BMP9 is a key player in endothelial identity and its loss is sufficient to induce arteriovenous malformations

AIMS

BMP9 is a high affinity ligand of ALK1 and endoglin receptors that are mutated in the rare genetic vascular disorder Hereditary Hemorrhagic Telangiectasia (HHT). We have previously shown that loss of Bmp9 in the 129/Ola genetic background leads to spontaneous liver fibrosis via capillarization of liver sinusoidal endothelial cells (LSEC) and kidney lesions. We aimed to decipher the molecular mechanisms downstream of BMP9 to better characterize its role in vascular homeostasis in different organs.

METHODS AND RESULTS

For this, we performed a RNAseq analysis on LSEC from adult WT and Bmp9-KO mice and identified over 2000 differentially expressed genes. Gene ontology analysis showed that Bmp9 deletion led to a decrease in BMP and Notch signaling, but also LSEC capillary identity while increasing their cell cycle. The gene ontology term "glomerulus development" was also negatively enriched in Bmp9-KO mice versus WT supporting a role for BMP9 in kidney vascularization. Through different imaging approaches (electron microscopy, immunostainings), we found that loss of Bmp9 led to vascular enlargement of the glomeruli capillaries associated with alteration of podocytes. Importantly, we also showed for the first time that the loss of Bmp9 led to spontaneous arteriovenous malformations (AVMs) in the liver, gastro-intestinal tract and uterus.

CONCLUSIONS

Altogether, these results demonstrate that BMP9 plays an important role in vascular quiescence both locally in the liver by regulating endothelial capillary differentiation markers and cell cycle but also at distance in many organs via its presence in the circulation. It also reveals that loss of Bmp9 is sufficient to induce spontaneous AVMs, supporting a key role for BMP9 in the pathogenesis of HHT.

Regression of renal vascular fibrosis by endothelin receptor antagonism

In previous studies, we have observed that endothelin participates in the progression of renal vascular and glomerular fibrosis during hypertension by activating collagen I gene synthesis. The present study investigated whether administration of endothelin receptor antagonists leads to the regression of renal sclerotic lesions. Experiments were performed in transgenic mice harboring the luciferase gene under the control of the collagen I-alpha2 chain promoter. Hypertension was induced by long-term inhibition of nitric oxide synthesis by N(G)-nitro-L-arginine methyl ester (L-NAME); systolic pressure gradually increased, reaching a plateau of 165 mm Hg after 10 weeks of hypertensive treatment. At the same time, collagen I gene expression was increased 2- and 5-fold compared with control animals in afferent arterioles and glomeruli, respectively (P<0.01). This increase was accompanied by the appearance of sclerotic lesions within the renal vasculature. When renal vascular lesions had been established (20 weeks of L-NAME), animals were divided into 2 subgroups: the one continued to receive L-NAME, whereas in the other, bosentan, a dual endothelin antagonist, was coadministered with L-NAME for an additional period of 10 weeks. Bosentan coadministration did not alter the increased systolic pressure at 30 weeks; in contrast, collagen I gene activity returned almost to control levels in renal vessels and glomeruli. In this subgroup of animals, renal vascular lesions (collagen and/or extracellular matrix deposition) and mortality rates were substantially reduced compared with untreated mice. These data indicate that endothelin participates in the mechanism(s) of renal vascular fibrosis by activating collagen I gene. Treatment with an endothelin antagonist normalizes expression of collagen I gene and leads to the regression of renal vascular fibrosis and to the improvement of survival, thus providing a complementary curative approach against renal fibrotic complications associated with hypertension.

