|Home | About | Journals | Submit | Contact Us | Français|
The redox potential of the major thiol/disulfide couple, cysteine (Cys) and its disulfide cystine (CySS), in plasma (EhCys) is oxidized in association with oxidative stress, and oxidized EhCys is associated with cardiovascular disease risk. In vitro exposure of monocytes to oxidized EhCys increases expression of the proinflammatory cytokine, interleukin-1β (IL-1β), suggesting that EhCys could be a mechanistic link between oxidative stress and chronic inflammation. Because cell membranes contain multiple Cys-rich proteins, which could be sensitive to EhCys, we sought to determine whether EhCys specifically affects proinflammatory signaling or has other effects on monocytes. We used microarray analysis and mass spectrometry–based proteomics to evaluate global changes in protein redox state, gene expression, and protein abundance in monocytes in response to EhCys. Pathway analysis results revealed that in addition to IL-1β-related pathways, components of stress/detoxification and cell death pathways were increased by oxidized EhCys, while components of cell growth and proliferation pathways were increased by a reduced potential. Phenotypic studies confirmed that a cell stress response occurred with oxidized Eh and that cell proliferation was stimulated with reduced Eh. Therefore, plasma EhCys provides a control over monocyte phenotype, which could contribute to cardiovascular disease risk and provide a novel therapeutic target for disease prevention.
Redox-regulated signaling pathways respond to reactive oxygen species (ROS) and reactive nitrogen species and also to the redox potential (Eh) of thiol/disulfide couples (e.g., thioredoxin [reduced/oxidized], glutathione disulfide [GSH]/GSSG) (Jones, 2008). Specificity is provided by localized oxidant generation (D'Autreaux and Toledano, 2007) counterbalanced by thiol/disulfide couples, where redox balance of specific protein cysteine (Cys) residues is determined by relative rates of oxidation and reduction. A failure of regulation in specific redox signaling pathways can contribute to oxidative pathology even without free radical–mediated macromolecular damage (Jones, 2006).
Numerous studies show that signaling molecules essential for cell growth and death are regulated by redox states at transcriptional and translational levels. Redox regulation of transcription factors, including NF-κB, Nrf-2, p53, HIF-1α, and AP-1, has been well defined in a wide range of cell and organ systems (Liu et al., 2005). For example, nuclear factor (erythroid-derived 2)-like 2 (Nrf-2) activation by tert-butylhydroquinone, observed by its phosphorylation and nuclear translocation, was dependent on the redox states of GSH/GSSG and Trx1Red/Trx1Ox in the cytoplasm and nuclei, respectively (Hansen et al., 2004). Such studies establish that the redox states are critical in control of gene and protein function. However, less information is available concerning the extracellular redox couples, which can affect intercellular communication and cell signaling through redox elements on the cell surface.
While the intracellular redox milieu is mainly controlled by GSH and Trx systems, the most abundant low–molecular weight thiol/disulfide couple in plasma is cysteine (Cys) and cystine (CySS). The redox potential of plasma Cys/CySS, EhCys, and that of plasma GSH/GSSG, EhGSH, are weakly associated, and human studies show that EhCys and/or EhGSH are oxidized in association with known risk factors for cardiovascular disease, including age (Hildebrandt et al., 2002; Jones, 2002), type 2 diabetes (Samiec et al., 1998), carotid intima media thickness (Ashfaq et al., 2006), brachial artery reactivity (Ashfaq et al., 2008), smoking (Moriarty et al., 2003), and alcohol abuse (Brown et al., 2007). Although EhGSH has been shown to be critical for platelet activation (Essex, 2004), EhCys also appears to be important because mammalian cells in culture regulate extracellular Eh to the Cys/CySS value (Jonas et al., 2002). In addition, in vitro studies showed that an oxidized EhCys triggers monocyte adhesion to vascular endothelial cells (Go and Jones, 2005) and controls inflammatory cytokine interleukin-1β (IL-1β) levels in a monocyte cell line (Iyer et al., 2009).
