Lack of the Antioxidant Enzyme Glutathione Peroxidase-1 Accelerates Atherosclerosis in Diabetic Apolipoprotein E–Deficient Mice
Background— Recent clinical studies have suggested a major protective role for the antioxidant enzyme glutathione peroxidase-1 (GPx1) in diabetes-associated atherosclerosis. We induced diabetes in mice deficient for both GPx1 and apolipoprotein E (ApoE) to determine whether this is merely an association or whether GPx1 has a direct effect on diabetes-associated atherosclerosis.
Methods and Results— ApoE-deficient (ApoE−/−) and ApoE/GPx1 double-knockout (ApoE−/−GPx1−/−) mice were made diabetic with streptozotocin and aortic lesion formation, and atherogenic pathways were assessed after 10 and 20 weeks of diabetes. Aortic proinflammatory and profibrotic markers were determined by both quantitative reverse-transcription polymerase chain reaction analysis after 10 weeks of diabetes and immunohistochemical analysis after 10 and 20 weeks of diabetes. Sham-injected nondiabetic counterparts served as controls. Atherosclerotic lesions within the aortic sinus region, as well as arch, thoracic, and abdominal lesions, were significantly increased in diabetic ApoE−/−GPx1−/− aortas compared with diabetic ApoE−/− aortas. This increase was accompanied by increased macrophages, α-smooth muscle actin, receptors for advanced glycation end products, and various proinflammatory (vascular cell adhesion molecule-1) and profibrotic (vascular endothelial growth factor and connective tissue growth factor) markers. Quantitative reverse-transcription polymerase chain reaction analysis showed increased expression of receptors for advanced glycation end products (RAGE), vascular cell adhesion molecule-1, vascular endothelial growth factor, and connective tissue growth factor. Nitrotyrosine levels were significantly increased in diabetic ApoE−/−GPx1−/− mouse aortas. These findings were observed despite upregulation of other antioxidants.
Conclusions— Lack of functional GPx1 accelerates diabetes-associated atherosclerosis via upregulation of proinflammatory and profibrotic pathways in ApoE−/− mice. Our study provides evidence of a protective role for GPx1 and establishes GPx1 as an important antiatherogenic therapeutic target in patients with or at risk of diabetic macrovascular disease.
Received September 11, 2006; accepted February 21, 2007.
Diabetes mellitus (DM) is a major risk factor for the development of atherosclerosis, with a 2- to 4-fold higher incidence of cardiovascular disease in diabetic patients than in the general population.1 Other known risk factors, such as dyslipidemia, hyperglycemia, hypertension, and obesity, only partly explain the more advanced lesions2 and increased incidence observed in these patients. Although the underlying mechanisms that accelerate diabetic atherosclerosis remain unknown, various pathways have been implicated, including the biochemical process of advanced glycation3 and, in particular, the receptor for advanced glycation end products (AGEs), RAGE.4 Furthermore, various proteins implicated in the atherosclerotic process are upregulated in the diabetic condition, including the adhesion molecule vascular cell adhesion molecule-1 (VCAM-1), the chemokine monocyte chemotactic protein-1, and the profibrotic growth factor connective tissue growth factor (CTGF).5 In addition, strong evidence suggests the involvement of reactive oxygen species (ROS) derived from the hyperglycemia-driven increase in mitochondrial electron transport chain activity,6 glucose autoxidation,7 and enzymes such as NAD(P)H oxidase.8 ROS, in turn, are known to upregulate a number of proatherogenic processes, such as monocyte infiltration, platelet activation,9 smooth muscle cell (SMC) migration,10 cell adhesion,11 release of CTGF,12 and increased production of AGEs.7 Indeed, many of these processes are enhanced in a murine model of DM-associated atherosclerosis.5,13 Importantly, a heightened state of oxidative stress has been observed in diabetic patients14 and postulated to play a role in a range of diabetic complications.7 Eukaryotic cells have evolved an extensive array of antioxidant defenses to regulate the flux of ROS and to limit oxidative damage.15 Blood vessels are known to express antioxidant enzymes to counteract oxidant stress.16 However, reductions in antioxidant defense, which lead to increased ROS accumulation, can elicit pathophysiological consequences.17 Recent evidence from apolipoprotein (Apo) E–deficient mice suggest that the levels of several antioxidant enzymes decline during atherogenesis,18 which implies a link between reduced antioxidant capacity and increased lesion formation. Conversely, overexpression of the antioxidant catalase, known to remove hydrogen peroxide, reduced the severity of lesions in ApoE-deficient mice,19 which implies a role for hydrogen peroxide–scavenging enzymes in atherosclerotic processes. However, limited information is available on the involvement of individual antioxidant enzymes in DM-associated atherosclerosis.
