Protein Modification by Oxidized Phospholipids and Hydrolytically Released Lipid Electrophiles: Investigating Cellular Responses

Abstract

Oxygen is essential for the growth and function of mammalian cells. However, imbalances in oxygen or abnormalities in the ability of a cell to respond to oxygen levels can result in oxidative stress. Oxidative stress plays an important role in a number of diseases including atherosclerosis, rheumatoid arthritis, cancer, neurodegenerative diseases and asthma. When membrane lipids are exposed to high levels of oxygen or derived oxidants, they undergo lipid peroxidation to generate oxidized phospholipids (oxPL). Continual exposure to oxidants and decomposition of oxPL results in the formation of reactive electrophiles, such as 4-hydroxy-2-nonenal (HNE). Reactive lipid electrophiles have been shown to covalently modify DNA and proteins. Furthermore, exposure of cells to lipid electrophiles results in the activation of cytoprotective signaling pathways in order to promote cell survival and recovery from oxidant stress. However, if not properly managed by cellular detoxification mechanisms, the continual exposure of cells to electrophiles results in cytotoxicity. The following perspective will discuss the biological importance of lipid electrophile-protein adducts including current strategies employed to identify and isolate protein adducts of lipid electrophiles as well as approaches to define cellular signaling mechanisms altered upon exposure to electrophiles.

1. Lipid Electrophile Generation, Clearance and Reactivity

1.a. Lipid Peroxidation and the Formation of Reactive Lipid Electrophiles

Membrane phospholipids undergo enzymatic and non-enzymatic oxidation of their polyunsaturated fatty acid (PUFA) side chains to generate a variety of oxidized phospholipid products, including hydroperoxides and cyclic peroxides [1]. Enzymatic oxidation of membrane phospholipids is mediated by lipoxygenase (LOX) and cycloxygenase (COX). Oxidation of fatty acids derived from membrane phospholipid, such as arachidonic acid, results in the formation of hydroperoxyeicosatetraenoic acid (HPETE), prostaglandins, thromboxane and leukotrienes. Alternatively, exposure to free radicals, such as hydroxyl radicals, lipid oxyl or peroxyl radicals, singlet oxygen, and peroxynitrite formed from nitrogen oxide, can induce non-enzymatic oxidation of membrane lipids [2].

The products of both enzymatic and non-enzymatic oxidation of membrane lipids can undergo further metabolism to generate a number of products with varied physiological functions. Oxidized phospholipids can undergo oxidative fragmentation or decomposition to form a number of biologically active molecules, including the aforementioned derivatives of PUFA side chains. The critical initiation step of lipid peroxidation (LPO) involves the oxidant mediated abstraction of a bis-allylic hydrogen atom from ω-3 and ω-6 unsaturated fatty acids [3]. This initiates a series of radical chain reactions, resulting in the formation of lipid hydroperoxides, intramolecular rearrangement and chain-breaking reactions [3]. The peroxidation of membrane lipids containing ω-3 and/or ω-6 polyunsaturated fatty acids results in the formation of several classes of reactive aldehydes, including malondialdehyde (MDA), acrolein, and 4-hydroxy-alkenals [4-8]. Some of the aldehydes are not cleared from the cell and exhibit cytotoxic effects [9-11]. The most extensively studied of the lipid electrophiles is HNE, an α,β-unsaturated aldehyde that is highly reactive and readily binds to proteins, DNA and phospholipids. Due to the presence of a conjugated double bond between the α and β carbons, the β carbon of these aldehydes is electron-deficient, rendering HNE readily reactive with nucleophilic amino acid side chains on target proteins through Michael addition to thiols and amines. Specifically, Michael addition results in the formation of covalent adducts between electrophiles and amino acid side chains, such as lysine, histidine, and cysteine residues. The resulting products can then undergo cyclization and hemi-acetal formation [6, 12-14]. Additionally, the carbonyl component of HNE forms a Schiff base with lysine residues to generate pyrrole adducts and fluorescent crosslinks [15-18].

