Participation of Renal and Circulating Endothelin in Salt-Sensitive Essential Hypertension

Home , Endothelin

Journal of Human Hypertension (2002) 16, 459–467  2002 Nature Publishing Group All rights reserved 0950-9240/02 $25.00 www.nature.com/jhh REVIEW ARTICLE Participation of renal and circulating endothelin in salt-sensitive essential hypertension

F Elijovich, and CL Laffer Department of Medicine, College of Human Medicine, Michigan State University, Medical Education and Research Center of Grand Rapids, MI, USA

Salt sensitivity of blood pressure is a cardiovascular depending on their site of generation and binding to dif- risk factor, independent of and in addition to hyperten- ferent receptors. We review the available data on endo- sion. In essential hypertension, a conglomerate of clini- thelin in salt-sensitive essential hypertension and con- cal and biochemical characteristics defines a salt-sensi- clude that abnormalities of renal endothelin may play a tive phenotype. Despite extensive research on multiple primary role. More importantly, the salt-sensitive patient natriuretic and antinatriuretic systems, there is no may have blood pressure-dependency on endothelin in definitive answer yet about the major causes of salt-sen- all states of salt balance, thus predicting that endothelin sitivity, probably reflecting the complexity of salt- receptor blockers will have a major therapeutic role in balance regulation. The endothelins, ubiquitous pep- salt-sensitive essential hypertension. tides first described as potent vasoconstrictors, also Journal of Human Hypertension (2002) 16, 459–467. doi: have vasodilator, natriuretic and antinatriuretic actions, 10.1038/sj.jhh.1001419