Clinical Perspective p 2187
Glutathione peroxidase-1 (GPx1) is a major and ubiquitously expressed antioxidant enzyme present in the cytosol and mitochondria. It is involved in the detoxification of hydrogen and lipid peroxides15 and acts as a peroxynitrite reductase.20 In the absence of this antioxidant enzyme, a buildup of ROS ensues that is known to damage DNA, proteins, and lipids.20 Clinical evidence now suggests a potential role for GPx1 in DM-associated atherogenesis.21,22 Polymorphisms identified within the GPx1 gene that result in reduced GPx1 activity have been linked with increased intima-media thickness of carotid arteries and an increased risk of cardiovascular and peripheral vascular disease in type 2 DM patients.21 Furthermore, additional studies suggest a protective role for GPx1 in the atherogenic process per se. For example, reductions in red blood cell GPx1 activity were associated with an increased risk of cardiovascular events in a prospective cohort study,23,24 and atherosclerotic plaques of patients with carotid artery disease have reduced GPx1 activity.25 In animal studies, reduced GPx1 expression increased cell-mediated oxidation of LDL26 and decreased the bioavailability of nitric oxide, leading to endothelial dysfunction.27 These findings imply that GPx1 is a key enzyme for the protection of vessels against oxidative stress and atherogenesis and that GPx1 may be of particular importance in the highly prooxidant diabetic environment. However, to date, no study has directly tested the impact of reduced GPx1 function on the development of atherosclerosis in a diabetic context. Accordingly, the aims of the present study were (1) to evaluate the effects of a lack of GPx1 on the development of DM-associated atherosclerosis and (2) to determine whether a lack of GPx1 impacts on known pathways implicated in DM-associated atherosclerosis.
Animal Groups and Experimental Design
GPx1-deficient mice on a C57Bl/J6 genetic background and described by us previously28 were mated with C57Bl/J6 ApoE-deficient (ApoE−/−) mice to establish a line of ApoE−/−GPx1−/− double-knockout mice. ApoE−/− mice were generated through the breeding program and were maintained as a separate line. All studies were performed in male age-matched ApoE−/− or ApoE−/−GPx1−/− mice.
Eight-week-old ApoE−/− and ApoE−/−GPx1−/− mice were rendered diabetic by 2 intraperitoneal injections of streptozotocin (Sigma-Aldrich, St. Louis, Mo) on consecutive days, at a dose of 100 mg · kg−1 · d−1. Sham-injected ApoE−/− and ApoE−/−GPx1−/− mice served as nondiabetic controls. Nonfasted blood glucose readings were measured every 3 to 4 weeks from the tail vein. Animals had unrestricted access to water and standard mouse chow. All studies were performed with approval of the Institutional Ethics Committee.
Animals were maintained for 20 weeks (20-week study) for enface aortic lesion analysis. In addition, to assess the status of other antioxidants and various pathways implicated in the progression of atherosclerosis, separate groups were assessed after 2 or 10 weeks of DM. Lesions were quantitated within the aortic sinus at both the 10- and 20-week time points. The study specifically involved enface assessment after 20 weeks of DM on the basis of previous studies in our laboratory in which total plaque area was shown to be significantly increased in diabetic ApoE−/− mice compared with nondiabetic controls.5 We specifically chose to investigate the expression of known proatherogenic markers after 10 weeks of DM to establish whether lack of GPx1 affected proatherogenic pathways at an earlier stage, thereby gaining valuable information about potential molecular pathways involved in the progression toward highly developed lesions. Because oxidative stress and activation of redox-sensitive transcription factors have been shown to occur within hours after glucose treatment,29 and therefore as an early event of the diabetic milieu, we also believed that analysis at the earlier 2-week time point would give additional insight into the effects of a lack of GPx1.
Blood Sampling, Plasma Biochemistry, and Tissue Collection
Mice were anesthetized by an intraperitoneal injection of 2,2,2-tribromoethanol (Sigma Chemical Co., St. Louis, Mo) after food was withheld for 4 hours. The thoracic cavity was opened rapidly and blood drawn by direct puncture of the right ventricle. Blood samples were placed into heparinized tubes and centrifuged at 2500 rpm, and plasma was frozen at −20°C until analyzed.
Plasma glucose, cholesterol, HDL, and triacylglycerols were measured with commercial enzymatic kits. The Friedewald calculation was used to determine levels of LDL.
Hearts with aortas attached were rapidly removed and placed in cold saline. The aorta was severed close to the heart and cleaned of extraneous fat under a dissecting microscope. Cleaned aortas from the 20-week study were stored in 10% neutral buffered formalin for assessment of atherosclerotic lesions and subsequent immunohistochemical analysis. Aortas from the 2- and 10-week study were snap-frozen in liquid nitrogen and stored at −80°C for protein and gene expression studies, respectively. Hearts, cut at a plane parallel to the tips of the atria, were fixed in 4% paraformaldehyde and frozen in OCT (Sakura, Torrance, Calif) for lesion assessment within the sinus region.
