**4. Lipoprotein oxidation**

We now turn to the specifics of lipoprotein oxidation with reference to the different lipid and protein components.

#### **4.1. Oxidation of lipid components**

Of the different lipid components that may be potentially oxidized in various lipoprotein particles, we will focus on the oxidation of fatty acyl chains, including those on phospholipids, cholesterylester and triglycerides, and on cholesterol derivatives. The reader is referred to a comprehensive review by Subbaiah and colleagues (Levitan et al., 2010) for an introduction to the role of ceramides in Ox-LDL and of sphingosine 1-phosphate in HDL

#### *4.1.1. Oxidation of fatty acyl chain*

One group of biological targets that are highly vulnerable to attack by ROS and aldehydes are the lipids: their abundance in lipoproteins and the ease with which their unsaturated bonds are oxidatively modified make them susceptible to damage. Products of lipid peroxidation have been associated with the pathophysiology of numerous disease states, including atherosclerosis, diabetes and cancer. Although found in low concentrations in normal healthy tissues, they are found to be enriched in pathological cells and tissues, including macrophage foam cells and atherosclerotic lesions (Olkkonnen, 2008; Brown & Jessup, 1999; Olkkonen & Lehto, 2004; Javitt, 2008; Tsimikas et al., 2005; Berliner & Watson, 2005). PUFA peroxidation products cause further damage to proteins by oxidative modification of amino acid side chains and formation of protein carbonyl groups (Refsgaard et al., 2000).

Briefly, lipid peroxidation is initiated by the hydroxyl radical abstracting a hydrogen from the methylene group adjacent to a double bond of fatty acids, **Figure 5**. The fatty acid may also be part of the phospholipid at the *sn*-2 position or esterified to the –OH group of cholesterol in a lipoprotein. This process gives rise to an unstable lipid radical, which undergoes rearrangement of double bonds and addition of oxygen to form a peroxyl radical. The lipid radical or the peroxyl radical could react with a neighbouring fatty acyl chain as well, thereby propagating the peroxidation process. The chain reaction may be terminated by eventual formation of a lipid hydroperoxide.

392 Lipoproteins – Role in Health and Diseases

*3.3.3. Lipoxygenase* 

(Hiltunen et al., 1995).

**4. Lipoprotein oxidation** 

**4.1. Oxidation of lipid components** 

*4.1.1. Oxidation of fatty acyl chain* 

et al., 2000).

and protein components.


Lipoxygenases are iron-containing enzymes that catalyze the dioxygenation of PUFA in lipids. Specifically, they form hydroperoxides from fatty acids and molecular oxygen. 15 lipoxygenase is the main protein in this family; it is involved in the metabolism of eicosanoids (prostaglandins, leukotrienes), which function as secondary messengers. 15 lipoxygenase has been shown to be involved in LDL oxidation (Bailey et al., 1995), with considerably greater 15-lipoxygenase activities in atherosclerotic compared to normal aortas

We now turn to the specifics of lipoprotein oxidation with reference to the different lipid

Of the different lipid components that may be potentially oxidized in various lipoprotein particles, we will focus on the oxidation of fatty acyl chains, including those on phospholipids, cholesterylester and triglycerides, and on cholesterol derivatives. The reader is referred to a comprehensive review by Subbaiah and colleagues (Levitan et al., 2010) for an introduction to the role of ceramides in Ox-LDL and of sphingosine 1-phosphate in HDL

One group of biological targets that are highly vulnerable to attack by ROS and aldehydes are the lipids: their abundance in lipoproteins and the ease with which their unsaturated bonds are oxidatively modified make them susceptible to damage. Products of lipid peroxidation have been associated with the pathophysiology of numerous disease states, including atherosclerosis, diabetes and cancer. Although found in low concentrations in normal healthy tissues, they are found to be enriched in pathological cells and tissues, including macrophage foam cells and atherosclerotic lesions (Olkkonnen, 2008; Brown & Jessup, 1999; Olkkonen & Lehto, 2004; Javitt, 2008; Tsimikas et al., 2005; Berliner & Watson, 2005). PUFA peroxidation products cause further damage to proteins by oxidative modification of amino acid side chains and formation of protein carbonyl groups (Refsgaard

