E ndothelial cells play a central role in maintaining vascular

Fraga-Silva et al    ACE2 Activation and Endothelial Function    1235

was absent in Mas

−/−

 mice when compared with Mas

+/+

 mice 

(Figure 2F). The initial vasodilatory effect observed in Mas

−/−

 

mice was not associated with Ang II degradation by ACE2 

because incubation with losartan did not affect the vasodilator 

response of XNT (Figures 2G and 2H).

XNT Attenuates Oxidative Stress

Consistent evidences indicate that ROS play an important 

role in the development of endothelial dysfunction. Thus, we 

investigated the participation of the oxidative stress in the 

effects of XNT on the endothelial function of diabetic Wistar 

rats. Specifically, we evaluated the ROS production in aor-

tic vessels of diabetic animals. Diabetes mellitus caused an 

increase in the generation of ROS, which was significantly 

reduced by XNT treatment (Figure 3). No significant changes 

were observed in the expression of catalase, superoxide dis-

mutase, and NOX2 in aorta of diabetic rats treated or not with 

XNT (Figure S7).

In addition, we tested whether XNT is able to reduce the 

Ang II–inducing ROS production in human aortic endothelial 

cells. It was observed that XNT treatment attenuated the ROS 

production stimulated by Ang II. XNT alone did not change 

the basal level of ROS (Figure S8).

Discussion

The beneficial role of ACE2 in the pathophysiology of cardio-

vascular and metabolic diseases is currently under intensive 

investigation. In fact, the involvement of this enzyme in car-

diac function, hypertension, atherosclerosis, and other cardio-

vascular diseases has recently been demonstrated.

6,17–20,23–27

 The 

most significant finding of the present study is that pharmaco-

logical ACE2 activation using XNT exerts protective effects on 

endothelial function. Furthermore, we showed that this action 

involves the Mas receptor and reduction of ROS production.

As the main ACE2 enzymatic function is degrading Ang 

II with consequent production of Ang-(1–7), the likely 

mechanism underlying the XNT effects is balancing the 

bioavailability of these 2 peptides. Indeed, we observed 

that the Mas antagonist D-pro7-Ang-(1–7) attenuated the 

vasorelaxant response elicited by XNT, thereby indicating the 

involvement of Ang-(1–7)/Mas in this effect. Unexpectedly, 

A-779, a classical Mas antagonist, did not affect this action. In 

spite of these apparent contradictory results, these findings are 

in keeping with previous studies showing that the vasorelaxant 

effect of Ang-(1–7) in aortic rings of Sprague-Dawley rats 

was blocked by D-pro7-Ang-(1–7) but not by A-779.

28

 

Nowadays, there is only 1 Ang-(1–7) receptor identified (ie, 

Mas receptor).

8

 Usually, this receptor is blocked by A-779 

and D-pro7-Ang-(1–7).

29

 However, as mentioned above, in 

certain situations one of these antagonists is not efficient or 

is only partially effective in blocking the Ang-(1–7) effects.

28

 