Evaluation of Atherosclerotic Lesions
Atherosclerosis was assessed by 2 approaches, first by the enface technique described by us previously5 to determine total and regional aortic plaque distribution, and second, by cross-sectional assessment through the aortic sinus. After enface analysis, aortic tissue was embedded in paraffin, and cross-sectional serial sections 3-μm thick were prepared for immunohistochemical analysis. Lesions within the aortic sinus region were assessed on the basis of the method of Paigen et al30 and described by us previously.31 Briefly, 10-μm-thick cryostat sections, selected 80 μm apart and covering 320 μm of the sinus, were stained with oil red O to detect lesions. Images were quantitated with Optimas software (Media Cybernetics, version 6.2; Bethesda, Md). For each mouse, lesion size (μm2) was determined from the average of 5 cross sections, and the results of each group are expressed as lesion size (μm2) ±SEM. (See also the expanded Methods section in the online Data Supplement).
Serial aortic paraffin sections were stained for α-smooth muscle actin (α-SMA), RAGE, the macrophage marker F4/80, nitrotyrosine, gp91phox, CTGF, and vascular endothelial growth factor (VEGF). Frozen aortic sinus sections were stained for VCAM-1. Primary and secondary antibodies and dilutions are described in Data Supplement Table I. Immunohistochemical methods were described by us previously5 (see also the expanded Methods section in the Data Supplement). Images were visualized under light microscopy and quantitated with Optimas (Media Cybernetics, version 6.2). Analysis was performed in a blinded fashion. On average, 3 sections were assessed per mouse, and 7 to 9 mice were analyzed per group. Results were calculated as percentage of positively stained tissue in the aortic media (for all except VCAM-1, for which the percentage of positive staining was determined within the sinus region) and expressed relative to nondiabetic ApoE−/− mice that were arbitrarily assigned a value of 1.
Quantitative Reverse-Transcription Polymerase Chain Reaction
Total RNA extraction, preparation of DNA-free RNA, and RNA reverse transcription have been described by us previously.31 Gene expression of VCAM-1; CTGF; VEGF; RAGE; α-SMA; nuclear factor (NF)-κB subunit p65; the antioxidant enzymes GPx1, GPx3, GPx4, catalase, and superoxide dismutase-1 (Sod1) and -2 (Sod2); and the NAD(P)H oxidase subunit gp91phox were analyzed by quantitative reverse-transcription polymerase chain reaction as described by us previously.31 (See also online Data Supplement Methods and Table II, which describe probes and primers for VEGF, Sod1, Sod2, gp91phox, NF-κB, and RAGE).
Immunoblot Analysis of Individual Aortas
Individual aortas were homogenized in 20 mmol/L cold HEPES buffer containing 1 mmol/L EGTA, 210 mmol/L mannitol, and 70 mmol/L sucrose. Cytosolic preparations were prepared by 2-step centrifugation; the first at 1500g for 5 minutes followed by a 10 000g spin for 15 minutes. Protein concentration was determined, and equivalent amounts of protein were loaded per lane. We used rabbit polyclonal anti-GPx1 antibody, anti-Cu/Zn superoxide dismutase (SOD) antibody, and anti-catalase antibody (all from Abcam Ltd, Cambridge, United Kingdom) as the primary antibodies. Pierce (Rockford, Ill) ImmunoPure goat anti-rabbit IgG, peroxidase conjugated, was used as secondary antibody. Mouse monoclonal anti-β-actin antibody (Abcam Ltd; ab6276) and goat anti-mouse IgG, peroxidase conjugated (Dako, Glostrup, Denmark), were used as primary and secondary antibodies to detect β-actin. Proteins were detected by ECL (Pierce SuperSignal West Pico Chemiluminescence Substrate) and visualized on a ChemiDoc documentation system (Bio-Rad Laboratory, Milan, Italy). Band intensities were determined by densitometry. Four aortas were analyzed per group, and the results are expressed as mean±SEM (see also Data Supplement Methods section).
Data were analyzed by 2-way ANOVA with diabetic status and GPx genotype as the 2 variables. Pairwise multiple comparisons were made with the Student-Newman-Keuls method to detect significant differences between groups (SigmaStat version 2.03, Systat Software, San Jose, Calif). Results are expressed as mean±SEM. P<0.05 was considered statistically significant.
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
Phenotypic Assessment of ApoE−/−GPx1−/− Mice and Metabolic Parameters
We generated mice lacking both GPx1 (GPx1−/−) and ApoE (ApoE−/−). Double-knockout mice (ApoE−/−GPx1−/−) appeared healthy, bred normally, and exhibited standard mendelian transmission of both mutant alleles. No obvious pathological changes were observed in any organ investigated.
In the absence of DM, lack of GPx1 did not affect the lipid profile of any of the major lipoprotein groups investigated in sham-injected ApoE−/− mice analyzed after 20 weeks, as is shown in Data Supplement Table III. Similarly, body weight and plasma glucose were unaffected by the lack of GPx1 in sham-injected ApoE−/−GPx1−/− mice (Data Supplement Table III).
