Atheroprotective action of a modified organoselenium compound: in vitro evidence

Oxidation of low-density lipoprotein (LDL) has been strongly suggested to play a significant role in the pathogenesis of atherosclerosis. Thus, reducing LDL oxidation is a potential approach to decrease the risk of the atherosclerosis. Organoselenium compounds have demonstrated promising atheroprotective properties in experimental models. Herein, we tested the in vitro atheroprotective capability of a modified organoselenium compound, Compound HBD, in protecting isolated LDL from oxidation as well as foam cells formation. Moreover, the glutathione peroxidase (GPx)-like activity of Compound HBD was analyzed in order to explore the mechanisms related to the above-mentioned protective effects. The Compound HBD in a concentration-dependent manner reduced the Cu-induced formation of conjugated dienes. The protein portion from LDL were also protected from Cu-induced oxidation. Furthermore, the Compound HBD efficiently decreased the foam cell formation in J774 macrophage cells exposed to oxidized LDL. We found that the atheroprotective effects of this compound can be, at least in part, related to its GPxlike activity. Our findings demonstrated an impressive effect of Compound HBD against LDL-induced toxicity, a further in vivo study to investigate in more detail the antioxidant and antiatherogenic effects of this compound could be considered.

LDL, the main blood cholesterol carrier, becomes atherogenic after undergoing oxidative modifications (Brown et al. 1981, Itabe et al. 2011, Ganini and Mason 2014. The hypothesis that oxidative modification of LDL contributes to the progression of atherosclerosis is supported by an impressive body of in vitro findings and by persuasive results in animal models of atherosclerosis (Bird et al. 1998, Berrougui et al. 2006, Pirillo et al. 2013. All major vascular cell types are capable of oxidizing LDL, and several lines of evidence have demonstrated the in vivo occurrence of oxidized lipoproteins in atherosclerotic lesions (Steinberg and Witztum 2010, Yoshida andKisugi 2010).
Oxidized LDL is intensively taken up by macrophages through scavenger receptors that subsequently promote foam cell formation which compose fatty streaks (hallmark of early atherogenesis) followed by the development of fibrous and atheromatous plaques (Steinberg 1997, 2002, Miller et al. 2010. Furthermore, oxidized LDL has been shown to enhance atherogenesis by other mechanisms, such as cytotoxicity towards endothelial cells and macrophages and stimulation of thrombotic and inflammatory events (Witztum 1993).
One possible method to prevent atherosclerotic diseases would be the administration of antioxidant substances thereby making LDL less sensitive to this oxidative process. In fact, it has been evidenced that the antioxidant capability of LDL can be increased by dietary antioxidant supplementation, i.e. LDL can incorporate endogenous and exogenous antioxidants in its supramolecular structure, decreasing its susceptibility to be oxidized. In fact, many endogenous and exogenous compounds have been reported to display beneficial effects against LDL oxidation (Noguchi et al. 2000, Chu and Liu 2005, Barcelos et al. 2011. However, the strategies with antioxidants supplementation had generated both "positive" and "no response" effects in decreasing atherogenesis (Rahman et al. 2014).
Reports have shown that selenium-containing organic molecules are generally more potent antioxidants than ''classical'' antioxidants, and this fact serves as an impetus for an increased interest in the rational design of synthetic organoselenium compounds (Mugesh et al. 2001, Nogueira et al. 2004, Nogueira and Rocha 2011. In the last years our laboratory have been studying the in vitro and in vivo antioxidant and anti-inflammatory properties of a simple diorganoselenium compound, diphenyl diselenide (PhSe) 2 (a prototype of this class of compounds) in models of atherosclerosis (de Bem et al. 2008, Hort et al. 2011, Straliotto et al. 2013b, Mancini et al. 2014. Notably, this compound inhibited LDL oxidation induced by copper ions (Cu 2+ ) (de Bem et al. 2008) as well as potently reducing the formation of atherosclerotic lesions in hypercholesterolemic low density lipoprotein receptor knockout (LDLr -/-) mice (Hort et al. 2011). Importantly, chemical modifications in this organoselenium compound could confer higher pharmacological efficiency and less toxicological effects (Nogueira and Rocha 2011). In this regard, we recently demonstrated the powerful effect of the disubstituted diaryl diselenides, p-methoxyldiphenyl diselenide and p-chloro-diphenyl diselenide against LDL-induced toxicity (Straliotto et al. 2013a).
Considering that, the purpose of the present study was to investigate the potential beneficial effects of a modified organoselenium compound, Compound HBD (Fig. 1a), i.e. 2-((1-(2-(2-(2-(1-(2-hydroxybenzylideneamino) ethyl) phenyl) diselanyl) phenyl) ethylimino) methyl) phenol, in protecting in vitro isolated LDL from oxidation, as well as foam cells formation, which are the main elements involved in the early steps of atherogenesis. Moreover, the GPx-like activity of Compound HBD was also evaluated in an attempt to delve into molecular mechanisms related to the aforementioned protective effects.

