E ndothelial cells play a central role in maintaining vascular

1234  Hypertension  June 

2013

primarily catalyzes the conversion of Ang II into Ang-(1–7),

10

 

thereby contributing to the balance between the 2 peptides 

and, consequently, it is a key modulator of the 2 axes of the 

renin–angiotensin system.

6

The role of the ACE/Ang II/AT

1

 axis in the development 

and progression of the endothelial dysfunction is well recog-

nized, especially in terms of ROS production by endothelial 

and vascular smooth muscle cells.

3,4,7

 Treatment with free 

radical scavengers, such as superoxide dismutase, catalase, 

and tempol, reduces blood pressure and vascular damage in 

response to Ang II.

12,13

 Moreover, it was reported that activa-

tion of the ACE2/Ang-(1–7)/Mas axis ameliorates the endo-

thelial function in many animal models.

14

 Indeed, short-term 

infusion of Ang-(1–7) improved endothelial response to ace-

tylcholine,

15

 and Mas deficiency caused endothelial dysfunc-

tion and increases in blood pressure associated with elevation 

of the ROS production.

16

Altogether, these findings led us to postulate that activa-

tion of intrinsic ACE2 would improve endothelial dysfunction 

by decreasing the production of ROS. To test this hypothesis, 

we evaluated the effects of 1-[[2-(dimetilamino)etil]amino]-

4-(hidroximetil)-7-[[(4-metilfenil)sulfonil]oxi]-9H-xantona-9 

(XNT), a small molecule that has been reported to activate 

intrinsic ACE2,

17

 on endothelial dysfunction. XNT was dis-

covered on the basis of crystal structure of ACE2 using a vir-

tual screening strategy.

17

 Administration of this compound in 

spontaneously hypertensive rats (SHRs) decreased blood pres-

sure, improved cardiac function, and reversed myocardial and 

perivascular fibrosis through a mechanism involving reduc-

tions in extracellular signal–regulated kinases expression.

17,18

 

Moreover, XNT reduced pulmonary hypertension induced by 

monocrotaline,

19

 attenuated thrombus formation and platelet 

attachment to vessels of hypertensive rats,

20

 and ameliorated 

cardiac and autonomic function of diabetic rats.

21,22

Methods

The procedures used for the measurement of ACE2 activity, iso-

lated aortic ring preparation, radioimmunoassay, Western blotting, 

and immunohistochemistry are described in the online-only Data 

Supplement.

Animals

Male Sprague-Dawley rats and SHRs (12 weeks of age) were pur-

chased from Charles River Laboratories (Wilmington, MA). Male 

Wistar rats (12 weeks of age) were obtained from the CEBIO-Federal 

University of Minas Gerais (Belo Horizonte, MG, Brazil). Male wild-

type (Mas

+/+

) and Mas knockout (Mas

−/−

) mice (FVB/N background, 

12 weeks of age) were bred at the transgenic animal facility of the 

Laboratory of Hypertension, Federal University of Minas Gerais 

(Belo Horizonte, MG, Brazil). All animals were kept in tempera-

ture-controlled rooms with 12/12-hour light/dark cycle and had free 

access to water and food. All experimental protocols were performed 

in accordance with the University of Florida (Gainesville, FL) and 

the Federal University of Minas Gerais (Brazil) Institutional Animal 

Care and Use Committees, which are in compliance with the National 

Institutes of Health guidelines.

Statistical Analysis

The results are presented as mean±SEM. Two-way ANOVA with 

Bonferroni multiple comparison post test was used to compare the 

curves obtained in the ACE2 activity and aortic ring preparation pro-

tocols. In addition, 1-way ANOVA followed by the Bonferroni post 

test was used to analyze the Western blotting, immunohistochemistry, 

and ROS production data, and Student t test was used to analyze the 

radioimmunoassay results. All statistical analyses were considered 

significant when P<0.05.

Results

XNT Improves Endothelial Function of SHRs  

and Diabetic Rats

In accordance with previous studies by Hernández Prada  

et al,

17

 we observed that incubation of rhACE2 with XNT 

in vitro increased the activity of ACE2, thereby confirming 

the ability of XNT to activate this enzyme (Figure S1 in the 

online-only Data Supplement). Similar data were observed 

when aortic samples from normal and diabetic animals were 

incubated with the fluorogenic substrate (Figure S2). As a 

consequence, ACE2 activation significantly increased the con-

centration of Ang-(1–7) in plasma of diabetic animals, but not 

in aorta (Figure S3). Although 

≈30% of decrease in plasma 

Ang II levels was observed in diabetic-treated rats, it did not 

reach statistical significance (Figure S4). Importantly, the 

ACE2 activity in diabetic rats was significantly lower when 

compared with normal animals (Figure S2). Treatment with 

XNT was unable to change the ACE2 protein expression in 

diabetic rats (Figures S5 and S6).

To examine the effects of chronic XNT treatment on the 

endothelial function, the vasorelaxant responses to ACh (ace-

tylcholine) and SNP (sodium nitroprusside) were evaluated in 

aortic rings from hypertensive and diabetic rats. The vasodilatory 

responses to ACh were markedly enhanced in both SHRs (Figure 

1A) and diabetic Wistar rats (Figure 1B) treated with XNT. In 

contrast, the endothelial-independent responses to SNP were not 

affected by XNT when compared with vessels from untreated 

SHRs (Figure 1C) and diabetic rats (Figure 1D). These results 

showed that the endothelium-dependent vascular responses were 

improved by ACE2 activation in SHRs and diabetic rats.

XNT Produces Vasorelaxant Responses Associated 

With Mas Activation

XNT caused concentration-dependent vasorelaxation in aor-

tic rings of Sprague-Dawley rats preconstricted with phenyl-

ephrine (Figure 2A). Using the submaximal concentration of 

10 

μmol/L, XNT promoted a time-dependent vasorelaxation 

with maximal effect reached after 7 minutes (Figure 2B). To 

evaluate the participation of the endothelium in the vasorelax-

ant effects of XNT, aortic rings of rats with or without intact 

endothelium were incubated with this compound. It was found 

that the XNT effects were dependent on the endothelial cells 

(Figure 2C).

To address the mechanism by which XNT produces 

vasorelaxation, the vessels were preincubated with 2 different 

Mas antagonists. It was observed that the vasorelaxant effect of 

XNT was attenuated by D-pro7-Ang-(1–7) (10 

μmol/L; Figure 

2D). Interestingly, preincubation with A-779 (10 

μmol/L), 

a classical Mas antagonist, did not change its vasorelaxant 

activity (Figure 2E). On the basis of possible participation of 

Mas in the vasorelaxant effects of XNT, we further evaluated 

the actions of this compound in vessels of Mas

−/−

 mice. We 

found that XNT at 10 

μmol/L caused similar vasorelaxation 

in Mas

−/−

 and Mas

+/+

 mice during the first 5 minutes of 

incubation. However, after this initial period the XNT effect 

 by guest on November 6, 2017

http://hyper.ahajournals.org/

Downloaded from 

---