Endothelin as a natriuretic hormone: the case for a paracrine action mediated by nitric oxide

Time for primary review 26 days. There is a widely held view that function of the kidney is the key to understanding human essential hypertension. Indeed, all of the currently recognised human monogenic causes of hypertension reflect abnormalities in the genes that determine ion transport in the distal portion of the nephron [1]. It is also recognised that permanent elevation of blood pressure in experimental models of hypertension, and both primary and secondary hypertension in man, requires the shifting of the natriuretic response curve of the kidney towards higher pressures [2]. However, the mechanism underlying pressure-natriuresis, and its resetting in hypertension, remains poorly understood. In this review we have analysed experiments indicating a role for endothelin (ET)-1 as a key natriuretic hormone and propose a new and testable hypothesis for a paracrine action of ET-1 through NO that accommodates the existing experimental findings. ET-1 is a 21 amino acid peptide, synthesised mainly by endothelial cells (ECs) and produced from an inactive precursor peptide by several rounds of specific proteolysis. The last step is performed by endothelin converting enzyme (ECE), which releases ET-1 from big ET-1. ET-1 acts in an autocrine and paracrine fashion on two subtypes of ET receptors, termed the ETA (ETAR) and ETB receptors (ETBR). Both ETAR and ETBR are located on vascular smooth muscle cells and binding of ET-1 to these receptors results in sustained vasoconstriction. ETBR are also present on ECs where their activation results in the production of NO and dilator prostanoids, and subsequent vasodilatation. The ET system is now known to play an important role in the regulation of vascular tone and blood pressure [3,4]. However, the role of ET-1 produced and secreted in the kidney may be independent of the systemic vascular effects. Analysis of different tissues found the highest concentrations of immunoreactive ET-1 in the inner medulla of the kidney, while the concentrations in kidney cortex are very low [5]. Kohan was the first to show that high levels of ET-1 are produced by inner medullary collecting duct cells (IMCDs) in culture [6]. Later, ETARs and ETBRs were characterised in IMCD [7]. On this basis, an autocrine model of endothelin action in the kidney was proposed [8]. There is now growing evidence for an intra-renal ET system that is central to the control of water and salt homeostasis, and to the regulation of blood pressure [9]. Abnormalities of this system have been found in experimental forms of hypertension, and may also be important in the pathogenesis of human hypertension, heart failure and kidney failure, in all of which sodium balance is also critical [4]. Indeed, there is evidence that endothelin receptor antagonists effectively reduce blood pressure in human essential hypertension [10]. Extensive functional studies support a natriuretic and diuretic action of ET-1 in the renal medulla, related to sodium transport functions. By intravenous infusion of ET-1 in anaesthetised rats, it has been shown that low doses cause natriuresis due to reduced sodium transport in the proximal and distal nephron segments and that only higher doses result in sodium retention due to glomerular vasoconstriction [11]. In experiments conducted on anaesthetised dogs with ETA and ETB specific antagonists, it has been shown that the ETAR predominantly accounts for the renal vasoconstriction, whilst the ETBR is largely responsible for diuresis and natriuresis [12]. ET-1 and anti-diuretic hormone (ADH) have opposing effects in the control of salt and water regulation in the nephron. Nanomolar concentrations of ET-1 reduce ADH-stimulated increases in water permeability of the collecting ducts in rats [13]. Studies in primary cultures of IMCDs also demonstrate that ET-1, acting via ETBRs, down-regulates the activity of the epithelial sodium channel (ENaC) and Na/K ATPase [14]. This effect is mediated by an ETBR-stimulated reduction in cAMP levels [15]. The synthesis and secretion of ET-1 from a collecting duct cell line was upregulated by hypertonic extracellular media [16]. In turn, ADH down regulates ETBR, but not ETAR, density in isolated IMCDs [17]. Several studies document increased ET-1 production in IMCDs induced by increased osmolarity, and specifically by increased concentrations of sodium ions [18,19]. In addition, hypoxia