HIV-1, Reactive Oxygen Species and Vascular Complications

Abstract

Over 1 million people in the United States and 33 million individuals worldwide suffer from HIV/AIDS. Since its discovery, HIV/AIDS has been associated with an increased susceptibility to opportunistic infection due to immune dysfunction. Highly active antiretroviral therapies (HAART) restore immune function and, as a result, people infected with HIV-1 are living longer. This improved survival of HIV-1 patients has revealed a previously unrecognized risk of developing vascular complications, such as atherosclerosis and pulmonary hypertension. The mechanisms underlying these HIV-associated vascular disorders are poorly understood. However, HIV-induced elevations in reactive oxygen species, including superoxide and hydrogen peroxide, may contribute to vascular disease development and progression by altering cell function and redox-sensitive signaling pathways. In this review, we summarize the clinical and experimental evidence demonstrating HIV- and HIV antiretroviral therapy-induced alterations in reactive oxygen species (ROS) and how these effects likely contribute to vascular dysfunction and disease.

Introduction

Human immunodeficiency virus type 1 (HIV-1) infection and acquired immunodeficiency syndrome (AIDS) pose one of the greatest challenges to global public health. Since the development of highly active antiretroviral therapies (HAART), mortality and the incidence of opportunistic infections in people living with HIV-1 have declined substantially [1, 2]. As HIV/AIDS patients live longer, however, serious non-AIDS events occur and are associated with a greater risk of death than opportunistic AIDS-related events [3]. As such, vascular complications including coronary heart disease, pulmonary hypertension (PH), and atherosclerosis [4–7] are some of the most widely recognized [8, 9] uncharacteristic AIDS phenomena recorded in HIV-infected patients. In addition to the increase in susceptibility, clinical data also reveal that vascular complications in HIV-1 patients progress much more rapidly than in non-infected individuals [6, 10]. The exact mechanisms by which HIV-1 promotes the development and progression of these disorders remain unknown and are likely multi-factorial. Current research has identified increased vascular oxidative stress as a contributing underlying pathway. This review summarizes the clinical and experimental data correlating HIV-1 and vascular disease, the in vitro and in vivo evidence of HIV-mediated oxidative stress, and the sources of reactive oxygen species (ROS) and antioxidants impacted by HIV-1 and HIV-1 antiretroviral agents. This review also discusses how ROS promote the development and/or progression of vascular complications and why the strategic targeting of HIV-induced ROS may be of potential therapeutic value.

HIV-associated Vascular Disorders

Noninfectious complications occur much more frequently in AIDS patients than opportunistic infections and are now associated with a greater risk of death than AIDS-related events [3]. Atherosclerosis and pulmonary hypertension (PH) are two distinct and extensively-studied examples of HIV-associated vascular disorders. Although the exact underlying mechanisms remain unknown, the pathogenesis of these diseases is strongly associated with HIV-1 infection.

Pulmonary Hypertension

Pulmonary hypertension (PH) is a persistent elevation of pulmonary artery pressure and pulmonary vascular resistance. The symptoms of PH are nonspecific and include dyspnea, syncope, fatigue, chest pain, and nonproductive cough. Chronic PH increases the load on the right ventricle (RV) causing RV hypertrophy, right heart failure, the clinical syndrome of cor pulmonale, and ultimately, death [11]. Numerous studies report an increased frequency of PH in the HIV-infected population, with a prevalence of approximately 1 case per 200 (0.5%). More recent studies suggest, however, that this number is increasing and estimate that up to 1.0% of HIV-1 patients will develop PH [12]. This increase in incidence is likely as HIV-1 positive patients are not routinely examined for PH, and PH is often misdiagnosed resulting in an inaccurate assessment of incidence among HIV-1 patients [13]. Overall, the current data suggests that more than 10,000 HIV-1-infected individuals in the U.S. alone will develop PH. This incidence of PH in the HIV-1-infected population is extremely high compared to the 1 to 2 cases per million recorded in the general population [14]. The reason HIV-infected patients develop PH at this alarming rate remains unknown. It has been suggested that the increased lifespan of HIV-1 patients on antiretroviral therapy increases the likelihood of exposure to the “multiple hits” believed to be required to develop PH [15].

Recent studies investigating HIV-PH suggest that HAART fails to prevent the development of HIV-PH or improve the hemodynamic parameters in HIV-PH patients [16]. In addition, although HAART regulates viral replication and improves survival, patients with well-controlled HIV infection still develop PH. These observations underscore the severity of this disorder and also the need for further investigation of this disease.

Coronary Heart Disease and Atherosclerosis

HIV-1 positive patients also have a higher prevalence of atherosclerotic lesions [17–20], and elevated markers of subclinical atherosclerosis including increased carotid artery intima-media thickness [6, 21–28], increased arterial stiffness [29, 30] and endothelial dysfunction [31–34]. Clinical studies examining cardiovascular disease in HIV-1-positive people prior to the era of HAART are relatively few, yet there is evidence of serious cardiovascular anomalies in these patients. Seminal work by Joshi revealed coronary arteriopathy in 3 of 6 HIV-1-infected children at autopsy; it also described vasculitis and perivasculitis with infiltration of lymphocytes and mononuclear cells in vessel walls [35]. Other post-mortem analyses described major atherosclerotic lesions in proximal coronary arteries in 6 out of 8 HIV-infected patients who were 23–32 years of age [36]. The high frequency of abnormalities in these early studies is striking considering that cardiovascular pathologies are normally rare in these age groups. Vasculitis in small blood vessels [37, 38], aneurysms in medium or large arteries [39], and significantly lower levels of high density lipoprotein cholesterol (HDLc) in the bloodstream [40] of untreated HIV-1-positive individuals indicate that HIV-1 infection increases cardiovascular complications. These findings provide initial evidence of vascular dysfunction in HIV-1 patients, and support the premise that HIV-1 viral proteins have a role in the development of cardiovascular disease in this population.

Antiretroviral-naïve HIV-1-positive patients are found to have markers of endothelial activation including elevated plasma levels of von Willebrand factor, plasminogen activator inhibitor-1 antigen, and tissue-type plasminogen activator [41, 42]. These elevations in markers of endothelial dysfunction were found to correlate with anti-p24 antibodies and disease severity [41]. Antiretroviral-naïve HIV-1-positive people have also been found to have higher levels of soluble vascular cell adhesion molecule-1 (VCAM-1) [43], intracellular adhesion molecule-1 (ICAM-1) [44], and E-selectin [42, 44, 45] compared to healthy controls. This up-regulation of cell adhesion markers suggests that HIV-1 increases endothelial cell activation and dysregulation. These derangements may contribute to the increased incidence of pulmonary and systemic vascular disease.

There is also indirect clinical evidence showing that the presence of the HIV-1 virus potentiates cardiovascular risk. An ongoing retrospective analysis by the Kaiser Permanente Medical Care Program of Northern California has determined hospitalization rates for coronary heart disease and myocardial infarction in 4159 HIV-1-positive male members [46]. The authors did not find a correlation between antiretroviral therapies (ARTs) and hospitalization rates in the HIV-1 positive group after a 4 year follow-up. However, they demonstrated significantly higher hospitalization rates in the infected group when comparing them to age- and sex-matched HIV-1-negative controls during this same timeframe. They were unable to establish correlations between the increase in hospitalization rate with other known risk factors (i.e. smoking, hypertension, diabetes, and hyperlipidemia) in the HIV-1-positive group, thus postulating that HIV-1 infection itself increases the hospitalization rate for coronary heart disease and myocardial infarction. This conclusion has been supported by other clinical studies as well. A 2007 study examined acute myocardial infarction in patients (3851 HIV-1-positive and 1,044,589 HIV-1-negative) at 2 large Massachusetts hospitals [47]. The authors found a significantly increased risk for heart attack in the HIV-1-positive population at all ages examined, even when adjusted for the presence of other traditional risk factors in this group. Interestingly, the risk for myocardial infarction was roughly tripled in HIV-1-positive women compared to uninfected women, a group generally considered to be at lower risk of developing cardiovascular disease compared to men. However, the impact of HAART on this observation could not be adjusted for due to insufficient data. Another long-term multi-institution analysis, the Strategies for Management of Antiretroviral Therapy (SMART) Study Group, concluded that cessation of antiretroviral therapy (ART) in HIV-1-positive patients increases their short-term risk of developing cardiovascular disease [48]. Because prolonged ART has been associated with major metabolic and cardiovascular disorders, the authors of the study had hoped to evaluate the effectiveness of episodic ART in 2,720 HIV-1-positive patients using a treatment paradigm that administered HAART to maintain CD4+ lymphocyte levels. Unfortunately, interruption of antiretroviral therapy did not benefit this cohort and actually increased the incidence of major cardiovascular events.