COMPOUND HBD
Compound HBD, 2-((1-(2-(2-(2-(1-(2-hydroxybenzylideneamino) ethyl) phenyl) diselanyl) phenyl) ethylimino) methyl) phenol (Fig. 1a), was synthesized according to literature methods (Braga et al. 2005 with little modifications. Analysis of the 1 H NMR and 13 C NMR spectra showed that the compound obtained (with 99.9% purity) presented analytical and spectroscopic data in full agreement with their assigned structure. LDL ISOLATION AND OxIDATION LDL from human plasma was isolated by discontinuous density-gradient ultracentrifugation, as previously indicated by de Bem et al. (2008). The protein concentration in LDL preparation was measured using the method of Lowry et al. (1951). To prepare oxidized LDL (oxLDL), isolated LDL  . After 10 min, CuSO 4 5 µM was added to the reaction medium and the reaction was monitored for 6 h for evaluating CD production. The oxidation was continuously monitored by measuring the increase in absorbance at 234 nm due to CD formation as previously described (Esterbauer et al. 1989). In addition, the value of the lag phase was determined as the intercept of the tangent of the slope of the absorbance curve in propagation phase with the time axis, and was expressed in min. The oxidation rate (V max ) was obtained from the slope of the absorbance curve during the propagation phase (Gieseg and Esterbauer 1994).

MEASUREMENT OF LDL-TRP FLUORESCENCE
The time course of tryptophan (Trp) fluorescence emission intensity is used to monitor Cu 2+ induced apoB-100 LDL oxidation. The fluorescence spectra of native LDL display a single band centered at approximately 332 nm, which is assigned to the Trp residues in apoB-100 (Giessauf et al. 1995).
Loss of Trp fluorescence is a marker for oxidations at the protein core of LDL (Reyftmann et al. 1990, Giessauf et al. 1995. The kinetics of LDL oxidation was followed by measuring the decrease of Trp-fluorescence, corresponding to the decomposition of this amino acid, after the addition of 3.3 µM CuSO 4 , in absence or presence of different Compound HBD (0 to 15 µM) concentrations. Trp fluorescence was measured at different time points (0 to 360 min) (excitation at 282 nm and emission at 331 nm). Then the parameter ''half-time'' (t max/2 ) was used to characterize the fluorescence changes in quantitative terms for practical purposes. It is defined as the time needed to observe a reduction in fluorescence of 50% of the difference between initial and residual fluorescence intensity (Jerlich et al. 2000).

