Vascular Smooth Muscle Emilin-1 Is a Regulator of Arteriolar Myogenic Response and Blood Pressure
Objective—Emilin-1 is a protein of elastic extracellular matrix involved in blood pressure (BP) control by negatively affecting transforming growth factor (TGF)-β processing. Emilin1 null mice are hypertensive. This study investigates how Emilin-1 deals with vascular mechanisms regulating BP.
Methods and Results—This study uses a phenotype rescue approach in which Emilin-1 is expressed in either endothelial cells or vascular smooth muscle cells of transgenic animals with the Emilin1−/− background. We found that normalization of BP required Emilin-1 expression in smooth muscle cells, whereas expression of the protein in endothelial cells did not modify the hypertensive phenotype of Emilin1−/− mice. We also explored the effect of treatment with anti-TGF-β antibodies on the hypertensive phenotype of Emilin1−/− mice, finding that neutralization of TGF-β in Emilin1 null mice normalized BP quite rapidly (2 weeks). Finally, we evaluated the vasoconstriction response of resistance arteries to perfusion pressure and neurohumoral agents in different transgenic mouse lines. Interestingly, we found that the hypertensive phenotype was coupled with an increased arteriolar myogenic response to perfusion pressure, while the vasoconstriction induced by neurohumoral agents remained unaffected. We further elucidate that, as for the hypertensive phenotype, the increased myogenic response was attributable to increased TGF-β activity.
Conclusion—Our findings clarify that Emilin-1 produced by vascular smooth muscle cells acts as a main regulator of resting BP levels by controlling the myogenic response in resistance arteries through TGF-β.
The Emilin/Multimerin family of extracellular matrix proteins (for nomenclature see online-only Data Supplement material) comprises of 4 members with a common multidomain structure consisting of an amino-terminal elastin microfibril interface domain, a central sequence with high probability of coiled-coil conformation, and a carboxy-terminal gC1q-like domain.1,2 All 4 proteins are expressed in the cardiovascular system, however, with significant differences concerning their fine distribution.1,3,4 While Emilin-1 is expressed by endocardium and right ventricle myocytes and by cells of the entire blood vessel wall (endothelial cells [ECs], smooth muscle cells [SMCs] and adventitial fibroblasts), the other genes have a more restricted expression pattern: Emilin2 is active in myocardium, whereas Mmrn1 and Mmrn2 are expressed only by ECs in blood vessels and, in addition, in platelets and endocardial cells, respectively.
The function of these proteins in the cardiovascular system has not been fully disclosed. One prominent function of Emilin-1 is the regulation of systemic blood pressure (BP).4 The protein, through its elastin microfibril interface domain, inhibits transforming growth factor (TGF)-β biosynthesis by blocking the proteolytic cleavage of the proTGF-β precursor into the latency associated peptide/TGF-β complex. The increase in TGF-β signaling induced by Emilin-1 deficiency results in systemic hypertension, accompanied by narrowing of arterial tree and structural alterations of the wall of elastic arteries.4,5
In spite of this important information on alterations induced by Emilin-1 deficiency, it is still unknown which cell types mediate protein effects. In particular, it has not been investigated whether hypertension is causally related to abnormal behavior of vascular SMCs (VSMCs) or ECs. Gaining insight into this issue not only would add new knowledge to the biology of Emilin-1 but could also be of potential relevance in investigations addressing the function of EMILIN1 gene in human hypertension. In fact, recent association studies have suggested that specific haplotypes of EMILIN1 are useful genetic markers of essential hypertension in Japanese men and that the interaction of age and genotype variation of specific single nucleotide polymorphisms might increase the risk of hypertension in a northern Han Chinese population.6,7
Here, we have addressed the question of the cellular context of Emilin-1 activity using a phenotype rescue approach by expressing the protein in either ECs or VSMCs or in all cells of the arterial wall of Emilin1−/− mice. Our results indicate that Emilin-1 expression in VSMCs is specifically required for BP control and further clarify that the protein regulates the arteriolar myogenic response through TGF-β.