can mediate ET-1 secretion in IMCDs [20], and this may suppress renin gene transcription during long term ischaemia, leading to glomerular vasodilatation. However, the opposite effect was seen following short-term ischaemia. Thus, at present, there probably remains insufficient evidence from which to make any firm conclusions regarding the hypoxia-mediated effects of ET-1 in the kidney [21]. Recent work utilising transgenic animal models with dysfunctional ETBR genes has confirmed that the predominant role of the ETBR in the kidney is one of tonic inhibition of sodium reabsorption, possibly via inhibition of ENaC [22]. A naturally occurring 301 base pair deletion of the coding sequence of the ETBR gene is found in spotting lethal rats (sl) [23]. Gariepy and colleagues ‘rescued’ homozygous sl rats by expression of an ETBR transgene under a dopamine-β-hydroxylase promoter (DβH) [24]. The resultant DβH-ETB; sl/sl rats only express functional ETBRs in ganglionic cells of the intestine (preventing the development of fatal intestinal obstruction) and in adrenergic tissues. The lack of functional vascular ETBRs in DβH-ETB; sl/sl rats is supported by an absence of the normal early ETBR-mediated transient hypotensive response following bolus ET-1 administration [22]. On either a normal or high (8% NaCl) salt diet, DβH-ETB; sl/sl rats were found to develop severe hypertension (170 mmHg systolic). The rise in blood pressure was reversed to normal by the specific ENaC inhibitor, amiloride. ETAR blockade with the selective antagonist FR139317 only modestly reduced blood pressure. Although it is most likely, that ETBR deficiency in this model is translated into reduced ability to excrete sodium, normalisation of blood pressure by amiloride of itself does not prove it. Also, elevated blood pressure caused by volume expansion or systemic vasoconstriction can be normalised by diuretics like amiloride. In addition, increased plasma ET-1 concentrations in rescued ETBR−/− mutants complicates interpretation of the results. Indeed, plasma ET-1 concentrations in DβH-ETB; sl/sl rats were between 5 and 6 times higher than those of DβH-ETB; ETB+/+ rats. The authors attribute this to the absence of ‘scavenging’ endothelial ETBRs in DβH-ETB; sl/sl rats. Nevertheless, such a rise in plasma ET-1 concentrations may have caused a significant increase in vascular resistance through activation of ETAR on vascular smooth muscle cells, including those of the glomerulus. Indeed, the observed partial normalisation of blood pressure in DβH-ETB; sl/sl rats by acute injection of an ETAR antagonist supports this possibility. In further support of this interpretation, it was recently reported that chronic treatment with ABT-627, an ET(A) receptor antagonist, completely suppresses DOCA-salt-induced hypertension in sl/sl and wild type rats [25]. Two mouse models of ETBR deficiency exist. One is similar to the DβH-ETB; sl/sl rat, but with knocked out ETB alleles. Salt sensitive hypertension was reported in DβH-ETB; ETB−/− mice. No data on plasma ET-1 concentrations are available [26]. Another mouse model was produced by breeding ETBR+/− animals to piebald (s) mutants, which are a mouse model similar to the sl rat. This resulted in 1/8 of the normal expression of ETBR in s/− animals. Mean BP in ETB s/− animals was 20 mm higher than in s/+ or wild type controls on a normal salt diet. Plasma concentrations of ET-1 in s/− and s/+ animals were not significantly different but were higher than in wild type controls [27]. A potent inhibitory action of NO on tubular sodium reabsorption is well described [28]. Given that components of the ET and NO systems are present within the renal tubular system, and that an important interaction between ET and NO is already known to exist in the peripheral vasculature, it would not be unreasonable to propose an interaction between these systems in the kidney. Indeed, a number of possible interactions between the renal ET and NO system were recently summarised by Plato and Garvin [29]. The same authors have demonstrated inhibition of chloride transport in the thick ascending limb of Henle by endogenous nitric oxide [30]. Since then, important new data have been published showing that the natriuretic action of ET-1 is mediated by NO in the renal medulla. In this study, the NO synthase inhibitor, N-nitro-l-arginine methyl ester (l-NAME), and the specific ETBR antagonist, A-192621, both acted independently to abolish the diuretic