Because monocytes have a critical function in early events of cardiovascular disease (CVD), such responses of monocytes to extracellular redox potential could provide a mechanistic basis for association between plasma EhCys and CVD risk. In the present study, we used global profiling approaches as an unbiased means (i.e., no preconceived hypothesis concerning pathways) to determine effects of extracellular EhCys on THP1 monocytes. Cells were exposed to oxidized (0 mV) or reduced (−150 mV) EhCys by controlled addition of Cys and CySS to the culture media, and gene expression by microarray, protein redox states by mass spectrometry (MS)–based redox proteomics, and protein abundance by quantitative MS–based proteomics were determined. Results show that oxidized extracellular Eh activated cell stress and detoxification systems, while reduced Eh increased cell proliferation. These responses of monocytes to Eh could provide a novel mechanistic link between diet, environmental and genetic factors, and associated increased risk of CVD.
Human monocytes THP1 cells purchased from ATCC (Manassas, VA) were cultured in 10% fetal bovine serum in Rosewell Park Memorial Institute (RPMI) (37°C, 5% CO2). EhCys was obtained by appropriate Cys and CySS addition to CySS-free RPMI (Mediatech, Inc., Manassas, VA) with a constant total Cys equivalents (200μM) (EhCys of −150 mV: Cys [150μM], CySS [50μM]; EhCys of 0 mV: Cys [0.7μM], CySS [99.65μM]) (Go and Jones, 2005; Jonas et al., 2002). Cells were treated with EhCys for 3 or 36 h.
Cells exposed to extracellular EhCys were harvested, and RNA was extracted (Qiagen, Valencia, CA). After verification of RNA integrity (28S:18S ratio), diluted RNA (1 μg/ml) was sent to Expression Analysis (Durham, NC) for DNA microarray hybridization and scanning (Affymetrix gene chips HG-U133plus2.0). File processing included Robust Multichip Average in Bioconductor (http://www.bioconductor.org) for probe analysis and normalization of the microarray data. File and hybridization quality control checks were performed. The array files were quantitatively compared by fold change. Greater than twofold changes (increase or decrease) were recorded. Gene ontology of the significant gene list was performed in Ingenuity Pathways Analysis (http://www.ingenuity.com).
To validate the microarray results, we selected several genes that were changed more than twofold including ATP-binding cassette, subfamily C, member 3 (ABCC3), NAD(P)H dehydrogenase, quinone 1 (NQO1), prostaglandin-endoperoxide synthase 2 (Ptgs2), aldo-keto reductase family 1, member C2 (Akr1c2), and TIMP metallopeptidase inhibitor 3 (Timp3) and measured the gene expression changes by quantitative real-time (RT)-PCR. The complementary DNA of THP1 cells treated with extracellular EhCys (−150 or 0 mV) was created using reverse transcriptase (Qiagen). Quantitative RT-PCR was performed on an icycler (Bio-Rad, Hercules, CA) using the SyberGreen master mix (Sigma-Aldrich, St Louis, MO) in triplicate and analyzed using a standard curve for each gene, which was standardized to 28S. The primers were picked using Primer3 software (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi); forward (F) and reverse (R) primers for each gene are listed as follows: 28S (NM_001031), F: 5′-CGATCCATCATCCGCAATG-3′, R: 5′-AGCCAAGCTCAGCGCAAC-3′; Abcc3 (NM_003786), F: 5′-GCTCCAAGATCCTTTTAGCCAA-3′, R: 5′-GCCAAGATGAGGGCAGAGAGTA-3′; Nqo1 (NM_000903), F: 5′-TGAAGAAGAAAGGATGGGAGG-3′, R: 5′-AGGGGGAACTGGAATATCAC-3′, Ptgs2 (NM_000963), F: 5′-AAGTGCGATTGTACCCGGAC-3′, R: 5′-ACTGTGTTTGGAGTGGGTTTCA-3′; Akr1c2 (NM_001354), F: 5′-AATTCCAGTTGACTTCAGAGG-3′, R: 5′-ACCAGCATAGAGCCATCC-3′. Timp3 (NM_000362), F: 5′-CCACCAAGCACAGTCAAG-3′, R: 5′-AACCAGAACCAACTAACACC-3′.