Materials and Methods
Transgenic mice were generated by pronuclear microinjection of DNA into fertilized eggs using the constructs described in Figure 1. Animals were anesthetized with a single intraperitoneal injection of avertin8 (250 mg/kg) and killed by cervical dislocation. All procedures involving animals conformed to the Directive 2010/63/EU of the European Parliament and have been approved by the Ethics Review Board of our Institutions (University of Padova, approval reference number 63TER/2010, and IRCCS Neuromed). The procedures used for the synthesis of constructs, the generation and breeding of lines, and the characterization of transgenic mouse lines using immunohistochemistry, electron microscopy, histology, and echocardiography are described in detail in the Methods section of the online-only Data Supplement.
Evaluation of BP
BP was evaluated using implanted radiotelemetry pressure transducer or noninvasively by tail cuff plethysmography in conscious animals.
Vascular Function Evaluation
Mesenteric arteries were isolated from mice euthanized with an overdose of sodium pentobarbital (250 mg/kg), placed in a pressure myograph in Krebs solution and myogenic response and contraction to phenylephrine and angiotensin II analyzed as described in the Methods section in the online-only Data Supplement.
Infusion of Antibodies to TGF-β
Neutralizing antibodies to TGF-β (0.5 µg/kg per day; R&D Systems Inc., Minneapolis, MN, diluted in PBS) or preimmune IgG were infused for 24 days into Emilin1−/− and eNOS−/− mice or the corresponding wild-type control littermates through osmotic minipumps (Alzet model 2004, Durect Corporation, Cupertino, CA) implanted subcutaneously on the right side of the back of the mice. BP was evaluated before and during treatment. The antibody dose used corresponds to the lowest dose effective in reducing BP of Emilin1−/− mice to normal levels, as determined in preliminary experiments.
For BP measurements, vascular reactivity parameters and percentage of phospho Smad3 (P-Smad3)-positive nuclei data are expressed as mean±SEM. Multiple comparisons were evaluated by 1-way ANOVA for factorial design or by 2-way ANOVA for repeated measures, accordingly to study design, followed by Bonferroni post hoc test.
Transgenes Expression Levels
The constructs used for tissue-specific expression of Emilin1 cDNA are sketched in Figure 1. The expression of transgenic mouse cDNA transcripts relative to the mRNA levels produced from 1 copy of endogenous Emilin1 gene is reported in the Table for the different lines. For further studies, the lines with the highest expression in aorta were chosen for constructs with the Emilin1 and SM22a promoters (lines Emilin1-Emilin1.73 and SM22a-Emilin1.165, respectively). Among the various Tie2-Emilin1 lines, 2 were chosen: one (number 73) among the high expressing and the other (number 71) among the low expressing ones.
The levels of Emilin1 mRNA produced in the aorta of some transgenic mouse lines were comparable with those of the endogenous gene. This was the case for the Emilin1-Emilin1.73 line, with about the same expression of 1 chromosomal gene copy, and the Tie2-Emilin1.71 line. The latter produced about one half of the mRNA expressed by 1 copy of endogenous gene; however, it should be considered that the normal gene is synthesized by both ECs and VSMCs, while the Tie2 transgene is active only in ECs. Assuming an equivalent production of Emilin1 mRNA by ECs and VSMCs on a per cell basis, and a VSMCs/ECs number ratio of ≈5 (the number of VSMCs layers in aorta is 5–6, while there is only 1 ECs layer), the calculated actual excess of transgene Emilin1 mRNA compared with 1 normal gene copy in ECs is 2.5 times, a range close to the physiological levels expressed by wild-type mice. On the other hand, the level of expression of transgenic Emilin1 mRNA attained with the SM22a promoter was low in aorta; however, the immunofluorescence analysis (see below) showed a stronger positive staining for Emilin-1 in resistance arteries when compared with aorta, suggesting a higher level of expression of the protein in vessels that give a major contribution to BP regulation.
