Stenosis coexists with compromised α1-adrenergic contractions in the ascending aorta of a mouse model of Williams-Beuren syndrome

Williams-Beuren syndrome (WBS) is a rare disorder caused by a heterozygous deletion of 26–28 contiguous genes that affects the brain and cardiovascular system. Here, we investigated whether WBS affects aortic structure and function in the complete deletion (CD) mouse model harbouring the most common deletion found in WBS patients. Thoracic aortas from 3–4 months-old male CD mice and wild-type littermates were mounted in wire myographs or were processed for histomorphometrical analysis. Nitric oxide synthase (NOS) isoforms and oxidative stress levels were assessed. Ascending aortas from young adult CD mice showed moderate (50%) luminal stenosis, whereas endothelial function and oxidative stress were comparable to wild-type. CD mice showed greater contractions to KCl. However, α1-adrenergic contractions to phenylephrine, but not with a thromboxane analogue, were compromised. Decreased phenylephrine responses were not affected by selective inducible NOS blockade with 1400 W, but were prevented by the non-selective NOS inhibitor L-NAME and the selective neuronal NOS inhibitor SMTC. Consistently, CD mice showed increased neuronal NOS expression in aortas. Overall, aortic stenosis in CD mice coexists with excessive nNOS-derived NO signaling that compromises ascending aorta α1-adrenergic contractions. We suggest that increased neuronal NOS signaling may act as a physiological ‘brake’ against the detrimental effects of stenosis.

mouse models carrying chromosome microdeletions affecting the WBS critical region have been generated to mimic the molecular defects present in patients 16,17 . Mice carrying a heterozygous distal deletion (DD) (from Limk1 to Trim50) are hypertensive, show increased aortic oxidative stress levels, cardiac hypertrophy, decreased aortic compliance, and histological alterations characterized by aortic wall thickening and fragmentation of elastin lamellar units 18,19 . However mice carrying a complete deletion (CD) (from Gtf2i to Fkbp6) show mild hypertension and cardiac hypertrophy at 8 months of age 17 . Notably, aortic function has never been explored in the ascending aorta of either CD mice or other mouse models carrying WBS chromosome microdeletions. The identification of early aortic dysfunction in WBS may help predict and prevent disease-related cardiovascular problems.
Previous evidence suggests a role of aortic valve endothelium in the maintenance of valvular homeostasis 20 . The endothelium releases a number of vasoactive substances including nitric oxide (NO), which is a signaling molecule that has many beneficial and sometimes detrimental effects on cardiovascular function 21,22 . NO plays an important role in the decreased ability of a vessel to respond to persistent stimulation by α1-adrenoceptor agonists, a process called 'NO-mediated desensitization' 23 . Moreover, in aorta, a modulatory role of neuronal NOS (nNOS) associated to activation of the α1 A -adrenoceptor subtype has been proposed as a physiological 'brake' against the detrimental effects of excessive adrenergic vasoconstriction 24 . Remarkably, an impairment of NO responsiveness is reported in patients with aortic stenosis [25][26][27] , and a proportion of WBS patients manifest higher levels of oxidative stress 28 ; effects that might be associated with aortic dysfunction. Deep understanding of development of functional aortic anomalies in animal models of WBS and exploration of the mechanisms potentially involved would identify proper targets for human therapy.
In the present study, we provide the first comprehensive assessment of thoracic aorta structure and reactivity in young CD mice, a mouse model carrying the most common deletion found in WBS patients, which often manifest cardiovascular disease early in life. We conclude that ascending aortas from young CD mice show stenosis and impaired phenylephrine-induced α1-adrenergic contractions because of a NO-mediated desensitization.

Methods
Animals. Three-to four-months old male heterozygous CD (n = 25) mice and age-matched wild-type (WT, n = 26) littermates were used. Genotyping was made as previously reported 29  Blood pressure measurements in conscious mice. Measurement of systolic blood pressure was performed in conscious WT and CD mice using the tail-cuff method (NIPREM 645; Cibertec, Madrid, Spain). The average systolic blood pressure of each mouse was determined from six consecutive measurements after habituation, as described 30 . Tissue preparation. Segments (Fig. 1A) of the ascending aorta (histomorphometry, reactivity, elastin autofluorescence, Western blot, and qRT-PCR), aortic arch (immunofluorescence and dihydroethidium), and the descending aorta (histomorphometry) were dissected free of fat and connective tissue and placed in ice-cold Krebs Henseleit solution gassed with a 95% O 2 -5% CO 2 mixture. Aortic segments used for immunofluorescence studies were fixed with 4% phosphate buffered paraformaldehyde for 1 h, washed in three changes of phosphate buffered saline solution, and processed as described 31 . Plasma (200 μl) was collected after decapitation and was kept at −70 °C until the day of analysis of circulating superoxide anion levels by high-performance liquid chromatography (HPLC).