1.b. Mechanisms to Detoxify Lipid Electrophiles in vivo

The extent and severity of cellular damage elicited by electrophilic lipids is highly dependent on the ability of cells to clear and detoxify these aldehydes. Once formed, the cell attempts to clear electrophiles through mechanisms involving glutathione (GSH), aldehyde dehydrogenases, aldo-ketoreductases and P450 enzymes [19-28]. The first line of defense that cells utilize to eliminate reactive aldehydes is conjugation with glutathione (GSH) [21]. HNE is able to spontaneously form Michael adducts with GSH to form GSH conjugates. Alternatively, the conjugation of GSH to the α,β-unsaturated carbonyl can be catalyzed by glutathione-S-transferases (GSTs), including hGSTA4-4 and hGST5.8, in a reaction that proceeds several hundred times faster than the spontaneous reaction [29-31]. The resulting product, GSH-HNE, is water soluble and able to be pumped from cells by glutathione conjugate export pumps and, ultimately, excreted from the body in urine [32]. Additionally, the aldehyde moiety of HNE can either be reduced into alcohol or oxidized into acid in reactions catalyzed by alcohol dehydrogenase, aldo-keto reductases and aldehyde dehydrogenase, respectively. Specifically, alcohol dehydrogenase and aldo-keto reductases (AKRs) catalyze the reduction of 4-hydroxyalkenals and their GSH-conjugates to 1,4-dihydroxy-2-nonene (DHN) or GSH-DHN, which are less toxic and able to be efficiently cleared from the cell [21-26]. Reduction of GSH-HNE stabilizes the Michael adduct and improves its likelihood of clearance. Aldo-keto reductases have been shown to be upregulated in response to oxidative damage, emphasizing the importance of AKRs in detoxification during oxidative stress [33, 34]. Alternatively, aldehyde dehydrogenases catalyze the oxidation of HNE to form 4-hydroxy-2-nonenoic acid (HNEA). Cytochrome P450 can also contribute to the clearance of reactive aldehydes [27, 28]. At low levels, α,β-unsaturated aldehydes are oxidized by mammalian P450 enzymes to their carboxylic acid derivatives [27]. Additionally, P450 has been shown to catalyze the reduction of α,β-unsaturated aldehydes to their alcohol form in vivo [28]. Thus, redundant mechanisms for the detoxification and clearance of reactive lipid electrophiles exist to efficiently scavenge such reactive species and prevent cytotoxicity.

The enzymes responsible for the regulation of cellular lipid electrophiles efficiently control the intracellular levels of HNE under basal conditions. However, during oxidative stress the threshold of the enzymes is overcome by the increased production of lipid electrophiles, thus exposing cellular macromolecules to the deleterious effects of HNE. Since GSH conjugation is the primary and major step in the detoxification of HNE, it is not surprising that depletion of GSH exposes cellular macromolecules to lipid electrophiles [35, 36]. The reduction in cellular GSH levels results in a concomitant increase in circulating HNE [37]. HNE can be catabolized to acetyl-CoA, propionyl-CoA, and formate through β-oxidation [38]. The increase in HNE and its catabolic intermediates can result in increased cytotoxicity. Therefore, during oxidative stress, the level of reactive lipid electrophiles becomes so high that the cell is unable to compensate with detoxification mechanisms and, as a result, deleterious effects are observed.

The mechanisms controlling the detoxification of lipid electrophiles vary depending on cell type. Recent studies of prostate cancer cell lines show distinct differences in the ability of individual cell lines to metabolize HNE by the aforementioned detoxification methods. In this study, the susceptibility of each cell line to HNE-mediated toxicity was directly related to the efficiency of HNE removal through reactions with GSH [39]. Interestingly, the prostate cancer cell lines analyzed in this study differed in their GSH content and GST activity [39]. Given the major differences observed between cell lines derived from the same cancer of origin, it is clear that each respective cell type will have varying levels of detoxification machinery and, thus, differences in their susceptibility to lipid electrophile exposure. This could especially be true when considering the responsiveness of primary cells versus immortalized cell lines with respect to HNE cytotoxicity.

1.c. Chemical Reactivity and Toxicity of Lipid Electrophiles

Recently, the chemical reactivity and cellular effects of specific lipid electrophile species was examined in model cell systems. While HNE is the most extensively studied lipid electrophile, its oxidation product, 4-oxononenal (ONE), is detectable in vivo [40]. Additionally, HNE and ONE can be further metabolized to their carboxylic acid derivatives, HNEA and ONEA respectively, through oxidation of their aldehyde moieties (figure 1)[41, 42]. A recent study by McGrath et al, one of the first to comprehensively examine the individual roles of lipid electrophiles in a biologically relevant system, examined the reactivity of each compound toward N-acetylcysteine and compared reactivity to the toxicity of the compounds in colon cancer and macrophage cell lines [42]. Consistent with previous reports suggesting that ONE is more reactive than HNE, the ONE analogues exhibited higher cytotoxicity compared to HNE analogues. However, in this study, the parent compounds ONE and HNE exhibited similar cytotoxicity, with an IC₅₀ of 20 μM in both cell lines tested. Additionally, the chemical reactivity of each class of compounds towards N-acetyl cysteine followed trends similar to their toxicity, with the more toxic ONE analogues exhibiting higher chemical reactivity than HNE analogues. However, the rate of reactivity of ONE with N-acetyl cysteine was 238 M⁻¹s⁻¹ while the reactivity of HNE was much slower at 1.22 M⁻¹s⁻¹[42]. This disparity between the chemical reactivity and biological effects (i.e. toxicity) with parent compounds HNE and ONE suggests that, perhaps, the rate-limiting event in the cellular response to lipid electrophiles may be of a biological rather than a chemical nature. Thus, it is imperative that a biological approach must be employed to determine the relevance of lipid electrophile modification.