Mechanisms mediating the renal profibrotic actions of vasoactive peptides in transgenic mice

Transgenic mice are useful tools to investigate the mechanisms of the renal profibrotic actions of endothelin and angiotensin II. The overexpression of angiotensinogen and renin genes induces renal sclerosis independently of changes in systemic hemodynamics. The same results are observed when the endothelin-1 gene is overexpressed. Transgenic mice harboring the luciferase gene, under the control of the collagen I alpha2 chain promoter, and made hypertensive by induction of a nitric oxide (NO) deficiency have been studied. In this strain of mice, luciferase activity is an early index of renal and vascular fibrosis. Luciferase activity was increased in preglomerular arterioles and glomeruli when mice were treated with N:(omega)-nitro-L-arginine methyl ester, an inhibitor of NO synthases. Bosentan (an endothelin receptor antagonist) was as efficient as losartan (an AT1 receptor antagonist) in preventing renal fibrosis, although it did not decrease BP. In short-term experiments, angiotensin II produced an increase in luciferase activity that was entirely prevented by losartan but also by bosentan. It can be concluded that, during chronic inhibition of NO, the collagen I gene is activated, which contributes to the development of nephroangiosclerosis and glomerulosclerosis. Angiotensin II plays a major role in this fibrogenic process, and its effect is at least partly independent of systemic hemodynamics and mediated by the profibrotic action of endothelin-1.

Angiotensin II activates collagen I gene through a mechanism involving the MAP/ER kinase pathway

Vascular remodeling and rearrangement of the extracellular matrix formation are among the major adaptive mechanisms to chronic increase in blood pressure. In previous studies we have found that angiotensin II (Ang II) participates in the hypertension-associated aortic and renal vascular fibrosis by stimulating collagen type I formation. The purpose of the present study was to gain insight into the molecular events that lead from the Ang II receptor to collagen I gene activation. To this end, we used a novel strain of transgenic mice harboring the luciferase gene under the control of the collagen I-alpha(2) chain promoter [procolalpha(2)(I)]. Ang II produced an early (1 hour) 2- to 3-fold stimulation of procolalpha(2)(I) activity in freshly isolated aortas and renal cortical slices (P:<0. 01) followed by similar increase in procolalpha(2)(I) mRNA aortic levels. This effect of Ang II was inhibited by AT1-receptor antagonism (candesartan) and blockade of the MAPK/ERK cascade (PD98059); in contrast, inhibition of the P38 kinase pathway (SB202190) and blockade of the release of the transcription factor NFkappaB (PDTC) did not have any effect in the Ang II-induced activation of the collagen I gene. In addition, Ang II induced a rapid (5 minutes) increase of the MAPK/ERK activity that was accompanied by increased expression (3-fold) of the c-fos proto-oncogene. This increase of c-fos mRNA expression was blocked by PD98059; in addition, curcumin, a blocker of the transcriptional factor AP-1, canceled the effect of Ang II on the collagen I gene. Decorin, a scavenger of the active form of transforming growth factor-beta (TGF-beta), canceled the Ang II effect on collagen I gene, whereas inhibition of the MAPK/ERK pathway had no effect on the TGF-beta-induced activation of procolalpha(2)(I). These data indicate that the cellular events after AT1 receptor stimulation and leading to activation of collagen I gene expression require activation of both the MAPK/ERK and TGF-beta signaling pathways.

Angiotensin II activates collagen type I gene in the renal vasculature of transgenic mice during inhibition of nitric oxide synthesis: evidence for an endothelin-mediated mechanism

BACKGROUND

Hypertension is frequently associated with renal vascular fibrosis. The purpose of this study was to investigate whether angiotensin II (Ang II) is involved in this fibrogenic process.

METHODS AND RESULTS

Experiments were performed on transgenic mice harboring the luciferase gene under the control of the collagen I-alpha(2) chain promoter [procolalpha(2)(I)]. Hypertension was induced by chronic inhibition of NO synthesis (N(G)-nitro-L-arginine methyl ester, L-NAME). Procolalpha(2)(I) activity started to increase in the renal vasculature after 4 weeks of L-NAME treatment (P<0.01) and at 14 weeks reached 3- and 8-fold increases over control in afferent arterioles and glomeruli, respectively (P<0.001). Losartan, an AT(1) receptor antagonist, given simultaneously with L-NAME prevented the increase of procolalpha(2)(I) levels and attenuated the development of renal vascular fibrosis without normalizing systolic pressure increase. Because we found previously that endothelin mediated renal vascular fibrosis in the L-NAME model, the interaction between Ang II, endothelin, and procolalpha(2)(I) was investigated in ex vivo and short-term in vivo experiments. In both conditions, the Ang II-induced activation of procolalpha(2)(I) in renal cortex was blocked by an endothelin receptor antagonist.