Briefly, lipid peroxidation is initiated by the hydroxyl radical abstracting a hydrogen from the methylene group adjacent to a double bond of fatty acids, **Figure 5**. The fatty acid may also be part of the phospholipid at the *sn*-2 position or esterified to the –OH group of cholesterol in a lipoprotein. This process gives rise to an unstable lipid radical, which undergoes rearrangement of double bonds and addition of oxygen to form a peroxyl radical.

**Figure 5.** Lipid peroxidation. The lipid peroxidation process may be divided into the initiation, propagation and termination steps; only the double bond containing segment of a fatty acid is shown.

In the context of a lipoprotein, it was postulated that LDL oxidation was initiated through cellgenerated ROS formation, with the involvement of the lipoxygenase pathway (Cyrus et al, 1999). Both apoE-null and LDLr-null mice genetically deficient in the 12/15-lipoxygenase were found to have significantly less atherosclerosis. Other groups developed mice overexpressing the 12/15-lipoxygenase, and found spontaneous aortic fatty streak lesions on a chow diet (Reilly et al., 2004). Finally, it was shown that the endothelial cells within the vessels required 12/15-lipoxygenase to generate oxidized phospholipids. It was also shown that the MPO pathway contributed to Ox-LDL in human atherosclerotic lesions (Savenkova et al., 1994).

LDL oxidation may occur within the arterial endothelial cells, which have high levels of precursor molecules like linoleic acid and arachidonic acid - fatty acids that are involved in producing eicosanoids. Hydroperoxide reaction with linoleic acid produces 13(*S*) hydroperoxy-octadecadienoic acid (13(*S*)-HPODE), and reaction with arachidonic acid produces 15(*S*)-hydroperoxy-eicosatetraenoic acid (15(*S*)-HPETE). These molecules are present as part of LDL surface phospholipids and trigger further oxidation of phospholipids with arachidonic acid. These are early events occurring prior to apoB-100 modification and constitute the 'minimally-modified' LDL (Navab et al., 2001). Another consequence of phospholipid oxidation is fragmentation of the fatty acyl group at *sn*-2 position resulting in short chain fatty acids, which structurally and functionally resemble platelet activating factors with chemotactic activity.

Although less understood, *in vitro* oxidation of HDL is also shown to generate oxidized lipids and proteins. It is likely that HDL lipids may be oxidized initially even before LDL lipids, upon exposure of human plasma to peroxyl radicals, Cu2+ ions or lipoxygenase

(Garner et al., 1998). These studies also show that Met residues are oxidized in apoAI and apoAII. Further, 1H and 31P NMR analysis indicate a loss of the unsaturated system with appearance of epoxides on fatty acyl chains and 5,6-epoxide derivatives of cholesterol indicating significant modification of HDL (Bradamante et al., 1992).

### *4.1.2. Oxysterols*

Oxysterols are molecules that are formed from the oxidation of cholesterol, which can occur at several sites (Vaya & Schipper, 2007). They display higher water solubility compared to cholesterol. The major oxysterols isolated from plasma LDL include 7α-OH and 7β-OH cholesterol. In addition, 7-keto cholesterol, which impairs cholesterol efflux and reduces cell membrane fluidity, and 5,6-epoxide derivatives of cholesterol, have also been identified in Ox-LDL (Levitan et al., 2010). Other types of oxysterols have been localized in atherosclerotic lesions, including those wherein the side chains of cholesterol are oxidized (for example, 27-hydroxycholesterol). The cytochrome P-450 system is largely responsible for generating the hydroxylated derivatives. They also function as transcriptional effectors and attenuate the Liver X Receptor (LXR). A majority of PUFA in LDL is esterified as cholesterylester; thus, hydroperoxide and hydroxide derivatives of cholesterylester are found in abundance in human atherosclerotic lesions.