Figure 1.

Generation of lipid electrophiles from membrane polyunsaturated fatty acid (PUFA) side chains and their downstream effects.

While it is clear that lipid electrophiles exhibit considerable cytotoxicity and effects on signaling cascades, the contribution of unhydrolyzed-PL to cellular effects remains uncertain. Preliminary data show that HNE or ONE covalently bound to PC (9-hydroxy-12-oxo-10-dodecenoyl-PC, HODA-PC or 9,12-oxo-10-dodecenoyl-PC, ODA-PC, respectively) has much less cellular toxicity compared to the diffusible electrophiles alone (figure 2). Specifically, in THP-1 macrophages, HODA-PC exhibited 8.5-fold less toxicity compared to HNE (177 μM HODA-PC versus 21 μM native HNE) and ODA-PC exhibited 11.7-fold less toxicity compared to ONE (246 μM ODA-PC versus 20.3 μM native ONE) [42][C. Lawrence, unpublished results]. Additionally, the biological effects of HODA-PC are far less striking than diffusible electrophiles. Cytokine array analysis showed that HODA-PC exhibits fewer effects on inflammatory pathways compared to HNE [A. Jacobs, unpublished results]. Heme oxygenase (HO-1) induction, which occurs following 5 μM treatment with HNE, is only induced at very high concentrations of HODA-PC [60 μM, A. Jacobs, unpublished results][9]. Thus, diffusible oxidation products of membrane phospholipids appear to be the major source of bioactive compounds that chemically modify cellular proteins, DNA and macromolecules. In the following discussion, we will review current literature examining protein and gene modification by oxidized phospholipids with a focus on their biochemically significant lipid electrophile byproducts.

Figure 2.

Generation of oxidized phospholipids from parent glycerophosphotidylcholine molecules.

2. Methods to Identify Electrophile Protein Adducts

As we have previously discussed, phospholipids (PL) may be metabolized to release their fatty acid components, which are then metabolized to generate biologically active lipid species. Phospholipases hydrolyze PLs at the sn1 or sn2 position to release fatty acids, which can then function as important and biologically relevant secondary lipid signaling molecules. Specifically, phospholipase A2 (PLA2) selectively hydrolyzes oxPL at the sn2 position, releasing arachidonic acid (AA) or lysophospholipids. The AA released by PLA2 hydrolysis can undergo lipid peroxidation with cellular free radicals to generate highly reactive compounds, including HNE and ONE. Diffusible electrophiles are likely the main source of chemical modification of cellular proteins and have been extensively studied. As a result, we will focus on the diffusible electrophiles released from PL cleavage in this review.

2.a. Methods Development for Global Analysis of Proteins Adducted by Phospholipid-Derived Lipid Electrophiles

Modification of cellular proteins is considered to play a significant contribution to the role of lipid electrophiles in the progression of disease. Additionally, in order to completely understand the effects of HNE on cell biology, it is important to clearly define its protein targets. While HNE has been identified as a significant modulator of cell function and signaling, limited work has been done to in identify the critical protein mediators underlying HNE’s multifarious cellular effects. Previous studies from our laboratory and others show that exposure of cells to HNE results in extensive protein modification [43, 44]. Early studies to identify protein adducts relied largely on 2D-gel electrophoresis coupled by Western blotting using antibodies specific for HNE. This approach is useful for determining if a particular protein of interest is adducted. However, the limitation of Western blotting is that it does not allow the global analysis of potential electrophile protein targets. Thus, a considerable obstacle in defining the role of electrophile adduction in cell biology is the limitation in our ability to monitor electrophile adduction in cellular systems. Identification and characterization of electrophile protein targets in biological samples is significantly complicated by the relatively low levels of adducts in comparison to the large abundance of unmodified proteins. Global analysis of electrophile adducts requires methodology to perform high affinity capture of adducted proteins. Presently, the most effective methods to study HNE susceptible protein involve a mass spectrometry (MS)-based proteomic approach. Such methodologies not only allow detection of protein targets at high resolution but also permit the subsequent identification of specific amino acid residues modified by lipid electrophiles.

