All protocols with animal study were conducted in accordance with the Brazilian Council of Animal Research Guidelines (CONCEA), reviewed and approved by the Ethics Committee on Animal Use of Federal University of Minas Gerais (UFMG) under the protocol number 225/2013. Male Balb/c mice, 8 weeks of age, were obtained from the Center of Bioterism of Biological Sciences Institute of UFMG and kept under controlled conditions of temperature and luminosity (light-dark cycle of 12 h), with free access to water and food.

The animals were randomly divided into two groups: control and HC. The control group received a standard diet (Nuvilab CR-1), while the HC group received a refined carbohydrate enriched diet (HC diet) for 4 weeks. The HC diet was prepared using 45% (395 g) of the standard diet (powder), added to 45% (395 g) of condensed milk and 10% (83.79 g) of refined sugar, mixed until it forms a homogeneous mass to make small pellets. The macronutrient composition of the standard diet (4.0 kcal/g) was 65.8% carbohydrate, 3.1% fat, and 31.1% protein, obtained from the manufacturer’s information, while the macronutrient composition of the HC diet (4.4 kcal/g) was 74.2% carbohydrate, 5.8% fat, and 20% protein, obtained from the nutritional analysis carried out by Oliveira et al. (2013).

Animals were weighed once a week and the food intake was measured twice a week. Samples of epididymal, retroperitoneal, and mesenteric adipose tissues were weighed to evaluate the adiposity index, according to the equation below (Oliveira et al., 2013).

Animals were euthanized by decapitation and the thoracic aorta was carefully isolated and sectioned into two rings with 3 mm length each. In one of the rings the PVAT was completely removed while the other was kept intact. The aortic rings were placed between two stainless-steel stirrups and connected to an isometric tension transducer (World Precision Instruments, Inc., Sarasota, FL, United States). The vessels were placed in organ chambers containing modified Krebs–Henseleit physiological solution (mmol/L: NaCl 135.0; KCl 5.0; KH₂PO₄ 1.17; CaCl₂ 2.5; MgSO₄ 1.4; NaHCO₃ 20.0; glucose 11.0) at 37°C with a stable pH 7.4 and gassed with carbogenic mixture (95% O₂ and 5% CO₂) (White Martins, Brazil). After 1 h of stabilization at a basal tension of 4.9 mN (0.5 g), the vessels were stimulated with potassium chloride (9 × 10^–2 mol/L) in order to determine its viability. Subsequently, the aortas were previously contracted with phenylephrine (EC₅₀ PE: 10^–7 mol/L) and the presence of a functional endothelium was verified by the addition of acetylcholine (EC₅₀ ACh: 10^–6 mol/L). The endothelial integrity was considered in a minimum of 80% relaxation for acetylcholine. To assess the effect of HC diet on endothelium-dependent vasodilation, cumulative concentration-response curves for acetylcholine (ACh 10^–10–10^–4 mol/L) were obtained in aortas previously contracted with phenylephrine (EC₅₀ PE: 10^–7 mol/L) in the presence or absence of PVAT. Acetylcholine-induced vasodilation was expressed in percentage of relaxation. The effect of HC diet on vascular contractility was assessed in cumulative concentration-response curves for phenylephrine (PE 10^–10–10^–4 mol/L) obtained in endothelium-intact aortas in the presence or absence of PVAT. Phenylephrine-induced vasoconstriction was expressed in mN.

