The mechanisms whereby testosterone increases cardiovascular risk are not clarified. However, oxidative stress and inflammation seem to be determinants. Herein, we sought to determine whether exogenous testosterone, at physiological levels, induces leucocyte migration, a central feature in immune and inflammatory responses and the mediating mechanisms. We hypothesized that testosterone induces leucocyte migration via NADPH oxidase (NADPHox)-driven reactive oxygen species (ROS) and cyclooxygenase (COX)-dependent mechanisms. Sixteen-week-old Wistar rats received an intraperitoneal injection (5 ml) of either testosterone (10−7 mol/l) or saline. Rats were pre-treated with 5 ml of sodium salicylate (SS, non-selective COX inhibitor, 1.25×10−3 mol/l, 1 h prior to testosterone or saline), flutamide (androgen receptor antagonist, 10−5 mol/l), apocynin (NADPHox inhibitor, 3×10−4 mol/l), N-[2-Cyclohexyloxy-4-nitrophenyl]methanesulfonamide (NS398, COX2 inhibitor, 10−4 mol/l) or saline, 4 h before testosterone or saline administration. Leucocyte migration was assessed 24 h after testosterone administration by intravital microscopy of the mesenteric bed. Serum levels of testosterone were measured by radioimmunoassay. NADPHox activity was assessed in membrane fractions of the mesenteric bed by dihydroethidium (DHE) fluorescence and in isolated vascular smooth muscle cells (VSMC) by HPLC. NADPHox subunits and VCAM (vascular cell adhesion molecule) expression were determined by immunoblotting. Testosterone administration did not change serum levels of endogenous testosterone, but increased venular leucocyte migration to the adventia, NADPHox activity and expression (P<0.05). These effects were blocked by flutamide. SS inhibited testosterone-induced leucocyte migration (P<0.05). Apocynin and NS398 abolished testosterone-induced leucocyte migration and NADPHox activity (P<0.05). Testosterone induces leucocyte migration via NADPHox- and COX2-dependent mechanisms and may contribute to inflammatory processes and oxidative stress in the vasculature potentially increasing cardiovascular risk.

CLINICAL PERSPECTIVES

  • Leucocyte migration is a key process in immune response, inflammation and some cardiovascular diseases such as atherosclerosis. The mechanisms by which testosterone contributes to inflammation, cardiovascular diseases and immune response are unknown.

  • Testosterone induces leucocyte migration by COX2- and NADPHox-dependent pathways.

  • COX2 and NADPHox are potential therapeutic targets to decrease testosterone-driven cardiovascular risk.

INTRODUCTION

The effects of anabolic androgenic steroids, including testosterone, in the cardiovascular system remain controversial. Although a beneficial role for testosterone in cardiovascular and metabolic functions in humans has been reported [1,2], other studies indicate that older men with higher testosterone levels are more likely to have cardiovascular diseases [3]. In addition, increased testosterone levels are positively associated with higher blood pressure levels in women [4]. Studies with animal models show that testosterone can lead to cardiac hypertrophy in rats [5] and can enhance vasoconstriction in rabbits [6]. Accordingly, exogenous, as opposed to endogenous, testosterone has been associated with increased cardiovascular risk, coagulatory activation and accelerated progression of coronary artery disease [7].

Oxidative stress is intimately involved in vascular dysfunction, a characteristic feature of many cardiovascular disorders, such as atherosclerosis, hypercholesterolaemia, arterial hypertension and heart failure [8]. Although many studies have investigated the role of testosterone in vascular function, only a few have focused on the effects of testosterone on redox signalling and oxidative stress. We earlier reported that testosterone directly regulates reactive oxygen species (ROS) generation in cultured mesenteric vascular smooth muscle cells (VSMCs), by activating the NADPH oxidase (NADPHox) enzyme, a main source of ROS in the vasculature [9]. However, the effects of testosterone on NADPHox activity and ROS generation in vivo require further investigation.

Leucocyte migration is an important process in the inflammatory response and plays a key role in atherosclerosis [10] and thus may contribute to cardiovascular risk. NADPHox and ROS modulate leucocyte migration via regulation of adhesion molecules such as VCAM (vascular cell adhesion molecule) and P-selectins [11,12].

The effects of testosterone on inflammatory processes are not fully understood and this steroid hormone has been shown to have both pro- [13] and anti-inflammatory [14] actions. In addition, it is unknown whether testosterone-induced ROS generation can also trigger inflammatory processes, which would further contribute to testosterone-induced vascular dysfunction. Accordingly, we sought to determine whether exogenous testosterone activates NADPHox in vivo and whether this effect is related to leucocyte migration. Considering that cyclooxygenases (COXs) are key mediators of inflammation and a pharmacological target for anti-inflammatory agents, we also investigated the role of this enzyme in testosterone-induced leucocyte migration.

METHODS

Chemicals

Chemicals were of the highest purity grade available. Testosterone, flutamide (selective androgen receptor antagonist), sodium salicylate (SS, non-selective COX inhibitor), apocynin (antioxidant/NADPHox inhibitor), N-[2-Cyclohexyloxy-4-nitrophenyl]methanesulfonamide (NS398, selective COX type II inhibitor), DNA-cellulose double-stranded from calf thymus and NADPH were obtained from Sigma. Dihydroethidium (DHE) was purchased from Invitrogen. Ketamine and xylazine were purchased from Agibrands do Brasil. Krebs–Henseleit and phosphate buffer components were from Merck S/A. Antibodies were as follows: NADPH oxidase 1 (NOX1) was bought from ABCAM, NADPH oxidase beta subunit gp91phox (gp91phox, hereafter NOX2) from Upstate, cytochrome b-245 light chain (p22phox) from Santa Cruz Biotechnology, VCAM-1 from Santa Cruz Biotechnology and β-actin from Sigma.