It is suggested that HIV-1 infection elicits endothelial dysfunction in patients, as measured by flow-mediated dilation (FMD) of the brachial artery. A controlled case-study of 4 HIV-1-positive patients suggested that viral load inversely correlated with endothelium-dependent FMD without any relation to antiretroviral regimens [49]. Solages monitored FMD in 75 HIV-1-positive and 223 control subjects, and found significantly impaired endothelial function in the infected population. This study also found that viral load was a significant predictor of FMD [50]. The authors did not observe an association between endothelial dysfunction and the use of HAART, which could potentially be explained by the small sample size and unrepresentative demographic characteristics of this specific population. A smaller study in patients from 3.5–19.5 years of age also showed that HIV-1-infected children had significantly reduced FMD, increased wall stiffness, and lower cross-sectional compliance and distensibility of the carotid artery than non-infected children [51]. Interestingly, no differences in these parameters were observed when comparing HIV-1-positive children on HAART with those who were HAART-naïve, suggesting that the HIV-1 viral infection increased endothelial dysfunction. Other data also suggest that HIV, independent of HAART, can induce vascular dysfunction [32, 52]. However, a 2007 study published by Lorenz concluded that HIV-1 infection and HAART are both independent risk factors for the development of atherosclerosis in adults [53]. They found that intima media thickness of the carotid bifurcation, a predictor of subclinical atherosclerosis, was 24.8% higher in an HIV-1-positive/antiretroviral-naïve group compared to an uninfected control group. They also observed significantly greater IMT of the carotid bifurcation and the common carotid artery associated with HAART treatment in HIV-1-positive individuals. This effect of HIV-1 proteins was confirmed in a 2009 study which demonstrated that HIV-1 infection is independently associated with carotid intima media thickening, a measure of sub-clinical atherosclerosis [54].

HIV-1 Infection and Elevated ROS

Clinical Evidence

HIV-1 patients consistently demonstrate marked increases in ROS production as well as significant reductions in antioxidant availability and activity. In a study with treated and untreated HIV-infected subjects, the oxidative stress marker, d-ROM (derivatives of reactive oxygen metabolites) was shown to be greater in serum of HIV-1 patients than that of healthy controls [55]. Malondialdehyde (MDA), an index of lipid peroxidation, was also significantly elevated in the serum from both symptomatic and asymptomatic HIV-1 patients [56–58]. The marked increases in ROS biomarkers, d-ROM and MDA in HIV-1 patients demonstrate a HIV-induced dysregulation of ROS. These alterations are likely attributable to numerous mediators. However, several studies suggest that increased ROS production in HIV-1 patients results from diminished antioxidant expression and activity (Figure 1). For example, a dramatic attenuation in the total antioxidant capacity including vitamin A and C serum concentrations has been noted in HIV-1 seropositive patients [56, 59, 60]. Glutathione (GSH), the predominant antioxidant in the lung, is also significantly altered in HIV-1 patients. Studies demonstrate that total and reduced GSH in the epithelial lining fluid of symptom-free HIV-seropositive individuals was 60% less than those in normal subjects [61]. GSH levels are also reduced in the blood of HIV-1 patients [62–64]. Plasma of HIV-infected patients displays a 30% reduction in glutathione when compared to healthy controls. There was no difference in glutathione levels in the untreated HIV-infected group when compared to subjects on ART regimens [65]. Interestingly, however, HIV-1 produces a contradictory effect in the antioxidant, thioredoxin (Trx) which is significantly elevated in the plasma of HIV-infected healthy volunteers [66]. Studies show that Trx functions as an antioxidant in both the cytosol and mitochondria and contributes to cell growth, DNA repair and transcription factor regulation [67]. The HIV-induced increase in Trx may, therefore, function as a cellular compensatory mechanism or an attempt to normalize antioxidant capacity.

Figure 1. Effects of HIV-1 and Antiretroviral Therapies on ROS Sources and Scavengers.

1. NADPH Oxidases (Noxes), the primary producer of ROS in vascular cells, are dramatically up-regulated by HIV-1 [69, 92, 96] and ART [133]. 2. Uncoupled eNOS produces superoxide, instead of nitric oxide. Superoxide may then couple with NO to generate the highly reactive radical peroxynitrite (ONOO⁻) [173], which oxidizes tetrahydrobiopterin and causes lipid peroxidation [174]. HIV-1 [56, 68] and ART [158] stimulate elevations in lipid peroxidation markers such as MDA and nitrotyrosine. 3. HIV-1 [77, 89] and ART [131, 137–139] promote ROS release by inducing mitochondrial dysfunction. 4. HIV [71] and ART [153, 293] reduce SOD expression and activity. 5. HIV-1 negatively modulates catalase expression and activity. 6. GSH levels are significantly decreased in HIV-1 patients [65, 128]. In addition, HIV-1 proteins alter GSH release [86] and regulation [88, 116]. ART also deplete cellular GSH [131, 158]. 7. Mitochondrial antioxidant, SOD2 is decreased in HIV-1 [80, 89, 293].

These studies demonstrate an obvious association between HIV-1 and altered ROS production and antioxidant availability. These alterations are noted in numerous tissues as elevated ROS biomarkers and decreased antioxidant levels are observed in both serum and epithelial lining fluid from HIV-1 patients. The attenuation in cellular antioxidant systems provide a likely explanation for the considerable elevation of ROS documented in HIV-1 patients. However, while clinical studies are unable to determine the cellular sources of ROS, experimental HIV models have shed more light on specific mediators of HIV-induced ROS.

Experimental Models

In vitro models of HIV-1 infection mirror clinical studies demonstrating both increases in ROS and declines in antioxidant activity. HIV-1 infection of human primary macrophages produces a 6-fold increase in malondialdehyde (MDA) [68]. This finding implicates HIV-1 infection as the principal cause for elevated macrophage ROS levels. However, it remains controversial whether HIV-1 infection or HIV-1-induced mediators contribute to the increased ROS production and antioxidant depletion seen in infected patients. Data from several recent studies strongly suggest that HIV-induced mediators, independent of HIV-1 infection, are sufficient to alter cellular ROS generation. In an in vitro model of HIV-induced oxidative stress, podocytes expressing the NL4-3 HIV-1 construct with a deleted gag/pol region exhibit a marked increase in ROS generation over a 3-hour interval of HIV exposure. The NL4-3-induced increase in podocyte ROS generation was attenuated by diphenyleneiodonium (DPI) administration, implicating flavin-containing enzymes such as NADPH oxidase as the primary source for ROS increases [69]. This study suggests that virus infectivity due to gag and pol function is not necessary for HIV-induced ROS release.

In vivo models studying the effect of HIV-1 proteins on oxidative stress reach similar conclusions. Mice expressing the HIV-1 Tat protein (HIV-1 Tat+) exhibit a significant reduction in total intracellular GSH content in both the liver and erythrocytes. Additionally, glutathione synthetase activity in HIV-1 Tat+ mouse liver was decreased to 73% of control levels [70]. Similarly, studies from our group employing an HIV-1 transgenic rat model indicate that the expression of HIV-1 proteins is sufficient to augment ROS production and alter antioxidant expression. These animals express a HIV-1 provirus that encodes for the viral genes env, tat, nef, rev, vif, vpu, and vpr. However, due to the deletion of the gag and pol regions, the HIV-1 transgene is both nonreplicative and noninfectious. Studies using this model show that HIV-1 transgenic (Tg) rat aortas display significant increases in superoxide and 3-nitrotyrosine levels compared to wild-type controls. HIV-1 Tg rats also exhibit marked decreases in circulating nitric oxide (NO) and total GSH as well as reductions in aortic SOD1 expression and activity [71]. HIV-1 transgene expression also induces marked elevations in rat lung superoxide, hydrogen peroxide (H₂O₂), and NO metabolite levels as well as concomitant decreases in lung lavage fluid GSH when compared to wild-type rats [72, 73].

