an allylic hydrogen from an unsaturated lipid (LH), e.g. the sn-2 fatty acyl group of a phospholipid (PL), or the C7 hydrogen of cholesterol (Ch). The resulting lipid radical (L^●) reacts rapidly with ground state O₂ to give a peroxyl radical (LOO^●), which can abstract H from a proximal other lipid, L’H, to give L’^● and a hydroperoxide species (LOOH); in the process, other lipid hydroperoxides will arise. As discussed below, several different immediate fates are possible for these LOOHs, some with pathophysiological potential.

Fig. 1.

Generation of damaging lipid hydroperoxide (LOOH) species in membranes and lipoproteins, and possible fates of these species. Examples of free radical and non-radical LOOH formation are shown; also shown for newly formed LOOHs is damage-enhancing 1-electron reduction, damage-suppressing 2-electron reduction, and translocation to acceptor lipid compartments, where either type of reductive turnover can ensue.

2. One electron reduction of lipid hydroperoxides

In the presence of a reductant such as NADH and a redox metal ion such as iron, a newly formed LOOH can undergo one-electron reduction to give an oxyl radical (LO^●) intermediate. The latter may initiate rounds of membrane- or lipoprotein-damaging chain LPO directly or, more likely (9), after undergoing rearrangement with O₂ addition to give an epoxyallylic peroxyl radical (OLOO^●) (Fig. 1). Initiation of LPO by the non-radical, ¹O₂, gives a LOOH directly, with an allylic shift in the double bond of the target LH and retention of its hydrogen [10]. In the presence of catalytic iron and reductants, this LOOH can undergo one-electron reduction to initiate chain LPO as described for free radical initiators [1–4] (Fig. 1).

Like PLs, Ch is found in all membrane compartments of mammalian cells, most of it located in the plasma membrane (~45 mol % of total lipid) [11]. However, unlike natural PLs, Ch exists as a single molecular species, making its hydroperoxide intermediates (ChOOHs) and alcohol products (ChOHs) easier to separate and characterize than those of PLs. Non-esterified Ch is also found in the surface monolayer of lipoproteins, amounting to 10 wt.% of total lipid in low density lipoprotein (LDL) or ~25% the cholesteryl ester content [12]. Being a monounsaturated lipid, Ch is less susceptible to autoxidation than polyunsaturated PLs, although Ch oxidation gives rise to a variety of potentially mutagenic, metabolically disabling, and cytotoxic species known collectively as “oxysterols” [13,14]. These species include various redox-active ChOOHs and redox-inactive products (ChOHs, epoxides, and a ketone) [13]. In ¹O₂ reactions, three ChOOH isomers arise: 5α-OOH, 6α-OOH and 6β-OOH (Fig. 2), the latter two in much lower yield [ 15]. In free radical-initiated/propagated reactions, two ChOOH epimers predominate, 7α-OOH and 7β-OOH (Fig. 2). These species (occasionally denoted collectively as 7-OOH) have been detected in oxidized LDL (oxLDL) and could play a key role in oxLDL-stimulated atherogenesis [16,17]. They can also impair Ch homeostasis in macrophages and hormone synthesis in steroidogenic cells, both of which will be discussed subsequently.

Fig. 2.

Structures of cholest-5-en-3β-ol (Ch) and Ch hydroperoxide (ChOOH) species generated by free radical (R^●)-mediated reactions: 3β-hydroxycholest-5-ene-7α-hydroperoxide (7α-OOH), 3β-hydroxycholest-5-ene-7β-hydroperoxide (7β-OOH); or by singlet oxygen (¹O₂)-mediated reactions: 3β-hydroxycholest-6-ene-5-hydroperoxide (5α-OOH), 3β-hydroxycholest-4-ene-6α-hydroperoxide (6α-OOH), and 3β-hydroxycholest-4-ene-6β-hydroperoxide (6β-OOH).