The discovery of protein adducts has been greatly accelerated by the use of alkynylated electrophiles that mimic their native counterparts in the adduction of target proteins [43]. Alkynylated electrophiles display similar toxicity and react with target protein nucleophiles with similar chemistry compared to non-alkynylated electrophiles. Alkynylated electrophiles are also stable in cells and in vivo, which makes them good probes for specifically labeling HNE modified proteins [43]. Furthermore, the use of click chemistry, or Huisgen 1,3-dipolar cycloaddition, to conjugate an azido-biotin molecule to the alkynyl electrophile tag improves the specificity of isolated protein adducts by enriching them for alkyne-containing products [45]. Proteomic analysis of affinity-purified biotin-conjugated proteins using alkynylated-HNE probes identified a broad spectrum of electrophile protein targets in an unbiased fashion [43]. Groups of proteins involved in stress signaling and redox regulation were identified as targets of HNE, emphasizing specific pathways which should be further characterized with respect to their activity following electrophile modification. The combination of internal electrophile probes with affinity purification was the first major step toward achieving a global understanding of the proteins modified by lipid electrophiles during oxidative stress.

Affinity tags are essential for the identification of proteins modified by lipid electrophiles. Post-labeling strategies that use biotin hydrazide to biotinylate carbonyl-containing compounds in HNE-treated cells allow the enrichment of adducted proteins when coupled with streptavidin capture [44]. As decribed above, one of the major reactions that electrophiles, including HNE, can undergo with protein nucleophiles is Michael addition. Electrophiles are able to form Michael adducts with cysteine thiols, histidine imidazoles and lysine ε-amines on target proteins. Michael addition of HNE to target proteins results in the formation of a residual carbonyl group. Biotin hydrazide reacts with protein carbonyls, including the residual carbonyl moieties formed during Michael addition of HNE to protein nucleophiles, to generate stable hydrazone derivatives (figure 3) [44, 46-48]. The biotinylated proteins can be captured on a streptavidin affinity column allowing the isolation of tagged-proteins from complex cellular mixtures. Once isolated, the proteins can be removed from the streptavidin matrix by boiling, resulting in the isolation of a purified population of biotinylated proteins. This isolation strategy, coupled with LC-MS-MS, allowed the identification of 417 discernible proteins with a statistically significant increase in adduction with increasing HNE exposure [44]. Western blot analysis of a subset of the streptavidin captured proteins showed adduction at HNE concentrations as low as 1 μM. Thus, the sensitivity of the biotin hydrazide-streptavidin capture and release approach allows a global analysis of electrophile-adducted proteins at physiologically relevant concentrations of HNE. Furthermore, protein interaction network analysis of the proteomics data showed enrichment in several important subsystems following HNE treatment, including oxidative stress response, proteasomal degradation, protein folding, and ribonucleoproteins, among others [44]. This study highlights the relevance of capture and release approaches involving biotin and streptavidin with respect to gaining a global perspective of cellular systems impacted by lipid electrophile adduction of proteins.

Figure 3.

Method for adduct capture and analysis utilizing click chemistry compatible alkynyl-HNE analogues and a biotinylated linker.

The use of biotin hydrazide was crucial in the discovery of patterns of protein damage within the proteome. However, these studies were limited by the presence of high background due to the lack of an efficient way to elute adduction proteins from affinity columns as well as to crossreactivity of biotin hydrazide with endogenous carbonyls [43, 44]. Additionally, the inability to remove the biotin moiety from target protein adducts makes mass spectrometry analyses of isolated proteins difficult. This occurs because the biotin appendage alters MS/MS fragmentation patterns and prevents the identification of peptide adduction sites [44, 49]. In order to reduce high background proteins in proteomic adduct inventories, it was necessary to synthesize a biotin linker that could be easily released from target protein adducts to improve MS/MS identification of adducted proteins. A newly synthesized click reagent containing azido and biotin groups separated by a photocleavable linker improves upon these preliminary affinity purification studies and allows the isolation of a protein mixture enriched in adducted proteins (figure 4) [50]. Importantly, by removing the biotin moiety, this approach allowed the LC-MS/MS detection of specific modification sites on proteins. Thus, such affinity purification-photorelease strategies improve the capability to isolate and identify specific lipid electrophile modifications to proteins from complex biological mixtures.

Figure 4.

Method for adduct capture and analysis utilizing click chemistry compatible HNE analogues and photocleavable biotin linker.