In mice administered streptozotocin, nonfasted tail-vein blood glucose levels were significantly increased for the duration of the study compared with sham-injected animals (data not shown). Importantly, similar blood glucose levels were achieved for diabetic ApoE−/− and ApoE−/−GPx1−/− mice, thereby eliminating differences in hyperglycemia as potentially contributing to experimental outcomes. Plasma glucose levels from fasted animals were also significantly and similarly increased in diabetic ApoE−/− and ApoE−/−GPx1−/− mice compared with nondiabetic counterparts at the conclusion of the 10-week (P<0.01; data not shown) and 20-week (P<0.01; Data Supplement Table III) studies. Body weights were not significantly affected by DM in either group, as shown in Data Supplement Table III. Twenty weeks of DM did not affect triglycerides, HDL, LDL, or the LDL/HDL ratio of ApoE−/− mice in the present study. A trend toward elevated total cholesterol levels occurred in diabetic ApoE−/− mice, although this did not reach significance (P=0.06). Similarly, lack of GPx1 did not affect lipid profiles of diabetic ApoE−/−GPx1−/− mice after either 10 (data not shown) or 20 (Data Supplement Table III) weeks of DM.
Assessment of Aortic Atherosclerotic Lesions
En Face Analysis of Aortic Tree
In the absence of DM, lack of GPx1 did not affect the regional or total plaque deposition in aortas of sham-injected ApoE−/− mice (Figure 1A, 1B, 1E, and 1F). Diabetes was associated with a 2.5-fold increase in total plaque in ApoE−/− mice (P<0.01; Figure 1C and 1E). Regional evaluation revealed a 3.0-fold increase in the highly susceptible arch region (P<0.001) together with a significant increase in thoracic plaque (P<0.05) in diabetic ApoE−/− aortas compared with nondiabetic ApoE−/− aortas (Figure 1C and 1F). A similar trend, albeit not significant, was observed in the abdominal region of these aortas. Diabetic mice lacking GPx1 displayed a further 2.0-fold increase in total plaque compared with diabetic ApoE−/− mice (P<0.001; Figure 1D and 1E). Similarly, the aortic arch (P<0.001), thoracic (P<0.01), and abdominal regions (P<0.01) showed significant increases in plaque deposition in diabetic ApoE−/−GPx1−/− compared with diabetic ApoE−/− aortas (Figure 1D and 1F).
Aortic Sinus Region
In the absence of DM, lack of GPx1 did not affect lesion size or composition within the sinus region of mice, 10 or 20 weeks after sham injection (P>0.05; Figure 2A, 2B, and 2E and Figure 2F, 2G, and 2J, respectively). An approximately 2.5- to 3-fold increase in lesion size was observed within the aortic sinus of ApoE−/− mice after 10 (P<0.01; Figure 2A, 2C, and 2E) and 20 (P<0.001; Figure 2F, 2H, and 2J) weeks of streptozotocin-induced DM. Importantly, lack of GPx1 augmented this a further 1.5- to 2.5-fold after 10 (P<0.001; Figure 2C, 2D, and 2E) and 20 (P<0.001; Figure 2H, 2I, and 2J) weeks of DM.
Proatherogenic Markers in the Absence of DM
In the absence of DM, lack of GPx1 did not significantly affect any of the proatherogenic markers investigated in the present study, either at the transcriptional or protein level (P>0.05 for all; VCAM-1: Figure 3E and 3F; CTGF: Figure 4E and 4F; VEGF: Figure 4K and 4L; α-SMA: Figure 5A and 5B; RAGE: Figure 5C and 5D; and F4/80: Figure 5E).
Proatherogenic Markers in the Presence of DM
VCAM-1 gene expression (Figure 3E) and protein levels (Figure 3A through 3D and 3F) were significantly increased (P<0.01) within the aortic sinus of diabetic ApoE−/−GPx1−/− mice compared with diabetic ApoE−/− controls after 10 weeks of DM. Interestingly, VCAM-1 levels were mainly increased within the vessel wall (Figure 3D). Similarly, the macrophage marker F4/80 was significantly increased in diabetic ApoE−/−GPx1−/− aortas compared with diabetic ApoE−/− controls (P<0.01; Figure 5E). Gene expression and protein levels were significantly increased for CTGF (Figure 4E and 4F; P<0.001 and P<0.01, respectively), VEGF (Figure 4K and 4L; P<0.01 and P<0.05, respectively), α-SMA (Figure 5A and 5B; P<0.001 and P<0.05, respectively), and RAGE (Figure 5C and 5D; P<0.01) in diabetic ApoE−/−GPx1−/− compared with diabetic ApoE−/− aortas. Photomicrographs for α-SMA, RAGE and F4/80 immunostaining are provided in Data Supplement Figure I.
Antioxidant Gene Expression and Immunoblot Analysis
Because DM and atherosclerosis can alter the expression of various antioxidant enzymes,32 and deletion of 1 enzyme such as GPx1 may result in compensatory changes in other antioxidant enzymes,17 we investigated the expression of a range of antioxidants with the potential to remove ROS in vascular tissue.16 Consequently, we investigated the expression of (1) GPx1; (2) the extracellular lipid and hydrogen peroxide–reducing isoform, GPx3; (3) the phospholipid hydroperoxide-reducing isoform, GPx4; (4) the peroxisomal hydrogen peroxide–specific reducing enzyme, catalase; and (5) the cytosolic and mitochondrial isoforms (Sod1 and Sod2, respectively) of the SOD enzyme family.