Altogether, these studies highlight the ability of HIV-1 proteins to independently alter ROS release and antioxidant activity. Indeed, details regarding whether HIV-1 protein concentrations in these model are physiologically relevant are needed. Additionally, further investigation to ascertain if these proteins potentiate the effect of HIV-1 infection is warranted. Nonetheless, the contribution of HIV-1 proteins to altered ROS production and regulation is clear.

Increased ROS – Contribution of HIV-1 Proteins

Considerable research indicates that HIV-1 significantly alters the cellular oxidant/antioxidant balance. However, research utilizing HIV- and Tat-Tg animal models argues that virus-induced mediators such as HIV-1 proteins may serve as sufficient inducers of oxidative stress. As a result, the investigation of HIV-1 proteins and their effects on ROS release and regulation has increased substantially. This research sheds light on the imbalance between oxidants and antioxidants, and strongly suggests that HIV-1 proteins contribute to both increased production of ROS and diminished antioxidant activity.

HIV-1 proteins are encoded by 9 genes located within the virion capsid. Three of these genes, gag, pol, and env, are found in all retroviruses and are vital to the structure of HIV-1. For example, the gag and pol regions encode for the reverse transcriptase and integrase enzymes necessary for efficient HIV-infection and replication. The env gene encodes for gp160, the precursor for the envelope proteins gp120 and gp41, which are necessary for virus entry into cells. The 6 remaining “accessory” genes are unique to HIV-1. Two of these, tat and rev, perform a regulatory function and are essential for viral replication [74]. However, the roles of HIV-1 genes, vpr, vpu, vif, and nef are less fully understood [75]. The following section will focus on the role of HIV-1 proteins that have been shown to affect ROS levels – Tat, Nef, gp120, and Vpr.

Tat

Of all the HIV-1 proteins, the early viral protein, Tat is the most widely studied. Composed of 86–101 amino acids, Tat serves as a transcriptional transactivator of viral gene expression by binding to a transactivation-responsive region in the HIV long terminal repeat (LTR). The expression of Tat is critical for productive HIV infection, as Tat-deficient viruses are non-infectious. In HIV-1 patients, Tat can be secreted from infected T cells and monocytes [76] and following its release, circulates in the bloodstream. From the circulation, Tat enters uninfected cells [76–79] and alters cellular physiology by positively or negatively affecting gene expression. In 1995, Westendorp et al reported plasma Tat levels between 1–3 ng/mL in HIV-infected patients [80]. More recently, however, Tat serum levels in HIV-1-infected patients were estimated to fall between 2 and 40 ng/mL [81]. It is also suggested that Tat concentrations are higher around HIV-infected perivascular cells and in the proximity of endothelial cells [82]. This effect is thought to occur because macrophages and monocytes act as viral reservoirs and secrete Tat as well as cytokines and oxidants near endothelial cells.

Extensive research demonstrates that HIV-1 Tat increases ROS levels and decreases antioxidant levels. For example, Tat causes a dose-dependent increase of ROS in cultured brain microvascular cells [83] and significantly induces ROS production and lipid peroxidation in rat brain endothelial cells [84]. In the HIV indicator (HeLa-CD4-LTR-B-gal), or MAGI cells, transfection with a Tat-expressing plasmid for 48 hours significantly increases ROS levels and reduces intracellular GSH levels by 50%. This study also showed that the Tat-induced alterations are reversed by pretreatment with the antioxidant, N-acetylcysteine (NAC) [85]. Murine fibroblasts expressing the full-length HIV-1 Tat protein exhibit a similar 50% decrease in cellular GSH concentrations [86]. Additionally, in vivo studies demonstrate that the intravenous injection of Tat protein decreases mouse brain GSH levels by 85% [87]. Tat is also shown to affect other cellular antioxidant enzymes as Tat over-expression in HeLa cells results in a 3-fold reduction in the glutathione peroxidase (GPx) mRNA ratio as well as a 2.5 fold decrease in GPx activity [88]. Moreover, HeLa cells stably producing the Tat protein express 48% less SOD2 compared to control cells [80, 89], which may be caused by Tat-induced disruption of Sp1 and Sp3 binding in the SOD2 basal promoter [90]

These studies demonstrate that Tat alters cellular ROS and antioxidant regulation. Yet, the exact mechanism and source of Tat-induced oxidative stress remain unclear. Recent studies, however, have demonstrated Tat-induced activation of several ROS-producing enzymes. For example, Gu et al showed that Tat acutely increases intracellular oxidant levels in ECV-304 cells. This Tat-induced oxidant activity is decreased by pretreatment with two NADPH oxidase inhibitors, DPI and apocynin [91]. Co-culture of human umbilical vein endothelial cells with HeLa-Tat cells also significantly induces endothelial H₂O₂ production via Nox 4 activation [92]. These studies implicate NADPH oxidases as potential mediators of Tat-induced ROS. However, other oxidases may contribute to HIV-induced ROS release as DPI and apocynin are somewhat nonspecific inhibitors.

These studies provide remarkable evidence of the independent effects of Tat on ROS levels and antioxidant availability. Recent studies examining novel anti-AIDS therapies have attempted to target Tat activity and binding. However, the available data regarding the effectiveness of these agents is controversial. Although further investigation is needed to better understand the mechanism underlying its effects, Tat clearly alters cellular function and may likely contributes to the vascular dysfunction and disease associated with HIV-1 infection.

Nef

Several studies implicate the HIV-1 protein Nef as a potential mediator in HIV-induced oxidative stress and vascular disease. Nef, “the negative factor” is an HIV viral accessory protein with a molecular weight ranging between 27–34 kDa. Although normally found within the cytoplasm, Nef associates with the cellular membrane upon activation via myristoylation. Nef expression has been shown to down-regulate the cell-surface levels of both CD4 and MHC-1 molecules. It also interferes with numerous intracellular pathways, leading to the dysregulation of cellular signaling and activation [93]. In vitro studies indicate that Nef influences HIV-1 pathogenesis through its ability to increase viral replication and infectivity in primary lymphocytes and macrophages. In addition, in vivo studies show that Nef is essential for high virus replication and disease progression to AIDS in HIV-infected individuals [94].

Research investigating ROS elevations in response to Nef is limited. In 2002, it was shown that Nef protein expression does not independently induce microglial NADPH oxidase [95]. However, Nef significantly enhanced superoxide release by NADPH oxidase following challenge with the calcium ionophore, formyl peptide or lipopolysaccharide (LPS) [95]. Other studies reveal that Nef regulates superoxide production in a biphasic manner. Research by Olivetta et al demonstrates that human monoblastic cells (U937) stably transfected with a vector expressing a Nef-ER fusion protein produce greater ROS than controls at 1- and 4- hours post-transfection. However, Nef-expressing cells exhibit a complete ablation in ROS production at later time points. In more recent studies, exposure of neutrophils from healthy donors to Nef for one hour increased superoxide production. DPI administration significantly reduced the Nef-induced superoxide production, implicating activation of a flavin-containing enzyme. Also, studies performed with neutrophil cellular lysates demonstrate that Nef associates with p22-phox, but not any other NADPH oxidase subunits [96]. Similarly, in ex vivo studies, exposure of porcine pulmonary arteries or pulmonary artery endothelial cells (HPAEC) to Nef markedly increases superoxide release by 54% and 70%, respectively. In addition to these effects on ROS, Nef also concomitantly decreased eNOS expression and NO production in porcine arterial rings and HPAEC [97].

Collectively, these reports implicate Nef as a mediator of HIV-induced ROS and vascular dysfunction. Although in vitro data suggest a potential cell-type dependent effect, ex vivo studies underscore the potential physiological relevance of Nef in HIV-related vascular disease. Moreover, studies performed by the Flores group demonstrate that HIV-1 Nef contributes to HIV-associated PH by promoting vascular remodeling [98–100]. Altogether, HIV-1 Nef may contribute to the vascular dysfunction documented in the HIV-1 population via increased ROS production and effects on the nitric oxide synthase pathway.

gp120

The HIV-1 protein, gp120 also promotes ROS production. The envelope glycoprotein gp120 is expressed on the surface of HIV-1 virions and facilitates the receptor binding and subsequent membrane fusion required for HIV-1 infection [101]. In addition, soluble gp120, estimated to exist between 12–92 ng/mL in the serum of HIV-1 patients [102], can be shed from virus particles or infected cells into the circulation. As a result, gp120 can cause extensive cellular damage by stimulating oxidative stress pathways [103, 104], inflammatory cytokine release [105], apoptosis [106, 107], and tight junction injury [108, 109].