CELL CULTURE AND FOAM CELL FORMATION ASSAY
Murine macrophage cell line J774A was purchased from the American Type Culture Collection (ATCC, Rockville, MD, USA). These cells were maintained with Dulbecco's Modified Eagle Medium (DMEM) supplemented with 2 mM glutamine, 10 mM HEPES, 100 U/mL penicillin, 100 mg/mL streptomycin and 10% fetal bovine serum (FBS) in a 5% CO 2 humidified atmosphere at 37 °C. For the foam cells formation, J774 cells were plated in 12-well plate at equal density (2 x 10 5 cells per well) in DMEM medium supplemented with 10% FBS for 24 h. After that, the medium was replaced by DMEM medium without FBS and the cells were incubated with Compound HBD 1µM or vehicle in a 5% CO 2 humidified atmosphere at 37 °C. After 24 h, oxLDL (100 µg/ mL) was added to the medium for additional 3 h. Foam cell formation assay was performed with the Oil-Red O staining method (Koopman et al. 2001). Briefly, after oxLDL incubation, cells were fixed with 4% paraformaldehyde, washed in phosphate buffer, then stained by 0.3% Oil-Red O for 10 min.
Hematoxylin was used as counterstaining. Images of cells were acquired using a light microscopy (Olympus, Bx-41). Six images were captured from each group, and the total pixels intensity was determined using NIH ImageJ 1.36b imaging software (National Institutes of Health, Bethesda, MD), and lipid content was expressed as arbitrary units (a.u.).

GLUTATHIONE PEROxIDASE (GPx) LIKE ACTIVITY
The GPx-like activity of Compound HBD was measured according to a method previously described by Wilson et al. (1989). Compound HBD at different concentrations (1 to 30 µM) was incubated at 37 °C in a medium containing 50 mM potassium phosphate buffer, pH 7.0, 1 mM ethylene diamine tetraacetic acid (EDTA), 1 mM reduced glutathione (GSH), 1 mM azide, 0.2 U of GR and 0.25 mM NADPH. The reaction was initiated by addition of 0.2 mM of H 2 O 2 . The activity was followed by the decrease of NADPH absorption at 340 nm.

STATISTICAL ANALYSIS
Values are presented as mean ± SEM. The statistical analyses were performed by using oneway analysis of variance (ANOVA) followed by the post hoc Duncan's multiple range test. Linear regression analysis was also used to test concentration-dependent effects. A value of p < 0.05 was considered to be significant. All tests were performed using the Statistica software package (StatSoft Inc., Tulsa, OK, USA).

COMPOUND HBD EFFECTS ON CU 2+ -INDUCED LDL OxIDATION
LDL was subjected to oxidation with 5 µM Cu 2+ . The Cu 2+ -ions exerted peroxidative modification of LDL polyunsaturated fatty acids and led to a molecular rearrangement, thus forming CD. The kinetic profile of the LDL oxidation was characterized by an initial lag time followed by a propagation period, where the rate of CD formation was maximal, and then by a decomposition phase. Notably, Compound HBD inhibited Cu 2+induced generation of CD in a concentration-dependent manner (Fig. 1b). Compound HBD in a concentration-dependent manner prolongs the lag period (β = 0.932, and p < 0.001), and decreases the oxidation rate -V max (β = -0.962, and p < 0.05), evidenced by changes in the propagation phase slope. It is noteworthy that Compound HBD presented maximal efficacy at 20 µM.

COMPOUND HBD EFFECTS ON LDL-PROTEIN OxIDATION FLUORESCENCE KINETICS
Probably, oxidative modification of the lipid part of the LDL particle is followed by a modification of apoB-100 (Esterbauer et al. 1987). Fig. 2 shows that protein fraction of LDL are also oxidized as function of time in the presence of Cu 2+ (3.3 µM), resulting in a decrease in the kinetic of tryptophan fluorescence. When LDL was incubated with Compound HBD, the LDL-protein oxidation was prevented in a concentration-dependent manner (β = 0.856, and p < 0.05), evidenced by 50% inhibition of fluorescence tryptophan (T max/2 ), reaching more than 50% of inhibition in the higher concentrations (10 and 15 µM).