3. Two electron reduction of lipid hydroperoxides

As an alternative to one-electron reduction, newly formed ChOOHs or PLOOHs may undergo enzyme-catalyzed two-electron reduction to redox-inactive ChOHs or PLOHs. This is a detoxification process that acts in opposition to toxicity-enhancing one-electron reduction. Although the radical initiator, HO^●, has no physiological scavengers, its potential precursors, superoxide (O₂^●-) and hydrogen peroxide (H₂O₂) can be detoxified by superoxide dismutases and catalase, respectively. In contrast, the non-radical initiator, ¹O₂, has no known enzymatic scavengers. For ¹O₂, therefore, secondary, or back-up defenses need to be engaged, viz. those that can inactivate ¹O₂-generated LOOHs. Such defenders might also inactivate downstream (chain LPO) LOOHs as well as those arising from free radical initiators. Several enzymes with direct or indirect LOOH detoxifying ability are known, Type I glutathione peroxidase (GPx1) and Type 4 glutathione peroxidase (GPx4) being among the most prominent. Each of these enzymes has an active site selenocysteine residue that participates in the glutathione (GSH)-dependent two-electron reduction of organo-peroxides. Monomeric GPx4 (M_r ~20 kDa) can act directly on LOOHs in membrane or lipoprotein environments [18]. However, more abundant tetrameric GPx1 (M_r ~82 kDa) lacks this ability, and for PLOOHs, requires prior fatty acyl-OOH hydrolysis to generate free FAOOHs [18]. At present, GPx4 is the only enzyme known to be capable of catalyzing the direct detoxification of ChOOHs in membranes or lipoproteins [19]. Rate constants for the different isomers with GPx4 varied as follows: 6β-OOH > 7-OOH ~ 6α-OOH >> 5α-OOH [20]. In an early study, a breast tumor cell line overexpressing GPx4 in mitochondria was found to be much more resistant to 7-OOH-induced cytotoxicity than a wild-type control, thus demonstrating the enzyme’s importance in cytoprotection against mitochondrial peroxidative damage [21]. More recently, GPx4 has been found to protect cells against a novel iron/LPO-dependent form of cell death known as ferroptosis [22].

4. Lipid hydroperoxide translocation

As first demonstrated about 20 years ago, damaging chain LPO is not necessarily restricted to a LOOH’s membrane or lipoprotein of origin, but can be disseminated to other lipophilic acceptors via LOOH translocation [23–26]. Spontaneous translocation is favored by factors such as (i) LOOH polarity being greater than that of parent lipids; and (ii) overall lipid mass of an acceptor membrane/lipoprotein being significantly greater than that of a donor. Although all LOOHs can translocate on their own, ChOOHs do so much more rapidly than PLOOHs; however, both classes depart from donors at greater rates than their parent lipids [26]. An early experiment with photoperoxidized red cell membranes as LOOH donors and small unilamellar liposomes in ~10-fold lipid molar excess as acceptors showed that the transfer rate constant for total ChOOH was at least 60-times greater than that for parent Ch [23]. Transfer was found to occur via an aqueous pool, departure from the membrane donors being rate-limiting (23]. For individual ChOOH isomers, first-order transfer rate constants decreased in the following order: 7-OOH > 5α-OOH > 6α-OOH > 6β-OOH, which corresponded with their decreasing hydrophilicity on reverse-phase HPLC [24]. If acceptor membranes are relatively rich in reductants and properly chelated iron, they may be at greater risk of damaging chain peroxidation than donor counterparts. Conversely, if the GSH/GPx4 system is more abundant in acceptors, translocation might serve to detoxify damaging LOOHs. Up to now, however, no definitive evidence in line with either of these mechanisms has been reported.

Lipid transfer (trafficking) proteins are known to play essential roles in lipid metabolism, biogenesis, and membrane structure/function [27,28]. Sterol carrier protein-2 (SCP-2) is a small (13.2 kDa) intracellular protein that mediates translocation of Ch and other sterols as well as various phospholipids [29]. Because of its broad recognition of low polarity ligands, SCP-2 is often referred to as a non-specific lipid transfer protein (ns-LTP). A groundbreaking early study by Vila et al. [30] was the first to demonstrate that SCP-2 not only facilitates Ch and PL transfer from one membrane to another, but ChOOH and PLOOH transfer as well. Using a recombinant SCP-2, these investigators showed that first-order rate constants for SCP-2-accelerated ChOOH transfer from red cell membranes to liposomes decreased in the following order: 7-OOH > 5α-OOH > 6α-OOH > 6β-OOH [30], i.e. the same as observed for spontaneous transfer (above).