THP1 cells (2.5 × 105) were cultured at −150 or 0 mV EhCys for 36 h (Go and Jones, 2005). Cell growth and proliferation were quantified by measuring absorbance of the dye product from the nonradioactive quantitative reagent WST-1 (Roche, Basel, Switzerland) and by cell counts using a hemacytometer.
To quantify effects on protein abundance, 100 μg protein samples were taken from simultaneous cultures treated with −150 and 0 mV. Following reduction of each sample with TCEP (tris-[2-carboxyethyl phosphine]), the −150 mV sample was labeled with heavy (H; 13C) and the 0 mV sample was labeled with light (L; 12C) Isotope Coded Affinity Tag (ICAT) reagent, for 2 h at 37°C. H- and L-labeled proteins were mixed and digested by trypsin for 18 h. Tryptic peptides were purified by cation exchange and avidin columns following instruction provided by manufacturer (Applied Biosystems) and analyzed by nanoLC-MS/MS (Ultimate 3000 nanoHPLC [Dionex, Sunnyvale, CA] and QSTAR XL MS/MS [Applied Biosystems]) system. Proteins were identified with H:L ratio as a comparison of protein abundance, and data are shown as fold change of protein abundance. All quantification was performed by the ProteinPilot V2.0.1 software using the Swiss-Prot database. Quantification for proteins of interest was manually validated by examination of the raw data.
In addition to quantify protein abundance, the ICAT approach has been used to identify oxidant-sensitive proteins (Sethuraman et al., 2004). To examine redox sensitivity of proteins, we performed redox ICAT analysis following the procedures described in the previous report (Go et al., 2009). This procedure differed from the above ICAT method for protein abundance in that precipitated proteins were not initially reduced. Instead, 100 μg of protein from each culture conditions was first treated with the H-ICAT reagent for 2 h, which allowed labeling of only Cys residues in the reduced form at the time of sampling. Protein was precipitated by 5% trichloroacetic acid for 30 min on ice, pelleted by centrifugation, washed with acetone, and resuspended in denaturing buffer provided by the manufacturer (Invitrogen, Carlsbad, CA/Applied Biosystems). Unlabeled disulfides in the proteins were then reduced to thiols by TCEP, and these thiols were labeled with L-ICAT for 2 h. Each sample was digested with trypsin for 18 h, purified, and analyzed by MS as described above. Using this approach, proteins from each treatment (−150 or 0 mV) were identified with an H:L ratio as a measure of the oxidized state of protein. These are expressed as percentage values and labeled as “% oxidized” because the efficiency of trapping redox states could differ for different peptides. Because all samples were treated equivalently, values of oxidized state are comparable even though they may not precisely reflect percentage oxidation in vivo.
Functional genomic and proteomic networks affected by extracellular EhCys changes were analyzed using Ingenuity Pathway Analysis (IPA) (www.ingenuity.com) as previously described (Calvano et al., 2005; Li et al., 2007). IPA uses a curated database of known functional interactions to aid in identification of networks to elucidate biological mechanisms altered by different experimental conditions. Significant pathways were determined by overlap between the experimental genes with a change of expression and the genes in the annotated pathways of the IPA curated database. We input the gene expression data to the IPA system based on the gene chip operating software analysis. We specifically explored networks that had more than twofold changes in gene expression. We input proteomics data for proteins that had a significant p value of identification through the ICAT software and were changed in abundance by at least 30%.
THP1 cells treated with EhCys for 3 h were washed by PBS, incubated with dichlorofluorescin diacetate (50μM; Invitrogen), or MitoSOX (5μM; Invitrogen) to quantify ROS production in the cytoplasm (mainly hydrogen peroxide [H2O2]) and mitochondria (mitochondria-derived superoxide [O2·−), respectively. Cells labeled with these reagents were washed by PBS and fluorescent dichlorofluorescin (DCF), and MitoSOX were measured as indicators of cellular and mitochondrial ROS production, respectively, following the procedures provided by manufacturer (Invitrogen) (Go and Jones, 2005).