EFFECTS OF COMPOUND HBD IN THE OxLDL MEDIATED FOAM CELL FORMATION
The oxLDL uptake by macrophages and consequent foam cell formation was induced in macrophages exposed to oxLDL. The pretreatment with Compound HBD (1 µM) significantly reduced the oxLDL uptake by macrophages, a critical step involved in the atherogenic process ( Fig. 3a and b; p < 0.01). It is important mentioning that we preliminarily performed a MTT assay to analyze the cell viability after Compound HBD incubation. We exposed the macrophages cells to different concentrations of Compound HBD during 24 hours to attest that 1 µM is a safe concentration (data not show).
GPx-LIKE ACTIVITY The next step was to investigate the mechanisms by which Compound HBD was able to prevent JADE DE OLIVEIRA et al.
LDL oxidation. The GPx is critically involved in cell protection against oxidative stress by reducing of numerous reactive species, at the expenses of GSH. It is well documented the GPx-like activity of organoselenium compounds, such as ebselen and (PhSe) 2 (Wilson et al. 1989). Therefore, using a GR-coupled assay, we investigated the GPx-like activity of Compound HBD. Notably, the Compound HBD displayed a concentrationdependent GPx-like activity in an in vitro system ( Fig. 4; β = 0.98, and p < 0.001).