Several studies indicate that gp120 stimulate ROS release in numerous cell types. For example, exposure to gp120 for 24 hours causes an almost 6-fold elevation in MDA levels in astroglial cell homogenates. This effect was significantly antagonized by pretreatment with the antioxidant, NAC [110]. gp120 exposure also induces marked increases in human retinal epithelial cell MDA and NO production, as well as inducible nitric oxide synthase expression over a 72-hour interval when compared to untreated controls [111]. Additionally, gp120 induces marked staining for HNE, an indicator of lipid peroxidation, in cells expressing the endothelial cell marker, CD31. These increases in ROS production were associated with elevations in MMP-9, and gene delivery of the antioxidant enzymes GPx and SOD1 returned MMP-9 to control levels [112]. Research also demonstrates that low concentrations of gp120 promote ROS release. Recombinant gp120 at a concentration of 340 nM increases ROS production in human monocyte-derived macrophages [113]. Also, gp120 concentration of 40 nM was also shown to increase intracellular H₂O₂ in lymphoid cells [114]. Picomolar concentrations of gp120 induce ROS release in U937 cells, whereas co-administration of catalase and SOD decreased the gp120-induced oxidative damage by 81% [115]. Also, in vivo studies also established that injecting 500 ng of gp120 significantly increases MDA levels.

In addition to the increases in ROS, gp120 has also been shown to alter antioxidant regulation. Seventy two hours of gp120 exposure significantly decreases the mRNA expression of the Nrf2 transcription factor in the L2 epithelial cell line [116]. The decrease in Nrf2 mRNA expression, however, did not produce a significant reduction in Nrf2 protein expression. Moreover, HIV-1 transgene expression in rat alveolar epithelial cells produced a 30% attenuation in Nrf2 mRNA expression [116]. Overall, these data suggest a role of gp120 in ROS release. In addition, the effects of gp120 on Nrf2 expression may act as a possible mechanism underlying HIV-induced antioxidant depletion.

Vpr

HIV-1 Viral Protein R, or Vpr is a 14 kDa HIV-1 accessory protein that is highly conserved in HIV-1 and simian immunodeficiency virus [117]. In addition to the HIV-1 proteins Tat and gp120, Vpr is linked to several pathways associated with HIV-1 infection and replication as well as cellular function. Studies show that Vpr transactivates long terminal repeat (LTR), promotes cell cycle arrest, induces DNA damage and apoptosis, and regulates nuclear factor-kappaB (NF-kB) activity [118]. Moreover, when exogenously administered, Vpr induces viral reproduction in latently infected cells [119]. Research also indicates that Vpr is present in the plasma of infected patients at nanomolar concentrations [120, 121].

Recent studies conducted by Deshmane showed that infection with a Vpr-expressing adenovirus induces a 4-fold elevation in H₂O₂ generation and an approximate 15-fold increase in ROS production in the mitochondrial compartments of infected microglial cells. These effects are consistent with results showing that infection with the JR-FL strain of HIV-1 generates very similar increases in microglia H₂O₂ release following 24-hours of infection [122]. Studies performed using human monocyte-derived macrophages also demonstrate that recombinant Vpr administration induces mitochondrial dysfunction, release of the pro-inflammatory cytokine IL-6, and oxidized phosphatidylcholine generation. These effects were significantly inhibited by the administration of the antioxidant, NAC [123].

Collectively, these data confirm that HIV-1 proteins significantly contribute to HIV-induced ROS generation and reduce antioxidant activity. While it is unclear whether the concentrations of HIV-1 proteins tested in vitro are physiologically relevant or appropriately effective to induce these changes, it is clear that specific HIV-1 proteins significantly impact ROS levels. Research indicates that HIV-infected monocytes and macrophages serve as reservoirs for HIV. Whether these reservoirs have the ability to continuously secrete HIV-1 proteins into the bloodstream remains controversial and more studies are needed to determine the effect of HIV-1 HAART on circulating HIV-1 proteins.

Increased ROS - Contribution of Antiretroviral Therapies

Numerous studies implicate antiretrovirals (ARVs) as a major contributor to the increased ROS levels documented in HIV-infected patients. ARVs are divided into 5 major classes: the nucleoside reverse transcriptase inhibitors (NRTIs), non-nucleoside reverse transcriptase inhibitors (NNRTIs), protease inhibitors (PIs), integrase inhibitors, and entry inhibitors [124]. In order to better understand the effects of ARVs on ROS, numerous investigators have used both in vitro and in vivo models to examine individual and combined HIV-1 therapies. Similar to clinical studies, the effects of ARVs on ROS were dependent on the cell studied, drugs examined, and duration of treatment. The following section summarizes the effects of ARVs such as NRTI and/or PI on ROS generation and how these agents may induce these alterations.

Nucleoside Reverse Transcriptase Inhibitors (NRTIs)

NRTIs were the first drugs approved for the treatment of HIV-1 (in 1987) and constitute the majority of prescribed ARVs [124]. NRTIs block HIV-1 replication by inhibiting the action of reverse transcriptase, the enzyme necessary for production of the viral double-stranded DNA, which is subsequently integrated into the genetic material of infected cells. The antiviral activity of the nucleoside analogues requires phosphorylation to their respective tri-phosphorylated moieties by a variety of nucleoside- and nucleotide-specific host cell kinases. NRTIs are well-documented to have adverse clinical toxicities which include lactic acidosis and hyperlactatemia, peripheral neuropathy, skeletal myopathy, cardiomyopathy, pancreatitis, and lipodystrophy [125]. Toxic effects of NRTIs are commonly ascribed to competitive inhibition with endogenous nucleotides for cellular mitochondrial DNA polymerase-γ (pol-γ), the sole enzyme responsible for mitochondrial DNA (mtDNA) replication and repair [128]. However, mounting evidence indicates that the effects of NRTIs are not exclusively due to actions on pol-γ but may rely on ARV-induced ROS release and antioxidant dysregulation. This evidence is discussed below.

In a study assessing the effect of NRTIs on oxidative stress, HIV-1 patients receiving ARV therapies demonstrated significantly greater plasma F₂ isoprostane concentrations when compared to untreated HIV-infected subjects [126]. In a more recent study, 6 months of ARV treatment with stavudine (d4T), lamivudine (3TC), and the NNRTI, nevirapine (NEV) in HIV seropositive patients reduced serum GSH and increased MDA levels compared to non-treated HIV-infected subjects [127]. HIV-1 HAART-treated subjects also show significant elevations in serum GPx as well as a reduction in glutathione reductase (GR) following 12 months of HAART treatment compared to baseline levels [128]. Furthermore, although glutathione disulfide (GSSG) concentrations were similar among baseline, 6-, and 12-month time points, GSH levels in the epithelial lining fluid of asymptomatic HIV-1 patients are dramatically decreased following 6 months of ARV, though CD4+ lymphocyte counts are only moderately altered [129].

In addition, long-term exposure to NRTIs promotes endothelial ROS production. Bovine aortic endothelial cells (BAEC) exposed to 1 μM azidothymidine (AZT) for 14 days produce 30% greater amounts of superoxide than untreated controls [130]. In 2009, our group showed that 3- and 5-weeks of AZT exposure concomitantly decrease human aortic endothelial cell (HAEC) GSH levels and increase both total intracellular and mitochondrial superoxide production. Interestingly, d4T had no significant effect on HAEC levels of GSH or superoxide, emphasizing that different NRTIs may have divergent effects on ROS levels [131].

Animal studies investigating the effects of NRTIs on oxidative stress make similar conclusions. Mice administered 100mg/kg of AZT for 5 days display 50–60% reductions in aortic aconitase activity, a marker for ROS production due to its high susceptibility to inactivation by ROS [132]. Aortas from mice treated with AZT for 35 days show marked increases in superoxide staining when compared to vehicle-treated controls [130]. AZT administration for 8 months markedly reduces plasma and heart homogenate antioxidant levels, increases gp91phox expression, and stimulates membrane-associated p47phox levels in Wistar-Kyoto rats. These AZT-induced effects were attenuated by treatment with vitamin C, further supporting a ROS contribution [133].