In the CNS, the neurons convert cholesterol to 24S-hydroxycholesterol (also known as cerebrosterol), **Figure 6**; the conversion facilitates its movement out of the CNS, since cholesterol as such does not cross the blood brain barrier. Almost all circulating 24Shydroxycholesterol originates from the brain (Lutjohann et al., 1996; Lutjohann et al., 2000; Bjorkhem et al., 1998) and may reflect CNS cholesterol turnover.

**Figure 6.** 24S-hydroxycholesterol (obtained from: http://pubchem.ncbi.nlm.nih.gov)

#### **4.2. Oxidation of protein components**

#### *4.2.1. ApoB-100*

From a historical perspective, apoB-100 was the earliest apolipoprotein to be identified as being a target for oxidative modification on lipoproteins (Steinberg et al., 1989). It is implied in the discussions involving Ox-LDL hypothesis, along with oxidation of the lipid moieties in LDL. The term Ox-LDL is used to identify LDL that has been modified to an extent that it is not recognized by the LDLr anymore; instead it becomes a ligand for the scavenger receptors family of proteins. On the other hand, the term 'minimally-modified' LDL has been adopted to encompass the different preparations that have been modified enough to be chemically distinguishable from, but recapitulates the LDLr binding feature of, unmodified LDL. ApoB-100 is one of the oxidizable targets on lipoproteins: one of the earliest *in vitro* studies demonstrate that incubation of human LDL with 4-HNE (Haberland et al., 1984; Jurgens et al., 1986), results in modification of 45 lysines, 7 histidines, 23 serines and 51 tyrosine residues on apoB-100.

4-HNE forms covalent adducts with lysine residues on apoB-100, thereby blocking its ability to recognize the macrophage LDLr. Evidence for *in vivo* LDL oxidation was provided by immunocytochemical staining and immunoblot analysis of extracts from atherosclerotic lesions of LDLr-deficient rabbits using antibodies against Ox-LDL, MDA-lysine or 4-HNElysine, (Palinski et al., 1990). This study also demonstrated higher titers of autoantibodies against MDA-conjugated LDL in the human and rabbit antisera. 4-HNE-modified LDL is an efficient ligand for scavenger receptors (Hoff & O'Neil, et al, 1993; Rosenfeld et al., 1990). ApoB-100 can also be modified by MPO, which leads to formation of chlorotyrosine and nitrotyrosine derivatives.

#### *4.2.2. ApoE*

394 Lipoproteins – Role in Health and Diseases

*4.1.2. Oxysterols* 

(Garner et al., 1998). These studies also show that Met residues are oxidized in apoAI and apoAII. Further, 1H and 31P NMR analysis indicate a loss of the unsaturated system with appearance of epoxides on fatty acyl chains and 5,6-epoxide derivatives of cholesterol

Oxysterols are molecules that are formed from the oxidation of cholesterol, which can occur at several sites (Vaya & Schipper, 2007). They display higher water solubility compared to cholesterol. The major oxysterols isolated from plasma LDL include 7α-OH and 7β-OH cholesterol. In addition, 7-keto cholesterol, which impairs cholesterol efflux and reduces cell membrane fluidity, and 5,6-epoxide derivatives of cholesterol, have also been identified in Ox-LDL (Levitan et al., 2010). Other types of oxysterols have been localized in atherosclerotic lesions, including those wherein the side chains of cholesterol are oxidized (for example, 27-hydroxycholesterol). The cytochrome P-450 system is largely responsible for generating the hydroxylated derivatives. They also function as transcriptional effectors and attenuate the Liver X Receptor (LXR). A majority of PUFA in LDL is esterified as cholesterylester; thus, hydroperoxide and hydroxide derivatives of cholesterylester are

In the CNS, the neurons convert cholesterol to 24S-hydroxycholesterol (also known as cerebrosterol), **Figure 6**; the conversion facilitates its movement out of the CNS, since cholesterol as such does not cross the blood brain barrier. Almost all circulating 24Shydroxycholesterol originates from the brain (Lutjohann et al., 1996; Lutjohann et al., 2000;

From a historical perspective, apoB-100 was the earliest apolipoprotein to be identified as being a target for oxidative modification on lipoproteins (Steinberg et al., 1989). It is implied in the discussions involving Ox-LDL hypothesis, along with oxidation of the lipid moieties in LDL. The term Ox-LDL is used to identify LDL that has been modified to an extent that it is not recognized by the LDLr anymore; instead it becomes a ligand for the scavenger

indicating significant modification of HDL (Bradamante et al., 1992).

found in abundance in human atherosclerotic lesions.