Animals

These studies were approved by the Animal Ethics Committee (CEEA) of the Institute of Biomedical Science–University of Sao Paulo (ICB-USP). The experiments were conducted in accordance with the guidelines from The Brazilian College of Animal Experimentation (COBEA), affiliated with the International Council for Laboratory Animal Science (ICLAS) and with Institutional guidelines. Animals were housed under standard laboratory conditions with free access to food and water. Sixteen-week-old, male Wistar rats were used. On the first day of the experiment, rats received a single 5 ml intraperitoneal injection of one of the following: saline, apocynin 3×10−4 mol/l, NS398 10−4 mol/l or flutamide 10−5 mol/l, as appropriate. After 4 h, animals received a single 5 ml intraperitoneal injection of testosterone (10−7 mol/l) or saline. When animals were pre-treated with SS (1.25×10−3 mol/l), testosterone injection was given 1 h after SS administration due to the fast elimination rate of SS [15]. Chemical concentrations were chosen based on previous studies from our group [9]. Collection of samples or assessment of leucocyte behaviour was performed 24 h after testosterone administration. Testosterone plasma levels were determined by radioimmunoassay according to the manufacturer's instructions (Siemens).

Analysis of leucocyte migration by intravital microscopy

Experimental protocols were performed as previously described [16]. Briefly, male Wistar rats were anesthetized with an intramuscular injection of ketamine (113 mg/kg) and xylazine (7.4 mg/kg) and the mesentery was exteriorized for microscopic examination in situ. Rats were maintained on a special thermostatically-controlled board (37°C), which included a transparent platform on which the tissue to be transilluminated was placed. The preparation was superfused with Krebs–Henseleit solution (in mmol/l: 113.0 NaCl, 4.7 KCl, 2.5 CaCl2·2H2O, 25 NaHCO3, 1.1 MgSO4, 1.1 KH2PO4, 0.03 EDTA and 5.5 glucose; bubbled with 95% N2 an 5% CO2; pH 7.4). A digital camera (Leica DFC 300 FX, Leica) was incorporated into a trinocular microscope (DM LFS, Leica) with a 40× water immersion objective lens with 0.65 numerical aperture to facilitate observation of the enlarged image (2500×) on the video screen. Single unbranched mesenteric venules (length 150 μm) were transilluminated and the images were captured by a computer system (IM50, Leica) for posterior analysis.

ROS production and NADPH oxidase activity

Membrane fractions were isolated from frozen mesenteric vessels. Vessels were homogenized in lysis buffer (50 mM phosphate buffer, pH 7.4; 2 mM EDTA; 5 mM EGTA; 4% NP40; 2% sodium deoxycholate; 04% SDS; 2 mM sodium orthovanadate; 50 mM PMSF and leupeptin/aprotinin/pepstatin 1 mg/ml) and the homogenate was centrifuged at 15 000 g. The supernatant was collected and centrifuged at 100000 g. The pellet (membrane fraction) was dissolved in lysis buffer (50 mM phosphate buffer, pH 7.4; 2 mM EDTA; 5 mM EGTA; 1% Triton; 2 mM sodium orthovanadate; 50 mM PMSF and leupeptin/aprotinin/pepstatin 1 mg/ml). NADPHox activity was assessed by the kinetics of DHE oxidation in a plate fluorimeter (96-wells plates were used). Five microlitres of the membrane extract were added to each well, along with 0.12 μl of DHE (10 mM), 25 mg/ml of DNA and 101.88 μl of phosphate buffer (50 mM, pH 7.4). Fluorescence was read for 30 min (λexc=485 nm, λem=595 nm). Three microlitres of 2×10−3 mol/l NADPH (50 μmol/l) were then added and a new reading was performed after 30 min. The delta for the area under the curve (ΔAUC) was calculated (ΔAUC=AUC after NADPH − AUC before NADPH). ΔAUC was normalized by protein content in the sample. The results are expressed as percentage of control (saline-treated animals).

Superoxide production was assessed in cultured VSMCs by separating DHE-derived oxidation products by HPLC as previously described [17]. Briefly, mesenteric VSMCs were stimulated with testosterone (10−7 mol/l) for 2 h in the presence or absence of flutamide (10−5 mol/l). Cells were then washed with Hank's Buffer [in mmol/l, 1.3 CaCl2, 0.8 MgSO4, 5.4 KCl, 0.4 KH2PO4, 4.3 NaHCO3, 137 NaCl, 0.3 Na2HPO4 and 5.6 glucose, pH 7.4, containing DTPA (diethylene triamine pentaacetic acid) (100 mmol/l)] and incubated with DHE (10−5 mol/l) in Hanks/DTPA for an additional 30 min. Cells were washed twice with cold Hanks' buffer, harvested in acetonitrile (0.5 ml/well) and centrifuged (12000 g for 10 min at 4°C). Supernatants were dried under vacuum (Speed Vac Plus model SC-110A, Thermo Savant) and pellets maintained at −20°C until analysis. Samples were re-suspended in 120 μl of Hanks/DTPA and injected (100 μl) into the HPLC system. Simultaneous detection of DHE and its derived oxidation products [2-hydroxyethidium (2-OHE) and ethidium] was made by ultraviolet (245 nm) and fluorescence detection (λexc=510 nm and λem=595 nm) respectively. DHE was used as an internal control during organic extraction of each sample. Results were expressed as ratios of 2-OHE and ethidium generated per DHE consumed (initial DHE concentration − remaining DHE; 2-OHE/DHE and ethidium/DHE). Chromatographic separation was carried out using a NovaPak C18 column (3.9×150 mm, 5 μm of particle size) in a HPLC system (Waters) equipped with a rheodyne injector and photodiode array (W2996) and fluorescence (W2475) detectors. Quantification was performed by comparison of integrated peak areas between the experimental samples and standard solutions under identical chromatographic conditions.

Immunoblotting

Frozen mesenteric bed vessels or mesenteric VSMCs were homogenized in lysis buffer and proteins were extracted. Fifty microgram of proteins were separated by electrophoresis and transferred to a nitrocellulose membrane. Membranes were incubated with specific primary antibodies (NOX1, NOX2, p22phox, VCAM-1 or β-actin). Non-specific binding sites were blocked with 5% skim milk in Tris-buffered saline solution with 1% of Tween for 1 h at room temperature. After incubation with appropriate secondary antibodies, signals were revealed with chemiluminescence, visualized by autoradiography and quantified densitometrically. Values were normalized for expression of β-actin and results are expressed as percentage of control (saline-treated animals).

Statistical analysis

Data are presented as mean ± S.E.M. Groups were compared using one-way ANOVA or Student t test, as appropriated. Newman–Keuls post-test was used to compensate for multiple testing procedures. P<0.05 was considered statistically significant.