Total cellular EhGSH was measured by high-performance liquid chromatography (HPLC; Jones, 2002), and cytoplasmic and nuclear thioredoxin (Trx1) and mitochondrial Trx2 were measured by redox Western analyses following procedures described previously (Halvey et al., 2005). THP1 cells treated with dithiothreitol (DTT) (5mM) and H2O2 (5mM) were used for reduced and oxidized controls, respectively.
Cells (1 × 106) were cotransfected with plasmids containing ARE4 luciferase (the ARE4 region of GCLC inserted into a pT81 luciferase vector) (Mulcahy et al., 1997) and lacZ (Invitrogen) using Lipofectamine LTX following instruction provided by the manufacturer. On the next day, cells were treated with EhCys (−150 or 0 mV) for 36 h and used with the ARE4-luciferase assay as a measure of Nrf-2 activity. Luciferase activity was normalized by β-galactosidase activity. To initiate luciferase activity measurement, cell lysates (20 μl) were added to 100 μl of reaction buffer (Promega, Madison, WI) and luminescence was recorded using a luminometer (Packard, Ramsey, MN). β-Galactosidase activity was quantified by monitoring cleavage of o-nitrophenyl-β-D-galactopyranoside.
Previous research showed that extracellular EhCys has little effect on cellular EhGSH or EhTrx1 in endothelial cells but alters the redox states of membrane proteins (Go and Jones, 2005). To determine whether extracellular EhCys altered thiol/disulfide redox states of the proteins in THP1 monocytes, cells were incubated at EhCys of −150 mV (highly reduced) or EhCys of 0 mV (highly oxidized), redox value found in human plasma, and protein extracts were analyzed by redox ICAT. Percentage of oxidized state represented by 504 protein peptides for −150 mV and 314 protein peptides for 0 mV are summarized in a histogram (Fig. 1). Results show that redox states of many proteins were shifted by oxidized extracellular EhCys (0 mV), suggesting that EhCys has a global effect on protein thiol/disulfide redox states in THP1 cells (−150 mV [black], mean = 57.2 ± 16.6% and 0 mV [red], Mean = 69.3 ± 18.1%). The results indicate that multiple signaling pathways in monocytes could be affected by oxidized extracellular EhCys.
To test for effects of extracellular redox potential on gene expression, THP1 cells (1 × 107) were exposed to EhCys at −150 or 0 mV and examined using microarray analysis. The results identified 62 genes that were more than twofold different between −150 and 0 mV (Supplementary Data 1). The original gene array data were provided in Supplementary Data 2 and 3. Because the sample size was two, statistical methods could not be used to test significance and fold change was used to provide the gene list. Because we could not assure significance of low fold changes (< 2.0), we may have missed expression changes that may have been important. The top 10 transcripts for EhCys of −150 and 0 mV are shown in Table 1. For −150 mV, these include increases associated with cell growth and proliferation. In contrast, 0 mV increased a variety of genes, including NQO1, suggestive of an oxidative stress response. Since the microarray chips were only used as a global consideration, several genes that changed more than twofold were further examined by RT-PCR to validate the microarray results. The subset of transcripts measured by RT-PCR included ABCC3, Akr1c2, NQO1, PTGS2, and TIMP3. Relative messenger RNA (mRNA) levels were examined by comparing differences in copy numbers between reduced and oxidized treatments. Figure 2 shows that mRNA levels of ABCC3 (9.6 ± 2.6), AKR1C2 (10.0 ± 2.9), and NQO1 (5.2 ± 0.9) were higher in 0 mV compared to −150 mV (Fig. 2, black bars). mRNA levels of PTGS2 (7.7 ± 0.1) and TIMP3 (2.0 ± 1.0) were higher at −150 than at 0 mV (Fig. 2, gray bars). These results confirmed the microarray analysis data showing that ABCC3 (2.8-fold), AKR1C2 (3.2-fold), and NQO1 (3.7-fold) are higher at 0 mV, while PTGS2 (4.1-fold) and TIMP3 (2.2-fold) are higher at −150 mV.