CONCLUSIONS

During chronic inhibition of NO, the collagen I gene becomes activated, leading to the development of renal vascular fibrosis. Ang II is a major player in this fibrogenic process, and its effect on collagen I gene is independent of systemic hemodynamics and is at least partly mediated by the profibrogenic action of endothelin.

Vascular endothelin-1 gene expression and synthesis and effect on renal type I collagen synthesis and nephroangiosclerosis during nitric oxide synthase inhibition in rats

BACKGROUND

The progression of hypertension during NO deficiency is associated with renal vascular fibrosis due to increased extracellular matrix (mainly collagen I) formation. The purpose of the present study was to investigate whether endothelin-1 (ET-1) is involved in this pathophysiological process.

METHODS AND RESULTS

Treatment of rats for 4 weeks with the NO synthase inhibitor Nomega-nitro-L-arginine methyl ester (L-NAME) 50 mg. kg-1. d-1 increased systolic blood pressure to 159+/-12 mm Hg. In animals treated with L-NAME, histological evaluation of renal sections revealed an increased formation of extracellular matrix (Masson's trichrome), and specifically of collagens (Sirius red). A part of this fibrosis was attributed to abnormal collagen I presence, because mRNA expression of the collagen I alpha1 chain (reverse transcription-polymerase chain reaction) and procollagen I formation (radioimmunoassay) were increased 3- and 2.5-fold, respectively, in the renal resistance vessels of hypertensive animals. In subsequent experiments, we examined whether ET-1 was involved in activation of collagen I formation. mRNA expression (RNase protection assay) of preproET-1 and ET-1 content (radioimmunoassay) were 10-fold and 3-fold increased, respectively, in renal microvessels of rats treated with L-NAME. Interestingly, in these vessels, ET-1 (immunostaining) was colocalized with sudanophilic lesions. Bosentan, an ET receptor antagonist (20 mg. kg-1. d-1), coadministered with L-NAME canceled the increased mRNA expression and synthesis of collagen I and attenuated the severity of renal vascular lesions without affecting L-NAME-induced high blood pressure.

CONCLUSIONS

These data demonstrate that ET-1 synthesis is increased in renal microvessels when NO production is suppressed. In this model of hypertension, ET-1 is a major activator of collagen I formation in renal resistance vessels and participates in the development of renal fibrosis without affecting systolic blood pressure.

[Interactions between the renin-angiotensin system, nitric oxide and endothelin]

The renin-angiotensin system plays a major role in the regulation of blood pressure and sodium balance. Nitric oxide (NO) and endothelin (ET-1) are involved in the regulation of renin release and modulate the vasoconstrictive and fibrogenic effects of angiotensin II. the mechanisms that activate renin production are less effective when endogenous NO synthesis is inhibited. In the absence of NO, ET-1 prevents renin secretion. Angiotensin II stimulates the production of NO and ET-1 by endothelial cells. The vascular effects of angiotensin II are inhibited by NO reinforced by ET-1. The stimulation of ET-1 secretion could partly explain the long-term effects of angiotensin II on vascular remodelling.

Evidence for functional ANP receptors in cultured alveolar type II cells

Because atrial natriuretic peptide (ANP) is considered to play a role in lung physiology and pathology, our aim was to characterize natriuretic peptide receptors in cultured rat alveolar type II (ATII) cells. Guanylate cyclase A- and B-receptor but not clearance-receptor mRNAs were detected by reverse transcription-polymerase chain reaction. The absence of clearance-receptor expression in ATII cells was confirmed by competitive inhibition of ANP binding; ANP (0.1-100 nM) decreased the binding of 125I-ANP, whereas C-ANP-(4-23), a specific ligand of clearance receptors, was ineffective. ANP induced a dose-dependent increase in guanosine 3',5'-cyclic monophosphate (cGMP) production, with a threshold of 0.1 nM, whereas the response to C-type natriuretic peptide was weak and was observed only at high concentrations (100 nM). In ATII cells cultured on filters, 1) ANP receptors were present on both the apical and basolateral surfaces and 2) cGMP egression was polarized, as indicated by the greater ANP-induced cGMP accumulation in the basolateral medium, and was partially inhibited by probenecid, an organic acid transport inhibitor. Influx studies demonstrated that ANP decreased the amiloride-sensitive component of 22Na influx but did not change ouabain-sensitive 86Rb influx. In conclusion, ATII cells behave as a target for ANP. ANP activation of guanylate cyclase A receptors produces cGMP, which is preferentially extruded on the basolateral side of the cells and inhibits the amiloride-sensitive Na-channel activity.