RESULTS

Testosterone induces leucocyte migration in a time-dependent manner

To elucidate whether testosterone modulates leucocyte behaviour, we determined the number of migrated leucocytes in the mesenteric venous bed of testosterone-treated animals by intravital microscopy. Preliminary experiments from our group showed that topically administered testosterone (in situ administration to exposed mesenteric vessels) does not affect the number of rolling, adherent or migrated leucocytes. In addition, leucocyte migration is not observed at earlier time-points when administered intraperitoneally (result not shown). Leucocyte migration was then assessed 24 h after testosterone administration. Testosterone increased the number of migrated leucocytes, an effect blocked by pre-treatment with flutamide, an androgen receptor antagonist (Figure 1). Intraperitoneal injection of exogenous testosterone did not alter its endogenous plasma levels (Supplementary Figure S1).

Testosterone increases leucocyte migration

Figure 1
Testosterone increases leucocyte migration

(A) The number of migrated leucocytes (evaluated in a 2500 μm2 standard area of connective tissue adjacent to a postcapillary venule) was determined 24 h after the intraperitoneal injection of testosterone or saline in rats pre-treated with saline or flutamide. Results are mean ± S.E.M. *P<0.05 compared with saline; **P<0.05 compared with testosterone (n of at least 6 animals/group). (B) Representative images of migrated leucocytes in the adventia of rats treated with saline (control), testosterone and flutamide + testosterone. Images captured by a computer system (IM 50, Leica) with a 40× longitudinal distance objective with 0.65 numerical aperture and visualized through an intravital microscope (DM LFS, Leica) with a 40× water immersion objective lens.

Figure 1
Testosterone increases leucocyte migration

(A) The number of migrated leucocytes (evaluated in a 2500 μm2 standard area of connective tissue adjacent to a postcapillary venule) was determined 24 h after the intraperitoneal injection of testosterone or saline in rats pre-treated with saline or flutamide. Results are mean ± S.E.M. *P<0.05 compared with saline; **P<0.05 compared with testosterone (n of at least 6 animals/group). (B) Representative images of migrated leucocytes in the adventia of rats treated with saline (control), testosterone and flutamide + testosterone. Images captured by a computer system (IM 50, Leica) with a 40× longitudinal distance objective with 0.65 numerical aperture and visualized through an intravital microscope (DM LFS, Leica) with a 40× water immersion objective lens.

Testosterone-induced leucocyte migration depends on NADPH activity

To clarify whether testosterone-induced leucocyte migration is a redox-sensitive phenomenon, we evaluated the effects of apocynin, an antioxidant/NADPHox inhibitor. Pre-treatment with apocynin reduced testosterone-induced leucocyte migration to levels similar to those observed in saline-treated animals (Figure 2). Apocynin did not alter testosterone serum levels (Supplementary Figure S1).

Testosterone-induced leucocyte migration depends on ROS

Figure 2
Testosterone-induced leucocyte migration depends on ROS

The number of migrated leucocytes (determined in a 2500 μm2 standard area of connective tissue adjacent to a postcapillary venule) was determined 24 h after the intraperitoneal injection of testosterone or saline in rats pre-treated with saline (control) or apocynin. Results are mean ± S.E.M. *P<0.05 compared with saline; **P<0.05 compared with testosterone (n of at least 5 animals/group).

Figure 2
Testosterone-induced leucocyte migration depends on ROS

The number of migrated leucocytes (determined in a 2500 μm2 standard area of connective tissue adjacent to a postcapillary venule) was determined 24 h after the intraperitoneal injection of testosterone or saline in rats pre-treated with saline (control) or apocynin. Results are mean ± S.E.M. *P<0.05 compared with saline; **P<0.05 compared with testosterone (n of at least 5 animals/group).

Testosterone positively regulates NADPH oxidase

To further investigate the role of NADPHox on testosterone actions in the mesenteric bed, we assessed protein expression of the NADPHox subunits NOX1, NOX2 and p22phox by immunoblotting. NADPHox activity was determined in membrane-isolated fractions of mesenteric vessels by DHE fluorescence. Testosterone increased NADPHox activity (Figure 3A) and protein levels of NOX2 (Figure 3B), p22 phox (Figure 3C) and NOX1 (Figure 3D). These effects were blocked by flutamide (Figures 3A–3D). In addition, testosterone increased superoxide anion production in VSMCs isolated from the mesenteric bed of Wistar rats, as assessed by HPLC (Supplementary Figure S2).

Testosterone positively regulates NADPHox

Figure 3
Testosterone positively regulates NADPHox

(A) Testosterone increases NADPHox activity in mesenteric vessels. NADPHox activity was assessed by membrane-fraction DHE fluorescence from the mesenteric vessels of saline-, testosterone- and flutamide + testosterone-treated animals. Testosterone increases protein expression of (B) NOX2, (C) p22phox and (D) NOX1. Protein expression was assessed in mesenteric vessels from rats treated with saline (S), testosterone (T) or flutamide + testosterone (F + T). *P<0.05 compared with control (saline-treated animals); **P<0.05 compared with testosterone (n of at least 3 animals/group). A.U., arbitrary units.

Figure 3
Testosterone positively regulates NADPHox

(A) Testosterone increases NADPHox activity in mesenteric vessels. NADPHox activity was assessed by membrane-fraction DHE fluorescence from the mesenteric vessels of saline-, testosterone- and flutamide + testosterone-treated animals. Testosterone increases protein expression of (B) NOX2, (C) p22phox and (D) NOX1. Protein expression was assessed in mesenteric vessels from rats treated with saline (S), testosterone (T) or flutamide + testosterone (F + T). *P<0.05 compared with control (saline-treated animals); **P<0.05 compared with testosterone (n of at least 3 animals/group). A.U., arbitrary units.

Type 2 cyclooxygenase contributes to testosterone-induced leucocyte migration

To determine whether COX enzymes contribute to testosterone-induced leucocyte migration, the effects of testosterone were determined in the presence of SS, a non-selective COX inhibitor. We first assessed whether pre-treatment with SS for 4 h before single testosterone injection had any effect on testosterone-induced leucocyte migration. Within this time-frame, testosterone effects on leucocyte migration were not significantly inhibited (result not shown). Considering that the concentration of plasma salicylic acid (SS metabolite) decreases to less than half within 2 h of its administration [15], we shortened the pre-treatment period to 1 h prior to testosterone single injection. As shown in Figure 4(A), pre-treatment with SS for 1 h, moderately inhibited testosterone effects on leucocyte migration. To investigate whether COX2 is the isoform mediating this effect, we assessed leucocyte migration in animals pre-treated with NS398. As shown in Figure 4(B), treatment of animals with NS398, a selective COX2 inhibitor, abolished testosterone effects on leucocyte migration. NS398 did not change testosterone serum levels (Supplementary Figure S1).