Transgenes Expression Patterns
The appropriate tissue-specific expression of the transgenes carrying the mouse Emilin1 cDNA was tested by immunohistochemistry in an Emilin1−/− background. The expression of the transgenic Emilin1 promoter in embryos reproduced the complexity expected from the distribution of endogenous gene products (Figure I in the online-only Data Supplement).1,3,9 Matching of this promoter activity with that of the endogenous gene was investigated in mice harboring the human EMILIN1 cDNA in Emilin1+/− embryos. As shown in Figure II in the online-only Data Supplement, distribution of the transgenic protein extensively overlapped with that produced from the chromosomal gene. Analysis of adult blood vessels also revealed a good correspondence of transgenic and endogenous protein deposition in both conductance and resistance arteries (Figure III in the online-only Data Supplement). Expression of the Tie2 promoter was strictly confined to ECs in both the mouse lines investigated, namely Tie2-Emilin1.71 (Figure IV in the online-only Data Supplement) and Tie2-Emilin1.73 (data not shown). As for the SM22a.165 line, the transgenic protein was poorly expressed during development (data not shown), while it could be detected in adult blood vessels where labeling was not associated with the endothelium (Figure V in the online-only Data Supplement and Table). Taken together, these data show that the transgenes generated with the Tie2 and SM22a promoters correctly target the expression of Emilin1 cDNA in ECs and VSMCs, respectively, while the construct containing Emilin1 promoter directs production of recombinant protein to the whole arterial wall.
To test whether expressed transgenes were functional, we searched for reversal of blood vessels morphological alterations induced by Emilin1 deficiency. Although no ultrastructural defects were detected in resistance arteries (data not shown), aorta showed irregular outline and fragmentation of elastic lamellae, frequent blebbing, and detachment of ECs from the subendothelial extracellular matrix and atrophic, sometimes necrotic, VSMCs surrounded by enlarged extracellular space and disrupted adhesions of VSMCs to the elastic lamellae5 (Figure VI in the online-only Data Supplement). The presence of the Emilin1-Emilin1.73 transgene reversed the mutant morphology to a normal-looking ultrastructure of the vessel wall (Figure VI in the online-only Data Supplement). In particular, alterations of ECs and VSMCs dropped from 32% and 14% to 0.4% and 0.3%, respectively (Table I in the online-only Data Supplement). Re-expression of Emilin-1 in the endothelium (line Tie2-Emilin1.71) considerably reduced alterations of ECs (Table I and Figure VI in the online-only Data Supplement) and also attenuated the defects of the media to some extent (Table I and Figure VI in the online-only Data Supplement). The morphology of the internal elastic lamella was slightly improved, whereas the alterations of media elastic lamellae remained (Table I in the online-only Data Supplement). A similar result was obtained with the line Tie2-Emilin1.73 (data not shown). Production of Emilin-1 only in VSMCs (SM22a-Emilin1.165;Emilin1−/−) significantly reduced the abnormal morphology of the media, although not as efficiently as the Emilin1-Emilin1 transgene (Table I and Figure VI in the online-only Data Supplement). Interestingly, it also ameliorated ECs alterations and improved internal elastic lamella morphology (Table I and Figure VI in the online-only Data Supplement). These data show that the transgenic proteins are functional.
Transgene Expression in VSMCs, but Not ECs, Rescues the Hypertensive Phenotype of Emilin1−/− Mice
As shown in Figure 2 and in Figure VII in the online-only Data Supplement, telemetric and tail cuff monitoring of hemodynamics indicated that Emilin1−/− mice displayed higher BP, confirming our previous observation.4 Expression of transgenic Emilin1 cDNA under Emilin1, Tie2, or SM22a promoters in the wild-type background did not have any influence on BP (data not shown). On the contrary, when cDNA was expressed in the Emilin1 null background, the effect on BP depended on the cell type in which the promoter was active. In particular, BP was normalized when the VSMC-specific SM22a promoter (line SM22a-Emilin1.165) was used, but not when the transgene was driven by the EC-specific Tie2 promoter (lines Tie2-Emilin1.71 and Tie2-Emilin1.73) (Figure 2 and Figure VII in the online-only Data Supplement). Furthermore, we showed that, as expected, the Emilin1−/− mice hypertensive phenotype was rescued when the transgene was driven by the Emilin1 promoter (Figure 2 and Figure VII in the online-only Data Supplement). Interestingly, the percentage of P-Smad3–positive VSMCs nuclei, a marker for TGF-β signaling, followed a pattern of variation similar to that of hypertension, being higher in Emilin1−/− and Tie2-Emilin1.73;Emilin1−/− lines and comparable to controls in Emilin1-Emilin1.73;Emilin1−/− and SM22a-Emilin1.165;Emilin1−/− mouse lines (Figure VIIIA in the online-only Data Supplement). Instead, the P-Smad3–positive nuclei of ECs increased not only in the Emilin1 knockout but also in the line SM22a-Emilin1.165;Emilin1−/−, while they remained similar to controls in lines Tie2-Emilin1.73 and Emilin1-Emilin1.73;Emilin1−/− (Figure VIIIA in the online-only Data Supplement). Because the difference among groups was rather low (20%–25%), global P-Smad2 was also assayed by immunoblotting in Emilin1-deficient and control arteries to assess the maximum variation. The average induction was ≈50% (Figure VIIIB and VIIIC in the online-only Data Supplement). The difference obtained with the 2 methods is likely attributable to higher background of the former (immunoperoxidase) or lower induction of P-Smad3 compared with P-Smad2 detected by immunoperoxidase and Western blotting, respectively. Overall, the above data strongly suggest that Emilin-1 must be expressed in VSMCs to modulate BP levels and further link hypertension of Emilin1−/− animals to increased TGF-β signaling.