Activation of renin synthesis is dependent on intact nitric oxide production

The present study investigated whether or not nitric oxide (NO) synthesis mediates mechanisms regulating activation of renin formation. Studies were performed on afferent arterioles freshly isolated from the rat kidney. We have shown previously that this preparation is a useful model to study regulation of renin synthesis and secretion. The expression of renin mRNA was assessed by ribonuclease protection assay, and total renin content and renin secretion by radioimmunoassay. In afferent arterioles isolated from rats treated with the angiotensin-converting enzyme inhibitor ramipril, renin mRNA levels, total renin content and renin secretion were increased threefold compared to untreated controls. Inhibition of NO-synthase by NG-nitro-L-arginine methyl ester (L-NAME) in the ramipril-treated rats, abolished the increase in renin mRNA levels, total renin content and renin secretion. In other animals furosemide, a diuretic acting on macula densa cells, activated renin synthesis to a level similar to that found in the ramipril-treated group. Addition of L-NAME to the furosemide-treated rats suppressed the increases in renin mRNA levels, total renin content and renin secretion, suggesting that NO acts on renin activation by a mechanism independent of angiotensin II. In separate experiments, the inhibitory effect of L-NAME on the activation of renin secretion was abolished when afferent arterioles were treated with nicardipine, an L-type Ca2+ channel blocker, suggesting that the suppression of renin activation during NO inhibition is due to increased Ca2+ entry. Since endothelin is a potent mediator of Ca2+ influx and an inhibitor of renin release, we tested whether or not endothelin could be involved in the inhibitory effect of L-NAME on renin secretion. Application of the endothelin receptor antagonist, bosentan, in vitro mimicked the effect of nicardipine. In addition, bosentan coadministered with L-NAME in vivo blunted the inhibitory effect of L-NAME and restored the increases in renin mRNA level, synthesis and secretion. These data indicate that the physiological mechanism(s) regulating activation of renin synthesis and secretion are impaired during NO inhibition, probably because of increased Ca2+ influx. This increase in calcium flux is mediated at least partially by the action of endothelin.

EGF and TGF-beta regulate neutral endopeptidase expression in renal vascular smooth muscle cells

We recently reported that neutral endopeptidase (NEP) expression on renal vascular smooth muscle cells (VSMC) was downregulated in the presence of serum. Here we examine the role of epidermal growth factor (EGF) and transforming growth factor-beta 1 (TGF-beta) in this downregulation and the consequences of the changes in NEP activity on their mitogenic effects. EGF inhibited NEP activity, whereas TGF-beta was stimulatory. Expression of the enzyme was studied by measuring the binding of [125I]RB-104, a specific NEP inhibitor, and the fluorescence intensity of NEP-labeled cells. Both parameters were decreased by EGF and were increased by TGF-beta. NEP mRNA expression in EGF-treated cells was reduced after 48 h. In contrast, it was increased in TGF-beta-treated cells. Interestingly, NEP inhibition influenced the mitogenic effect of EGF. Indeed, thiorphan, an NEP inhibitor, and an anti-NEP antibody decreased EGF-dependent [3H]thymidine incorporation and cell proliferation by approximately 50%. TGF-beta had no effect on VSMC growth. These results indicate that EGF but not TGF-beta participates in the downregulatory potency of serum on NEP expression in VSMC. They also demonstrate that the full effect of EGF on VSMC proliferation depends on intact NEP activity.