Numerous studies demonstrate mitochondrial-specific ROS generation in response to NRTIs. Since mitochondria serve as the primary intracellular site of oxygen reduction, they have the greatest potential for ROS formation and are highly susceptible to ROS toxicity. Extensive research demonstrates that exposure to the NRTI AZT induces significant mitochondrial dysfunction and ROS production in a variety of cell types including endothelial cells, cardiomyocytes, and lymphoid cells [134, 135]. AZT concentrations up to 10 μmol mg⁻¹ induce an approximate 3-fold increase in H₂O₂ production in mitochondria from rat liver. However, AZT co-administration with mildronate, a drug shown to protect mitochondrial membranes from damage by free fatty acids and prevent NF-κB activation in cardiac tissues, reduces H₂O₂ production from rat liver mitochondria by 55% [136]. Treatment with AZT has also been shown to significantly elevate mitochondrial ROS generation in human primary cardiomyocytes. These studies show that increases in AZT-induced mitochondrial ROS stimulate caspase-3 and -7 activation as well as PARP-1-mediated cell death pathways [137]. Studies in cultured fibroblasts also demonstrate that the NRTIs, d4T, and AZT induce mitochondrial dysfunction and elevate ROS generation. AZT and d4T increase fibroblast mitochondrial mass by 3- to 4-fold when assessed by Mitotracker Red fluorescence. AZT and d4T administration also significantly increases fibroblast ROS production when compared to untreated cells. Exposure to the NRTIS abacavir, lamivudine, didanosine, or tenofovir produced no change in fibroblast mitochondrial mass or ROS production [138]. Furthermore, 40 nM concentrations of the NRTI, 2′, 3′-dideoxycytidine (ddC) induces significant increases in isolated mitochondria protein oxidation levels after 6 hours, but similar concentrations of ddC fail to affect mitochondrial lipid peroxidation levels [139]. In an in vivo study of AZT-induced oxidative stress, Kohler demonstrated that daily AZT administration induces significant increases in heart mitochondrial H₂O₂ and reduced cardiac aconitase activity in SOD2 knockout mice, however, transgenic mice over-expressing SOD2 or mitochondrially-targeted catalase demonstrate no alterations in heart mitochondrial H₂O₂ release following AZT administration [140]. This suggests that mitochondria-specific ROS generation and antioxidant depletion both play an important role in AZT-induced oxidative stress.

At least In the case of AZT, data suggest that the pol-γ effect is not the sole mechanism for ROS generation. Data indicate that AZT and AZT-MP directly interact with Complex I of the mitochondrial electron transport chain and prevent its cAMP-dependent phosphorylation independently of pol-γ inhibition [141]. In addition, researchers have found that AZT and its azido-containing derivatives (AZT-MP, AZT-TP, and glucoronidated-AZT) have direct pro-oxidant activities in an in vitro cell-free chemical system, while non-azido-containing derivatives do not [142]. Furthermore, AZT and AZT-MP induce lipid peroxidation in membrane preparations devoid of cellular constituents and presumably most secondary targets [143]. These findings suggest that a secondary target likely contributes to the pro-oxidative properties of NRTIs.

Non-nucleoside Reverse Transcriptase Inhibitors (NNRTIs)

NNRTIs (Efavirenz, Nevirapine and Delavirdine) are non-competitive inhibitors of reverse transcriptase, and, in contrast to NRTIs, do not require phosphorylation to achieve their anti-viral effects. The first drug of this class was approved by the FDA in 1996. Data indicate that NNRTIs likely cause any significant effect on ROS generation alone. However, since the efficacy of NNRTI drugs is impaired by rapid emergence of drug-resistance mutations, NNRTIs are often prescribed with agents of other ARV classes. The high-level resistance to NNRTIs arises due to the development of point mutations within the allosteric binding site of the reverse transcriptase [144]. Since NNRTIs share similar binding sites, mutations commonly cause cross-class resistance [145]. Although second generation NNRTIs, such as etravirine and rilpivirine, hold great promise for their resistance to mutations and possible use in ARV naïve patients, further investigation is needed to examine their effects on ROS and antioxidant levels.

Protease Inhibitors (PI)

PIs such as, saquinavir, ritonavir, indinavir, lopinavir, nelfinavir, amprenavir, and darunavir were first approved for HIV-1 therapy in 1995. These agents prevent HIV-1 replication by inhibiting the cleavage of viral polypeptides into mature and functional proteins [124]. In vitro studies examining PIs demonstrate significant increases in oxidative stress in endothelial cells [146, 147], adipocytes [148, 149], and macrophages [150]. For example, PIs, either alone or as part of a HAART regimen, increase the generation of ROS in human aortic endothelial cells [151]. The PIs ritonavir and/or lopinavir markedly increase ROS production in human coronary artery endothelial cells after 30 days of exposure. However, NAC or MnTBAP administration during the final 24 hours of the 30-day exposure dramatically reduced PI-induced ROS [152]. Studies also report very similar effects in other in vitro models. The PI nelfinavir, but not saquinavir or atazanavir, increases ROS production in the rat pancreatic insulinoma cell line, INS-1, after 24 hours of exposure. Nelfinavir also decreases GSH content, as well as SOD1 expression and activity [153]. Additionally, macrophage-derived foam cells exposed to RTV produce significantly greater amounts of superoxide than untreated controls [154].

In ex vivo studies, endothelial cells from coronary vessels also demonstrate decreased eNOS mRNA levels following RTV, APV, and saquinavir (SQV) exposure. PIs are shown to produce similar effects in porcine coronary arteries as exposure to ritonavir (RTV) and amprenavir (APV) increases superoxide levels in vessel endothelial layers by 47% and 52%, respectively, when compared to untreated controls. Vessel treatment with RTV and ATV also reduces nitrite release by approximately one-third when compared to controls. These alterations likely contribute to the altered endothelial-dependent relaxation in vessel rings treated with 15 μM concentrations of RTV, ATV, and SQV [155]. Interestingly, in later studies, treatment with curcumin [156] and the ginsenosides, Rb1, Rc, and Re [157] reversed the ritonavir-induced effects on superoxide generation, eNOS expression, nitrite release, and endothelial-dependent relaxation.

In addition to the independent effects of NRTI and PI on ROS levels, several studies have examined the combined effect of NRTIs and PIs on cellular ROS regulation. Exposure to the HAART drug combination of azidothymidine (AZT), a NRTI, and indinavir (IDV), a PI, for 48-hours dose dependently increases ROS production in human brain microvascular endothelial cells (hBMEC). Moreover, 72-hours of AZT and IDV exposure dose-dependently reduce endothelial intracellular GSH levels and increase the lipid peroxidation metabolite, MDA [158].

Other ARVs

Integrase inhibitors (i.e. raltegravir), fusion inhibitors, and entry inhibitors (i.e. enfuvirtide, maraviroc) are relatively new to HIV-1 therapy and are typically given to HAART-experienced patients demonstrating therapeutic failure. The effects of these newer agents on ROS are currently unknown. Future studies may provide information regarding the effect of these agents on ROS release and/or antioxidant activity.

Overall, these data demonstrate how ARVs increase ROS and ROS-mediated effects. More importantly, however, the data demonstrates how the effects of HIV-1 and ARV on ROS production may have combined effects on the development and progression of HIV-induced vascular disorders.