2b. Identification of Specific Residues Adducted by Lipid Electrophiles

Once a protein is recognized as a target of lipid electrophile adduction, identifying the specific sites of modification can provide critical information to define the role the adduct plays in altering the biochemical and biological functions of the protein as well as to define specific patterns of adduct formation that may be useful in biomarker method development. One potential approach to isolate individual proteins and identify the specific adduction site(s)may be to utilize protein-selective affinity capture approaches to perform targeted analysis of electrophile adducts and their consequent biological effects. An example where such an approach has proved useful is in the isolation and identification of HNE adducts of Hsp90 [51]. Recently, a novel non-covalent affinity capture method was employed to isolate Hsp90 using a biotinyl-geldanamycin probe. Geldanamycin is a benzoquinone ansamycin antibiotic that specifically binds to Hsp90 and, thus, can be used to isolate a purified population of Hsp90 protein from cell lysates when combined with biotin-based affinity capture methodology [52, 53]. This strategy was useful in isolating both alpha and beta isoforms of Hsp90 from RKO cells. In vitro and in vivo treatment with HNE coupled with LC-MS/MS identified seven histidine residues that are adducted by HNE that were not previously identified as sites of HNE adduction [51]. Additionally, this inhibitor-based capture method coupled with LC-MS/MS analysis improved both yield and coverage of the protein compared to previous studies. Thus, the use of high-affinity, small molecule based capture methods presents a valuable strategy for the targeted analysis of protein modifications by lipid electrophiles.

2.c. Identification of Protein Adducts of Oxidized Phospholipids

The majority of studies in the literature that examine protein adduction by oxidized lipids are focused on the decomposition products of oxidized phospholipids. However, a limited number of studies identify direct protein targets of oxPL species. Approaches to study the effects of oxPL on their capacity to bind cellular proteins largely rely on model PLs that mimic in vivo oxidative lipid byproducts. The synthesis of biotinylated phospholipid analogs has greatly accelerated the isolation and identification of oxPL protein adducts. Studies by Tallman et al synthesized, characterized, and identified potential protein targets of the glycerylphosphatidylcholine (PC) analogs, including biotinylated 1,2-dipalmitoyl-glycerylphosphatidylcholine (DPPB), biotinylated 1-palmitoyl-2-linoleoylglycerylphosphatidylcholine (PLPB), and the biotin-sulfoxide analog of 1-palmitoyl-2-linoleoylglycerylphosphatidylcholine (PLPBSO)(figure 5A)[54]. The compounds consist of modified headgroups containing a biotin moiety, which allows them to be employed for affinity purification strategies. Free radical-mediated oxidation of the modified phospholipids generates electrophiles that are able to adduct model peptides as well as isolated proteins, including human serum albumin (HSA)[54]. These compounds can be incorporated into human plasma and, thus, will be useful in the identification of proteins targeted by oxPL adduction in vivo.

Figure 5.

Structures of biotinylated glycerophospholipids used to capture adducted proteins for identification and analysis.

Recently, a biotinylated derivative of oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphatidylcholine (Ox-PAPC) was used in proteomics analyses to identify target adducted proteins. The derivative, oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphatidyl-(N-biotinylethanolamine) (Ox-PAPE-N-biotin, unoxidized structure shown in figure 5B), has similar biological activity to its parent compound (Ox-PAPC) and was shown to stimulate the upregulation of inflammatory response genes, including interleukin-8 (IL-8), hemoxygenase-1 (HO-1), and activating transcription factor-3 (ATF-3) [55]. Preliminary studies using ox-PAPE-N-biotin to capture and purify/enrich for adducted protein targets successfully identified 29 adducted proteins in endothelial cells. The proteins remained biotinylated throughout the denaturing conditions of SDS-PAGE, suggesting a covalent interaction between the Ox-PAPE-N-biotin and target proteins, similar to the Michael addition and Schiff base formation observed for the hydrolytic components of oxPL [55]. However, as discussed previously with HNE-adduct isolation strategies, the methods employed within this study hindered accurate proteomics identification due to the presence of the bulky biotin tag. Thus, future studies are required to obtain a more comprehensive analysis of the protein adduction capabilities of oxPL.

Another category of phospholipid that has been studied with respect to oxPL-protein adduction is phosphatidylcholine hydroxyalkenal (PC-HA). PC-HAs are oxPLs generated from the oxidation of arachidonate or linoleate at the sn2 position to form terminal γ-hydroxy, α,β-unsaturated aldehydes (figure 2). The resulting products, 5-hydroxy-8-oxo-6-octenoic acid-PC (HOOA-PC) and 9-hydroxy-12-oxo-10-dodecenoic acid-PC (HODA-PC) have been detected in vivo and, because they contain a γ-hydroxy-α,β-unsaturated terminal aldehyde, they can form Michael adducts with the thiol group of cysteine residues in target proteins [56]. The biological significance of HOOA-PC and HODA-PC was recently reviewed extensively [57, 58]. However, with respect to protein modification, the most biologically relevant protein adduct of PC-HAs identified to date is cathepsin B, a lysosomal cysteine protease [56]. The adduction of cathepsin B by PC-HAs in mouse peritoneal macrophages corresponds to a concomitant reduction in cathepsin B enzyme activity, suggesting that modification of the cathepsin B by oxPL could reduce the capacity of macrophages to perform lysosomal degradation of internalized material. While these initial studies are promising, a more thorough analysis of HOOA-PC and HODA-PC targets is necessary and could be achieved by the aforementioned methods using affinity purification of protein adducts followed by proteomics.