5. High performance techniques for analyzing LPO and LOOHs

LPO in general and LOOHs in particular can be detected and quantified by a wide variety of techniques, ranging from the relatively simple to the rather complex. The first category includes classical “bulk-type” spectrophotometric methods such as (i) the thiobarbituric acid (TBA) assay, which detects aldehyde by-products of free radical-mediated LPO, and (ii) the iodometric assay, which determines total LOOH content of an experimental sample [31]. More recent approaches based on gas chromatography (GC), high-performance thin layer chromatography (HPTLC), and high-performance liquid chromatography (HPLC) have allowed LOOHs to be separated and quantified at much higher levels of sensitivity and specificity. Two HPLC-based techniques for LOOH identification and determination are especially noteworthy, one involving post-column chemiluminescence detection (HPLC-CL) [32] and the other, reductive mode electrochemical detection (HPLC-EC) [33–35]. The latter approach has become more important, largely because it’s sensitivity and specificity limits far exceed those of the former. Reverse-phase HPLC with reductive mode (mercury drop) electrochemical detection [HPLC-EC(Hg)] is a prime example. About 30 years ago, the authors of this review were the first to introduce HPLC-EC(Hg) for direct LOOH analysis in biological samples [33]. This technique involves a C₁₈ reverse-phase column, an electrochemical detector in amperometric mode, a uniquely designed flow cell for column effluents, a capillary dispensed mercury drop hanging within the flow cell, and an O₂-purged mobile phase [33–35]. Individual ChOOH species (7α/β-OOH, 5α-OOH, 6α-OOH, 6β-OOH) are well-separated by HPLC-EC(Hg) and their detection limits are in the 100-200 fmol range [33]. PLOOH families (e.g. PCOOH, PEOOH, PSOOH, PIOOH) can also be resolved, each with a detection limit in the 20-30 pmol range [33]. HPLC-EC(Hg) analysis has been proven to be extremely valuable for monitoring different LOOH fates, as described in preceding sections 2–4.

In addition to HPLC-EC(Hg), the authors developed a novel analytical approach based on detection of radiolabeled Ch oxides, i.e. “ChOX” species, which could comprise ChOHs and other products as well as ChOOH intermediates. In this approach, [¹⁴C]Ch is incorporated into a membrane or lipoprotein and used as probe of LPO activity, the [¹⁴C]ChOX produced being separated and quantified by high-performance thin layer chromatography with phosphor-imaging detection (HPTLC-PI) [36]. For an experiment involving iron-catalyzed, one-electron reduction of a non-labeled “priming” peroxide (5α-OOH) in [¹⁴C]Ch-labeled liposomal membranes, the generated [¹⁴C]ChOX included redox-active 7α/β-OOH intermediates and redox-silent products (7α/β -OHs, 7-ketone, 5,6-epoxides) [36]. As 5α-OOH decreased exponentially during incubation, [¹⁴C]7-OOHs formed rapidly, reached an apparent steady-state level, and then decayed as [¹⁴C]7-OHs and other radiolabeled ChOX continued to accumulate. This approach has also been used to detect free radical-mediated (chain) LPO in cells. For example, when L1210 leukemia cells labeled in plasma membrane with [¹⁴C]Ch were exposed to 5α-OOH-containing liposomes, a progressive increase in cellular [¹⁴C]ChOX levels was observed along with a loss of viability [36]. When selenium-depleted cells exhibiting low GPx4 activity were treated similarly, [¹⁴C]ChOX levels and viability loses were much greater, demonstrating GPx4 cytoprotection via detoxification of 5α-OOH and/or downstream LOOHs [36].