DM was associated with a significant 2-fold increase in GPx1 mRNA levels in ApoE−/− aortas (Figure 6A; P<0.001), most likely as a response to DM-induced increases in oxidative stress. Homozygous disruption of the GPx1 gene resulted in significant reductions in GPx1 mRNA expression in both diabetic and nondiabetic ApoE−/−GPx1−/− aortas, in agreement with previous data.28 The approximately 10% to 20% residual GPx1 expression in GPx1-deficient aortas most likely reflects background or nonspecific primer amplification, because we were unable to detect any GPx1 protein after immunoblot analysis of individual ApoE−/−GPx1−/− aortas (Figure 6A), consistent with homozygous knockout of the GPx1 gene28 Diabetic ApoE−/− aortas also showed significant increases in GPx3 expression, with similar trends, albeit nonsignificant, for GPx4 (Figure 6A), catalase, and Sod2 (Figure 6B) compared with nondiabetic ApoE−/− controls, most likely a consequence of enhanced oxidative stress known to accompany DM.7,14 Similarly, GPx3, GPx4, catalase, and Sod1 and Sod2 expression was significantly increased in diabetic ApoE−/−GPx1−/− aortas compared with nondiabetic counterparts (P<0.001, except Sod2, for which P<0.05) and was significantly greater than that seen in diabetic ApoE−/− aortas (GPx3 and GPx4, P<0.05; catalase and Sod1, P<0.01). Immunoblot analysis revealed that Sod1 and catalase (Figure 6C) are significantly increased in diabetic ApoE−/−GPx1−/− aortas compared with nondiabetic ApoE−/− aortas (P<0.05, respectively) and that these changes were already apparent after 2 weeks of DM.
Prooxidant NAD(P)H Oxidase and NF-κB Gene Expression
We were interested in the potential impact that lack of GPx1 might have on the expression of the superoxide-generating NADPH oxidase within a diabetic milieu, because lipid peroxides are implicated in the upregulation of NADPH oxidase.33 Lack of GPx1 did not significantly affect gene expression or protein levels of gp91phox (the subunit associated with phagocytic NADPH oxidase) in the absence of DM (Figure 7E and 7F). Gene expression and protein levels were significantly increased in diabetic ApoE−/−Gpx1−/− compared with diabetic ApoE−/− aortas (Figure 7E and 7F; P<0.001 and P<0.05, respectively). In light of recent data that NF-κB mediates oxidant-induced upregulation of NADPH oxidase,34 we also investigated the expression of the p65 subunit of NF-κB. We found p65 mRNA expression to be significantly increased in diabetic ApoE−/− aortas compared with nondiabetic counterparts (Figure 8F; P<0.05) and importantly, significantly increased a further 2.5-fold in diabetic ApoE−/−Gpx1−/− aortas (P<0.001) compared with diabetic ApoE−/− aortas.
Detection of Nitrotyrosine in Aortas
Lack of GPx1 caused a significant increase in nitrotyrosine levels in diabetic ApoE−/−Gpx1−/− aortas (Figure 8A through 8E; P<0.01 versus nondiabetic counterparts and diabetic ApoE−/−), suggestive of increased peroxynitrite-mediated damage to proteins in these aortas.20 No change was observed in nondiabetic ApoE−/−Gpx1−/− or diabetic ApoE−/− aortas.
In the present study, we demonstrate that a deficiency in GPx1 accelerates DM-associated atherosclerosis in the ApoE-knockout mouse. The present data are highly supportive of a recent clinical study in which diabetic patients with reduced GPx1 function, as a consequence of polymorphisms within the GPx1 gene, were shown to have an increased risk for cardiovascular disease.21 The present study has also shown that accelerated DM-driven atherosclerosis is accompanied by the presence of increased nitrotyrosine, a marker of free radical damage,20 and by transcriptional changes of important inflammatory and profibrotic pathways. These molecular changes translated into increased levels of α-SMA, profibrotic CTGF and VEGF, proinflammatory VCAM-1, and RAGE and the increased presence of proinflammatory macrophages. The present study therefore unequivocally links a lack of the important intracellular ROS-removing antioxidant enzyme GPx1 with increased DM-associated atherosclerosis and identifies areas in which antioxidant defense, and GPx1 in particular, plays an important role in regulating atherogenic processes within a diabetic milieu.
The present study also shows that GPx1 does not appear as important in preventing atherosclerosis in ApoE−/− mice in the absence of DM. This difference between diabetic and nondiabetic mice may arise as a consequence of the higher prooxidant environment of DM. Because hyperglycemia is known to greatly increase oxidative stress through various mechanisms,7,8 the beneficial antioxidant function of GPx1 is likely to be of particular importance within a diabetic context. However, the impact of a lack of GPx1 on atherosclerosis in a nondiabetic context needs to be more fully explored in models that more directly address atherosclerosis than that used in the present study. We have recently shown that high-fat feeding of mice results in greatly increased GPx1 expression, which most likely is a consequence of the enhanced oxidative stress known to accompany such diets.31 Therefore, feeding prooxidant, proatherogenic diets should give a clearer indication of GPx1 function in the protection against atherosclerosis in ApoE−/− mice in the absence of DM.