Testosterone-induced leucocyte migration depends on COXs

Figure 4
Testosterone-induced leucocyte migration depends on COXs

The number of migrated leucocytes (determined in a 2500 μm2 standard area of connective tissue adjacent to a postcapillary venule) was determined 24 h after the intraperitoneal injection of testosterone or saline in rats pre-treated with (A) SS or (B) NS398. Results are mean ± S.E.M. *P<0.05 compared with saline-treated animals; **P<0.05 compared with testosterone (n of at least 5 animals/group).

Figure 4
Testosterone-induced leucocyte migration depends on COXs

The number of migrated leucocytes (determined in a 2500 μm2 standard area of connective tissue adjacent to a postcapillary venule) was determined 24 h after the intraperitoneal injection of testosterone or saline in rats pre-treated with (A) SS or (B) NS398. Results are mean ± S.E.M. *P<0.05 compared with saline-treated animals; **P<0.05 compared with testosterone (n of at least 5 animals/group).

Type 2 cyclooxygenase is a regulator of testosterone effects on NADPH oxidase activity

To investigate the mechanisms whereby COX2 contributes to testosterone effects in the vasculature we assessed testosterone-induced NADPHox activation in the presence of NS398. As shown in Figure 5, pre-treatment of animals with NS398 abolished the effects of testosterone on NADPHox activity.

COX2 regulates testosterone-induced NADPHox activity

Figure 5
COX2 regulates testosterone-induced NADPHox activity

NADPHox activity was assessed by membrane-fraction DHE fluorescence from the mesenteric vessels of testosterone- and NS398 + testosterone-treated animals. *P<0.05 compared with saline-treated animals; **P<0.05 compared with testosterone (n of at least 4 animals/group). A.U., arbitrary units.

Figure 5
COX2 regulates testosterone-induced NADPHox activity

NADPHox activity was assessed by membrane-fraction DHE fluorescence from the mesenteric vessels of testosterone- and NS398 + testosterone-treated animals. *P<0.05 compared with saline-treated animals; **P<0.05 compared with testosterone (n of at least 4 animals/group). A.U., arbitrary units.

DISCUSSION

Findings from our study demonstrate that testosterone induces leucocyte migration by NADPHox-driven ROS and COX2-related mechanisms. These processes may represent important mechanisms of testosterone effects in the cardiovascular system, especially in conditions where inflammation and oxidative stress are key features, such as atherosclerosis, arterial hypertension, obesity and diabetes.

We provide evidence that testosterone-induced leucocyte migration may contribute to the pro-inflammatory effects of this steroid hormone in the vasculature and thus could contribute to increased cardiovascular risk. This suggestion is supported by several observations: (1) testosterone induces leucocyte migration; (2) testosterone increases ROS production; (3) testosterone increases NADPHox activity and expression; (4) COXs are mediators of leucocyte migration; (5) COX type 2 is a determinant in testosterone-induced leucocyte migration; (6) COX2 is required for NADPHox activation by testosterone.

Testosterone regulation of leucocyte behaviour was observed only 24 h after administration, suggesting a genomic effect and/or activation of indirect pathways. We have previously shown that testosterone induces only long-term ROS generation in VSMC from male normotensive animals (Wistar–Kyoto rats), as opposed to acute- and long-term ROS generation in VSMC from male spontaneously hypertensive rats (SHR) [9]. Testosterone-induced ROS production is regulated by phosphorylation of the non-receptor tyrosine kinase, c-src [9], which mediates vascular contraction and hypertrophy [18] and whose activity is up-regulated in various pathological conditions, such as arterial hypertension [19]. Testosterone also induced VSMC migration, an effect that was augmented in VSMC from SHR [9]. These observations not only indicated that the genetic environment, as well as the physiopathological conditions to which cells are exposed, may influence the effects of testosterone, but also that the cellular processes activated by testosterone might further aggravate vascular dysfunction associated with cardiovascular diseases, e.g. arterial hypertension.

Herein, we demonstrate that testosterone also induces NADPHox activity in vivo (Wistar rats) and that NADPHox-driven ROS is involved in testosterone-induced leucocyte migration. These findings are supported by previous reports describing a role for ROS in the regulation of vascular inflammation [20]. It has been reported that NADPHox-derived ROS are required for VCAM-dependent leucocyte migration [11] and have been implicated in the generation of interleukin (IL)-8 production by IL-1β-activated mast cells [21]. The NADPHox pathway is also critical for lipopolysaccharide (LPS)-induced conversion of endothelial cells to a pro-inflammatory phenotype [22]. Additionally, the oxidase activity of vascular adhesion protein-1 induces E- and P-selectins and leucocyte binding [12]. Moreover, as ROS activate both mast cells and macrophages [23,24], they can also amplify the inflammatory response. In the present study we focused our investigation on the effects of testosterone in rat vascular cells, which express mainly NOX1, NOX2 and NOX4 isoforms of NADPHox. Circulating macrophages also express NOX1 and NOX2 in addition to the androgen receptor. Therefore, circulating leucocytes might be a potential source of ROS production by testosterone. It would be very challenging to distinguish if NADPHox ROS generation in our in vivo model is derived only from vascular cells or from both vascular and circulating cells. Testosterone effects on isolated macrophages could be tested but it would not necessarily represent what is happening in vivo. One complication of such an experiment is that usually monocytes are isolated and then treated with cytokines or PMA to differentiate them into macrophages. In this context, the agent used to differentiate the isolated cells is also capable of activating ROS production and we would not be able to evaluate the effects of testosterone per se. Such effects would have to be studied in combination with a pro-oxidative agent. Reports in the literature about testosterone effects in in vitro differentiated macrophages are still controversial. Chao et al. [25] have shown that testosterone acts in a concentration-dependent manner in isolated peritoneal macrophages and might induce ROS production when cells are stimulated with testosterone in concentrations higher than 10−9 mol/l. On the other hand, Mohan and Jacobson [26] reported that testosterone does not affect rat peritoneal macrophage superoxide production. In contrast, Juliet et al. [27] described that testosterone can in fact decrease superoxide production in macrophages derived from THP-1 cells (THP-1 cell line; human acute monocytic leukemia cell line). An alternative approach for evaluating whether macrophage-localized NADPHox could contribute to testosterone effects in leucocyte migration would be to use the Cre-loxP system to delete the androgen receptor from macrophages and then assess NADPHox activity in such cells.