Reactive Oxygen Species – Sources and Regulation

Research demonstrates that HIV-1 and HIV-1 ART severely alters ROS sources and regulation pathways, creating an environment of oxidative stress. These data underscore the possible role of oxidative stress in HIV-1 pathogenesis and the development of HIV-associated vascular pathologies. ROS, such as superoxide, hydrogen peroxide (H₂O₂), and hydroxyl radical (HO·) as well as the reactive nitrogen species nitric oxide (NO) and peroxynitrite (ONOO⁻) are biologically active species known to play important roles in vascular biology via redox signaling pathways [159–161]. These oxidants are produced by numerous sources such as the NADPH oxidases (Noxes), xanthine oxidase, cytochrome P450, uncoupled endothelial nitric oxide synthase (eNOS) and as byproducts of the mitochondrial respiratory chain [162]. To balance ROS levels and combat their toxic effects, cells employ several antioxidant enzymes including the superoxide dismutases (SOD), catalase, glutathione peroxidase (GPx), thioredoxins (Trx), and peroxiredoxins (Prx). Non-enzymatic antioxidant mechanisms also exist, including the vitamins E and C as well as glutathione, which acts as a reducing substrate for glutathione peroxidase [163]. These antioxidant systems are localized throughout the cell and function either in an independent or complementary manner to scavenge ROS (Figure 1). Overall, the balance between ROS generation and antioxidants is essential for normal cell function. However, research demonstrates marked alterations in ROS and antioxidant levels in HIV-1 patients. These data suggest a potential correlation between HIV-1 and oxidative stress, and raise the question of whether the oxidant/antioxidant imbalance found in HIV-1 patients contributes to the characteristic pathologies associated with this population.

Increased ROS in Vascular Disorders

Although cells utilize a variety of mechanisms to regulate ROS generation and inactivation, ROS are essential for normal vascular function and act as key second messengers in numerous signal transduction pathways [164–166]. Moreover, ROS levels regulate the activity of important transcription factors implicated in vascular function including NF-κB, activator protein-1 (AP-1), and HIF-1α [167–169]. Therefore, the excess generation of and/or the reduced ability to remove ROS can lead to detrimental effects such as dysregulated apoptotic or proliferative states, vascular smooth muscle cell migration and endothelial dysfunction. For example, low concentrations of H₂O₂ induce cellular proliferation, whereas high concentrations promote apoptosis and cell cycle arrest [170]. ROS also contribute to TNF-α-induced activation of endothelial apoptosis [171] and stimulate vascular smooth muscle cell migration by modulating the matrix metalloproteinases (MMP) -2 and -9 [172]. Furthermore, superoxide can cause endothelial dysfunction by combining with nitric oxide to produce the highly reactive radical, peroxynitrite (ONOO^-) and decrease NO levels [173]. Peroxynitrite is then able to oxidize the essential eNOS cofactor, tetrahydrobiopterin (BH4), stimulating eNOS uncoupling and further contributing to endothelial dysfunction by increasing superoxide and reducing NO availability [174]. ROS-induced endothelial dysfunction is pivotal to vascular injury and the inflammatory response and endothelial dysfunction is known to be an early predictor of cardiovascular events in patients without [175] and with known vascular disease [176, 177]. The extensive effects of ROS on the vessel wall support a role for ROS in the development of numerous vascular disorders including PH and atherosclerosis. For example, increased superoxide production has been observed in experimental models of PH [178] and biomarkers of oxidative stress are elevated in PH patients [179]. PH patients exhibit low NO levels in their exhaled breath [180, 181], which suggests that scavenging by superoxide radicals may mediate the reductions in NO bioactivity [182]. Moreover, superoxide regulates characteristic PH pathologies such as modulating pulmonary vasoconstriction [183] and stimulating pulmonary smooth muscle cell proliferation [184]. Superoxide and other oxygen radicals also promote atherosclerosis by altering NO and activating redox-sensitive pathways that mediate vessel remodeling and plaque stability [185]. In addition, coronary arteries from CAD patients express greater levels of the NADPH oxidase subunits, p22phox, p67phox, and p47phox [186] and produce significantly larger amounts of superoxide [187].

In vivo studies demonstrate that apoE (-/-) mice deficient in NADPH oxidases have attenuated atherosclerosis due to decreased superoxide release via deficient NADPH oxidases [188]. Conversely, an increased production of ROS in vascular endothelial cells is implicated in the oxidation of LDL and the expression of ROS-sensitive inflammatory genes such as VCAM-1 [189]. For example, the overexpression of extracellular SOD suppresses endothelial cell-mediated LDL oxidation [190] and vitamin E treatment reduces endothelial VCAM-1 and ICAM-1 expression induced by oxidized LDL [191]. Furthermore, ROS increase monocyte chemotactic protein-1 [192, 193] and mediate endothelial monocyte adhesion and infiltration [194].

Oxidative stress also enhances vessel inflammation by stimulating the release of pro-inflammatory cytokines such as endothelin-1 (ET-1) and interleukin-6 (IL-6). Patients with atherosclerosis demonstrate elevations in plasma levels [195] and immunoreactive staining for ET-1 in the vasculature [196]. These increases in ET-1 and IL-6 [74, 197–199] are implicated in PH development. ROS mediate the secretion of ET-1 from endothelial cells [200, 201] and polymorphonuclear leukocytes [202]. Also, the Tat-induced release of IL-6 from macrophages is attenuated by NADPH oxidase inhibition [203]. Antioxidant administration, however, inhibits the IL-6 and ET-1 release in healthy volunteers [204], as well as, ET-1-induced release of IL-6 from vascular smooth muscle cells [205]. Altogether, ROS and redox-sensitive pathways greatly contribute to vessel injury and vascular disease pathogenesis which further support a role for ROS in HIV-1 pathogenesis and HIV-associated vascular disease development.

ROS, HIV-1 Infection and Replication

Elevated ROS can further potentiate the effects of HIV-1 infection because oxidative stress plays a crucial role in viral replication, inflammatory response, immune cell proliferation, loss of immune function, and increased sensitivity to drug toxicity [206–208]. While it remains largely unknown how HIV-1 promotes oxidative stress, research demonstrates that HIV-1 increases in the activity of ROS sources, and reductions in cellular antioxidant capacity lead to enhanced viral replication. For example, H₂O₂ stimulates activation of the HIV-1 LTR in patients with latent HIV-1 [209]. In vitro studies also show that lowering intracellular GSH levels decreases cell survival [210], increases HIV replication, and increases NF-κB activation [211, 212]. HIV replication is associated with an increase in MDA levels and nitrotyrosine formation, which are both recognized as indicators of lipid peroxidation from free radical overproduction. Therefore, the impact of HIV-induced ROS on replication can result in a further insult to the vascular wall. In addition, treatments that increase oxidative stress are also known to activate HIV-1 replication [213], which appears to be mediated by ROS-induced activation of NF-κB [214].

Conversely, antioxidants suppress HIV replication in chronically infected cells [215–217]. For example, antioxidant molecules such as GSH, glutathione ester, and NAC are able to suppress HIV expression in infected monocytic cells [218, 219]. Exogenous GSH administration also suppresses the release of virus particles and HIV-1 proteins from chronically infected cells [220]. Similarly, MnTBAP, a synthetic peroxynitrite decomposition catalyst, demonstrates antiviral activity in both acutely- and chronically-infected monocytes/macrophages [68]. The NAC pro-drug, I-152, and beta-mercaptoethylamine (MEA) increase intracellular GSH in human MDMs and significantly decrease viral replication in HIV-infected MDMs [221]. Moreover, infection of human primary monocytes is blocked or substantially reduced by GSH or NAC treatment. GSH and NAC administration reduce the amount of virus released by infected cells by 90% [222] as well as viral replication and disease progression in murine AIDS [223, 224]. In vivo studies demonstrate that the combination of GSH-loaded erythrocytes with AZT and DDI treatments significantly decreases proviral DNA content following 10 weeks of infection in bone marrow macrophages and brain samples [225].

Cellular redox state may also negatively affect HIV-1 gene expression and transcription by altering HIV-1 Tat internalization and function. Previous studies demonstrate that the oxidation state of the cysteine-rich region of Tat strongly influences its capacity to enter cells [226]. In addition, more recent studies indicate that the redox state of Tat alters protein uptake by macrophages and promotes protein aggregation thereby hindering hinders Tat biological activity [227]. Research also indicates that cellular redox state regulates Tat transactivation [228]. Normally, Tat stimulates transcriptional elongation from the viral LTR through a specific interaction with a 59-residue stem-loop RNA known as the transactivation-response element (TAR). Recent studies demonstrate that 1 hour of NAC exposure reduces Tat-induced HIV-1 LTR transactivation in MAGI cells by over 60% [85]. Selenium administration also inhibits Tat-dependent LTR activity in human MDM and in U937 cells [229].