Ex vivo gross examination of the proximal thoracic aorta. After dissection, images of the thoracic aorta were obtained ex vivo using a dissecting microscope (Leica, Wetzlar, Germany). The length of the proximal thoracic aorta (i.e. ascending aorta and aortic arch) was measured along its medial curvature from the ventricular-aortic junction to the distal aortic arch that finishes when the inner and outer curvature become parallel. Aortic length was measured from calibrated digital images using ImageJ 1.51j8 (National Institutes of Health, Bethesda, MD, USA) software.
Measurement of elastin autofluorescence and number of elastin laminae. Total elastin content was studied in aortic cross-sections (14 μm-thick) based on the autofluorescent properties of elastin, as described 32 . Values of fluorescence intensity were estimated as a measure of elastin concentration, following the assumption that the concentration of elastin has a linear relationship with fluorescence intensity 33 . All of the images were taken using a laser-scanning confocal microscope (×20 objective; Leica TCS SP5, Manheim, Germany) under identical conditions of zoom (×1), laser intensity, brightness, and contrast. Quantitative analysis of elastin autofluorescence and number of elastin laminae was performed with ImageJ 1.51j8 software. The average intensity of fluorescence signal (expressed as arbitrary units) and the number of elastin laminae were measured in at least three rings from each animal. RNA preparation and gene expression quantification by qRT-PCR. RNA was extracted from ascending aortas using RNeasy Micro Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. cDNA was prepared from 1 µg of total RNA using random hexamers and SuperScript II RNase H reverse transcriptase (Invitrogen). The expression of Adra1a (α1 A -adrenoceptors), Limk1 (a gene contained in the WBS commonly deleted region) and Rps28 (internal control) was evaluated by quantitative PCR (qRT-PCR), as described 17 , using the appropriate primers (Supplementary Table S1). Each PCR was made with triplicates from two different RTs. The expression values were relativized according to the average expression of the WT animals for each gene.

Analysis of circulating 2-hydroxyethidium (2-EOH).
Plasma levels of 2-EOH (Sigma-Aldrich, St. Louis, MO, USA) were assessed by HPLC with fluorescence detection, as a quantitative measure of plasma superoxide anion levels, as described [37][38][39] . 2-EOH present in the samples was quantified by comparing with a calibration curve based on the reaction xanthine-xanthine oxidase from the method described by Michalski and cols 40 . www.nature.com/scientificreports www.nature.com/scientificreports/ Measurement of aortic oxidative stress. The oxidative fluorescent dye dihydroethidium (DHE; Sigma-Aldrich) was used to evaluate in situ production of superoxide anion in non-fixed 14 μm-thick aortic sections, as described 41 . Quantitative analysis of DHE-derived fluorescence in images obtained using a laser-scanning confocal microscope (×20 objective; Olympus FluoView 1000; Olympus, Shinjuku, Tokio, Japan) was performed with ImageJ 1.51j8 software. At least three rings from each animal were measured and the results were expressed as arbitrary units.
Aortic reactivity. Segments (2 mm) of the ascending thoracic aorta were set up on an isometric wire myograph (model 410 A; Danish Myo Technology, Aarhus, Denmark) filled with KHS (37 °C; 95% O 2 and 5% CO 2 ), as described 42 . Optimal tension was assessed in preliminary experiments by subjecting arterial segments to different resting tensions and challenging with 100 mM KCl 41,43 . The optimal tension of the ascending aorta was the same for WT and CD mice (5 mN). Therefore, the vessels were stretched to 5 mN, washed and allowed to equilibrate for 45 min. The tissues were contracted twice (10-min interval) with 100 mM KCl. After washing, vessels were left to equilibrate for a further 30 min before starting the experiments. Endothelial-dependent vasodilatations to acetylcholine (ACh; 10 −9 to 10 −5 M) were performed in phenylephrine-precontracted vessels to induce 70-100% of 100 mM KCl contraction, and thus produce a similar level of precontraction in either group. The use of thromboxane A 2 mimetic 9,11-Dideoxy-9α,11α-methanoepoxyprostaglandin F 2α (U46619) to contract the vessels was discarded, because of U46619-induced precontractions greatly reduced ACh-induced vasodilatations. Agonist-induced contractile responses were studied by evaluating stimulation of α1-adrenoceptors with phenylephrine (10 −9 to 3 × 10 −5 M) or thromboxane A 2 receptors with U46619 (10 −9 to 3 × 10 −6 M). The effects of the non-selective nitric oxide synthase (NOS) inhibitor Nω-nitro-l-arginine methyl ester (L-NAME; 3 × 10 −4 M) 41 , the selective iNOS inhibitor 1400 W (10 −5 M) 41 , and the selective nNOS inhibitor SMTC (10 −6 M) 24 were determined by adding each treatment 30 min before phenylephrine-or ACh-induced responses.