and natriuretic effects of big ET-1 in the kidneys of anaesthetised rats [31]. Previously the same group has shown that ET-1 increases renal medullary blood flow in NO dependent manner [32]. Although these experiments demonstrate an NO mediated diuretic and natriuretic action of ET-1 in the kidney, it is not possible to conclude which cell type is the major source of NO, and which isoform of NO synthase is responsible for these effects. All three isoforms — endothelial (eNOS), neuronal (nNOS) and inducible (iNOS) — are present in the kidney. iNOS plays an important role during inflammation but, at present, there is no evidence that it can be activated by ET-1 or other vasoactive hormones, and, because it is not regulated, we have not considered iNOS further. It is, however, important to distinguish whether ETBR are linked to eNOS or nNOS. As there are no highly specific inhibitors of these isoforms, pharmacological approaches are unlikely to be helpful. There is substantial evidence that, in the cortical collecting ducts (CCD), NO inhibits ADH stimulated sodium reabsorption [33] and hydro-osmotic water permeability [34]. There is also a preliminary report that ET-1 stimulates NO release from IMCDs [35] and the only published work, showing increased ET-1/ETBR stimulated cGMP production in rat IMCDs, provides indirect support for this link [17]. The group that discovered high NOS activity in the IMCD [36], has also demonstrated that intra-medullary injection of antisense oligonucleotides specific to nNOS promotes a modest salt dependent rise in blood pressure in rats [37]. Given that the ETBR is expressed in IMCDs, nNOS in the same location could be the target for autocrine activation by ET-1. However, involvement of nNOS in the regulation of extracellular volume and blood pressure is not supported by knock out experiments [38,39]. In fact, mean arterial pressure in anaesthetised nNOS knock out mice (nNOS−/−) is slightly lower then in wild type controls [40]. In contrast, eNOS−/− mice are hypertensive [41–43], but no direct data on salt dependence of blood pressure in either knock out has been published. However, indirect evidence suggests that eNOS−/− animals, but not nNOS knock outs or wild type controls, accumulate salt, reflected in their lower mRNA levels and renal renin content. In addition, l-NAME reduced renal renin mRNA levels in wild type and nNOS−/− mice, but not in eNOS−/− [44]. The relative importance of eNOS over nNOS in sodium excretion by the thick ascending limb was also demonstrated in another study using knock out strains. Here, the NO synthase substrate, l-arginine, inhibited chloride transport in the isolated thick ascending limb in wild type as well as nNOS−/− mice, whereas in eNOS−/− mice, l-arginine has no effect [40]. Taken together, these data indicate that eNOS is likely to be the major regulator of renal sodium reabsorption, potentially under the regulation of ET-1. As with all G-protein coupled receptors ETBR can activate the effector enzyme (nNOS or eNOS) only when they are expressed in the same cell. The important question is: what are the target cells for ET-1 action that respond by major NO production in the renal medulla? To answer this question, we have compared cellular expression patterns of the elements of ET system and NOS. The results are summarised in Table 1. The major source of ET-1 in the renal medulla is IMCDs [5,8,45]. Detailed light and electron microscopic (EM) investigation of radioactively labelled ET-1 binding to ETAR and ETBR comes from two independent laboratories [46,47]. Both groups report predominant expression of ETBR over ETAR in renal medulla, with intense staining of the inner stripes of medulla and papillae. Binding sites were in the glomeruli and in the vasa recta, the peritubular capillaries that penetrate deep into the medulla [47]. Using in vivo [125I]ET-1 binding and EM-autoradiography, ETBR specific binding was localised almost exclusively to the fenestrated ECs of glomerular capillaries and peritubular capillaries, with no appreciable binding in cells of glomerular mesangium, renal tubules and the interstitium [48]. In vitro labelling also demonstrated ETBR specific binding to the glomeruli, proximal tubules and renal medulla interstitial cells [49]. Nevertheless, radioligand binding studies do show ETBR on IMCDs, and functional responses to ET-1 in cell culture [17]. There is close contact between the basolateral membrane of the IMCDs and the