Contributing and Alternative Mechanisms

While oxidative stress can directly impact vascular function, the effects of HIV-1 on the vasculature are likely multi-factorial and may include HIV-mediated chronic immune activation and inflammation. Chronic immune activation is a hallmark of HIV-1 and is associated with the depletion of CD4+ lymphocytes, elevations in lipopolysaccharide levels, proinflammatory cytokine release, deregulation of hematopoiesis pathways, CD8+ T cell activation and apoptosis, as well as accelerated immunosenescence [230]. Although these events provide a functional role in HIV-1 pathogenesis, they are also linked to the development of vascular complications. For example, HIV-1 infection increases the release of the toll-like receptor ligand, LPS. Whereas the injection of LPS alone is insufficient to cause atherosclerosis [231], it is able to exacerbate atherosclerosis-associated pathologies [232, 233]. Also, mice deficient in toll-like receptor 4 are resistant to chronic hypoxia-induced PH development [234]. Furthermore, elevations in plasma LPS levels stimulate the release of numerous pro-inflammatory cytokines such TNF-α, IL-6, and IL-1β from peripheral macrophages and dendritic cells. Abnormal levels of these cytokines have detrimental effects throughout the vasculature and are implicated in the development and progression of atherosclerosis [235–238] and PH [198, 199, 239–241]. Chronic immune activation in HIV-1 patients is also associated with a deregulation of hematopoiesis, leading to reduced numbers of progenitor cells and a decline in the generation of new cells [242–244]. This phenomenon may mediate PH development, as the direct injection of endothelial progenitor cells exhibits therapeutic effects in experimental models of PH [245, 246]. In addition, abnormalities in CD8+ T-cell function may also contribute to the increased susceptibility of HIV-1 patients to develop PH [247, 248] and atherosclerosis [249].

HIV-induced ROS may also alter the innate immune response of infected patients. Increased ROS release is shown to mediate both the effector and induction phases of the immune response. It is well established that immune cells generate ROS causing deleterious effects to nearby tissues and cells during immune activation. However, ROS may also regulate the immune response by altering key signaling pathways in antigen-presenting cells (APC). Research demonstrates that GSH depletion triggers a Th2 response pattern [250, 251]. These studies indicate that the redox state of APC not only regulate T-cell activation but may also suppress the T helper (Th1) immune response and induce the Th2 response. This Th2 response pattern is associated with significant increases in IL-4 and IL-10 production and the stimulation of various antibody responses which can impact vascular cell responses [252].

Additionally, the physiological declines seen in HIV-1 patients show incredible similarities to those noted in the uninfected elderly population, such as dementia, unintentional weight loss, and fatigue [253, 254]. These alterations suggest that HIV-1 infection may induce an accelerated aging phenotype in seropositive patients, potentially contributing to the hallmark immune deficiency of HIV-1 patients and increased occurrence of vascular disorders. Furthermore, other factors may mediate the oxidant/antioxidant alterations noted in HIV-1 patients. HIV-1 patients often suffer from co-morbidities such as alcohol and drug abuse, malnutrition, and nicotine addiction. In addition, ART induces hyperlipidemia, hyperglycemia, insulin resistance, and central obesity, which may also increase ROS levels [255]. Although these factors may prove difficult to account for in clinical and experimental studies, they may modulate the dysregulation of ROS production and antioxidant availability documented in HIV-1 patients.

Potential Therapies

Nutritional Antioxidants

HIV-1 reduces levels of plasma antioxidants [256], such as ascorbate, or vitamin C [257] and these decreases in antioxidant concentrations persist in HIV-1 patients although their dietary intake is sufficient for healthy individuals [258]. Also, low intakes of vitamin C [259] and low plasma concentrations of vitamin E [260] are associated with a greater risk of progression to AIDS in HIV-infected US subjects. These studies emphasize the importance of dietary antioxidant vitamins in HIV-1 seropositive individuals and suggest a potential therapeutic benefit for HIV-1 patients. However, research examining the effect of dietary supplementation in HIV-1 patients provides conflicting results. Multi-nutrient supplements containing vitamins C and E led to a lower risk of death due to HIV infection in Tanzanian women [261] and a small increase in CD4+ T lymphocyte counts in Kenyan women [262]. Vitamin C and E supplements were found to reduce oxidative damage and attenuate disease severity in HIV-positive Canadian adults [263]. Also, α-tocopherol, or vitamin E, (800 mg/day) administration decreased viral load in HIV-1 patients over a 60-day period [264]. In addition, research demonstrates that antioxidant molecules such as GSH, glutathione ester, and NAC are able to suppress HIV expression in infected monocytic cells [218, 219] as well as viral replication and disease progression in murine AIDS [223, 224]. These data argue that GSH administration may provide therapeutic benefit against HIV-1 replication in addition to its cardiovascular benefits.

Conversely, daily selenium administration has shown no significant effect on CD4+ cell counts or viral load in pregnant, HIV-infected women [265]. However, the results of a 9-month selenium supplementation study performed in 450 HIV-1 seropositive men and women may provide an explanation for the recent conflicting results and highlight the importance of treatment adherence. Study subjects with a selenium change less than or equal to 26.1 microgram/L, indicating poor subject compliance, were found to have an increase in HIV-1 viral load and a decrease in CD4+ lymphocyte counts after the 9-month treatment period. Conversely, subjects with an increase in serum selenium levels greater than 26.1 microgram/L demonstrate no change in viral load and increases in CD4+ cell counts [266]. The administration of alpha-lipoic acid, a glutathione-replenishing disulfide, three times daily increased total glutathione levels but failed to alter HIV RNA levels or improve CD4+ lymphocyte counts after 6 months [267]. These results of these intervention trials may not be completely attributable to the antioxidant actions of the supplements, but the data suggest that proper multi-nutrient and antioxidant supplementation may diminish the severity of HIV disease. Still, the varying outcomes of these studies underscore the need for further research in this area.

Nrf2 Activation

NF-E2 related factor 2 (Nrf2) is a ubiquitously expressed transcription factor that regulates antioxidant enzyme expression by binding to the antioxidant response element (ARE). Due to its function, the Nrf2 pathway is thought to play an essential role in cellular protection against ROS effects and oxidative stress [268]. Recent studies indicate that some vascular protective compounds act via the Nrf2 pathway. For example, resveratrol dose-dependently increases Nrf2 promoter activity and stimulates expression of Nrf2-regulated genes such as heme-oxygenase-1 in cultured primary human coronary arterial endothelial cells. Resveratrol also reduces mitochondrial and cellular ROS release following high glucose and TNF-α exposure in an Nrf2-dependent manner [269]. Additionally, Nrf2 is linked to the anti-atherogenic effects of the Ginkgo biloba extract (GBE). Studies indicate that GBE increases Nrf2 promoter activity and nuclear translocation in human aortic endothelial cells. Also, Nrf2 knockdown abolishes GBE-induced suppression of TNF-α-induced VCAM-1 expression in human aortic endothelial cells [270].

In addition, research indicates that Nrf2 activation protects vascular cells against ROS release and inflammation. For example, Nrf2 activation by sulforaphane reduces VCAM-1 signaling in human umbilical vein endothelial cells [271] and adenoviral Nrf2 overexpression prevents injury by ROS and inhibits monocyte adhesion in endothelial cells [272]. Nrf2 gene transfer via adenoviral transduction inhibits proliferation in human and rabbit smooth muscle cells. AdNrf2 also reduces inflammation and oxidized LDL accumulation following aortic balloon injury in rabbits [273]. Dh404 stimulates Nrf2 in primary cardiac myocytes and increases Nrf2 nuclear translocation and transcriptional activity in H9C2 cardiomyocytes. Although Nrf2-deficient mice display significant increases in liver MDA, Nrf2 knockdown in ApoE-deficient mice produced a decrease in aortic stiffness and a 61% reduction in plaque area after 20 weeks when compared to ApoE KO controls. Moreover, macrophages from ApoE/Nrf2 deficient mice demonstrate a reduced uptake of AcLDL, a commonly used indicator for OxLDL [274] and the Nrf2 activator, Dh404 inhibits basal and Angiotensin II-induced superoxide and peroxynitrite formation [275]. Also, treatment with the Nrf2 activator, CDDO-Imidazolide (CDDO-Im) prevents cigarette smoke-induced increases in RV pressures and alterations in RV diastolic and systolic functions and increases GSH levels and attenuates alveolar apoptosis and pulmonary oxidative damage [276]. Interestingly, numerous clinical trials are currently investigating the effect of Nrf2 activation and sulforaphane treatment in numerous diseases including cystic fibrosis, COPD, asthma, cancer, autism, and cardiovascular disease. As such, Nrf2 activation may serve as a therapeutic for HIV-associated vascular disorders.