IPA is a web-based software application (www.ingenuity.com) to identify the biological mechanisms, pathways and functions most relevant to experimental data sets, or genes of interest. Pathway analysis of the 62 genes that changed in response to EhCys showed that many of these are involved in networks controlling cell proliferation and cell stress. Figure 3A was obtained from networks identified by IPA and demonstrates eight genes that were highly expressed by EhCys of −150 mV (red) that are involved in the platelet derived growth factor (PDGF) pathway, indicating that the more reduced redox potential stimulates this pathway for cell growth and proliferation. Additional information of the molecular and cellular functions by IPA also indicates that 18 genes were involved in cellular growth and proliferation mechanisms (p value < 0.02). Figure 3B obtained from networks identified by IPA shows genes highly expressed by EhCys of 0 mV are involved in oxidative stress signaling mechanisms associated with IL-1β, H2O2, and retinoic acid signaling. Top Tox Lists identified by IPA include thyroid hormone receptor/retinoid X receptor activation (p value < 0.01), oxidative stress response mediated by Nrf-2 (p value < 0.01), and oxidative stress (p value < 0.01). Consequently, the IPA results suggest that reduced extracellular Eh activates mechanisms for cell growth and proliferation, while oxidized Eh controls inflammatory and oxidative stress signaling.
Based on the IPA results showing that EhCys regulates expression of genes for cell growth and proliferation mechanisms, THP1 cells exposed to EhCys were examined for cell growth and proliferation. Cell growth and proliferation analyzed by WST-1 staining, which largely measures mitochondrial activity, showed significant increase in staining for cells at −150 mV after 24 or 36 h (Fig. 3C). These results were confirmed by cell counts obtained using a hemacytometer, which also showed THP1 cell counts were significantly higher by −150 mV (24 h: −150 mV, 5.8 ± 0.5 × 105 vs. 0 mV, 5.1 ± 0.7 × 105 and 36 h: −150 mV, 6.5 ± 0.3 × 105 vs. 0 mV, 5.1 ± 1.5 × 105). Thus, increased abundance of cell growth and proliferation genes in THP1 cells at EhCys of −150 mV, as identified by gene expression array, was confirmed by cell proliferation and growth assay.
The top 10 genes from microarray that were increased under the oxidized EhCys (0 mV) did not clearly reveal pathway responses. However, the pathway analysis of 62 genes showed that oxidized Eh stimulated a Nrf-2-mediated oxidative stress response as a top canonical pathway reflected by the abundance of ferritin (FTH1), v-maf (MAF), and NQO1 (Table 2). Other pathways stimulated by oxidized EhCys are shown. These include GSH metabolism, xenobiotic metabolism, and androgen and estrogen metabolism. Additional study by quantitative proteomic analysis of EhCys on THP1 cells, as indicated below, also supported oxidized Eh stimulated Nrf-2-mediated oxidative stress responses.
To verify Nrf-2-mediated pathways are activated in oxidized conditions (EhCys), we examined whether Nrf-2 is activated in cells exposed to an oxidized EhCys (0 mV) by measuring the antioxidant response element (ARE)–conjugated luciferase activity (Fig. 2B). Nrf-2 activation stimulates expression of genes encoding detoxifying enzymes by activation of ARE as consequence of Nrf-2 binding to ARE. As expected, the results (Fig. 2B) showed that EhCys of 0 mV stimulated ARE-luciferase activity relative to EhCys of −150 mV, suggesting that an oxidized extracellular EhCys stimulates Nrf-2 activation.