Anti-TGF-β Treatment Rescues Emilin1−/− Hypertensive Phenotype
It was previously found that inactivation of 1 TGF-β1 allele rescued the hypertensive phenotype of Emilin1−/− mice.4 This result, however, could be the consequence of the reversal of a vascular developmental defect induced by increased TGF-β signaling because of lack of Emilin-1 or an enduring alteration of VSMCs behavior triggered by enhanced TGF-β activity. Treatment of Emilin1−/− mice with neutralizing antibodies to TGF-β should distinguish between these 2 possibilities, as a vascular developmental defect is expected to be irreversible, whereas a functional alteration of VSMCs is not.
When a neutralizing anti-TGF-β antibody was administered to Emilin1−/− mice, a BP lowering effect was already evident after 6 days, followed by a complete normalization of BP within 15 days (Figure 3). The BP effect of anti-TGF-β antibody was not detected in nitric oxide synthase–deficient mice, a different model of hypertension, indicating that the antihypertensive effect was selective for Emilin1−/− mice (Figure 3). The effectiveness of the neutralizing antibody in lowering TGF-β signaling was indicated by the reduction of percentage of P-Smad3–positive nuclei in VSMCs in both Emilin1-deficient and wild-type mice, although the difference did not reach statistical significance in the latter group (Figure IX in the online-only Data Supplement). The conclusion coming from the above data is that increased TGF-β activity is continuously required for maintenance of the hypertensive phenotype of Emilin1−/− mice.
Emilin-1 in VSMCs Is a Regulator of Arteriolar Myogenic Response Through TGF-β
Our previous data suggested that hypertension is a primary abnormality in Emilin1−/− mice, attributable to increased peripheral vascular resistance.4 Resistance arteries have the specific property, called arteriolar myogenic response, to constrict in response to stepwise increases in perfusion pressure.10 Arteriolar myogenic response is known to be mediated by Ca2+ handling in VSMCs and is a key element for BP maintenance.10 Thus, we examined the effect of increasing the intra-arterial pressure in 2 different resistance arteries, mounting in a pressure myograph second-order mesenteric arteries and gracilis muscle arteries from Emilin1−/− and Emilin1+/+ mice. The myogenic response was calculated for each pressure step as the difference between passive and active diameters (measured in Ca2+ free and in normal physiological salt solution, respectively). Despite slightly reduced passive diameters in a Ca2+-free bathing solution, Emilin1−/− arteries had much narrower diameters in the presence of Ca2+ when compared with control arteries (Figure X in the online-only Data Supplement), indicating increased myogenic response in the mutant arteries (Figure 4A and Figure XI in the online-only Data Supplement). Interestingly, the SMC-specific SM22a-Emilin1 transgene (in addition to the Emilin1-Emilin1 construct) restored to normal levels the elevated myogenic response developed by Emilin1−/− arteries, whereas the EC-specific Tie2-Emilin1 transgene maintained an elevated myogenic response (Figure 4). Thus, the same transgenes that rescued hypertension of Emilin1−/− mice also reversed the abnormal vascular myogenic response in mutant vessels, suggesting a direct association between increased myogenic response and development of hypertension.
Then, we examined the response of resistance arteries to neurohumoral vasoconstrictors. In contrast to myogenic tone, amplitudes of the contraction were comparable between Emilin1−/− and Emilin1+/+ mesenteric arteries, for both phenylephrine and angiotensin (Figure XIIA and XIIB in the online-only Data Supplement). Moreover, neither the SMC- nor the EC-specific Emilin1 transgenes affected the contractile response to vasoactive agents in Emilin1−/− arteries (Figure XIIA and XIIB in the online-only Data Supplement).