Leucocyte migration through the vessel wall is a crucial step in vascular inflammation. ROS-mediated vascular injury may result in endothelial barrier dysfunction [28] and increased expression of cell adhesion molecules [11,12], both important mediators of leucocyte migration. Interactions between leucocyte and endothelial cells are keys to leucocyte traffic through the vessel wall. This process is initiated by the tethering of leucocytes to the endothelium that will first lead to weak adhesive interactions between cells (leucocyte rolling), then strong cell–cell interactions (leucocyte adhesion) which will ultimately result in leucocyte transmigration through the vascular endothelium [2931] to the adventitia.

Previous work from our group reported that endogenous testosterone can partly modulate early leucocyte–endothelial cell interactions in SHR [13]. Castrated SHR had lower numbers of rolling leucocytes when compared with non-castrated animals. When challenged with leukotriene B4, castrated SHR presented lower numbers of adhered leucocytes when compared with non-castrated animals. However, no differences were observed in leucocyte migration in this previous study. In addition, castration of SHR was accompanied by a decrease in blood pressure, so it is unknown if the differences observed in castrated animals are secondary to blood pressure levels or if it is a direct effect of decreased testosterone levels. To our knowledge, the present study provides the first evidence that testosterone at physiological levels can regulate leucocyte migration through the vessel wall in normotensive animals.

As mentioned before, shorter periods of treatment with testosterone (4 h) did not alter leucocyte migration. The fact that this effect could be observed only 24 h after testosterone administration together with the finding that flutamide blocks testosterone-induced leucocyte migration indicates that this effect occurs via genomic pathways, probably by indirect mechanisms. Previous reports from our group showed that testosterone does not alter venular diameter, blood flow velocity, shear rate and leucocyte velocity in the vascular bed studied in the present study (mesentery bed) [13], making unlikely the possibility that changes in venular parameters or leucocyte velocity are involved in the effects of testosterone on leucocytes migration.

We speculate that testosterone induces the expression of adhesion proteins, as well as stimulating the arachidonic acid pathway. Testosterone induces VCAM expression in cultured VSMC (Supplementary Figure S3). In addition, this steroid hormone contributes to increased P-selectin and ICAM1 (intercellular adhesion molecule 1) expression in mesenteric venules of hypertensive (SHR) when compared with venules of normotensive (Wistar) rats [13]. Reports in the literature have shown that non-specific COX inhibitors are not effective in blocking IL1-β-induced VCAM-1 expression in primary dental pulp cells [32] and human umbilical vein endothelial cells (HUVECs) [33]. Whether testosterone effects on VCAM-1 expression are COX2- and/or ROS-dependent and intercorrelated is still unknown.

Importantly, we define a redox-sensitive pathway, modulated by COX2, as a key mechanism for testosterone-induced leucocyte migration. Previous studies have shown that in non-cardiovascular tissues, COX2 expression is positively correlated with testosterone levels [34,35]. However, to our knowledge, a similar phenomenon in the vasculature has not been described. As our treatment did not change serum levels of testosterone, it is also unlike that augmented testosterone levels determined COX2 activity. COX2 activity might be altered by intermediary signalling pathways or by post-translational modifications in the enzyme. Testosterone did not alter COX2 total protein expression in mesenteric vessels or isolated VSMC (result not shown).

COXs increase leucocyte migration [36] by unknown mechanisms. Choi et al. [37] have reported different responses for intracerebroventricular LPS-induced neuroinflammation as assessed by leucocyte migration in COX1 and COX2 knockout mice, describing that for that specific model, COX1 deletion was shown to be protective and COX2 was shown to be deleterious towards inflammation. In addition, different COX isoforms might be more relevant in specific steps of the leucocyte migration process. In a model of LPS-induced rat peritoneal inflammation, whereas indomethacin, a non-specific COX inhibitor, proved to be effective in inhibiting cell adhesion, celecoxib, a specific COX2 antagonist, exhibited inhibitory effects in leucocyte rolling and adhesion [38].

In the present study, we demonstrate that selective and non-selective pharmacological inhibition of COXs decreases testosterone-induced leucocyte migration. Non-selective inhibition of COXs results only in a partial decrease in testosterone effects on leucocyte migration, whereas selective COX2 inhibition leads to blockade of such an effect. The partial blockade by SS may be due to the fact that SS has a lower affinity for COX2. Alternatively it is possible that the shorter treatment with SS did not produce a complete blockade of COX2. In this case, COX2 would not be completely inhibited and a partial effect of testosterone could still be observed. Our data support COX2 as a key intermediate in testosterone effects on leucocyte behaviour. It is important to note that SS has already been described as an efficient inhibitor of leucocyte migration [39].

The exact mechanisms by which testosterone modulates COX2 is still unknown. Although testosterone increased mRNA COX2 levels after 2–6 h in cultured aortic VSMC, no differences in COX2 protein expression were observed in mesenteric vessels of rats treated with testosterone for 24 h or in cultured aortic VSMC (result not shown). Thus it is likely that modulation of COX2 activity by testosterone is an indirect effect that might be caused by activation of intermediary signalling pathways or via post-translational modifications.

In addition, we characterize COX2 as a regulator of NADPHox activation by testosterone. Evidence in the literature about a possible interaction between COX2 and NADPHox is still controversial. It has been described that in animal models of hypertension, there is a vicious regulatory cycle between NADPHox and COX2, in which COX2 regulates NADPHox-driven ROS production and ROS regulates COX2 expression [40]. On the other hand, a 28-day treatment of dogs with an experimental selective COX-2 inhibitor revealed no drug-associated changes in production of superoxide anion in plasma and isolated neutrophils [41]. Although our data show that COX2 modulates NADPHox in response to testosterone in vivo, further investigation is required to clarify a role for NADPH-driven ROS in the regulation of COX2 and whether a vicious cycle between COX2 and NADPHox happens in testosterone-treated rats.