**4.2. Oxidation of protein components** 

*4.2.1. ApoB-100* 

Bjorkhem et al., 1998) and may reflect CNS cholesterol turnover.

**Figure 6.** 24S-hydroxycholesterol (obtained from: http://pubchem.ncbi.nlm.nih.gov)

ApoE is a 299 residue, 34 kDa protein that is commonly associated with VLDL, CM remnants and a sub-class of HDL. It is a major cholesterol transport protein in the plasma and the CNS (Hatters et al., 2006). In humans, apoE is polymorphic; variation in the *APOE* gene results in three major alleles, ε2, ε3 and ε4, occurring at frequencies of 8%, 77% and 15%, respectively in the population. The products of the three alleles are the isoforms, apoE2, apoE3 and apoE4, which differ in the amino acids at positions 112 and 158: apoE2 has Cys while apoE4 has Arg at these locations; apoE3 has a Cys and Arg at these locations, respectively. ApoE3 is considered an anti-atherogenic protein; individuals homozygous for the *APOE* ε2 allele are prone to developing familial type III hyperlipoproteinemia and premature atherosclerosis. The inheritance of one or more of the *APOE* ε4 alleles predisposes the bearer to hypercholesterolemia, as well as Alzheimer's disease, affecting both the age of onset and the severity of these diseases.

*In vitro* oxidative modification of the receptor-binding domain of apoE3 by acrolein generates epitopes recognized by an antibody specific for acrolein-lysine adducts, mAb5F6 (Tamamizu-Kato et al., 2007) with formation of both intra- and inter-molecular cross-linked products. This modification resulted in severe impairment of three major functions of apoE3: (i) its ability to interact with the LDLr, a function mediated by specific lysines and arginines located on helix 4 of the receptor-binding domain; (ii) its ability to bind heparin, which is facilitated predominantly by two specific lysines (K143 and K146), **Figure 7**; (iii) its ability to bind lipids. These studies indicate that acrolein either directly modifies the lysines in helix 4 that are involved in LDLr and heparin binding or that modification of lysines elsewhere on apoE3 alters the conformation of lysines in helix 4, thereby disrupting its binding. Further evidence was provided by direct exposure of VLDL isolated from human plasma to acrolein or Cu2+, which disrupted its ability to bind and internalize the lipoprotein particle via LDLr, LDLrrelated protein or HSPG on hepatocytes. (Arai et al., 1999; 2005). Taken together, oxidative modification appears to compromise the functional integrity of apoE3.

**Figure 7.** Lysines relevant in LDLr and HSPG binding function of apoE3. Ribbon diagram of apoE3 receptor-binding domain is shown with the LDLr- and HSPG-binding sites represented in green and yellow, respectively.

From a physiological perspective, the loss of apoE function by oxidative modification has direct implications in CVD: (i) decreased uptake and internalization of apoE-containing lipoproteins leading to their accumulation in the blood; (ii) decreased ability of apoE to interact with heparan sulfate proteoglycans (HSPG) lining the blood vessels and cell surfaces, where apoE is believed to be stored; and (iii) impaired ability of apoE to interact with lipoproteins, which is an essential prerequisite for the receptor-binding domain to elicit LDLr binding.