Role of alpha v integrins in mesangial cell adhesion to vitronectin and von Willebrand factor

This study demonstrates (by flow cytometry and immunoprecipitation after cell surface radiolabeling and by using monoclonal antibodies to alpha v, beta 3, and alpha v beta 3 and alpha v beta 5 complexes) that alpha v beta 3, the vitronectin receptor, and alpha v beta 5 are expressed in vitro on cultured human mesangial cells (HMC) of the 5th to 8th passages. Antibodies to alpha v, beta 3 and alpha v beta 3 respectively precipitated an alpha beta heterodimer with molecular weights of 140 and 97 kDa. We analyzed the role of the various integrins in HMC interactions with vitronectin, and with fibronectin and von Willebrand factor (vWf), which are synthetized respectively by mesangial and endothelial cells. Cell adhesion increased in a dose dependent manner with the concentration of plastic-coated matrix protein and vWf. Inhibition of cell attachment with monoclonal antibodies to integrins indicated that HMC adhesion to vWf primarily involves alpha v beta 3, and that alpha v beta 5 may also contribute to cell binding to vWf. Adhesion to vitronectin involves both alpha v beta 3 and alpha v beta 5 complexes. In contrast, adhesion to fibronectin was not affected by monoclonal antibodies to alpha v beta 3 and alpha v beta 5 complexes. We propose that integrins alpha v beta 3 and alpha v beta 5, present on HMC, could mediate an interaction between mesangial and endothelial cells by binding to vWf, released at the basal site of endothelial cells.

Interest and limits of in vitro studies in renal vascular endocrinology

Various in vitro preparations have been utilized to study the cellular activity of vasoactive agents on renal cortical microvessels and on mesangial cells. The receptors and the transduction pathways of bradykinin and atrial natriuretic factor were characterized on cultured cortical vascular smooth muscle cells from the rabbit kidney. A preparation of afferent arterioles that had been freshly isolated from the rat kidney was used to study the NO-dependent regulation of renin release. The influence of endothelin and angiotensin II on mesangial cell proliferation was evaluated, using cocultures of human endothelial and mesangial cells. These appropriate in vitro preparations have provided new insights on renal vascular endocrinology. However, extrapolation of in vitro data to in vivo physiology must be cautious because the phenotype of vascular cells often changes in culture conditions.

Plasma atrial and brain natriuretic peptides in mitral stenosis treated by valvulotomy

1. In order to appreciate the effect of changes in left atrial pressure on plasma brain natriuretic peptide, 20 patients with mitral stenosis treated by percutaneous valvulotomy were studied 10 min before and 15 min after the first balloon inflation. They were also studied 24 h before and 48 h after the valvulotomy. At these times the effect of postural changes on brain natriuretic peptide secretion was examined. A group of 10 control subjects was also studied under basal conditions. In each case, plasma atrial natriuretic peptide was measured in parallel with plasma brain natriuretic peptide. 2. Similarly to plasma atrial natriuretic peptide, plasma brain natriuretic peptide was elevated in patients with mitral stenosis (32 +/- 2.9 and 32 +/- 2.8 pg/ml in the upright and supine position respectively versus 13.5 +/- 0.5 and 13.8 +/- 1.8 pg/ml in controls; P < 0.01). Changing from standing to lying did not modify plasma brain natriuretic peptide, whereas it produced an increase in plasma atrial natriuretic peptide in controls (13.3 +/- 1.6 versus 24.8 +/- 5.2 pg/ml; P < 0.01) and in patients 48 h after valvulotomy (52.5 +/- 4.6 versus 66.9 +/- 6.6 pg/ml; P < 0.01). Plasma brain natriuretic peptide also fell at this time (18.8 +/- 1.1 and 19.1 +/- 1.1 pg/ml in the upright and supine position respectively; P < 0.01) similarly to plasma atrial natriuretic peptide and cyclic GMP (P < 0.01). The acute left atrial mean pressure variation was significantly correlated with the parallel change in plasma atrial natriuretic peptide (P < 0.001) but not in plasma brain natriuretic peptide.(ABSTRACT TRUNCATED AT 250 WORDS)