Targeting Specific Antioxidant Pathways

Although extensive research implicates oxidative mechanisms in the pathogenesis of vascular diseases, studies using antioxidants as a major disease therapy produce controversial results regarding disease protection and reversal [277–280]. These outcomes, as well as recent advances in ROS detection, have led researchers to redefine the concept of oxidative stress and direct more attention to the balance of redox signaling, particularly the major thiol/disulfide couples such as glutathione (GSH)/glutathione disulfide (GSSG), reduced thioredoxin [Trx-(SH)₂]/oxidized thioredoxin (Trx-SS), and cysteine (Cys)/cystine (CySS) [281]. These redox couples regulate numerous biological functions within the cell and evidence suggests that specific redox states may work together to perform distinct functions in various cellular locations [282].

Interestingly, the cellular localization of these redox couples and the resulting biological events may have considerable consequences on vascular function and hence, the development and progression of vascular diseases. For example, the expression and DNA-binding of redox-sensitive transcription factors, including NF-κB and AP-1, may be severely altered in response to modulations in nuclear redox states. Moreover, mitochondrial redox states may regulate mitochondrial permeability transition and thereby, trigger cellular necrosis or apoptosis. Differences in redox states, particularly those across the plasma membrane, are implicated in cell proliferation as well as monocyte adhesion to endothelial cells [283]. Variations in extracellular Cys/CySS redox states and those found in human plasma also enhance oxidant-induced apoptosis and mediate decreases in cell number [284]. Altogether, these data encourage researchers to obtain a better understanding of redox signaling to allow for more effective antioxidant treatments.

Peroxisome Proliferator-Activated Receptors (PPAR) Agonists

Agents that restore NO levels and reduce ROS generation would likely have a favorable impact on HIV-mediated vascular disease. Peroxisome proliferator-activated receptors (PPARs) are ligand-activated transcription factors belonging to the nuclear hormone receptor superfamily [285]. PPARs regulate a variety of physiological processes ranging from lipogenesis to inflammation, and have been implicated in numerous disorders including diabetes and atherosclerosis. There are 3 PPAR isotypes: PPARα, PPAR β/δ, and PPARγ. These PPARs are expressed in multiple body tissues including the heart and vasculature. Of particular interest to vascular disorders is PPARγ, which regulates genes involved in characteristic vessel pathologies, such as cellular differentiation and growth, inflammation, ROS regulation, apoptosis, and angiogenesis.

Research shows that PPARγ agonists and PPARγ overexpression increase endothelial NO release [286]. PPARγ agonists also decrease the expression of Nox-2 and -4 and significantly stimulate Cu/Zn SOD expression and activity in human umbilical vein endothelial cells [287]. Reduced endothelial PPARγ expression attenuates NO production and produces significant increases in serum d-ROMs, derivatives of reactive oxygen metabolites [288]. Overexpression of PPARα and PPARγ reduces HIV-induced dysregulation of tight junction proteins in brain endothelial cells, effects mediated by alterations in matrix metalloproteinase [289]. Moreover, PPARγ activation via rosiglitazone administration in brain microvascular endothelial cells inhibits adhesion and transendothelial migration of HIV-1 infected monocytes [290]. The PPARγ agonist, rosiglitazone, also attenuates LPS-induced inflammation in vascular smooth muscle cells [291]. Additionally, PPARs decrease endothelial-leukocyte interactions in atherosclerosis models [292] and PPAR agonists demonstrate antiviral activity. PPAR agonists, rosiglitazone, PgJ2, ciglitazone, troglitazone and fenofibrate, inhibit HIV replication in HIV-1 infected peripheral blood mononuclear cells (PBMCs). These data indicate that decreases in PPAR activity may mediate endothelial dysfunction, ROS regulation, vascular injury, and HIV-1 replication. Although several have been unsuccessful due to patient safety concerns, PPAR agonists serve as an encouraging therapeutic for preventing HIV-associated vascular disorders.

Conclusion

In this review, we summarize how HIV- and ART-induced increases in ROS can impact vascular cell apoptosis, proliferation, migration and cytokine release signaling pathways. These data delineate how HIV-1 induced ROS may contribute to the cardiovascular events, such as atherosclerosis and PH that are observed in people living with HIV-1/AIDS. Due to the increased ROS generation and antioxidant dysregulation in HIV-1 patients, the idea of antioxidant supplementation offers promise. However, conflicting outcomes from these studies emphasize a need for an improved understanding of ROS and redox-sensitive pathways. Recent efforts have provided a fresh perspective of oxidative stress and a new focus on redox-responsive signaling and ROS compartmentalization [282].

Although data indicate that ROS play a prominent role in HIV-vascular disorders, the pathogenesis of HIV-associated vascular disorders is likely multi-factorial. An increase in ROS may serve as a “priming” mechanism preparing vascular cells for a “second hit”. HIV-1 patients often also suffer from other co-morbidities, such as cancers, as well as smoking and drug and alcohol abuse. These conditions combined with elevated ROS may play a role in the increased succeptibility of HIV-1 patients to develop vascular. It is obvious that HIV-1 infection models, either in vitro or in vivo, cannot replicate the events that occur in seropositive individuals. Having said this, further research is needed for a better understanding of these disorders which may then lead to improved therapies and preventative measures for HIV-1 mediated cardiovascular disease.

Highlights.

• Vascular diseases are one of the most recognized non-AIDS events in HIV-1 patients

• HIV-1 proteins and antiretrovirals impact antioxidants and superoxide production

• Increased reactive oxygen species (ROS) likely promote vascular disease

• ROS also potentiate HIV-1 effects by altering viral replication and immune function

• Targeting ROS sources or restoring antioxidants may reduce HIV vascular disease

Abbreviations

AIDS

Acquired Immunodeficiency Syndrome

AP-1

Activator Protein-1

APC

Antigen-presenting Cells

APV

Amprenavir

ARV

Antiretrovirals

AZT

Azidothymidine

BH4

Tetrahydrobiopterin

Cys

Cysteine

Cyss

Cystine

d4T

Stavudine

ddC

2′, 3′-dideoxycytidine

d-ROM

Derivatives of Reactive Oxygen Metabolites

DPI

Diphenyleneiodonium

eNOS

endothelial Nitric Oxide Synthase

ET-1

Endothelin-1

FMD

Flow-mediated Dilation

GPx

Glutathione Peroxidase

GR

Glutathione Reductase

GSH

Glutathione

GSSG

Glutathione Disulfide

H₂O₂

Hydrogen Peroxide

HAART

Highly Active Antiretroviral Therapies

HAEC

Human Aortic Endothelial Cells

HPAEC

Human Pulmonary Artery Endothelial Cells

HIV

Human Immunodeficiency Virus type 1

ICAM

Intracellular Adhesion Molecule

IL-1β

Interleukin-1beta

IL-6

Interleukin-6

IDV

Indinavir

iNOS

Inducible Nitric Oxide Synthase

LPS

Lipopolysaccharide

MDA

Malondialdehyde

MDM

Monocyte-derived Macrophages

MMP

Matrix Metalloproteinase

NAC

N-acetylcysteine

NF-kB

Nuclear factor-kappaB

NO

Nitric Oxide

Nox

NADPH Oxidase

NRTI

Nucleoside Reverse Transcriptase Inhibitors

NNRTI

Non-Nucleoside Reverse Transcriptase Inhibitors

ONOO⁻

Peroxynitrite

PBMCs

Peripheral blood mononuclear cells

PH

Pulmonary Hypertension

PI

Protease Inhibitors

PPAR

Peroxisome Proliferator-activated Receptors

ROS

Reactive Oxygen Species

RTV

Ritonavir

RV

Right Ventricle

SOD

Superoxide Dismutase

SQV

Saquinavir

Tg

Transgenic

TNF-α

Tumor Necrosis Factor-alpha

Trx

Thioredoxin

VCAM-1

Vascular Cell Adhesion Molecule-1