abluminal plasma membrane of ECs of the vasa recta, separated only by a basal membrane. Interestingly, the ETBR are localised exclusively on the abluminal surface of the capillaries, facing the IMCDs. In vitro binding detected ETBR in the basal infoldings of IMCDs, but only at levels 7-fold lower than in cells of the vasa recta [47]. Localisation of key elements of the endothelin system in the renal medullaa +++, high; ++, moderate; +, low; NP, not present; ND; not determined. IHC, immunohistochemistry; RIA, radioimmunoassay; RL, radioligand binding; RLEM, radioligand electron microscopy; RMIC, renal medullary interstitial Localisation of key elements of the endothelin system in the renal medullaa +++, high; ++, moderate; +, low; NP, not present; ND; not determined. IHC, immunohistochemistry; RIA, radioimmunoassay; RL, radioligand binding; RLEM, radioligand electron microscopy; RMIC, renal medullary interstitial As the highest levels of expression of ETBR are found in in the vasa recta and capillaries of the renal medulla. of renal the highest levels of NO synthase activity in vasa are than in glomeruli or and than higher than in and Both nNOS and eNOS mRNA were in vascular by data well with staining in transgenic animals a under a eNOS promoter It to be that eNOS expression is present where there is an between ECs and smooth muscle cells and, in the In this the vasa recta are a in showing eNOS In contrast, staining was not found in the renal tubular elements of the medulla. A very similar of staining has been shown using for eNOS but with this it is to distinguish of the of expression of eNOS and nNOS in the IMCDs, the of expression of ETBR in the renal medulla, is evidence has been found of NO in the collecting ducts by whereas it was detected in thick ascending by this and by nNOS was detected in and tubular by staining was intense in the By contrast, nNOS has been whereas eNOS was not in isolated of the rat No eNOS has been detected in the renal tubules of the rat whereas there is staining with an in of the kidney. In the same study, nNOS staining was present in and in the thick ascending limb of the nephron. data are with the staining in renal tubules and collecting ducts with a by an eNOS promoter of NO synthase activity in segments of the nephron and kidney has been performed The highest NOS activity mm of the was found in IMCDs and in vasa recta, with very low in medullary collecting and thick and segments of the of Both nNOS and eNOS mRNA was detected by in IMCD and vasa recta However, nNOS was the predominant isoform in IMCDs, and eNOS in vasa It has to be into that be in this type of and of collecting ducts with cells was found to be with specific for nNOS was demonstrated in isolated IMCDs of the rat and an increase in nNOS was detected by to a high sodium that ECs of vasa recta are a major of expression of ETBR and eNOS recta ECs are in close and are likely to be a major target ET-1 produced and secreted by experiments in knock out animals, that of either ETBR or of the eNOS but not iNOS or have major effects on blood pressure and ion transport in the nephron. evidence suggests that ET-1 produced within the kidney is in the regulation of sodium reabsorption, with ET-1 both natriuresis and Taken these to the following that ET-1 is a paracrine natriuretic and diuretic produced by IMCDs and acting mainly on ETBR of ECs of the vasa recta to activate eNOS and In turn, NO acts on IMCD cells to reduce sodium transport and on vasa recta to blood flow and the these effects to reduce sodium reabsorption and sodium In the kidney, ET-1 is produced by inner medullary collecting duct cells (IMCDs) and acts mainly on ETB receptors of the endothelial cells (ECs) of vasa recta capillaries, where it stimulates endothelial nitric oxide synthase to NO from l-arginine NO acts on in IMCDs to that acts in to reduce ENaC leading to sodium and water NO acts on smooth muscle cells and vasa recta to blood This of blood in turn, the of sodium chloride in the renal medulla and to natriuresis and The of this hypothesis the that ET-1 is a paracrine hormone secreted by IMCDs that ETBR mainly on the ECs of vasa This view with an autocrine model in which ET-1 is secreted by and ETBR of IMCD cells, while ETBR of the renal medullary are activated by ET-1 produced by It also that the is the major source of NO, acting as an between ECs of renal medullary capillaries and the tubular This hypothesis with the that the is by IMCDs and that the