High sensitivity/specificity HPTLC-PI for [¹⁴C]ChOXs and HPLC-EC(Hg) for LOOHs provide unique means of tracking free radical-mediated LPO as it progresses in membranes and lipoproteins. Because HPTLC-PI uses Ch as a natural in situ probe, it avoids potential artifacts associated with artificial probes such as spin traps or fluorophores, which could potentially perturb the systems being analyzed to give misleading findings. On the other hand, such approaches, unlike HPTLC-PI and HPLC-EC(Hg), do not require solubilization and extraction of membrane/lipoprotein lipids. However, this apparent advantage cannot overcome the high accuracy and reliability of the two chromatographic approaches, which were heavily used in studies described in the following sections.

6. Negative effects of 7-OOH trafficking on cholesterol homeostasis in macrophages

Vascular macrophages express scavenger receptors which initiate resistance to harmful Ch buildup by activating transport processes, whereby incoming Ch is delivered to mitochondria for 27-hydroxylase (CYP27A1)-catalyzed conversion to 27-hydroxycholesterol (27-OH) [37–39]. Ch delivery is mediated by StAR family trafficking proteins, including cytosolic StarD4/D5, which transport Ch to the mitochondrial outer membrane, where StarD1 (in consort with other proteins) moves it to the inner membrane for processing by CYP27A1 [40]. Elevated oxidative pressure leading to a stressful state underlies many types of cardiovascular disease, including atherosclerosis [41]. This condition is linked to a persistent unrestricted uptake of oxidized LDL (oxLDL) by macrophages via scavenger receptors [41]. Free and esterified Ch comprise a large portion of the lipids in LDL. Not surprisingly, therefore, oxLDL has been found to contain free radical-generated Ch oxides (oxysterols) such as redox-active 7α/β-OOH and inactive products such as 7α/β-OH and 7-ketone [42,43]. Although all of these species have been detected in atherosclerotic lesions, the 7α/β-OOH levels were relatively low, evidently because of their susceptibility to one-electron reduction to give 7α/β-OH [43]. To prevent adverse accumulation of Ch, macrophages are equipped to export it to extracellular acceptors such as high-density lipoprotein (HDL) or its apoprotein (apoA-I). This is an early step in reverse cholesterol transport (RCT) in which excess Ch is delivered to the liver for disposal [38,39]. CYP27A1-induced generation of 27-OH plays a key early role in RCT by activating nuclear LXR/RXR transcription factors, which induce expression of the plasma membrane ATP-binding cassette transporters ABCA1 and ABCG1 [44]. ABCA1 exports Ch to apoA-I while ABCG1exports it mainly to HDL [44]. Knowing that most oxLDL contains measurable levels of 7α/β-OOH, Korytowski and Girotti hypothesized that these species can be co-trafficked with Ch to macrophage mitochondria and upon arrival at the inner membrane, induce chain LPO that damages/disables CYP27AI [45]. The first support for this hypothesis was obtained by showing that 7-OOH uptake by mitochondria of cAMP-activated murine macrophages was StarD1/D4-dependent and set off membrane LPO with ultimate loss of ABCA1 expression [45]. Follow-up studies with human monocyte-derived THP-1 macrophages [46] confirmed these findings and showed, furthermore, that CYP27AI activity and 27-OH output of these cells were substantially reduced when Ch delivery was accompanied by 7-OOH (Fig. 3). In addition, ABCA1/G1 expression was substantially reduced, as was Ch exportation to apoAI or HDL [46]. Importantly, 7α/β-OH or 7-ketone (even at relatively high concentrations) had no negative effects on any of the indicated factors. The studies described [45,46] were the first to show how Ch homeostasis in macrophages can be disrupted by a redox-active homologue (7-OOH) which is trafficked similarly to Ch itself.

Fig. 3.

Effect of 7α-OOH exposure on CYP27A1 activity and 27-OH output of human THP-1 macrophages. Macrophages in 1% FBS/RPMI medium were stimulated with dibutyryl-cAMP, then incubated for 10 h with SUV-borne 7α-OOH at the indicated concentrations in bulk medium. POPC/Ch/7α-OOH (2:1:1 by mol) SUVs were used, along with POPC/Ch (3:1 by mol) SUVs as controls. After washing, cells were homogenized and the mitochondrial fraction was examined for CYP27A1 activity and net 27-OH output, starting with [¹⁴C]Ch and using HPTLC -PI for analysis. Inset shows observed [¹⁴C]27-OH levels at different starting 7α-OOH concentrations. Values are means ± SEM; n=3; *P<0.05 vs. zero-7α-OOH control. Adapted from Ref. 46; refer to this for additional experimental details.