In order to investigate the mechanisms underlying the effects of HC diet on the PVAT control of vascular tone, cumulative concentration-response curves for phenylephrine (PE 10^–10 –10^–4 mol/L) were only performed in the presence of PVAT previously incubated for 30 min with one of the following drugs: captopril (10^–5 mol/L - angiotensin converting enzyme inhibitor) (Kikta and Fregly, 1982; Su et al., 2008), A779 (10^–6 mol/L – selective Mas receptor antagonist) (Peiró et al., 2013), PD123,319 (10^–6 mol/L – selective AT₂ receptor antagonist) (Su et al., 2008), LY294,002 (10^–6 mol/L – PI3K inhibitor) (Jimenez et al., 2010), L-NAME (10^–4 mol/L - non-selective NOS inhibitor) (Araújo et al., 2012), L-NNA (10^–6 mol/L – selective eNOS inhibitor) (Araújo et al., 2012; Gonzaga et al., 2018; Nobrega et al., 2019), 1,400 W (10^–5 mol/L - selective iNOS inhibitor) (Garvey et al., 1997), 7Ni (10^–4 mol/L - selective nNOS inhibitor) (Babbedge et al., 1993), catalase (300 U/mL – catalyzes the decomposition of hydrogen peroxide) (Gonzaga et al., 2018) or tetraethylammonium (TEA, 10^–3 mol/L – non-selective blocker of potassium channels) (Bonaventura et al., 2011). All concentrations of drugs were based on previous studies abovementioned. Agonist potencies and maximal responses were analyzed and expressed as pD₂ (−log EC₅₀) and Emax (maximum effect elicited by the agonist), respectively.

Thoracic aortas were fixed in phosphate buffered formaldehyde solution for 48 h and then dehydrated in ascending concentrations (70, 80, and 90% and absolute I, II, and III) of ethyl alcohol, followed by diaphanization in xylol I, II, and III, and embedded in paraffin. 5 μm transversal sections were stained by hematoxylin-eosin for morphological analysis or picrosirius for quantification of collagen fibers. The area of the middle layer and the PVAT were quantified surrounding the region occupied by the middle layer in the thoracic aorta or the fractions of adipose tissue located around the adventitia, respectively, in a Leica microscope coupled to a Quantimet 500 image analysis system (Leica, Bannockburn, IL) using a × 5 lens magnification under a common light. Picrosirius-stained sections were examined in the same image analysis system aforementioned using a × 20 lens magnification. The area occupied by collagen was quantified by the color-detecting mode of the computer program in the adventitia. The aspect of collagen fibers was evaluated under a polarized light, allowing evaluation of the molecular disposition of collagen fibers (Borges et al., 2007; de Figueiredo Borges et al., 2008).

Frozen thoracic aortas of control and HC groups were serially cut in 10 μm transversal sections, fixed in cold 100% acetone and washed with phosphate buffered saline (PBS). The fixed cryosections were rinsed in wash buffer (4% BSA + 0.1%Triton X-100, in PBS). Following appropriate blocking procedures (3% BSA in PBS), the slides were incubated overnight at 4°C with rabbit monoclonal anti-Mas (Alomone Labs Cat# AAR-013, RRID:AB_2039972), rabbit anti-AT₂ (Alomone Labs Cat# AAR-012-AG, RRID:AB_2039724), mouse anti-eNOS (Santa Cruz Biotechnology Cat# sc-136977, RRID:AB_2267282), mouse anti-iNOS (Santa Cruz Biotechnology Cat# sc-7271, RRID:AB_627810), mouse anti-nNOS (Santa Cruz Biotechnology Cat# sc-5302, RRID:AB_626757), followed by incubation with goat anti-mouse secondary antibody conjugated with Alexa Fluor 488 (Santa Cruz Biotechnology Cat# sc-362257, RRID:AB_10989084) and goat anti-rabbit secondary antibody conjugated with Alexa Fluor 594 (Thermo Fisher Scientific Cat# A-11037, RRID:AB_2534095). The sections were examined under Nikon Eclipse Ti microscope (Nikon, United States) with excitation at 488/594 nm and emission at 520/600 nm. The fluorescence intensity emitted was measured in different fields with the same area and analysis parameters only in the PVAT of the control and HC groups using ImageJ^® software (NIH, Bethesda, MD, United States) and expressed as fold increase (Navia-Pelaez et al., 2017).