Statistical analysis. Data are presented as mean ± SEM of the number (n) of mice shown in figure legends.
Vasodilator responses to ACh were expressed as a percentage of the previous tone generated by phenylephrine. Sigmoid curve fitting (variable slope) was performed by non-linear regression to obtain maximal responses (E max ) and sensitivity (pEC 50 ). Preliminary testing for normality was performed before using Mann-Whitney U test or Student's t-test to compare the mean difference between two groups. Multiple t-test followed by Sidak-Bonferroni test as post hoc was used in the analysis of qRT-PCR data. Differences between concentration-response curves were also assessed by two-way repeated measures ANOVA followed by Bonferroni's test as post hoc. The software GraphPad Prism 8 (GraphPad Software Inc., La Jolla, CA, USA) was used to run statistical analyses. A value of P < 0.05 was considered as statistically significant.

Results
In agreement with a previous report 17 , young adult (3-4 months-old) CD mice showed significantly reduced (P < 0.001) body weight compared to WT mice ( Supplementary Fig. S1A). There were no significant differences in systolic blood pressure between groups (Supplementary Fig. S1B).
Young CD mice exhibited alterations in thoracic aorta morphology. The proximal thoracic aorta, including the ascending aorta and the aortic arch, from CD mice was longer (P < 0.05; Fig. 1B) and showed reduced (P < 0.01) elastin autofluorescence and increased (P < 0.01) number of elastin laminae in the ascending aorta (Fig. 1C). Compared to WT mice, CD mice ascending aortas showed smaller vessel (P < 0.01; results not shown) and lumen (P < 0.01; Fig. 2A,B) areas, whereas wall thickness and CSA were not significantly altered Endothelial function was preserved in aortas from CD mice. There were no differences in endothelium-dependent ACh-induced vasodilatation in the aorta of CD and WT mice ( Fig. 4; Table 1). The non-selective NOS inhibitor L-NAME (3 × 10 −4 M) greatly decreased (P < 0.001) ACh-induced vasodilatation in either group (Fig. 4) indicating that in both genotypes the ACh-induced vasodilatation is mediated by NO signaling. These results suggest that endothelial function is preserved in CD mice.

Contractile responses were altered in aortas from CD mice. Contractile responses to KCl (100 mM)
were significantly higher (P < 0.05) in CD compared to WT aortas (Fig. 5A). Despite enhanced KCl responses, concentration-dependent contractions to phenylephrine were markedly decreased in the aorta of CD compared to WT mice (Fig. 5B), since the E max of CD mice aortas was significantly lower (P < 0.001) than that of WT aortas (Table 1). In contrast, contractions to U46619 were similar between groups ( Fig. 5C; Table 1). These findings suggest that there is a lower aortic vasoconstriction specific to α1-adrenergic-mediated responses.
NO-mediated modulation of phenylephrine contractions was increased in CD mice. In the thoracic aorta, phenylephrine contractile responses are counteracted by NO release, as evidenced by treatment of arterial rings with L-NAME (3 × 10 −4 M), which enhanced (i.e. higher E max ; Fig. 6A, Table 2) the concentration-response curve to phenylephrine in both WT and CD mice. Notably, L-NAME treatment reversed the impairment of phenylephrine contractions observed in CD mice, as in these conditions, no contractile differences were observed between groups (Fig. 6A, Table 2). To elucidate the source of NO involved in contractile differences, we then evaluated the effects of the selective iNOS inhibitor 1400 W (10 −5 M) that had no effect on phenylephrine-induced contractions in either group (Fig. 6B, Table 2). In contrast, treatment with SMTC (10 −6 M), a selective nNOS inhibitor, increased (P < 0.05) the E max of phenylephrine contractions in WT and CD mice, an effect that prevented compromised aortic contractions in CD mice (Fig. 6C, Table 2). Taken together, www.nature.com/scientificreports www.nature.com/scientificreports/ these results suggest that NO-mediated negative feedback underlies impaired phenylephrine contractions in CD mice, an effect that depends on nNOS activation.