7. Negative effects of 7-OOH trafficking on steroid hormone synthesis

In addition to initiating its own export from vascular macrophages to prevent overloading, Ch plays an essential early role in the synthesis of cortisol, progesterone, testosterone, and other steroid hormones by steroidogenic cells [47,48]. This Ch can derive from extracellular donors such as LDL via a cell surface receptor [49]. It can also be supplied by intracellular sources such as the Ch-rich plasma membrane or by cytosolic lipid droplets, hydrolysis of abundant cholesteryl esters in the latter case giving free Ch. If necessary, Ch can be synthesized de novo in steroidogenic cells by reactions in the endoplasmic reticulum [47,48]. Hormone synthesis is initiated in mitochondria by inner membrane P450scc/Cyp11A1-catalyzed hydroxylation and cleavage of the Ch side chain to give pregnenolone. Steroidogenic cells, like vascular macrophages, rely on StAR family proteins to deliver Ch to mitochondria [50,51]. Cytosolic StrarD4 transports Ch to the outer mitochondrial membrane while StarD1, as part of a multi-protein complex, brings it from the outer to inner membrane [51,52]. It is now clear that hormone production by steroidogenic cells can be seriously impaired by the development and persistence of oxidative stress. Such stress often accompanies disorders such chronic inflammation, ischemia/reperfusion injury, and Type-2 diabetes, although advanced age can also bring it on [53,54]. Quite often, the underlying cause is a deficiency in antioxidants - either enzymatic (e.g. superoxide dismutase, catalase, GPx4) or non-enzymatic (e.g. glutathione, α-tocopherol, β-carotene [55]. Based on this and related evidence, Korytowski and Girotti [56] hypothesized that 7-OOH in oxLDL or other donors can (like Ch) be StarD4/D1-transported to mitochondria of steroidogenic cells, and upon arrival at the inner membrane, induce chain LPO which disables hormone synthesis. Strong initial support for this hypothesis was obtained by exposing testicular MA-10 Leydig cells to [¹⁴C]7α-OOH- and Ch-containing SUVs after cyclic-AMP stimulation [56]. Compared with non-stimulated controls, stimulated cells exhibited a strong upregulation of the StarD4/D1 proteins, and incorporated far more radioactivity from [¹⁴C]7α-OOH than the controls. Most of this radioactivity was detected in the mitochondrial fraction, consistent with StarD4/D1-mediated delivery. Correspondingly, after 7α-OOH exposure, stimulated cells exhibited a far greater loss of mitochondrial membrane potential (ΔΨ_m) than non-stimulated controls, implying that greater membrane damage had occurred in the former [56]. Finally, after contact with 7α-OOH, Elisa-assessed progesterone output of stimulated MA-10 cells was far lower than that of controls exposed to SUVs containing Ch but not the hydroperoxide. Cells were still attached after incubation times used in these experiments, so any significant loss of viability was ruled out as a possible cause for the effects described. On the other hand, prolonged exposure to SUV-7α-OOH did result in cell death, which increased with hydroperoxide concentration, stimulated cells being more sensitive than non-stimulated [56]. In striking contrast, stimulated and non-stimulated MA-10 cells were equally sensitive to killing by tert-butyl hydroperoxide, which is not a StAR protein ligand [56]. These findings were confirmed in a recent study which also showed that mitochondrial GPx4 was significantly upregulated in cAMP-primed MA-10 cells, and that this protected against a reduced progesterone output due to SUV-7α-OOH-induced chain LPO [57]. This is illustrated in Fig. 4, which shows that RSL3, a specific inhibitor of GPx4 activity [58], increased the extent of LPO while lowering the progesterone yield [57]. Thus, by overexpressing GPx4, these cells appeared to be preparing for a potential oxidative challenge such as imposed by peroxidized Ch ChOOH species.