discussiOn
Atherosclerosis is a multifactorial complication leading to cardiovascular diseases and stroke.
Elevated levels of total cholesterol and LDLcholesterol stand among the major risk factors of atherosclerosis. Specifically, the oxidized form of LDL is supposed to set the ground of this pathophysiology (Brown and Goldstein 1984, Ross 1999, Rahman et al. 2014). The antioxidants that can inhibit these oxidative processes might be useful in preventing atherosclerosis-related pathological conditions (Steinberg and Witztum 2010). Epidemiological studies show that there is a positive correlation between the consumption of antioxidants and a decrease of the incidence of coronary artery disease (Pandey and Rizvi 2009). In this regard, it has been suggested that organoselenium compounds have various pharmacological activities, such as antioxidant, anti-inflammatory, and cardioprotective activities (Nogueira and Rocha 2011). For instance, our previous studies demonstrated that (PhSe) 2 , the most simple organoselenium compound, presented atheroprotective properties in different in vivo and in vitro models (de Bem et al. 2008, Hort et al. 2011, Straliotto et al. 2013b. In this work, Compound HBD was chosen as a potential beneficial molecule against oxidation of human isolated LDL based on their chemical similarity with (PhSe) 2 . We hypothesized that the chemical modifications present in the Compound  It has been suggested that among all in vitro antioxidant capacity assays, measuring the LDL antioxidant activity is the most physiopathologically important and the most informative approach for screening antioxidant activity in compounds in order to prevent atherosclerosis (Katsube et al. 2004). LDL has multiple reduction sites with different affinities for Cu 2+ . A common method for measuring the inhibition of LDL oxidation in vitro is by determining the formation of CD in response to peroxidative modification of the polyunsaturated fatty acids (PUFAs) present in the LDL molecule mediated by Cu 2+ (Ziouzenkova et al. 1998). This approach was based on previous research findings that LDL oxidation at cell-free system by redox-active metal ions (copper and iron) is physiological and biochemically similar to that of cellular systems (Gunathilake and Rupasinghe 2014). Epidemiologic studies indicate the higher level of copper and iron ions in the arterial walls of the atherosclerotic individuals and these redoxactive metal ions have been implicated in playing a vital role in oxidizing the native LDL molecule both in vivo and in vitro (Rahman et al. 2014). In the present study, the Compound HBD prevented Cu 2+ -induced isolated LDL peroxidation in a concentration-dependent manner by decreasing the formation of CD, thus extending the lag phase and lowering the oxidation rate (V max ). It is noticeable that, the Compound HBD presented higher potential in inhibiting LDL oxidation when compared to (PhSe) 2 using the same in vitro model (de Bem et al. 2008). This evidence suggest that other organoselenium compounds, besides (PhSe) 2 and ebselen (the prototypes of these class) deserve to be evaluated as antiatherogenic agents.
Considering that during oxidation process the apoB-100 present in LDL is also modified, another significant result of our study was the ability of Compound HBD to prevent Cu 2+ -induced loss of Trp fluorescence in human isolated LDL. In this regard, it has been reported that the fluorescence spectrum of native LDL displays a single band centered at approximately 332 nm, which is assigned to the Trp residues in apoB-100, and the loss of Trp fluorescence is a marker for oxidations at the protein core of LDL (Reyftmann et al. 1990). The protective effect of Compound HBD against Cu 2+ -induced loss of Trp fluorescence indicates that, besides its beneficial effect against the oxidation of lipid moieties of LDL, this organoselenium compound also prevents the oxidation of protein part of human LDL, pointing to an additional mechanism that could contribute to inhibition of the atherogenic process.
It is worth mentioning that other wellestablished antioxidants also protect LDL from the oxidation induced by different insults in in vitro as well as in vivo assays. For instance, several studies using vitamin C e E demonstrated remarkable cardiovascular protective role of these molecules in prevent the LDL oxidation (Sabharwal and May 2008, Ghaffari et al. 2011, Shariat et al. 2013, Nadeem et al. 2012. The report of Nadeem et al. (2012) demonstrated that vitamin E derivatives, namely the tocopherols, become incorporated into LDL protecting this lipoprotein against oxidation. Specifically, preincubation with α-and γ-tocopherol (final concentration range, 0-5 μM) prevent the LDL oxidation induced by copper (II) chloride solution (CuCl2). Moreover, previously Galli et al. (2004) observed that vitamin E metabolites, carboxyethyl-6-hydroxychromans (concentration range 0.015-5 µM), exert concentration-dependent inhibition of the Cu 2+ -induced lipid oxidation of plasma. On the other hand, some molecules approved by US Food and Drug Administration for the use in the cardiovascular disease also present antioxidant properties. One interesting example is Rosuvastatin, a novel 3-Hydroxy-3-methylgutaryl CoA (HMG-CoA) reductase inhibitor widely used in the treatment of hypercholesterolemia, reducing the risk of myocardial infarction and stroke (MacDonald 2010). Rosuvastatin is able to deal with reactive species generated in the vascular environment and shows in vitro antioxidant capacity protecting against tissue lipid peroxidation induced by Fenton's reaction (Gómez-García et al. 2007, Ajith et al. 2008). In addition, a recent clinical study reported that rosuvastatin treatment reduce significantly plasma levels of oxLDL in hypercholesterolemic subjects (Homma et al. 2015).