Fluorescent probes 4-amino-5-methylamino-2′,7′-difluorescein diacetate (DAF-2DA) and 2′,7′-dichlorodihydrofluorescein diacetate (DCF-DA) were used to measure NO and H₂O₂ in situ, respectively, in the PVAT of the control and HC groups. For that, thoracic aortas with intact PVAT of both experimental groups were embedded in freezing medium (Tissue-Tek, Sakura Finetek, Torrance, CA, United States). Transversal sections (10 μm thick) of frozen thoracic aortas were incubated with DAF-2DA (2.5 μmol/L) or DCF-DA (2.5 μmol/L) at 37°C, protected from light. The images were captured on a Zeiss Axio Imager A2 fluorescence microscope where DAF-2DA was excited at 488/519 nm, and DCF-DA was excited at a 590/618 nm. The fluorescence intensity emitted was measured in different fields with the same area and analysis parameters only in the PVAT of the control and HC groups using ImageJ^® software (NIH, Bethesda, MD, United States) and expressed as fold increase (Campos-Mota et al., 2017).

Graphs and analysis were blinded performed in the GraphPad Prism version 8 (GraphPad Software, San Diego, CA, United States). Determinations of EC₅₀ and Emax were performed using the non-linear regression method of least squares (Meddings et al., 1989). The concentration values that produced half maximal contraction amplitude, which was determined after log transformation of the normalized concentration–response curves, were reported as negative logarithm (pD₂). The Emax values were considered as the maximal amplitude response reached in the concentration-response curves. Results were presented as standard mean ± error (SEM). After checking adherence to the normal distribution, statistical significance was determined using Student’s t-test for two group’s comparison or two-way analysis of variance (ANOVA) for multiple group comparisons as appropriate, followed by Holm–Sidak’s post hoc test. A value of p < 0.05 was considered statistically significant.

The drugs phenylephrine (PE), captopril, A779, PD123,319, LY-294,002, Nω-Nitro-L-arginine methyl ester hydrochloride (L-NAME), Nω-Nitro-L-arginine (L-NNA), 1,400 W, 7-Nitroindazole (7-Ni), catalase and tetraethylammonium (TEA) were purchased from Sigma-Aldrich (St. Louis, MO, United States). DAF-2DA and DCF-DA fluorescent probes were obtained from Invitrogen (Carlsbad, CA, United States).

Despite no change in cumulative food intake (Figure 1A) and body weight (Figure 1B and Table 1), mice fed HC diet exhibited considerable increase in visceral adiposity (Figure 1C and Table 1).

To verify whether HC diet induced an increase in the PVAT area, morphological analysis was performed. We verified in Figures 2A,B,D a significant increase in the area occupied by PVAT in the HC group when compared to the control group. No significant changes were verified in the middle layer area between the groups (Figures 2A–C).

Figure 2E shows the normal distribution of collagen fibers under the incidence of normal polychromatic light in the aorta of the control group. When evaluated under the incidence of polarized light, which allows analyzing the arrangement of the collagen fibers, Figure 2G shows that the collagen fibers were highly organized, exhibiting reddish color. Note the abundance of these fibers in adventitious layer (arrow). The same distribution and organization of the collagen fibers were observed in the aorta of the HC group (Figures 2F,H). Therefore, the HC diet did not induce vascular fibrosis, since no increase in the percentage of collagen fibers in the aorta of the HC group was found when compared to the control group (Figure 2I).

The endothelium-dependent vasodilation induced by ACh in the control group was similar in the presence or absence of PVAT (Figure 3A). The same result was found in the HC group (Figure 3B). The overlapping curves showed that the HC diet did not induce endothelial dysfunction when compared to the control group (Figure 3C). The Emax and pD₂ values can be visualized in Table 2.

In Figure 4A, the presence of PVAT did not alter the vasoconstrictor response induced by PE in aortas of the control group. However, after HC diet for 4 weeks, the presence of PVAT significantly attenuated PE-induced vasoconstriction (Figure 4B). The overlapping curves showed that, in the absence of PVAT, the vasoconstriction induced by PE was similar between both groups. Only in the presence of PVAT the contractile response induced by PE was significantly reduced in the HC group (Figure 4C).

Once the HC diet attenuated the vascular contraction induced by PE only in the presence of PVAT, the next experiments were performed in aortas with intact PVAT in order to identify the mechanisms underlying the anticontractile effect induced by HC diet. All the Emax and pD₂ values can be visualized in Table 3.