The uncoupling of agonist-induced vasoconstriction from the hypertensive phenotype of Emilin1−/− mice suggests a major role of arteriolar myogenic response in determining higher BP levels in resting conditions. We therefore investigated whether decrease of BP of Emilin1−/− mice by antibodies to TGF-β was also accompanied by the reduction of myogenic response. As shown in Figure 5A, anti-TGF-β treatment for 24 days restored the increased myogenic response to normal levels.
To further exclude that the effects on BP of chronic TGF-β antibody infusion were other than the ones in resistance arteries, we evaluated cardiac and large vessel function, finding no changes induced by the treatment. In particular, cardiac function was comparable among groups (Table II in the online-only Data Supplement), as was arterial pulse wave velocity of aorta and carotids (Table III in the online-only Data Supplement), an index of arterial stiffness, thus excluding that BP lowering effect could be ascribed to changes in vascular compliance.
To finally disclose whether the consequence of neutralizing antibodies on myogenic tone was attributable to a direct effect of TGF-β on resistance arteries or to indirect in vivo effects, we applied directly the antibody on mesenteric arteries during myogenic response measurements. As shown in Figure 5B, the treatment of mesenteric arteries of Emilin1−/− with neutralizing antibody to TGF-β rescued the increased myogenic response to perfusion pressure.
To gain insight into the mechanism of the Emilin1−/− phenotype, characterized by systemic hypertension, we used a phenotype rescue approach comprising re-expression of the protein in specific vascular cell types. The message coming from our study is that Emilin-1 expressed in VSMCs controls BP by regulating the arteriolar myogenic response to mechanical stress. The maintenance of increased myogenic tone is dependent on continuous higher levels of TGF-β activity in VSMCs, as indicated by the rescue of the hypertensive phenotype and enhanced myogenic tone by lowering TGF-β signaling using neutralizing antibodies.
The first finding of our work suggests that, for regulation of BP, Emilin-1 is strictly required in VSMCs. This deduction is straightforward in the light of 2 considerations. On one hand, the experimental setup used was appropriate to address the question regarding the cell type mainly involved in the regulation of BP by Emilin-1. Indeed, the regulatory sequences used to drive protein expression in the null mutant background exhibited a high degree of cell type specificity in both conductance and resistance arteries. On the other hand, BP measurement showed that the hypertensive phenotype of Emilin1−/− mice was only rescued by transgenes including VSMCs as expression targets (ie, those with Emilin1 and SM22a regulatory sequences). On the contrary, EC-specific production of recombinant Emilin-1 did not rescue BP mutant phenotype, although the expression levels of mRNA achieved in the 2 Tie2-Emilin1 lines tested were comparable to or higher than that of the endogenous gene.
In our previous article, we demonstrated that lowering TGF-β levels by a genetic approach, obtained by crossing Emilin1−/− with Tgfb1+/− mice, is sufficient to rescue the vascular defects and the hypertensive phenotype of Emilin1−/− animals.4 However, that strategy did not allow us to distinguish between a structural and a functional effect of TGF-β on BP control. Here, we report that the hypertensive phenotype is dependent on functional alterations induced by TGF-β activity, as afforded by BP lowering effects of pharmacological anti-TGF-β treatment. Compared with the genetic approach, the method used here tells us not only that TGF-β regulation is key to Emilin1−/− hypertension but also that increased TGF-β activity must be continuously operating to maintain higher BP. The difference of TGF-β signaling between wild-type and Emilin1 mutant vessels, assessed by the percentage of P-Smad3–positive nuclei after immunoperoxidase staining, may appear too small (20%–25%) to account for a considerable change in resistance vessel function (15 mm Hg–20 mm Hg difference in BP). Using immunoblotting to P-Smad2 on a relevant number of arterial extracts, variation of TGF-β signaling was found to be about double (50% higher). The small difference detected with immunoperoxidase staining may therefore be because of the high background of the method, which has the advantage of giving separate information for VSMCs and ECs, or the lower activation of Smad3 compared with Smad2, that were the targets in immunoperoxidase and Western blotting experiments, respectively. One may also wonder about the finding that anti-TGF-β treatment reduced BP only to wild-type levels in Emilin1−/− mice and had no significant effect in control animals. However, it should be pointed out that the dose of antibodies infused into animals was the lowest dose effective in normalizing BP, among several tested in trial experiments.