Fig. 4.

Effects of GPx4 inhibitor, RSL3, on sensitivity of MA-10 Leydig cells to mitochondrial membrane LPO with reduced progesterone output after exposure to Ch/7-OOH-containing SUVs. Cells at ~65% confluency in DME medium were stimulated with 1 mM dibutyryl-cAMP for 12 h, then treated with 0.5 or 1.5 mM RSL3 for 4.5 h in fresh DME, after which POPC/Ch/7-OOH (5:4:1 by mol) SUVs were introduced at 50 μM 7-OOH in bulk medium; POC/Ch (5:5 by mol) SUVs served as controls. After 2 h of incubation, and medium replacement with fresh DME, the LPO probe C11-BODIPY-581/591 (0.2 μM) was introduced and fluorescence was assessed using a fluorescence microscope and LED light source. After 2 h of additional incubation, progesterone level was determined by enzyme immunoassay. Plotted values are means ± SEM; n>3. Additional details are provided in Ref. 57, from which this is adapted.

One might ask how 7α-OOH could possibly survive the trip to mitochondria and not be deactivated along the way by cytosolic GPx4. There is unpublished evidence by the authors of this review that, like Ch itself, 7α-OOH is tightly sequestered by the StarD4/D1 transporters and that this protects them against deactivation during delivery to the mitochondrial inner membrane. This is a deleterious special characteristic of sterol hydroperoxides which non-sterol counterparts lack, e.g. fatty acid and phospholipid hydroperoxides.

8. Initiation of apoptosis by CLOOH translocation in mitochondrial membranes

It is now clear that intrinsic apoptotic cell death under oxidative stress conditions is initiated by at least three consecutive events involving cytochrome c (cyt c): (i) dissociation of this protein from the inner mitochondrial membrane (IM), (ii) migration through the intermembrane space, and (iii) release into the cytosol [59]. Underlying reactions are not completely understood, but oxidation of inner membrane diphosphatidylglycerol or cardiolipin (CL) to hydroperoxide species (CLOOHs) is now known to play a key role [60]. The sn-1 and the sn-2 fatty acyl groups of natural CL are typically both unsaturated, making it more susceptible to peroxidation than most conventional unsaturated phospholipids. About 12 years ago, Korytowski et al. [61] postulated that under oxidative stress conditions, CL-derived CLOOHs as well as cyt c are released from the IM. These CLOOHs would assist in apoptotic signaling by translocating to the outer membrane (OM), thereby promoting recruitment of pro-apoptotic tBid and Bax, for assembly of cyt c-traversing OM pores. Several factors favor IM-to-OM CLOOH transfer in oxidatively-challenged mitochondria, Including (i) the high CL starting concentration on the IM, (ii) CL’s high degree of unsaturation, making it readily peroxidizable, and (iii) the relatively short distance between IM and OM (~60 A°). Using in vitro liposome/liposome and liposome/mitochondria pairings to model IM-to-OM lipid transfer, Korytowski et al. [61 showed that CLOOH with ~1-OOH group per molecule transfers much faster than parental CL. Moreover, the rate was substantially increased by non-specific lipid transfer protein, SCP-2, which is known to exist in mitochondria of certain cells [62]. Korytowski et al. also found that tBid bound more avidly to CLOOH-containing liposomes than to CL-containing ones. Moreover, they showed that treating H9c2 cardiomyocytes with Antimycin A, which generates ROS by blocking Complex III, caused a large IM-to-OM redistribution of oxidized CL, most likely CLOOH species [61]. Importantly, this resulted in cyt c release and caspase-3/7 activation downstream of OM permeabilization by tBid/Bax. Collectively, these findings described an entirely novel mechanism for pro-apoptotic tBid/Bax recruitment under oxidative stress conditions, namely IM-to-OM translocation of CLOOH [61].