To expand on the mechanisms whereby a lack of GPx1 enhances atherosclerosis in a diabetic milieu, a number of cellular and molecular processes linked to DM-associated atherosclerosis were examined. First, we investigated the impact of reduced GPx1 function on the presence of macrophages and SMCs, because these cell types are integral components of plaque and are known to be increased in DM-associated atherosclerosis.5 The present study showed augmented staining for both cell types in the aortas of diabetic ApoE−/−GPx1−/− mice, which suggests that GPx1 plays an important role in limiting macrophage infiltration and SMC migration during DM-associated atherogenic processes. It is most likely that GPx1 limits such cellular involvement via its antioxidant function in the removal of DM-induced ROS, especially because ROS have been implicated in enhancing monocyte chemotaxis and SMC migration.10 In particular, Wang et al10 demonstrated evidence for VEGF-mediated vascular SMC migration via ROS-mediated activation of NF-κB, whereas pretreatment of these cells with the antioxidant N-acetylcysteine reduced ROS and cellular migration. Indeed, the present data, which show increased staining for nitrotyrosine is suggestive of increased ROS in diabetic ApoE−/−GPx1−/− aortas and is supportive of this notion. In agreement with the findings of Wang et al,10 vascular SMC migration may involve the growth factor VEGF, because we show significantly increased VEGF in ApoE−/−GPx1−/− aortas.
To further explore the underlying molecular mechanisms that may accelerate proinflammatory cell migration into the vascular wall of diabetic ApoE−/−GPx1−/− mice, we investigated the expression of an important mediator of the inflammatory response, namely, the cell adhesion molecule VCAM-1. Furthermore, we chose to focus on VCAM-1 based on strong supportive data that VCAM-1 is redox sensitive and upregulated by ROS.11,35 Indeed, in the present study, lack of GPx1 led to increased VCAM-1 mRNA and protein levels after 10 weeks of DM, which suggests that lack of GPx1 directly affected transcription and translation of this proinflammatory molecule in the progression toward highly developed lesions. Given the known antioxidant functions of GPx1, this most likely occurred via ROS-related mechanisms. Indeed, a number of pathways linked to the upregulation of VCAM-1 involve ROS. First, angiotensin II, the biologically active peptide generated by the renin-angiotensin system, upregulates VCAM-15,35 via ROS-mediated mechanisms.35 This effect was inhibited by a glutathione peroxidase mimic and enhanced by the catalase inhibitor aminotriazole.35 Similarly, RAGE upregulates VCAM-1 via ROS-mediated mechanisms.3 Therefore, a number of pathways may be activated by the lack of GPx1 that in turn increase VCAM-1 and contribute to proinflammatory cell migration. It is therefore also of particular interest that we found RAGE expression, both at the transcriptional and protein level, to be increased by a lack of GPx1 in diabetic ApoE−/−aortas. The present data are consistent with and add to previous studies that demonstrate ROS-mediated upregulation of RAGE within the vasculature.36,37 In particular, Mukherjee et al37 found that tumor necrosis factor-α induces the expression of RAGE via NADPH oxidase–mediated ROS production. In addition, Yamagishi and Takeuchi36 reported that a calcium antagonist with antioxidant properties prevented upregulation of RAGE through the inhibition of AGE-induced ROS generation. On the basis of the present study of diabetic ApoE−/−GPx1−/− aortas, we propose that increased RAGE, as a consequence of reduced GPx1 function, may contribute to the accelerated atherosclerosis detected in these animals, possibly via upregulation of VCAM-1.
In agreement with a previous study,5 the present study has once again shown the involvement of the profibrotic growth factor CTGF in DM-associated atherosclerosis. In addition, the present study now suggests that upregulation of CTGF is mediated in part via ROS, because lack of GPx1 further upregulated this cytokine in diabetic ApoE−/−mice, and the present study is supportive of Ruperez et al,12 who observed a significant ROS-mediated increase in CTGF expression within rat SMCs.