The rescue of the hypertensive phenotype of mutant mice using TGF-β neutralizing antibodies prompted us to look for functional alterations of mutated vessels that could be responsible for the hypertensive phenotype. Indeed, our previous data, showing narrower vessels in Emilin1−/− mice, suggested that the hypertensive phenotype was likely caused by developmental perturbation of vessel size.4 However, the present work clarifies that the narrowing in lumen size of Emilin1−/− vessels is recruited by the arteriolar myogenic response that is the functional contractile response to perfusion pressure. Indeed, we found that resistance arteries of Emilin1−/− mice exhibited an increased myogenic response to mechanical stress induced by stepwise pressure increase, in experimental setting completely devoid of any neurohumoral influence. Interestingly, the phenotype rescue experiment with SMC- and EC-specific transgenes revealed that the increased myogenic response of Emilin1−/− mice was dependent on the lack of Emilin-1 in VSMCs, in line with what was observed for the hypertensive phenotype. It is noteworthy to emphasize that arteriolar myogenic response to perfusion pressure is a pivotal element in BP control and is a peculiar feature of VSMCs.10 We should also note that in our previous data we found that lumen of Emilin1−/− arteries was reduced when compared with control and that this effect was more prominent with steady increase in perfusion pressure.4
We also demonstrated that the increased myogenic response of Emilin1−/− arteries is dependent on TGF-β signaling, because treatment with neutralizing antibody, both in vivo and ex vivo, was effective to restore this response to normal levels. Altogether, previous and present data demonstrate that lack of Emilin-1 in VSMCs, by deranging TGF-β signaling, increases the gain of vasoconstriction to perfusion pressure, thus resulting in higher BP levels. Because a main role in myogenic vascular contraction is played by Ca2+ influx into cells, it is conceivable that the role of Emilin-1 in VSMCs may be to control Ca2+ regulation through TGF-β signaling.
On the contrary to what was observed for myogenic response, the unaltered vasoconstriction to neurohumoral agents was comparable in Emilin1−/− and control vessels and remained unaffected by re-expression of either SMC- or EC-specific transgenes in Emilin1−/− animals, suggesting negligible contribution of this factor to hypertension of Emilin1 mutants and highlighting a major role of myogenic response for the establishment of resting BP levels. In conclusion, the present study clarifies that Emilin-1 produced by VSMCs, by modulating the bioavailability of TGF-β, acts as a main regulator of resting BP levels by controlling the myogenic response in resistance arteries.
Sources of Funding
The work was supported by grants from Agenzia Spaziale Italiana (DCMC Program, workpackage 1B1119, to G.M. Bressan), Telethon (Project number GGP06066, to G.M. Bressan), the Italian Ministero dell’Istruzione dell’Università e della Ricerca (PRIN 2007, to G.M. Bressan), and the Italian Ministero della Salute (2006, to G. Lembo).
The online-only Data Supplement is available with this article at http://atvb.ahajournals.org/lookup/suppl/doi:10.1161/ATVBAHA.112.254664/-/DC1.
- Received October 10, 2011.
- Accepted June 29, 2012.
- © 2012 American Heart Association, Inc.
- Zanetti M,
- Braghetta P,
- Sabatelli P,
- Mura I,
- Doliana R,
- Colombatti A,
- Volpin D,
- Bonaldo P,
- Bressan GM
- Shimodaira M,
- Nakayama T,
- Sato N,
- Naganuma T,
- Yamaguchi M,
- Aoi N,
- Sato M,
- Izumi Y,
- Soma M,
- Matsumoto K
- Nagy A,
- Gertsenstein M,
- Vintersten K,
- , Behringer R
- Fabbro C,
- de Gemmis P,
- Braghetta P,
- Colombatti A,
- Volpin D,
- Bonaldo P,
- Bressan GM
- Li L,
- Miano JM,
- Mercer B,
- Olson EN
- Schlaeger TM,
- Bartunkova S,
- Lawitts JA,
- Teichmann G,
- Risau W,
- Deutsch U,
- Sato TN