9. Cellular protection against redox damage by translocated LOOHs

As indicated in a previous section, selenoperoxidase GPx4 is the only antioxidant enzyme known to be capable of catalyzing the direct redox inactivation of 7α/β-OOH and other ChOOHs [19]. GPx4 can be found in cytosol, mitochondria, and other compartments of most mammalian cells, and its expression can be elevated in certain cells after metabolic activation, e.g. cAMP priming of MA-10 Leydig cells [57]. A highly relevant study about 15 years ago revealed that overexpression of GPx4 in all tissues of transgenic apoE(−/−) mice, including aorta, made them substantially more resistant to oxidative stress-induced atherogenesis than wild type controls [63]. Importantly, this resistance correlated with a greatly reduced level of systemic lipid peroxidation. At the mechanistic molecular level, GPx4’s ability to reductively inactivate atherogenic ChOOH species like 7-OOH [42] could have played a key role in the described resistance [63]. As described in this review, GPx4’s anti-ChOOH ability could also be crucial in protecting against impairment of (i) reverse cholesterol transport in vascular macrophages, and (ii) hormone synthesis in steroidogenic cells. Selenium deficiency has been reported to impair steroid hormone synthesis in animal models; however, this was attributed to generalized glutathione peroxidase deficiency, GPx4 not being specifically considered [64]. In addition to endogenous GPx4, administered antioxidants such as mitoquinone (MitoQ) could suppress mitochondrial LPO damage set off by any StAR-delivered ChOOHs. Driven by the electrochemical gradient, MitoQ accumulates in the mitochondrial inner membrane of live cells [65,66] and can scavenge damaging free radicals generated by electron transport inhibitors or by incoming oxidants, e.g. 7-O^● from turnover of StAR-delivered 7-OOH. Besides the indicated antioxidants, there are primary antioxidants such as superoxide dismutase and catalase, which respectively dispose of O₂^●- and H₂O₂, two ROS that can be initially involved in ChOOH formation. Deficiencies in a natural antioxidant such as GPx4 need to be guarded against, particularly in aged individuals or those susceptible to vascular or steroid hormone disorders. For GPx4, dietary selenium supplementation is relatively straightforward and often very effective [67].

9. Summary and Perspectives

When caught up in StAR protein-mediated Ch trafficking to mitochondria, oxidative stress-generated 7-OOHs can induce chain peroxidation of membrane lipids, thus disabling at least two physiological processes: steroid hormone synthesis and reverse Ch transport. Recent studies have indicated that other ChOOHs may also arise during Ch autoxidation under oxidative stress conditions, e.g. 4-OOH, 5α-OOH, and 6-OOH (68,69). However, their yields would be relatively minor compared with 7-OOH, making the latter much more important in trafficking-associated damage. In steroidogenic cells, early-stage metabolism involves Ch processing by inner membrane Cyp11A1, whereas in vascular macrophages, Ch homeostasis via reverse transport requires 27-OH formation by Cyp27A1. When tightly sequestered by StAR proteins, 7-OOHs would be protected against reductive turnover during transit. However, after 7-OOH release at an inner membrane, one-electron reduction would trigger free radical LPO with a progressive loss of Cyp11A1 or Cyp27A1 activity. There is a growing awareness that functionality of steroidogenic cells and Ch-regulating macrophages can be compromised by physiological conditions associated with oxidative stress, e.g. chronic obesity, diabetes, ischemia/reperfusion, or advanced age. The mitochondrial damage/dysfunction mechanism that we describe for in vitro experiments awaits validation at the in vivo level, and this is anticipated in due course as model studies progress. Increased recognition of deleterious ChOOH translocation, the first to be identified that operates via a natural lipid trafficking pathway, is expected to stimulate renewed interest in mitochondria-localizing antioxidants [66]. Like ChOOHs, phospholipid counterparts (PLOOHs) can induce/mediate membrane LPO, one example being mitochondrial CLOOHs (Sect, 8). Phospholipids are transported by relatively low specificity phospholipid transfer proteins (PLTPs) [70]. Whether these proteins can also recognize and translocate PLOOHs is currently unknown. If so, however, such activity might play a key role in oxidative stress disorders such as atherogenesis, which has been linked to the level of PLTP expression in vascular macrophages [71].