In light of recent data that lipid peroxides upregulate the superoxide-producing enzyme NADPH oxidase33 via the transcription factor NF-κB34, we investigated whether lack of GPx1 affects NADPH oxidase. Indeed, we demonstrate for the first time within the vasculature of diabetic mice that reduced GPx1 leads to the increased expression of gp91phox, one of the major subunits of NADPH oxidase, at both the mRNA and protein level. The present data therefore support the notion that NADPH oxidase is regulated by ROS.34 We also show increased expression of the p65 subunit of NF-κB. We speculate that lack of GPx1 within a diabetic context sets into motion a cyclical event that initially includes the formation of ROS, such as lipid peroxides. These ROS then upregulate NADPH oxidase, possibly via the ROS-sensitive transcription factor NF-κB, and in so doing increase superoxide formation. In turn, superoxide, the substrate of the SOD enzymes, is converted into hydrogen and lipid peroxide, thereby completing and enhancing the cycle. Indeed, the present data support this concept, because prooxidants (NADPH oxidase) and antioxidants (Sod1 and Sod2) upstream of GPx1 are affected by the lack of GPx1 in a diabetic milieu. The present study has shown that these antioxidant changes occur early after the initiation of DM and are still evident after prolonged exposure to the diabetic stimulus, as assessed by measurement of these enzymes at both the gene and protein level. In addition, superoxide can interact with nitric oxide, forming peroxynitrite, the potent cytotoxin known to attack various biomolecules in the vascular endothelium, vascular smooth muscle, and myocardium.38 Indeed, the present data showing enhanced staining of nitrotyrosine in GPx1-deficient diabetic aortas suggest that peroxynitrite may contribute to atherogenic mechanisms in this model, which is consistent with the notion that GPx1 functions as a peroxynitrite reductase.20
In summary, the present study has demonstrated that lack of GPx1 enhances atherosclerosis in a diabetic setting. This effect was associated with increased staining for SMCs and macrophages, consistent with increased SMC migration and macrophage infiltration. Furthermore, a range of molecules implicated in the progression and development of atherosclerosis, including adhesion molecules, growth factors, and RAGE, were increased by the absence of GPx1. These findings are consistent with recent clinical data implicating antioxidant defense, and GPx1 in particular, as playing a pivotal role in the progression of DM-associated atherosclerosis.
Sources of Funding
Dr de Haan is a recipient of a research grant from Merck Research Laboratories; Dr Allen is a holder of a joint Australian National Health and Medical Research Council (NHMRC)/Diabetes Australia Career Development Grant; and Dr Calkin is a holder of an Australian National Health and Medical Research Council postdoctoral fellowship. This work is supported by a Juvenile Diabetes Research Foundation grant, awarded to Dr Cooper.
At the time of the present study, Dr Kola was a Senior Vice President for Basic Research at Merck & Co, Inc. Merck & Co has substantial revenues from drugs that treat dyslipidemias; however, no activities fall within the specific area covered by this article. The remaining authors report no conflicts.
Basta G, Schmidt AM, De Caterina R. Advanced glycation end products and vascular inflammation: implications for accelerated atherosclerosis in diabetes. Cardiovasc Res. 2004; 63: 582–592.
Candido R, Jandeleit-Dahm KA, Cao Z, Nesteroff SP, Burns WC, Twigg SM, Dilley RJ, Cooper ME, Allen TJ. Prevention of accelerated atherosclerosis by angiotensin-converting enzyme inhibition in diabetic apolipoprotein E-deficient mice. Circulation. 2002; 106: 246–253.
Baynes JW. Role of oxidative stress in development of complications in diabetes. Diabetes. 1991; 40: 405–412.
Ruperez M, Lorenzo O, Blanco-Colio LM, Esteban V, Egido J, Ruiz-Ortega M. Connective tissue growth factor is a mediator of angiotensin II-induced fibrosis. Circulation. 2003; 108: 1499–1505.
Kislinger T, Tanji N, Wendt T, Qu W, Lu Y, Ferran LJ Jr, Taguchi A, Olson K, Bucciarelli L, Goova M, Hofmann MA, Cataldegirmen G, D’Agati V, Pischetsrieder M, Stern DM, Schmidt AM. Receptor for advanced glycation end products mediates inflammation and enhanced expression of tissue factor in vasculature of diabetic apolipoprotein E-null mice. Arterioscler Thromb Vasc Biol. 2001; 21: 905–910.
Davi G, Ciabattoni G, Consoli A, Mezzetti A, Falco A, Santarone S, Pennese E, Vitacolonna E, Bucciarelli T, Costantini F, Capani F, Patrono C. In vivo formation of 8-iso-prostaglandin f2alpha and platelet activation in diabetes mellitus: effects of improved metabolic control and vitamin E supplementation. Circulation. 1999; 99: 224–229.
De Haan JB, Crack PJ, Flentjar N, Iannello RC, Hertzog PJ, Kola I. An imbalance in antioxidant defense affects cellular function: the pathophysiological consequences of a reduction in antioxidant defense in the glutathione peroxidase-1 (Gpx1) knockout mouse. Redox Rep. 2003; 8: 69–79.
’t Hoen PA, Van der Lans CA, Van Eck M, Bijsterbosch MK, Van Berkel TJ, Twisk J. Aorta of ApoE-deficient mice responds to atherogenic stimuli by a prelesional increase and subsequent decrease in the expression of antioxidant enzymes. Circ Res. 2003; 93: 262–269.
Yang H, Roberts LJ, Shi MJ, Zhou LC, Ballard BR, Richardson A, Guo ZM. Retardation of atherosclerosis by overexpression of catalase or both Cu/Zn-superoxide dismutase and catalase in mice lacking apolipoprotein E. Circ Res. 2004; 95: 1075–1081.