This strongly suggests the existence of other unidentified 

Ang-(1–7) receptors. To further evaluate the role of Ang-(1–

7)/Mas in the effects of XNT, we measured the plasma and 

aortic Ang-(1–7) levels and tested the XNT effects in isolated 

aortic rings of Mas-deficient mice. It was observed that ACE2 

activation significantly increased the concentration of this 

peptide in the plasma. The results obtained in Mas

−/−

 mice 

showed that, during the first 5 minutes of incubation, XNT 

induced vasorelaxation in Mas

−/−

 and Mas

+/+

 mice. However, 

after 10 minutes the effects of XNT were absent in Mas

−/−

, 

indicating that Mas is involved in the vascular response of 

XNT. One may suggest that the initial vasodilatory effect 

observed in Mas

−/−

 mice is caused by degradation of Ang II by 

ACE2. Nevertheless, this hypothesis is not plausible because 

incubation of XNT associated with losartan did not block this 

vasodilatory response. Therefore, further experiments are 

required to explain this observation. Altogether, these findings 

Figure 1. 1-[[2-(Dimetilamino)etil]amino]-4-

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

xantona-9 (XNT) improves endothelial function 

of hypertensive and diabetic rats. Vasodilation 

produced by increasing cumulative concentrations 

of acetylcholine (

A) and sodium nitroprusside 

(

C) in aortic rings of XNT-treated or untreated 

hypertensive rats. Vasodilatory effects induced 

by increasing cumulative concentrations of 

acetylcholine (

B) or sodium nitroprusside (D) in 

aortic rings of XNT-treated or untreated diabetic 

rats. *P<0.05 (2-way ANOVA followed by the 

Bonferroni multiple comparison test). Each point 

represents the mean±SEM (n=7–10). n.s. indicates 

nonsignificant.

 by guest on November 6, 2017

http://hyper.ahajournals.org/

Downloaded from 

---

1236  Hypertension  June 

2013

indicated that the vasorelaxant effects of XNT are dependent 

on Ang-(1–7)/Mas axis.

The partial blockade (

≈50%) of the XNT actions by 

D-pro7-Ang-(1–7) was an expected finding because ACE2 

is an enzyme with a dual role within the renin–angiotensin 

system (ie, it degrades Ang II with consequent production of 

Ang-(1–7)). Thus, the residual effect of XNT in the presence 

of D-pro7-Ang-(1–7) might be caused by the reduction of the 

Ang II content. However, we did not observe any significant 

decrease in plasma Ang II levels in diabetic rats treated with 

XNT. Furthermore, increases in ACE2 protein and mRNA 

expression are a frequent finding when using ACE2 activa-

tors.

18

 This suggests that these compounds not only induce 

their beneficial effects by forming Ang-(1–7) and degrad-

ing Ang II, but a nonidentified mechanism is also present. 

Nevertheless, in our study, XNT was unable to increase the 

ACE2 protein expression maybe because of the duration of the 

experimental protocol. Thus, further investigations are neces-

sary to clearly identify the mechanisms of action of XNT.

The ROS are intracellular and intercellular second 

messengers that modulate the endothelial function.

30

 Under 

pathological conditions, elevation of the ROS content, the 

so-called oxidative stress, results in vascular dysfunction.

31

 

In fact, consistent evidences indicate that ROS play an 

important role in the development of organ damage in 

diabetes mellitus and hypertension.

30,32

 Ang II elicits many of 

its pathophysiological effects by stimulating ROS generation 

through the reduction of the nicotinamide adenine dinucleotide 

phosphate oxidase activity. Moreover, treatment with free 

radical scavengers, such as superoxide dismutase, catalase, 

Figure 2. 1-[[2-(Dimetilamino)etil]amino]-4-

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

xantona-9 (XNT) produces Mas-mediated 

vasodilation. Vasodilation produced by increasing 

cumulative concentrations of XNT (

A) and time 

curve of the submaximal concentration (10 

μmol/L) of XNT (B). The vasodilatory effects of 

XNT were dependent on the intact endothelium 

(

C). The effects of XNT (10 μmol/L) were partially 

blocked by D-pro7-Ang-(1–7) (

D) but not by 

A-779 (

E). Also, the vasorelaxant effect of XNT (10 

μmol/L) was absent in aortic rings of Mas

−/−

 mice 

after 5 minutes of incubation (

F). The AT

1

 receptor 

antagonist losartan did not affect the response of 

XNT in Mas

−/−

 (

G) and Mas

+/+

 (

H) mice. **P<0.01 

and ***P<0.001 (2-way ANOVA followed by the 

Bonferroni multiple comparison test). Each point 

represents the mean±SEM (n=8–10). n.s. indicates 

nonsignificant.

 by guest on November 6, 2017

http://hyper.ahajournals.org/

Downloaded from 

---