2.d. Activity Based Protein Profiling to Identify Groups of Potential Protein Adducts Based on Sequence Similarity

Cysteine residues are the major nucleophilic targets of lipid electrophiles. Specifically, the reactivity of HNE and ONE toward cysteine residues is 150 fold higher than the reactivity toward other amino acids [59, 60]. Therefore, in certain cases, it may be useful to identify a groups of proteins based on sequence similarities to enrich select proteins that may be more reactive towards electrophiles. In such instances, activity-based protein profiling (ABPP) may prove useful to identify groups of proteins that may be significant targets of electrophile modification. For example, Weerapana et al recently used ABPP to identify enzymes with cysteine functionality in their active sites [61]. This study was able to use a proteomics based approach to identify and quantitatively profile the reactivity of functional cysteine residues within a complex biological mixture. This is significant with regard to electrophile adduction studies because cysteine is the most nucleophilic of the amino acids and, as a result, is a significant site of oxidative modification [59, 60].

2.e. Protein Adducts as Biomarkers of Oxidative Stress

The protein adduction technologies presented within this review are on the threshold of revolutionizing the field of free radical biology. One of the outstanding needs in understanding oxidative stress from a physiological standpoint is the urgent need to develop reliable measures of oxidative stress. It is clear that certain lipid electrophile-protein adducts could provide such biomarkers to monitor the state of oxidative damage in humans. For example, apolipoprotein A1 (ApoA1), the major protein component of high density lipoprotein (HDL) is known to be highly modified by reactive lipids in human plasma [50, 62]. Characterization of the specific residues adducted on ApoA1 and other potential biomarker adducts will allow the development of multiple reaction monitoring (MRM)-based LC-MS-MS quantitative assays to be designed targeting specific protein adducts and, potentially, the identification of specific adducts isolated from human specimens. The recent progress in strategies to identify, isolate and characterize electrophile-protein adducts described herein will greatly improve the ability to utilize protein adducts as biomarkers of human disease.

3. Relationship Between Oxidized Lipid-Mediated Protein Adduction and Changes in Cellular Gene Expression

Oxidized phospholipids alter critical cellular signaling pathways including apoptosis, inflammation, and endoplasmic reticulum stress. However, the mechanism through which oxPL exhibit effects on signaling cascades remains unclear. In some instances where the oxPL possesses a terminal γ-hydroxy, α,β-unsaturated aldehyde, it is possible that oxPL could be adducting cell proteins to elicit changes in gene expression. Similarly, lipid electrophiles are known modulators of cell signaling and the ability of lipid electrophiles to alter cell-signaling pathways is believed to occur largely through their ability to modify cellular proteins. In this regard, it is important to note that changes in cellular signaling cascades by lipid electrophiles are highly related to the identity of the proteins targeted and adducted by lipid electrophiles. For example, treatment of cells with 4-HNE activates the heat shock response and pro-apoptotic signaling while reducing inflammatory processes. The aforementioned global analyses of electrophile adducts of proteins identified many proteins involved in the control of these significant regulatory mechanisms such as heat shock proteins, Keap-1 and IKK [63](figure 6). The adduction of such regulatory proteins has a variety of effects, including the degradation of the protein by the cellular proteasome or simply inhibiting the activity of the protein. It is clear that the relevance of electrophile protein adduction extends far beyond the simple chemical reaction between reactive lipids and nucleophilic amino acid residues. Clearly, a biological approach is necessary to understanding the role of protein adducts in cellular function.

Figure 6.

HNE-responsive signaling network [63]. Specific targets include cell cycle proteins, RNA splicing proteins, proteins involved in both cell cycle and RNA splicing, and proteins not involved in either process (colored in pink, cyan, green, and purple, respectively).