In the CNS, where apoE is the predominant apolipoprotein that has been identified so far, oxidative modification is expected to have serious implications in progression of neurological diseases such as Alzheimer's disease in an isoform-specific manner. About 65% of individuals with late-onset familial and sporadic Alzheimer's disease bear the *APOE* ε4 allele (Huang et al., 2004). The precise mechanism by which apoE4 is associated with Alzheimer's disease remains a contentious issue. While the role of apoE4 in aggravating the beta amyloid toxicity has received widespread attention, the inherent propensity of apoE4 to misfold noted under *in vitro* physiological conditions requires further scrutiny *in vivo*. Further, since oxidative damage plays a significant role in the pathogenesis associated with Alzheimer's disease (Perry et al., 2002), it is likely that oxidative modification of apoE4 further exacerbates its role in the etiology of the disease. Indeed, cerebrospinal fluid obtained by lumbar puncture in a limited number of Alzheimer's disease patients homozygous for *APOE3* or *APOE4*, and age-matched controls with or without dementia display a 50 kDa apoE-immunoreactive protein co-migrating with proteins immunoreactive for 4-HNE and MDA adducts (Montine et al., 1996; Bassett et al., 1999).

A similar 50 kDa apoE-immunoreactive protein was also reported in P19 neuroglial cultures differentiated into neurons and astrocytes subjected to oxidative stress. Interestingly, apoE3 appeared to be cross-linked to a greater extent than apoE4. The cross-linking has been attributed to the susceptible site provided by apoE3 in the form of Cys112, and the known reactivity of 4-HNE with sulfhydryl groups (Esterbauer et al., 1991). *In vitro* modification of apoE3 or apoE4 (isolated from plasma) by 4-HNE yielded cross-linked products with apparent molecular weights corresponding to dimeric and trimeric apoE (Montine et al., 1996). A similar trend was noted with MDA. It is possible that the greater susceptibility of apoE3 to crosslinking and oxidation than apoE4 is a reflection of its greater potency as an antioxidant.

#### *4.2.3. ApoAI*

396 Lipoproteins – Role in Health and Diseases

represented in green and yellow, respectively.

which disrupted its ability to bind and internalize the lipoprotein particle via LDLr, LDLrrelated protein or HSPG on hepatocytes. (Arai et al., 1999; 2005). Taken together, oxidative

**Figure 7.** Lysines relevant in LDLr and HSPG binding function of apoE3. Ribbon diagram of

From a physiological perspective, the loss of apoE function by oxidative modification has direct implications in CVD: (i) decreased uptake and internalization of apoE-containing lipoproteins leading to their accumulation in the blood; (ii) decreased ability of apoE to interact with heparan sulfate proteoglycans (HSPG) lining the blood vessels and cell surfaces, where apoE is believed to be stored; and (iii) impaired ability of apoE to interact with lipoproteins, which is an essential prerequisite for the receptor-binding domain to elicit LDLr binding.

In the CNS, where apoE is the predominant apolipoprotein that has been identified so far, oxidative modification is expected to have serious implications in progression of neurological diseases such as Alzheimer's disease in an isoform-specific manner. About 65% of individuals with late-onset familial and sporadic Alzheimer's disease bear the *APOE* ε4 allele (Huang et al., 2004). The precise mechanism by which apoE4 is associated with Alzheimer's disease remains a contentious issue. While the role of apoE4 in aggravating the beta amyloid toxicity has received widespread attention, the inherent propensity of apoE4 to misfold noted under *in vitro* physiological conditions requires further scrutiny *in vivo*. Further, since oxidative damage plays a significant role in the pathogenesis associated with Alzheimer's disease (Perry et al., 2002), it is likely that oxidative modification of apoE4 further exacerbates its role in the etiology of the disease. Indeed, cerebrospinal fluid obtained by lumbar puncture in a limited number of Alzheimer's disease patients homozygous for *APOE3* or *APOE4*, and age-matched controls with or without dementia display a 50 kDa apoE-immunoreactive protein co-migrating with proteins immunoreactive

A similar 50 kDa apoE-immunoreactive protein was also reported in P19 neuroglial cultures differentiated into neurons and astrocytes subjected to oxidative stress. Interestingly, apoE3 appeared to be cross-linked to a greater extent than apoE4. The cross-linking has been attributed to the susceptible site provided by apoE3 in the form of Cys112, and the known reactivity of 4-HNE with sulfhydryl groups (Esterbauer et al., 1991). *In vitro* modification of

for 4-HNE and MDA adducts (Montine et al., 1996; Bassett et al., 1999).

apoE3 receptor-binding domain is shown with the LDLr- and HSPG-binding sites

modification appears to compromise the functional integrity of apoE3.