As shown in Figure 5A, the anticontractile effect of PVAT induced by HC diet was lost when aortic rings were previously incubated with the angiotensin converting enzyme (ACE) inhibitor, captopril. In addition, we verified the involvement of RAS receptors in this anticontractile effect of PVAT. The antagonism of Mas (Figure 5B) and AT₂ (Figure 5C) receptors with A779 and PD123,319, respectively, also reestablished the contractile response induced by PE in the HC group to the level found in the control group.

Since the activation of Mas and AT₂ receptors was associated with the anticontractile effect of PVAT induced by HC diet, we further investigated if Mas and AT₂ receptors were expressed in the PVAT of control and HC groups. Immunofluorescence assays allowed to evaluate whether PVAT express the antigen of Mas and AT₂ receptors. The results demonstrated the presence of Mas (Figures 6B,E) and AT₂ (Figures 6I,L) receptors in the PVAT of animals fed standard or HC diet. However, the fluorescence intensity was markedly higher in the PVAT of the HC group when compared to the control group as shown in Figures 6G,N.

As the activation of Mas and AT₂ receptors can trigger the intracellular signaling cascade that activates PI3K-Akt pathway, we verified whether this pathway participates in the effect of HC diet on the control of vascular tone induced by PVAT. The inhibition of PI3K with LY294,002 reestablished the contractile response induced by PE in the HC group (Figure 7A). Knowing that the activation of PI3K-Akt pathway can lead to NOS phosphorylation, we evaluated the involvement of NOS in the anticontractile effect of PVAT induced by HC diet. As shown in Figure 7B, the non-selective inhibition of NOS with L-NAME reestablished the contractile response induced by PE in the HC group. Also, we verified in Figures 7C–E that basal NO availability was significantly higher in the PVAT from HC group when compared to the control group.

Once the involvement of NOS was confirmed, as well as an increase in basal levels of NO, we further investigated which NOS isoforms would be related to the anticontractile effect of PVAT induced by HC diet. The inhibition of eNOS with L-NNA was not able to reverse the contractile response induced by PE in the HC group (Figure 8A). The iNOS inhibition with 1,400 W reestablished the Emax of contractile response induced by PE in the HC group, but the difference in pD₂ values between both groups implies that the reversal of contractile response was only partial (Figure 8B and Table 3). However, the nNOS inhibition with 7Ni completely reestablished the contractile response induced by PE in the HC group (Figure 8C).

To verify whether or not PVAT express the antigen of NOS isoforms in the PVAT of control and HC groups, immunofluorescence assays were performed and revealed the presence of eNOS (Figures 9B,E), iNOS (Figures 9I,L), and nNOS (Figures 9P,S) in the PVAT of animals fed standard or HC diet. As shown in Figure 9G, the fluorescence intensity for eNOS was similar between both groups. However, the fluorescence intensity for iNOS (Figure 9N) and nNOS (Figure 9U) were markedly higher in the PVAT of the HC group. These findings were in agreement with the results found in vascular reactivity experiments, since only iNOS and nNOS isorforms were involved in the anticontractile effect of PVAT induced by HC diet.

Once we found that nNOS is the main isoform involved in the anticontractile effect of PVAT induced by HC diet, and as this isoform not only produces NO but also H₂O₂, we verified the involvement of H₂O₂, another potent vasodilator factor. The degradation of H₂O₂ with catalase reestablished the contractile response induced by PE in HC group (Figure 10A). As shown in Figures 10B–D, basal H₂O₂ availability was significantly higher in the PVAT from HC group when compared to the control group.

Also, to investigate if the vasodilator response induced by NO and H₂O₂ in the HC group involves hyperpolarization through the opening of potassium channels, aortic rings were previously incubated with TEA. The non-selective blockade of potassium channels with TEA reestablished the Emax of contractile response induced by PE in the HC group, but the difference in pD₂ values between both groups implies that the reversal of contractile response was only partial (Figure 10E and Table 3).