Testosterone has been used since 1930s for non-medical, athletic purposes. Testosterone abuse may lead to ventricular hypertrophy, myocardial infarction, coronary vasospasm and thrombosis [7]. Mechanisms by which testosterone may increase cardiovascular risk include coagulation activation as well as coronary artery disease. Here we add one more piece to the puzzle; testosterone induces leucocyte migration in normotensive rats by NADPHox- and COX2-dependent mechanisms in male normotensive rats. How these effects contribute to the development or accelerated progression of cardiovascular disease is still unknown.

As the absence of oestrogen in males or after menopause in women is independently associated with increased cardiovascular risk [42], it would be important to evaluate if such effects can happen in female rats. Such findings could also contribute to understanding vascular complications in some female-specific disorders, such as polycystic ovarian syndrome and pre-eclampsia in which testosterone levels are increased [43,44].

In conclusion, testosterone induces leucocyte migration in normotensive animals and this may represent one of the potential mechanisms whereby testosterone increases cardiovascular risk.

AUTHOR CONTRIBUTION

Andreia Chignalia conceived the study, designed and performed the experiments, analysed the data and wrote the manuscript. Maria Oliveira designed and performed the experiments, analysed the data and wrote the manuscript. Victor Debbas performed experiments. Randal Dull revised the manuscript. Francisco Laurindo, Rhian Touyz and Maria Carvalho, analysed the data and wrote the manuscript. Zuleica Fortes and Rita Tostes conceived the study, designed experiments, analysed the data and wrote the manuscript.

We thank Ana Rita Gonçalves for her technical support and Dr Maria Christina Avellar for donating the testosterone used in our studies.

FUNDING

This study was supported by the Fundação de Amparo à Pesquisa do Estado de São Paulo [grant numbers 2004/13796-9 and 2004/15968-1]; the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior [grant number 2667/06-1]; and the Conselho Nacional de Desenvolvimento Científico e Tecnológico [grant number 461598/2000-0].

Abbreviations

     
  • 2-OHE

    2-hydroxyethidium

  •  
  • COX

    cyclooxygenase

  •  
  • DHE

    dihydroethidium

  •  
  • DTPA

    diethylene triamine pentaacetic acid

  •  
  • IL

    interleukin

  •  
  • LPS

    lipopolysaccharide

  •  
  • NADPHox

    NADPH oxidase

  •  
  • ROS

    reactive oxygen species

  •  
  • SHR

    spontaneously hypertensive rats

  •  
  • SS

    sodium salicylate

  •  
  • VCAM

    vascular cell adhesion molecule

  •  
  • VSMC

    vascular smooth muscle cell

References

References
1
Hyde
 
Z.
Norman
 
P.E.
Flicker
 
L.
Hankey
 
G.J.
Almeida
 
O.P.
McCaul
 
K.A.
Chubb
 
S.A.
Yeap
 
B.B.
 
Low free testosterone predicts mortality from cardiovascular disease but not other causes: the health in men study
J. Clin. Endocrinol. Metab.
2012
, vol. 
97
 (pg. 
179
-
189
)
[PubMed]
2
Ruige
 
J.B.
Mahmoud
 
A.M.
De Bacquer
 
D.
Kaufman
 
J.M.
 
Endogenous testosterone and cardiovascular disease in healthy men: a meta-analysis
Heart
2011
, vol. 
97
 (pg. 
870
-
875
)
[PubMed]
3
Basaria
 
S.
Coviello
 
A.D.
Travison
 
T.G.
Storer
 
T.W.
Farwell
 
W.R.
Jette
 
A.M.
Eder
 
R.
Tennstedt
 
S.
Ulloor
 
J.
Zhang
 
A.
, et al 
Adverse events associated with testosterone administration
N. Engl. J. Med.
2010
, vol. 
363
 (pg. 
109
-
122
)
[PubMed]
4
Ziemens
 
B.
Wallaschofski
 
H.
Volzke
 
H.
Rettig
 
R.
Dorr
 
M.
Nauck
 
M.
Keevil
 
B.G.
Brabant
 
G.
Haring
 
R.
 
Positive association between testosterone, blood pressure, and hypertension in women: longitudinal findings from the study of health in pomerania
J. Hypertens.
2013
, vol. 
31
 (pg. 
1106
-
1113
)
[PubMed]
5
Nahrendorf
 
M.
Frantz
 
S.
Hu
 
K.
von zur Muhlen
 
C.
Tomaszewski
 
M.
Scheuermann
 
H.
Kaiser
 
R.
Jazbutyte
 
V.
Beer
 
S.
Bauer
 
W.
, et al 
Effect of testosterone on post-myocardial infarction remodeling and function
Cardiovasc Res.
2003
, vol. 
57
 (pg. 
370
-
378
)
[PubMed]
6
Ammar
 
E.M.
Said
 
S.A.
Hassan
 
M.S.
 
Enhanced vasoconstriction and reduced vasorelaxation induced by testosterone and nandrolone in hypercholesterolemic rabbits
Pharmacol. Res.
2004
, vol. 
50
 (pg. 
253
-
259
)
[PubMed]
7
Barton
 
M.
Prossnitz
 
E.R.
Meyer
 
M.R.
 
Testosterone and secondary hypertension: new pieces to the puzzle
Hypertension
2012
, vol. 
59
 (pg. 
1101
-
1103
)
[PubMed]
8
Heistad
 
D.D.
Wakisaka
 
Y.
Miller
 
J.
Chu
 
Y.
Pena-Silva
 
R.
 
Novel aspects of oxidative stress in cardiovascular diseases
Circ. J.
2009
, vol. 
73
 (pg. 
201
-
207
)
[PubMed]
9
Chignalia
 
A.Z.
Schuldt
 
E.Z.
Camargo
 
L.L.
Montezano
 
A.C.
Callera
 
G.E.
Laurindo
 
F.R.
Lopes
 
L.R.
Avellar
 
M.C.
Carvalho
 
M.H.
Fortes
 
Z.B.
, et al 
Testosterone induces vascular smooth muscle cell migration by NADPH oxidase and c-Src-dependent pathways
Hypertension
2012
, vol. 
59
 (pg. 
1263
-
1271
)
[PubMed]
10
Hopkins
 
P.N.
 
Molecular biology of atherosclerosis
Physiol. Rev.
2013
, vol. 
93
 (pg. 
1317
-
1542
)
[PubMed]
11
Cook-Mills
 
J.M.
 