The relative abundance of proteins for −150 and 0 mV was determined using nanoLC-MS/MS with the quantitative ICAT method. This identified 110 proteins that were more than 30% different between −150 and 0 mV (Supplementary Date 4). Pathway analysis results of the 110 proteins which changed by > 30% showed that oxidized extracellular EhCys stimulated toxicologic and canonical pathways mediated by Nrf-2 (p < 0.033), including ACTB (actin-beta, P60709), CAT (catalase, P04040), CCT7 (chaperonin containing TCP1, Q99832), GSTM3 (glutathione S-transferase M3, P21266), HSP90AB1 (heat-shock protein 90kD, P08238), and VCP (valosin-containing protein, P55072). Thus, the quantitative proteomics analyses confirmed that oxidized EhCys increased abundance of components linked to oxidative stress and GSH metabolism.
To provide a functional test for an oxidative stress response to oxidized EhCys, we examined ROS levels in THP1 cells with the DCF and MitoSOX assays. First, the oxidized extracellular EhCys significantly increased DCF fluorescence in THP1 cells, suggesting that oxidized EhCys signals to stimulate ROS formation in cells (0 mV; 252.2 ± 25.5% vs. −150 mV; 100 ± 8.5%, Fig. 3D, left). Increase in DCF fluorescence demonstrates an increase in total cellular ROS; however, DCF itself does not define the source of ROS since H2O2 is membrane diffusible. Thus, we examined whether oxidized EhCys increased ROS were associated with mitochondrial ROS production through measurement of mitochondrial O2·− (MitoSOX assay). Consistent with the result of the DCF assay, MitoSOX fluorescence was also significantly increased by oxidized EhCys (0 mV; 125 ± 3.4% vs. −150 mV; 100 ± 3.1%, Fig. 3D, right). Together, these results suggest a key role of mitochondria-derived ROS in extracellular EhCys-dependent oxidative stress signaling.
Analysis of the effects of the extracellular EhCys for 36 h on cellular thiol/disulfide systems showed no detectable differences between −150 and 0 mV on redox states of cellular GSH (Fig. 4A), cytoplasmic Trx1 (Fig. 4B, top), or nuclear Trx1 (Fig. 4B, bottom). However, Trx2, which is present in mitochondria, showed significant oxidation by oxidized EhCys (Fig. 4C), suggesting that responses to extracellular EhCys are signaled through a mitochondrial mechanism.
To evaluate whether GSH redox state in mitochondria was altered by oxidized EhCys, cells exposed to EhCys of −150 or 0 mV were treated with increasing concentrations of digitonin to release cell cytoplasm and leave mitochondria associated with the cytoskeleton (Jones, 1984). The GSH and GSSG contents of these digitonin-permeabilized cells were determined by HPLC. Cytoskeleton-associated GSH was ~1% of the total cellular GSH for both conditions (EhCys of 0 and −150 mV), while that of GSSG was about 20% of the cellular GSSG for both conditions (data not shown). The interpretation of these data is difficult because the secretory pathway, which contains high GSSG concentration, remains intact with digitonin permeabilization. Thus, even though the data suggest that the extracellular Eh does not have a major effect on the mitochondrial GSH content, the data are insufficient to conclude whether extracellular EhCys affects the mitochondrial EhGSH in the same way as affecting mitochondrial Trx2 redox state.
Plasma EhCys is oxidized in humans in association with aging, cigarette smoking, alcohol abuse, type II diabetes, and high body mass index, and this could reflect a susceptibility to environmental toxicants. These associations are noteworthy because the latter are risk factors for cardiovascular disease. Consequently, EhCys has been suggested as a possible biomarker of oxidative stress, which could be useful for assessment of cardiovascular and other disease risks. However, plasma EhCys varies as a function of time of day with greatest oxidation in the morning; moreover, the amplitude of redox variation increases with age (Blanco et al., 2007). Because of this, use of EhCys as a biomarker for oxidative stress requires consistency in collection, which is not always possible. On the other hand, biochemical systems that respond to EhCys would integrate responses to EhCys over time and may be easier to apply to routine clinical use. Hence, knowledge of responses to oxidized plasma EhCys is useful both for understanding underlying mechanisms of CVD and for development of novel biomarkers.