Increased levels of α1 A -adrenoceptor and nNOS were found in aortas from CD mice. A modulatory role of nNOS associated to activation of the α1 A -adrenoceptor subtype in aorta has been proposed as a physiological 'brake' against the detrimental effects of excessive adrenergic vasoconstriction 24 . We observed an increase (P < 0.05) in aortic Adra1a (α1 A -adrenoceptors) mRNA levels in CD compared to WT mice (Fig. 7A). On the other hand, aortic expression of nNOS was also increased (P < 0.05) in CD mice (Fig. 7B), and was detected in both endothelial and smooth muscle cell layers (Fig. 7C). The fluorescence signaling was more marked in CD mice, particularly in the endothelium (Fig. 7C). Consistently, quantitative analysis of fluorescence showed greater levels (P < 0.05) of nNOS expression in aortas from CD compared to WT mice (Fig. 7C). However, similar  www.nature.com/scientificreports www.nature.com/scientificreports/ levels of iNOS (equally distributed in all layers of the aortic wall) and eNOS (mainly located in the endothelium layer) were observed in WT and CD mice ( Fig. 7C; Supplementary Fig. S2).

Discussion
Several mice models carrying chromosome microdeletions affecting the WBS critical region have been generated 16,17 . These models are crucial to study potential alterations in vascular responses associated with WBS, a significant question that has not yet been fully addressed. The present study examined ascending aorta structure and reactivity in WT compared to CD mice harbouring the most common deletion found in WBS patients 17,44 .
The results show the presence of luminal stenosis and compromised contractile responses to α1-adrenoceptor activation associated with increased NO signaling in CD mice ascending aortas.
Approximately half of WBS patients develop hypertension 1,45 . This fact has been correlated with higher levels of oxidative stress due to increased expression of Ncf1 19,28 , which encodes the p47 phox subunit of the NADPH oxidase, the major source of superoxide anions in the vasculature 46 . However, a previous study of our group reported mild hypertension in 8-months old CD mice without increased levels of Ncf1 in the heart and the aorta 17 , suggesting that CD mice may develop slight increases in blood pressure later in life. Notably, CD mice showed cardiac hypertrophy 17,44 that was associated with increased levels of oxidative stress in the heart 44 . In the present study,  50 and E max values for phenylephrine (PHE)-and U46619-induced contraction (mN) and acetylcholine (ACh)-induced relaxation (%) in the ascending aorta from wild-type and CD mice. Data are shown as mean ± SE. The number of animals is shown in parentheses. ***P < 0.001 versus wild-type by unpaired Student's t-test. www.nature.com/scientificreports www.nature.com/scientificreports/ young (3-4 months-old) CD mice did not show high systolic blood pressure nor greater levels of circulating or in situ aortic superoxide anions.
The proximal thoracic aorta (i.e. ascending aorta and aortic arch) from CD mice was longer and showed decreased elastin content, which is compatible with the elastin gene haploinsufficiency. In WBS, stenosis can occur diffusely, the most common sites being the thoracic aorta and the pulmonary vascular bed 5 . Supravalvular aortic stenosis is the most frequent cardiovascular anomaly, affecting approximately 70% of patients 1,5 . Aortic stenosis is a serious disease that if left untreated can increase the risk of suffering life-threatening complications, such as stroke 6 and sudden death 7 . However, most experimental models of elastin haploinsufficiency do not show marked aortic luminal obstruction 47 . Notably, hemizygous deletion of the WBS critical region resulted in a moderate (50%) reduction in the ascending aorta lumen area compared to WT mice. Nevertheless, we did not measure histomorphometry in older animals to determine whether luminal obstruction subsequently progresses. Overall, young CD mice reproduce a crucial aspect of the human disorder, which is the presence of aortic stenosis close above the aortic root.
Relevant changes were observed in the functional characteristics of ascending aortas from CD mice. Thus, although endothelial-dependent relaxation to ACh was similar in either group, aortas from CD mice showed an increased contractile response to KCl-induced membrane depolarization together with compromised α1-adrenergic contractions to phenylephrine, despite contractions evoked by the thromboxane A 2 -analogue U46619 were unchanged. Notably, these attenuated α1-adrenergic contractions of WBS aortas were present in spite of higher contractile responses to KCl, which suggests that although the smooth muscle contractile phenotype is able to produce powerful contractions, there is a lower aortic vasoconstriction specific to α1-adrenergic-mediated responses. Given the importance of the α1-adrenergic control of vascular tone, the lower response to phenylephrine could represent an adaptive mechanism to counteract luminal stenosis. Although there is little information about the functioning of the autonomic nervous system in WBS patients, a previous study reported that WBS children show increases in both heart rate and wave reflection values during the night, which may suggest abnormal sympathetic nervous system hyperactivity in WBS 48 . The results of the present study suggest that this hyperactivity might be counteracted by a lower response to aortic α1-adrenoceptor activation. Nevertheless, additional evidence is needed to demonstrate that the results obtained in the present ex vivo study can be translated in vivo.