Hydrogen peroxide activation of endothelial cell-associated MMPs during VCAM-1-dependent leukocyte migration
Cell. Mol. Biol.
2006
, vol. 
52
 (pg. 
8
-
16
)
12
Jalkanen
 
S.
Karikoski
 
M.
Mercier
 
N.
Koskinen
 
K.
Henttinen
 
T.
Elima
 
K.
Salmivirta
 
K.
Salmi
 
M.
 
The oxidase activity of vascular adhesion protein-1 (VAP-1) induces endothelial E- and P-selectins and leukocyte binding
Blood
2007
, vol. 
110
 (pg. 
1864
-
1870
)
[PubMed]
13
Filgueira
 
F.P.
Lobato
 
N.S.
DosSantos
 
R.A.
Oliveira
 
M.A.
Akamine
 
E.H.
Tostes
 
R.C.
Fortes
 
Z.B.
Carvalho
 
M.H.
 
Endogenous testosterone increases leukocyte-endothelial cell interaction in spontaneously hypertensive rats
Life Sci.
2012
, vol. 
90
 (pg. 
689
-
694
)
[PubMed]
14
Ganesan
 
K.
Balachandran
 
C.
Manohar
 
B.M.
Puvanakrishnan
 
R.
 
Effects of testosterone, estrogen and progesterone on TNF-α mediated cellular damage in rat arthritic synovial fibroblasts
Rheumatol. Int.
2012
, vol. 
32
 (pg. 
3181
-
3188
)
[PubMed]
15
Yue
 
T.L.
Varma
 
D.R.
 
Pharmacokinetics, metabolism and disposition of salicylate in protein-deficient rats
Drug. Metab. Dispos.
1982
, vol. 
10
 (pg. 
147
-
152
)
[PubMed]
16
Fortes
 
Z.B.
Farsky
 
S.P.
Oliveira
 
M.A.
Garcia-Leme
 
J.
 
Direct vital microscopic study of defective leukocyte-endothelial interaction in diabetes mellitus
Diabetes
1991
, vol. 
40
 (pg. 
1267
-
1273
)
[PubMed]
17
Fernandes
 
D.C.
Wosniak
 
J.
Pescatore
 
L.A.
Bertoline
 
M.A.
Liberman
 
M.
Laurindo
 
F.R.
Santos
 
C.X.
 
Analysis of DHE-derived oxidation products by HPLC in the assessment of superoxide production and NADPH oxidase activity in vascular systems
Am. J. Physiol. Cell Physiol.
2007
, vol. 
292
 (pg. 
C413
-
C422
)
[PubMed]
18
Touyz
 
R.M.
Wu
 
X.H.
He
 
G.
Park
 
J.B.
Chen
 
X.
Vacher
 
J.
Rajapurohitam
 
V.
Schiffrin
 
E.L.
 
Role of c-Src in the regulation of vascular contraction and Ca2+ signaling by angiotensin II in human vascular smooth muscle cells
J. Hypertens
2001
, vol. 
19
 (pg. 
441
-
449
)
[PubMed]
19
Callera
 
G.E.
Montezano
 
A.C.
Yogi
 
A.
Tostes
 
R.C.
He
 
Y.
Schiffrin
 
E.L.
Touyz
 
R.M.
 
c-Src-dependent nongenomic signaling responses to aldosterone are increased in vascular myocytes from spontaneously hypertensive rats
Hypertension
2005
, vol. 
46
 (pg. 
1032
-
1038
)
[PubMed]
20
Kvietys
 
P.R.
Granger
 
D.N.
 
Role of reactive oxygen and nitrogen species in the vascular responses to inflammation
Free Radic Biol. Med.
2012
, vol. 
52
 (pg. 
556
-
592
)
[PubMed]
21
Kim
 
G.Y.
Lee
 
J.W.
Ryu
 
H.C.
Wei
 
J.D.
Seong
 
C.M.
Kim
 
J.H.
 
Proinflammatory cytokine IL-1beta stimulates IL-8 synthesis in mast cells via a leukotriene B4 receptor 2-linked pathway, contributing to angiogenesis
J. Immunol.
2010
, vol. 
184
 (pg. 
3946
-
3954
)
[PubMed]
22
Park
 
H.S.
Chun
 
J.N.
Jung
 
H.Y.
Choi
 
C.
Bae
 
Y.S.
 
Role of NADPH oxidase 4 in lipopolysaccharide-induced proinflammatory responses by human aortic endothelial cells
Cardiovasc. Res.
2006
, vol. 
72
 (pg. 
447
-
455
)
[PubMed]
23
Swindle
 
E.J.
Metcalfe
 
D.D.
 
The role of reactive oxygen species and nitric oxide in mast cell-dependent inflammatory processes
Immunol. Rev.
2007
, vol. 
217
 (pg. 
186
-
205
)
[PubMed]
24
Paul-Clark
 
M.J.
McMaster
 
S.K.
Sorrentino
 
R.
Sriskandan
 
S.
Bailey
 
L.K.
Moreno
 
L.
Ryffel
 
B.
Quesniaux
 
V.F.
Mitchell
 
J.A.
 
Toll-like receptor 2 is essential for the sensing of oxidants during inflammation
Am. J. Respir. Crit. Care Med.
2009
, vol. 
179
 (pg. 
299
-
306
)
[PubMed]
25
Chao
 
T.C.
Van Alten
 
P.J.
Walter
 
R.J.
 
Steroid sex hormones and macrophage function: modulation of reactive oxygen intermediates and nitrite release
Am. J. Reprod. Immunol.
1994
, vol. 
32
 (pg. 
43
-
52
)
[PubMed]
26
Mohan
 
P.F.
Jacobson
 
M.S.
 
Inhibition of macrophage superoxide generation by dehydroepiandrosterone
Am. J. Med. Sci.
1993
, vol. 
306
 (pg. 
10
-
15
)
[PubMed]
27
Juliet
 
P.A.
Hayashi
 
T.
Daigo
 
S.
Matsui-Hirai
 
H.
Miyazaki
 
A.
Fukatsu
 
A.
Funami
 
J.
Iguchi
 
A.
Ignarro
 
L.J.
 
Combined effect of testosterone and apocynin on nitric oxide and superoxide production in PMA-differentiated THP-1 cells
Biochim. Biophys. Acta
2004
, vol. 
1693
 (pg. 
185
-
191
)
[PubMed]
28
Makarenko
 
V.V.
Usatyuk
 
P.V.
Yuan
 
G.
Lee
 
M.M.
Nanduri
 
J.
Natarajan
 
V.
Kumar
 
G.K.
Prabhakar
 
N.R.
 