Sies H, Sharov VS, Klotz LO, Briviba K. Glutathione peroxidase protects against peroxynitrite-mediated oxidations: a new function for selenoproteins as peroxynitrite reductase. J Biol Chem. 1997; 272: 27812–27817.
Hamanishi T, Furuta H, Kato H, Doi A, Tamai M, Shimomura H, Sakagashira S, Nishi M, Sasaki H, Sanke T, Nanjo K. Functional variants in the glutathione peroxidase-1 (GPx-1) gene are associated with increased intima-media thickness of carotid arteries and risk of macrovascular diseases in Japanese type 2 diabetic patients. Diabetes. 2004; 53: 2455–2460.
Lapenna D, de Gioia S, Ciofani G, Mezzetti A, Ucchino S, Calafiore AM, Napolitano AM, Di Ilio C, Cuccurullo F. Glutathione-related antioxidant defenses in human atherosclerotic plaques. Circulation. 1998; 97: 1930–1934.
Guo Z, Van Remmen H, Yang H, Chen X, Mele J, Vijg J, Epstein CJ, Ho YS, Richardson A. Changes in expression of antioxidant enzymes affect cell-mediated LDL oxidation and oxidized LDL-induced apoptosis in mouse aortic cells. Arterioscler Thromb Vasc Biol. 2001; 21: 1131–1138.
Forgione MA, Weiss N, Heydrick S, Cap A, Klings ES, Bierl C, Eberhardt RT, Farber HW, Loscalzo J. Cellular glutathione peroxidase deficiency and endothelial dysfunction. Am J Physiol Heart Circ Physiol. 2002; 282: H1255–H1261.
de Haan JB, Bladier C, Griffiths P, Kelner M, O’Shea RD, Cheung NS, Bronson RT, Silvestro MJ, Wild S, Zheng SS, Beart PM, Hertzog PJ, Kola I. Mice with a homozygous null mutation for the most abundant glutathione peroxidase, Gpx1, show increased susceptibility to the oxidative stress-inducing agents paraquat and hydrogen peroxide. J Biol Chem. 1998; 273: 22528–22536.
Paigen B, Ishida BY, Verstuyft J, Winters RB, Albee D. Atherosclerosis susceptibility differences among progenitors of recombinant inbred strains of mice. Arteriosclerosis. 1990; 10: 316–323.
de Haan JB, Witting PK, Stefanovic N, Pete J, Daskalakis M, Kola I, Stocker R, Smolich JJ. Lack of the antioxidant glutathione peroxidase-1 does not increase atherosclerosis in C57BL/J6 mice fed a high-fat diet. J Lipid Res. 2006; 47: 1157–1167.
Wassmann S, Wassmann K, Nickenig G. Modulation of oxidant and antioxidant enzyme expression and function in vascular cells. Hypertension. 2004; 44: 381–386.
Anrather J, Racchumi G, Iadecola C. NF-kappaB regulates phagocytic NADPH oxidase by inducing the expression of gp91phox. J Biol Chem. 2006; 281: 5657–5667.
Pueyo ME, Gonzalez W, Nicoletti A, Savoie F, Arnal JF, Michel JB. Angiotensin II stimulates endothelial vascular cell adhesion molecule-1 via nuclear factor-kappaB activation induced by intracellular oxidative stress. Arterioscler Thromb Vasc Biol. 2000; 20: 645–651.
In the present study, we show the importance of the abundant cytosolic and mitochondrial antioxidant enzyme glutathione peroxidase-1 (GPx1) in limiting diabetes-associated atherosclerosis. Lack of GPx1 is shown to accelerate lesion formation within the entire aortic tree and the susceptible aortic sinus, as well as to affect known proatherosclerotic and proinflammatory pathways. In particular, we demonstrate increased cellular involvement (macrophages and smooth muscle cells), increased expression of inflammatory (vascular cell adhesion molecule-1) and prosclerotic (vascular endothelial growth factor and connective tissue growth factor) cytokines, and upregulation of receptors for advanced glycation end products in diabetic apolipoprotein E−/−GPx1−/− double-knockout mice. This occurred despite upregulation of other antioxidants (GPx3, GPx4, Sod1, Sod2, and catalase), which highlights the importance of GPx1 in limiting oxidative events in the diabetic milieu. In addition, we show increased staining for a marker of free radical damage, nitrotyrosine, which emphasizes the importance of GPx1 as a peroxynitrite reductase. These findings are particularly important in light of recent clinical studies that have focused attention on GPx1 in atherosclerosis. For example, a patient cohort showed blood GPx1 activity to be the strongest predictor of cardiovascular disease, whereas reduced GPx1 activity was associated with increased cardiovascular disease in a diabetic population. These clinical studies, although clearly of significant diagnostic importance, offered no functional role for Gpx1 in atherogenesis. The present study has clearly shown that lack of GPx1 affects important proatherogenic pathways in a diabetic setting, which suggests a significant antiatherogenic role for GPx1 and makes this antioxidant an important therapeutic target worthy of further study.
The online-only Data Supplement, consisting of methods, tables, and a figure, is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.106.664250/DC1.