3.b. Microarray Analysis Allows View of Global Changes in Gene Expression Resulting from Exposure to oxPL and Lipid Electrophile Protein Adduction

The identification of signaling pathways that are altered in response to oxPL-derived lipid electrophiles has been greatly accelerated by methods that examine global gene expression changes, including microarray analyses. One of the first studies to examine the effects of HNE on cellular gene expression used microarray technology to profile the response of ARPE-19 human retinal pigment epithelial to oxidants, including HNE [64]. This study examined the acute effects of HNE, H₂O₂ and tert-butylhydroperoxide on gene expression profiles. Results show the alteration of 35 genes in common amongst the three oxidants and changes in genes involved in important cell regulatory mechanisms including apoptosis, cell cycle regulation, cell-cell communication, signal transduction, and transcriptional regulation [64]. More recently, microarray-expression profiling experiments performed by West et al. examined both short-term (6h) and long-term (24h) exposure of RKO cells to HNE [65]. This comprehensive study identified pathways responding to oxidative injury, ER stress and extensive protein damage as being the most strongly induced by HNE. The results of this study suggest that protein adduction is one of the key determinants of downstream stress signaling.

Oxidized phospholipids are known to modulate cell-signaling cascades. Studies examining oxPL-induced changes in gene expression, similar to those described for diffusible electrophiles, show global changes in genes involved in inflammation, ER stress, and the antioxidant response [66, 67]. Results show that treatment of endothelial cells with oxPAPC induces proinflammatory gene expression, as evidenced by increased levels of interleukin-8 (IL-8), through a STAT3-dependent mechanism [68]. Furthermore, HOOA-PC, an autoxidation product of PAPC, induces endothelial cell expression of monocyte-chemotactic protein-1 (MCP-1) and IL-8 [69]. The results of studies employing oxPAPC to examine oxPL-induced changes in gene expression may be largely due to the diffusible, hydrolyzed oxidation products of PL. The fate of the oxPL was not determined once the autoxidation products were added to the cells so the identity of the species responsible for altered gene expression is uncertain. In the following sections we will focus on biologically significant changes in gene expression and their relationship to oxPL-derived lipid electrophile protein adducts.

3.c. HNE Upregulates HSF1 Transcriptional Activity Through Interactions with Heat Shock Response Chaperone Proteins

Protein adduction is critical in the activation of the heat shock response. Heat shock factor-1 (HSF1), the major transcription factor controlling the activation of downstream effectors of the heat shock response, is known to be upregulated following electrophile exposure [9]. The transcriptional activity of HSF1 is regulated through its association with a heat shock chaperone complex consisting of heat shock protein 90, heat shock protein 70 and HSF1 (expanded view, figure 6, and simplified, figure 7) [70]. Under conditions of heat shock or chemical stress, HSF1 dissociates from this complex of chaperones and the released transcription factor translocates into the nucleus where it trimerizes, binds to heat response elements (HREs) in the promoter of target genes and activates their transcription (figure 7) [71, 72]. HSF1’s targets include a number of heat shock proteins (hsps) that function as molecular chaperones to assist in the proper folding of other proteins that have been damaged as the result of thermal and/or chemical insult [73, 74]. Hsps, including hsp70 and hsp90, have been identified as significant targets of lipid electrophile modification [43, 44, 51, 75, 76]. A potential mechanism to explain the activation of heat shock response following electrophile stress could be that the modification of hsps results in the release of HSF1 and, as a result, HSF1 translocates to the nucleus where it is active in the transcription of downstream heat shock responsive genes. In support of this notion, the electrophilic fatty acid derivative nitro-oleic acid is known to induce a large number of HSF1 target genes, presumably through similar adduction of HSF1 chaperones [76]. Thus, electrophiles activate the expression of heat shock responsive genes through altering the chaperone-mediated control of HSF1 activity.

Figure 7.

Highlighted cellular signaling pathways that are altered in response to HNE, including heat shock, antioxidant response, and inflammation.

3.d. Keap1 Modification Releases nrf2 and Activates Antioxidant Response Signaling

The exposure of cells to reactive oxygen species or reactive intermediates activates the antioxidant response, which induces pro-survival signaling in an attempt to rescue the cell from the toxic insult. The transcription factor nuclear factor (erythroid-derived 2)-like 2 (nrf2) is a master regulator of antioxidant signaling. Upon activation, nrf2 translocates to the nucleus where it binds to antioxidant response elements (AREs) in the upstream promoter region of target antioxidant genes and stimulates their expression. Important antioxidant targets of nrf2 include heme oxygenase-1 (HO-1), glutathione-S-transferases (GSTs), NADH-dependent quinone oxidoreductase-1 (NQO1) and aldo-keto reductases, all of which play a significant role in the detoxification of reactive oxygen species [77]. Keap1, a protein that is considered one of the key oxidant sensors within the cell, represses Nrf2 transcriptional activity. Keap1 is an inhibitory binding protein that sequesters nrf2 in the cytoplasm under basal conditions and promotes the degradation of nrf2 through a mechanism involving Cullin3-dependent polyubiquitination [78]. Exposure of cells to reactive intermediates, including HNE and, potentially, oxPL, results in the modification of Keap1 at several important cysteine residues (figure 7) [79-82]. The result of such modification is the stabilization and release of nrf2, which then translocates to the nucleus and becomes active. Thus, adduction of Keap1 by lipid electrophiles appears to be a critical step in the activation of the antioxidant pathway.