by the same agonist (NO-mediated desensitization) 23 . In addition, intense or sustained α1-adrenoceptor activity upregulates an endothelium-dependent nNOS pathway that attenuates α1-adrenoceptor-mediated contractions in the rat aorta 24 . This modulatory role of nNOS is associated with activation of the α1 A -adrenoceptor subtype 24 located in endothelial cells 51 . This effect can be physiologically relevant, as it might prevent significant increases in cardiac afterload. Therefore, our next objective was to determine if this mechanism was involved in the lower α1-adrenergic response observed in CD versus WT mice. Firstly, CD mice aortas showed increased aortic α1 A -adrenoceptor expression, the subtype coupled to nNOS, and this change exhibited tissue specificity since it was not observed in the left ventricle (unpublished results). Secondly, our results pointed in the direction of a higher negative influence of NO on α1-adrenoceptor vasoconstriction in WBS, since compromised aortic contractions in CD mice were prevented by the non-selective NOS inhibitor L-NAME. In the vasculature, eNOS wild-type CD Control +L-NAME 3 × 10 -4 M +1400 W 10 -5 M +SMTC 10 -6 M Control +L-NAME 3 × 10 -4 M +1400 W 10 -5 M +SMTC 10 -6 M pEC 50 6.57 ± 0.09 (15) Table 2. pEC 50 and E max values for phenylephrine-induced contraction (mN) in the ascending aorta from wildtype and CD mice. Data are shown as mean ± SE. The number of animals is shown in parentheses. *P < 0.05, ***P < 0.001 versus Control; # P < 0.05, ### P < 0.001 versus the same group in wild-type by unpaired Student's t-test. www.nature.com/scientificreports www.nature.com/scientificreports/ and nNOS isoforms of NOS are generally involved in regulatory or signaling pathways and play important roles in the modulation of vascular tone, whereas the iNOS isoform is related with inflammatory responses 21 . Inhibition of nNOS with SMTC, but not iNOS inhibition with 1400 W, significantly increased phenylephrine contractions in both strains, an effect that removed contractile impairments in CD mice. In addition, we detected an increased expression of nNOS in aortas from CD mice, whereas eNOS and iNOS expressions were similar to WT mice. Altogether, these results suggest a higher negative influence of NO on α1-adrenoceptor constriction in WBS mice, and propose that nNOS is a relevant NO source in these conditions (Fig. 8).
Augmented negative influence of NO in aortic adrenergic contractions might prevent excessive increases in cardiac afterload in the face of supravalvular aortic stenosis. Notably, impairment of tissue NO responsiveness is a main predictor of aortic stenosis development [25][26][27] , which provides additional evidence that NO may exert protective effects in WBS. Furthermore, clinical studies in patients with aortic stenosis suggest that nitrates can be used to lower blood pressure without excessive risk of reducing cardiac output or end-organ perfusion 52 , confirming a beneficial role for NO in this clinical setting.
In conclusion, our study reveals that ascending aorta obstruction coexist with functional alterations in a mouse model carrying the most common chromosomal deletion present in WBS patients. These findings support the clinical relevance of young CD mice to model moderate aortic stenosis associated with human WBS. The results show excessive nNOS-derived NO signaling that compromises ascending aorta α1-adrenergic contractions. Therefore, increased nNOS signaling may act as a physiological 'brake' against the detrimental effects of stenosis. We suggest that early detection and monitoring of thoracic aorta dysfunction would offer the potential for timely intervention in this syndrome.

Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. Figure 8. Diagram illustrating the main findings of the study and the proposed mechanism involved in the observed alterations. The ascending aorta of complete deletion (CD) mice shows stenosis and attenuated phenylephrine-induced contractions compared to wild-type mice. The present study suggests that phenylephrine, besides activating α1 A/D -adrenoceptors located in smooth muscle cells (SMC) to induce contraction, could also stimulate endothelial cell (EC) α1 A -adrenoceptors that could trigger the release of nitric oxide (NO) by endothelial neuronal nitric oxide synthase (nNOS) 24 , whose expression is increased in CD mice. This increased release of NO could be responsible for the impairment of α1-adrenoceptor-mediated contractions in the ascending aorta of CD mice. WBS, Williams-Beuren syndrome.