Oxidized EhCys has been implicated as a causative factor in early atherogenic signaling because exposure of endothelial cells to oxidized EhCys increases expression of cell adhesion molecules and stimulates monocyte adhesion (Go and Jones, 2005). The current findings show that exposure of monocytes to oxidized EhCys stimulates genes associated with the IL-1β pathway and activates Nrf-2-dependent protective mechanisms. These results are consistent with results showing increased IL-1β production in U937 monocytes in response to oxidized EhCys (Iyer et al., 2009) and supports the possibility that oxidized EhCys contributes in a causal way to cardiovascular disease, either by enhanced proinflammatory cytokine production or by enhanced binding and transmigration through the endothelium. Importantly, the EhCys effects are associated with oxidation of mitochondrial but not cytoplasmic or nuclear redox systems, suggesting that mitochondrial oxidation may have a central role in signaling and could be a potential target to prevent early events of atherogenesis. This result supports the previous finding by Liu et al. (2003) showing a critical role of mitochondrial ROS generation in the regulation of vascular function.
The present redox proteomics analysis showed that exposure of monocytes to oxidized EhCys caused a shift of protein thiols to disulfides. Although only a small fraction of the total proteome is captured by this method, ~10% of the peptides detected were more oxidized. Despite the global nature of these proteomic changes, oxidation of the major cytoplasmic thiol systems, GSH and Trx1, was not detected, and microarray and proteomics analyses indicated specificity in activation of signaling pathways. In particular, oxidized EhCys initiated a stress response regulated by Nrf-2, which included the expression of detoxification enzymes. Expression levels for NQO1, GSTM3 and M5, and catalase were higher due to the oxidized EhCys. These findings are consistent with previous results documenting NF-κB activation in response to oxidized EhCys (Go and Jones, 2005). The NF-κB and Nrf-2 pathways are known to mediate similar outcomes and work coordinately in the regulation of cytoprotective genes. For instance, the promoter activity of GTP-binding protein (Giα) was enhanced by binding of both NF-κB and Nrf-2 transcription factors (Arinze and Kawai, 2005), demonstrating NF-κB and Nrf-2 work synergistically to activate target genes. Thus, although multiple proteins are oxidized, signaling responses occur through specific response pathways.
These observations indicate an important need to identify the signaling pathways whereby plasma membrane proteins are oxidized by extracellular EhCys, signal changes in transcription, and result in altered protein abundance. Because of the limited number of redox conditions in the present study, correlations between protein oxidation and gene expression were not possible. However, of the 110 proteins, which changed in abundance, none was present among the 62 transcripts, which changed more than twofold. These data may suggest that redox-dependent mechanisms other than transcription control protein abundance. Alternatively, more extensive proteomics coverage and systematic changes in experimental conditions may be needed to detect such associations.
In constant to the stress response observed with oxidized EhCys, results from previous studies show that enhanced cell proliferation is a common (but not universal) response to more reduced extracellular redox potential (Jonas et al., 2003). The present data for THP1 cells show that genes expressed highly under the reduced redox state are transcription factors and regulators (early growth response protein [EGR1, 2, 3], cyclin L1 [CCNL1], NR4A2), serine/threonine protein kinase (PLK2), and PTGS2 (cyclooxygenase-2) supporting positive regulation of cell growth and proliferation.
The in vitro monocyte culture model used in the current study relied on controlled variations in Cys/CySS redox state in tissue culture medium. Because of the artificial nature of in vitro tissue culture medium compared with plasma, it remains unclear whether response properties of cells in vivo would be quantitatively or qualitatively the same as presented here. Additional in vivo studies are needed to address this critical issue.
In summary, controlled oxidized and reduced extracellular EhCys resulted in different redox states of many cellular proteins and caused changes in both gene expression and protein abundance. These changes reflect phenotypic differences wherein cells cultured in the oxidized conditions exhibited a stress response and cells cultured in reduced conditions exhibited increased growth and proliferation. The results indicate that the extracellular EhCys influences cell fate of monocytes in patterns that could contribute to vascular disease risk. Finally, we propose the scheme for the extracellular EhCys-dependent cell signaling and control based on our current findings taken together with previous findings (Fig. 5).
National Institutes of Environmental Health Sciences (ES011195, ES009047).