ApoAI is a 243 residue, 28 kDa exchangeable apolipoprotein that is a major component of HDL. Like apoE, it is composed predominantly of amphipathic α-helices, with an Nterminal domain 4-helix bundle. Under normal physiological conditions, apoAI plays a critical role in promoting ATP Binding Cassette Transporter A1 (ABCA1)-mediated cholesterol efflux from macrophages. This aids in mobilizing cholesterol and phospholipids from peripheral tissues to the liver by the RCT process, for eventual disposal by biliary secretion. In atherosclerotic lesions, apoE plays a dominant role in RCT by virtue of the fact that cholesterol-laden macrophages secrete lipid-poor apoE, which in turn in promotes ABCA1-mediated cholesterol efflux (Huang et al, 1995).

With both apoAI and apoE, a nascent discoidal form of HDL is generated that is composed of a bilayer of phospholipids and cholesterol circumscribed by the α-helices of the protein. The discoidal HDL is an excellent substrate for LCAT, the enzyme that catalyzes the transfer of a fatty acyl chain from phospholipids to the free hydroxyl group of cholesterol to form cholesterylesters. The conversion of the amphipathic free cholesterol to the hydrophobic cholesterylesters promotes its transition to the core of the lipoprotein particle, thereby generating a spherical HDL containing a cholesterylester core. The HDL is targeted to the liver and steroidogenic tissues where they are recognized by the scavenger receptor class B Type 1 (SR-B1), which mediate selective uptake of cholesterylesters into the cells.

**Figure 8.** Immunohistochemical co-localization of apoAI and acrolein adducts in human atherosclerotic lesions. This research was originally published in The Journal of Biological Chemistry. Shao, B., Fu, X., McDonald, T. O., Green, P. S., Uchida, K., O'Brien, K. D., Oram, J. F. & Heinecke, J. W. Acrolein impairs ATP Binding Cassette Transporter A1-dependent cholesterol export from cells through site-specific modification of apolipoprotein A-I. J. Biol. Chem. (2005) Vol. 280, No. 43, pp. 36386-36396 © the American Society for Biochemistry and Molecular Biology.

There is strong evidence that MPO oxidizes HDL *in vivo* (Daugherty et al., 1994; Bergt et al., 2004; Pennathur et al., 2004; Zheng et al., 2004). In addition, aldehyde modification of HDL is also associated with the loss in ability of HDL to activate LCAT (McCall et al., 1995). Acrolein modifies apoAI specifically at Lys226 in helix 10 converting it to Nε-(3 methylpyridinium)lysine (MP-Lys). A corresponding decrease in ABCA1-mediated cholesterol efflux was also noted. In addition, immunohistochemical analysis demonstrated co-localization of acrolein-adducts and apoAI in human atherosclerotic lesions, **Figure 8**, confirming previous studies (Uchida et al., 1998b). Further chlorination of Tyr192 and oxidation of specific Met residues in apoAI via the MPO pathway impairs its ability to promote cholesterol efflux (Shao et al., 2010). Taken together, it appears that oxidative modification of HDL apoAI may be one of the contributory factors to atherogenesis.

**Figure 9** provides a simplified overview of the roles of apoB-100, apoAI and apoE in lipid distribution between liver and peripheral tissues. It also shows potential functional sites that are likely to be affected because of oxidative modification of lipoproteins.

**Figure 9.** Distribution cholesterol mediated by apoB-100, apoE and apoAI. The green block arrow shows the general direction of the RCT process. The red stars draw attention to processes that are affected by oxidative modification of protein and lipid components of lipoproteins.