expression of ETBR is on the It is also with the of NOS knock out animals. of salt and water can be by two that reduce sodium a reduction in tubular sodium transport and the an increase in blood flow to the vasa The ability of the kidney to excrete a on an of in the renal medulla. This is produced by a mechanism of sodium in the of Henle of the long and by transport from the collecting ducts into interstitium by specific blood flow through the vasa recta out the in the medulla, the mechanism for by and can to natriuresis and diuresis As was ET-1 can increase blood flow in the renal medulla, and this is NO dependent It was that of renal medullary interstitial cells blood flow in medullary capillaries As well as sodium there is a growing of evidence that the of the kidney plays an important role in important of the paracrine hypothesis is that it a mechanism between blood flow in the vasa recta and sodium reabsorption in the tubular Indeed, both systems can they respond to that is, ET-1 secreted by IMCD cells and NO produced by The primary for ET-1 secretion from IMCDs is an increased of the interstitial [16]. The effect of ET-1 secretion in renal medulla be on ETBR of vasa recta to NO to sodium On the other increased blood flow in the vasa recta caused by increased systemic blood pressure could be a primary for the activation of NO dependent blockade of ENaC and A of NO is its and high This that NO produced in endothelial or epithelial cells can its target — — in cells and ion transport in different segments of the nephron as well as medullary blood This of NO it an and would be in the autocrine A major of this hypothesis is that all of the interactions proposed can be in experiments using pharmacological and The key to the proposed interactions be mice with or genes for nNOS and recently mice with sites in and of the ETBR gene through in cells have confirmed the of sites in cell culture experiments through of and by of As shown this deletion to of function of animals with sites in both of ETBR are normal and do not show of A transgenic line in a manner in ECs has been in group A line in a manner in IMCDs has been by Kohan enzyme expression is highly specific and to cells of the collecting deletion of with sites by enzyme in the renal medulla has been confirmed using specific By these one can functional ETBR in a in the of the and this would ETBR in ECs and IMCDs models to the relative of ETBR expressed in ECs and IMCDs to the regulation of ET-1 stimulated of three transgenic be mice with ETBR expression down in ECs mice with ETBR expression down in IMCDs and the ‘rescued’ mice. of ET-1 stimulated NOS activity in these demonstrate the partial of cell type to NO In addition, the paracrine model is then down regulation of ETBR in ECs have a major effect on big ET-1 l-NAME sensitive sodium excretion [31]. on the the autocrine model is then NO produced by eNOS within ECs of the vasa recta not be in ET-1 mediated inhibition of tubular sodium In this mice with reduced expression of ETBR in the collecting duct ET-1 stimulated sodium the paracrine then the important question be of the role of ETBR on IMCDs. It is possible that these receptors ET-1 In transgenic animals with ETBR in IMCDs, ETBRs be expressed and we would not an increase in plasma ET-1 as in transgenic DβH-ETB; sl/sl rats and ‘rescued’ piebald mice. Here, and smooth muscle cell ETBR in the to the effect of tubular ETBRs on salt and water excretion the effects of deletion or blockade of vascular can also of ET-1 as a of ET-1 in and renal medulla during normal and high salt Analysis of in these of the relative of ETBR expressed in IMCDs and ECs in regulation of the acute response to volume expansion and to increased systemic blood pressure. NO from the vasa recta ECs are critical for sodium homeostasis, then it be very important to salt of blood pressure in eNOS−/− and nNOS−/− as well as in knock out animals, which have already been because inhibition of kidney nNOS with antisense oligonucleotides suggests that nNOS animals may salt sensitive hypertension [37]. In addition, ETBR knock out animals to eNOS−/− and nNOS−/− animals a further to determine the role of these isoforms in ET-1 mediated control of sodium The mechanism of of salt and water was by an and and responses have been by the of the The now is to is at in to the function of three genes eNOS and acting in two cell is currently supported by a from the by and on of the ETB receptor is supported by the