Intermittent hypoxia-induced endothelial barrier dysfunction requires ROS-dependent MAP kinase activation
Am. J. Physiol. Cell. Physiol.
2014
, vol. 
306
 (pg. 
C745
-
C752
)
[PubMed]
29
Butcher
 
E.C.
 
Leukocyte-endothelial cell recognition: three (or more) steps to specificity and diversity
Cell
1991
, vol. 
67
 (pg. 
1033
-
1036
)
[PubMed]
30
Springer
 
T.A.
 
Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm
Cell
1994
, vol. 
76
 (pg. 
301
-
314
)
[PubMed]
31
Alvarez
 
A.
Cerda-Nicolas
 
M.
Naim Abu Nabah
 
Y.
Mata
 
M.
Issekutz
 
A.C.
Panes
 
J.
Lobb
 
R.R.
Sanz
 
M.J.
 
Direct evidence of leukocyte adhesion in arterioles by angiotensin II
Blood
2004
, vol. 
104
 (pg. 
402
-
408
)
[PubMed]
32
Chang
 
M.C.
Lin
 
L.D.
Zwei-Ching Chang
 
J.
Huang
 
C.F.
Chuang
 
F.H.
Lee
 
J.J.
Jeng
 
P.Y.
Wang
 
T.M.
Jeng
 
J.H.
 
Regulation of vascular cell adhesion molecule-1 in dental pulp cells by interleukin-1β: the role of prostanoids
J. Endod.
2012
, vol. 
38
 (pg. 
774
-
779
)
[PubMed]
33
Lee
 
S.
Felts
 
K.A.
Parry
 
G.C.
Armacost
 
L.M.
Cobb
 
R.R.
 
Inhibition of 5-lipoxygenase blocks IL-1 beta-induced vascular adhesion molecule-1 gene expression in human endothelial cells
J. Immunol.
1997
, vol. 
158
 (pg. 
3401
-
3407
)
[PubMed]
34
Amalfi
 
S.
Velez
 
L.M.
Heber
 
M.F.
Vighi
 
S.
Ferreira
 
S.R.
Orozco
 
A.V.
Pignataro
 
O.
Motta
 
A.B.
 
Prenatal hyperandrogenization induces metabolic and endocrine alterations which depend on the levels of testosterone exposure
PLoS One
2012
, vol. 
7
 pg. 
e37658
 
[PubMed]
35
Wang
 
X.
Shen
 
C.L.
Dyson
 
M.T.
Eimerl
 
S.
Orly
 
J.
Hutson
 
J.C.
Stocco
 
D.M.
 
Cyclooxygenase-2 regulation of the age-related decline in testosterone biosynthesis
Endocrinology
2005
, vol. 
146
 (pg. 
4202
-
4208
)
[PubMed]
36
de Menezes
 
G.B.
dos Reis
 
W.G.
Santos
 
J.M.
Duarte
 
I.D.
de Francischi
 
J.N.
 
Inhibition of prostaglandin F(2alpha) by selective cyclooxygenase 2 inhibitors accounts for reduced rat leukocyte migration
Inflammation
2005
, vol. 
29
 (pg. 
163
-
169
)
[PubMed]
37
Choi
 
S.H.
Aid
 
S.
Choi
 
U.
Bosetti
 
F.
 
Cyclooxygenases-1 and -2 differentially modulate leukocyte recruitment into the inflamed brain
Pharmacogenomics J
2010
, vol. 
10
 (pg. 
448
-
457
)
[PubMed]
38
Menezes
 
G.B.
Rezende
 
R.M.
Pereira-Silva
 
P.E.
Klein
 
A.
Cara
 
D.C.
Francischi
 
J.N.
 
Differential involvement of cyclooxygenase isoforms in neutrophil migration in vivo and in vitro
Eur. J. Pharmacol.
2008
, vol. 
598
 (pg. 
118
-
122
)
[PubMed]
39
Asako
 
H.
Kubes
 
P.
Wallace
 
J.
Wolf
 
R.E.
Granger
 
D.N.
 
Modulation of leukocyte adhesion in rat mesenteric venules by aspirin and salicylate
Gastroenterology
1992
, vol. 
103
 (pg. 
146
-
152
)
[PubMed]
40
Martinez-Revelles
 
S.
Avendano
 
M.S.
Garcia-Redondo
 
A.B.
Alvarez
 
Y.
Aguado
 
A.
Perez-Giron
 
J.V.
Garcia-Redondo
 
L.
Esteban
 
V.
Redondo
 
J.M.
Alonso
 
M.J.
, et al 
Reciprocal relationship between reactive oxygen species and cyclooxygenase-2 and vascular dysfunction in hypertension
Antioxid. Redox Signal.
2013
, vol. 
18
 (pg. 
51
-
65
)
[PubMed]
41
Furst
 
S.M.
Khan
 
K.N.
Komocsar
 
W.J.
Fan
 
L.
Mennear
 
J.
 
Screening new drugs for immunotoxic potential: II. Assessment of the effects of selective and nonselective COX-2 inhibitors on complement activation, superoxide anion production and leukocyte chemotaxis and migration through endothelial cells
J. Immunotoxicol.
2005
, vol. 
2
 (pg. 
85
-
96
)
[PubMed]
42
Meyer
 
M.R.
Haas
 
E.
Barton
 
M.
 
Gender differences of cardiovascular disease: new perspectives for estrogen receptor signaling
Hypertension
2006
, vol. 
47
 (pg. 
1019
-
1026
)
[PubMed]
43
Ghorashi
 
V.
Sheikhvatan
 
M.
 
The relationship between serum concentration of free testosterone and pre-eclampsia
Endokrynol. Pol.
2008
, vol. 
59
 (pg. 
390
-
392
)
[PubMed]
44
Szydlarska
 
D.
Grzesiuk
 
W.
Kondracka
 
A.
Bartoszewicz
 
Z.
Bar-Andziak
 
E.
 
Measuring salivary androgens as a useful tool in the diagnosis of polycystic ovary syndrome
Endokrynol. Pol.
2012
, vol. 
63
 (pg. 
183
-
190
)
[PubMed]

Supplementary data