3.e. IκB Kinase

HNE modifies the immune response by adducting key proteins responsible for the regulation of inflammatory transcription factors. A major pathway identified as being altered in response to HNE is the nuclear factor kappa b (NFκB) proinflammatory signaling. IκB kinase (IKK) is an enzyme complex that is upstream of the NFκB signal transduction cascade. NFκB is a transcription factor whose expression is regulated by the inhibitor of kappa B (IκBα) protein. When the cell is in a basal, non-inflammatory state, IκBα sequesters NFκB in the cytoplasm, rendering the transcription factor inactive. IKK provides another level of regulation by controlling the activity of inhibitory IκBα. Specifically, IKK phosphorylates the IκBα protein, which results in the dissociation of IκBα from NFκB, thus allowing the nuclear translocation and activation of NFκB. Active NFκB induces the transcription of anti-apoptotic genes, including Bcl-2 family members Bfl-1 and Bcl-X_L, and inflammation, including IL-6. HNE has been shown to form covalent adducts with IKK at a critical regulatory cysteine residue, rendering the enzyme inactive [42, 83-85]. Thus, at high concentrations, HNE functions as an endogenous inhibitor of NFκB activation by preventing IKK activation, leading to the blockage in production of pro-inflammatory cytokines (figure 7) [84, 86]. Accordingly, recent studies showed a reduction in the production of the pro-inflammatory cytokines IL-6, IL-1β, and/or TNF-α in response to lipid electrophile exposure [42, 87]. Thus, it is clear that the adduction of regulatory IKK when HNE is abundant results in the inhibition of inflammatory cytokine production.

3.f. Using siRNA Technology to Define Relevant Pathways Whose Expression is Altered by Lipid Electrophiles

Combining siRNA technology with microarray analysis provides a complementary approach to defining connections between electrophile-protein adduction and changes in gene expression. Previous studies from our laboratory combine microarray analysis with siRNA silencing of transcription factors that are activated in response to treatment of HNE. This approach is a useful method to rapidly identify critical downstream pathways that are altered following exposure to lipid electrophiles. Initial studies focused on silencing of HSF-1, a transcription factor critical in the cellular response to HNE. Results showed that HSF1 is responsible for the transcriptional regulation of over 1000 unique genes following treatment with HNE, including genes that are involved in apoptosis, ER stress, and autophagy [10]. Importantly, silencing HSF1 revealed that cells deficient in heat shock gene expression are sensitized to the pro-apoptotic effects of HNE, suggesting that HSF1 is important in protecting cells from electrophile-mediated cell death [10]. Furthermore, microarray analysis of HSF1 silenced cells identified Bcl-2 associated athanogene domain 3 (BAG3) as a critical HSF1 target gene involved in the cellular response to HNE challenge. It is likely that the expression of many more downstream targets are altered following treatment with HNE, and the combination of siRNA silencing with microarray technology provides a fast and efficient method to identify such targets.

Given the success of defining the relevance of HNE in the expression and downstream signaling of HSF1, it is possible that a similar approach could be used in determining the role of other transcription factors in the cellular response to lipid electrophiles. For example, as mentioned previously, the transcription factor nrf2 is activated in response to treatment with HNE. While nrf2 plays a significant role in the control of the antioxidant response, it is possible that additional and unexpected pathways involved in regulation of the response to lipid electrophiles by nrf2 could exist. The combined siRNA-microarray approach would be extremely useful in elucidating such important pathways.

4. Conclusions

The development of methods to achieve a global understanding of the consequences of oxPL and oxPL-derived lipid electrophile exposure on cell function has greatly advanced our knowledge of the pleiotropic effects following macromolecule adduction. Results from the large-scale proteomic experiments described within demonstrate a complex pattern of protein adduction that leads to a variety of effects on cell signaling as established by microarray analysis of global gene expression changes. While we have only outlined three examples where protein adduction precedes alterations in gene expression, it is clear that the adduction of most protein targets of lipid electrophiles results in significant downstream changes. Ultimately, the current research on cellular lipid electrophile stress emphasizes the importance of adopting a biological approach to define the effects of electrophiles on cellular function.

Highlights.

• Proteomic analysis of electrophile consequences on cells shows severe protein adduction

• Protein adduction leads to significant global gene expression changes, as observed by microarray

• It is important to adopt a biological approach to define effects of electrophiles on cell function