Macrophage cholesterol accumulation is positively correlated with atherogenesis (Steinberg et al. 1989, Moore and Tabas 2011, Steinberg 2013. The oxidative modification of LDL resulted in diminished affinity for LDL receptors and increased affinity for macrophage scavenger receptors. The active uptake of the oxidized LDL by macrophages leads to their transition to foam cells, initiating plaque formation (Lv et al. 2014). This process is largely mediated and supported by the metabolism of endothelial and smooth muscle cells in response to oxidized LDL, concomitant with the release of proinflammatory cytokines from emerging foam cells (Nambiar et al. 2014). In this study, we have shown that pretreatment of J774 macrophages with Compound HBD significantly decreased oxidized LDL uptake and, consequently, foam cell formation. Consistent with this observation, we have reported that (PhSe) 2 (Straliotto et al. 2013b) as well as its derivatives (Straliotto et al. 2013a) potentially inhibited the foam cell formation induced by oxidized LDL. Indeed, the effect of Compound HBD in prevent foam cell formation could be through extracellular and intracellular mechanisms, and it is important to stablish this process in next studies. For instance, Bartolini et al. (2015a) showed that a microencapsulated formulation of PhSeZnCl (M-PhSeZnCl) was uptake through an endocytosis like mechanism in the MCF-7 cells (Bartolini et al. 2015a).
With the intention of understanding the mechanism by which the Compound HBD reduces the LDL oxidation, we further investigated the GPxlike activity of this compound. Notably, the Compound HBD displayed concentration-dependent GPx-like activity, as already demonstrated for other organoselenium compounds that efficiently prevent LDL oxidation (de Bem et al. 2008, Straliotto et al. 2013a. The antioxidant activity of organoselenium compounds have been attributed to its GPx-like activity (Sies 1993(Sies , 1994. In this sense, two pioneer studies showed that ebselen is efficient in reducing oxidative modification of LDL through its hydroperoxide-reducing activity (Noguchi et al. 1994, Lass et al. 1996. The decade of 1990s was also characterized by an enormous development of small synthetic organoselenium compounds that mimic GPx catalytic activity. One of them is (PhSe) 2 , the simplest structure in this series, that reacts very efficiently with hydroperoxides and organic peroxides, mimicking the reaction cycle of the GPx enzyme in the presence of reduced thiols. Importantly, the apparent GPx-like activity of (PhSe) 2 has been reported to be superior than that of ebselen (Mugesh 2000, de Bem et al. 2013. One possible explanation for this more efficient activity of (PhSe) 2 is the higher contribution of Se equivalents in the reaction medium (Bartolini JADE DE OLIVEIRA et al. et al. 2015a). Some new disubstituted diaryl-diselenides, such as Compound HBD and p-Cl-diphenyl diselenide (Straliotto et al. 2013a) display GPxlike activity comparable or even higher than that of (PhSe) 2 , while this enzymatic like effect was not evidenced for other compound of this series as p-metoxyl-diphenyl diselenide (Straliotto et al. 2013a). In line with this, Bartolini et al. (2015a) recently showed that phenylselenium zinc chloride (PhSeZnCl) presents higher apparent GPx-like activity than ebselen, but not superior than (PhSe) 2 (Bartolini et al. 2015a). Mechanistically, this result indicated that the GPx-like activity of Compound HBD display an important role in preventing LDL oxidation, pointing it as possible pathway involved in the antiatherogenic effect of this class of compounds. In this regard, clinical studies have suggested an important antiatherogenic role for the antioxidant enzyme GPx (Blankenberg et al. 2003). In animal studies, the lack of functional GPx1 has been shown to accelerate diabetes-associated atherosclerosis via the upregulation of proinflammatory and profibrotic pathways in ApoE -/mice (Lewis et al. 2007). Furthermore, reduced GPx1 expression has been associated with an increase in cell-mediated oxidation of LDL (Guo et al. 2001).
Considering our previous reports about antiatherogenic effect of the simple organoselenium compound (PhSe) 2 , we can not discard that other mechanisms can be involved in the antiatherogenic action of Compound HBD in in vitro and in vivo models (de Bem et al. 2013, Straliotto et al. 2013b). For instance, (PhSe) 2 promotes Nrf-2 activation in endothelial cells, which increases the expression of antioxidant genes leading to an improvement in cellular redox environment and consequently protecting these cells against oxidative insults (de Bem et al. 2013. Furthermore, Bartolini et al. (2015b) demonstrated that a new class of diselenides derived from (PhSe) 2particularly 2,2'-diselenyl dibenzoic acid -behave as mild thiol peroxidases leading to a moderate generation of oxidative stress, which ultimately stimulated Nrf-2 nuclear translocation and then the transcription of the same Nrf-2 gene as well as of glutathione-S-transferase and other phase II genes. This resulted in a higher degree of protection against H 2 O 2 cytotoxicity in cells (Bartolini et al. 2015b).