Keywords: endothelin; salt sensitivity; salt; blood pressure

Introduction biochemical characteristics that are common in salt- sensitive hypertensive humans. They include high It has long been known that blood pressure (BP) prevalence in African Americans,4,6,7 increasing responses to salt-loading and unloading are hetero- prevalence with aging,1,8 low PRA with blunted geneous and exhibit a Gaussian distribution in both responses to its physiological stimuli2 or normal normotensive and hypertensive human beings. It PRA in so-called ‘non-modulators’ of renal blood took several decades to produce the evidence that flow,9 a suppressed kallikrein-kinin system10,11 and salt-sensitivity and salt-resistance of BP are pheno- excess activation of the arginine-vasopressin pressor types, rather than a reflection of random variability system.12 Associations with gender,4,13 a hyperactive of BP responses to salt. Hence, these responses are 4,14 1 sympathetic system, a hypoactive dopaminergic reproducible over time, reproducible when meas- system,15 obesity and insulin resistance are contro- ured with different protocols,2 and closely associa- 3 versial, which may be explained if salt-sensitivity of ted in mono and dizygotic twins. BP is not a pure phenotype. This is supported by Different methods have been employed to study the finding of opposite alterations of insulin sensi- salt-sensitivity of BP in humans, including rapid tivity in the low-renin vs non-modulating groups of intravenous salt loading and pharmacologic salt-sensitive hypertensive patients.16 (furosemide) salt depletion,4 and prolonged dietary Perhaps the most relevant findings emerging from salt manipulation.2 The magnitude of BP responses the studies of salt-sensitivity of BP are the prognos- employed as a cut-off for definition of salt-sensi- tic ones. There is evidence that salt-sensitive hyper- tivity vs salt-resistance has been set at different arbi- tensive patients exhibit excess prevalence or sever- trary levels by different laboratories.4,5 ity of left ventricular diastolic dysfunction17 and Despite these methodological issues, there has hypertrophy18 (perhaps linked to decreased noctur- been remarkable agreement on several clinical and nal ‘dipping’ of BP19) and increased incidence of stroke20 and severity of microalbuminuria.21 This end-organ damage pattern resembles that of salt- Correspondence: F Elijovich, MD, FAHA, Professor of Medicine, CHM, MSU, Grand Rapids Medical Education and Research sensitive hypertensive rats, such as DOC-salt and Center, 1000 Monroe Avenue NW, Grand Rapids, Michigan Dahl-salt-sensitive (SS). Most recently, Weinberger 49503, USA. E-mail: fernandoelijovichȰgrmerc.net and co-workers have provided evidence that salt- Endothelin and salt-sensitive hypertension F Elijovich and CL Laffer 460 sensitivity of BP of normotensive individuals carries All research on mechanisms of salt-sensitivity of with it the same detrimental long-term prognosis BP in humans or animals can be conceptualised, (mortality) as hypertension in salt-resistant within this framework, as experiments in which patients.22 This emphasises the need for continued either nature or the investigator has ‘clamped’ one research on the cause(s) of salt-sensitivity or on its of these systems, avoiding its normal regulation by intermediate mechanisms, particularly if the results salt intake and therefore producing salt-sensitivity of these studies can be exploited for the tailoring of BP (ie, a ‘shift’ in the pressure natriuresis curve, and targeting of specific therapy to salt-sensitive Figure 1, right panel). patients. ‘Clamping’ of pressor-antinatriuretic systems at a stimulated level that is only appropriate for a low- Complexity of the search for cause(s) of salt intake can be produced by: (a) exogenous administration (eg, experimental angiotensin II- salt-sensitivity of BP infusion and DOC-salt hypertension in the rat); (b) The identification of the gene or genes involved in inbreeding of a biochemical abnormality (eg, determining salt-sensitivity of BP would be facili- increased vasopressin levels24 or endogenous inhibi- tated by a precise characterisation of the salt-sensi- tors of nitric oxide synthase25 in Dahl SS rats; over- tive intermediate phenotype. There have been production of vasoconstrictor 20-HETE26,27 or innumerable studies of alterations in pressor/ impaired downregulation of sodium retaining insu- depressor and natriuretic/antinatriuretic systems in lin receptors28 in spontaneous hypertensive rat (SHR) kidney; and mutations in adducin with salt-sensitive animals and humans, attempting to + + refine the definition of this phenotype, and some increased Na /K pump activity in Milan rats29); or candidate genes have been proposed by linkage and (c) spontaneous gene polymorphisms in humans (eg, association studies (see review by Luft23). However, shorter intronic repeats of the 11-HSD2 gene with the search remains somewhat elusive, which is most decreased activity of the enzyme and increased cor- likely accounted for by the complexity of the factors tisol-induced mineralocorticoid effect,30 and and interactions that regulate salt balance. Figure 1 mutations in the G-protein ␤3 subunit with (left panel) provides a scheme, derived from the increased activity of the salt-retaining antiporter classical studies of Guyton and coworkers, indicat- transport system31). ing that the intact animal, as opposed to an isolated Analogously, depressor-natriuretic systems may kidney, is salt-resistant because during a salt load, be ‘clamped’ at their suppressed level only appro- several pressor-antinatriuretic systems are inhibited, priate for a low-salt intake, thus leading to salt-sen- while several depressor-natriuretic systems are con- sitivity of BP when the animal or subject is given a comitantly stimulated. This permits excretion of the salt-load. Examples include: (a) exogenous inhi- salt-load without the need to resort to pressure natri- bition of the natriuretic sensory nervous system,32 uresis. nitric oxide,33–35 or natriuretic eicosanoids;36 (b) knockout of the proANP gene in mice;37 (c) inbreeding of biochemical abnormalities such as deficient renal medullary natriuretic 20-HETE38 or nitric oxide33,34,39 in Dahl SS rats; and (d) blunting of ANP,40 nitric oxide,41 kallikrein,10,11 dopaminergic,15 and ␤2 adrenergic42 systems as described in salt-sensitive hypertensive humans. Thus, alterations in any of the multiple mechanisms that regulate salt balance may lead to salt- induced hypertension. This multiplicity, com- pounded by the interactions between these regulatory systems, has produced lack of definitive answers regarding the major causes of salt-sensitivity of BP.

Kidney transplantation experiments and salt-sensitivity of BP Figure 1 Schematic representation of the mechanisms through Experiments conducted soon after the development which salt-sensitivity of BP is established. On the left, there is of the Dahl strains of salt-sensitive (SS) and salt- no change in BP from a low to a high salt intake because there is concomitant stimulation of depressor-natriuretic and inhi- resistant rats (SR), provided an important clue for bition of pressor-antinatriuretic systems. Thus, the sodium load research on the mechanisms of salt-sensitivity of BP, is excreted without the need for pressure natriuresis. In contrast, ie, that the recipient animal acquires the BP of the one or several of these salt-regulating mechanisms have been kidney donor.43 This is even observed when the kid- clamped in the right panel, at the level that is physiologic for a low salt intake. In the absence of these regulatory mechanisms, neys of young (8–9 weeks old) Dahl SR are trans- BP increases in response to a salt-load, to participate in the resto- planted into salt-fed, adult Dahl SS rats, with nor- ration of salt balance via pressure-natriuresis. malisation of the established hypertension of the

Journal of Human Hypertension Endothelin and salt-sensitive hypertension F Elijovich and CL Laffer

44 461 latter. Analogous experiments were conducted in lish whether disruption of the ETB gene contributes the Milan strain of salt-sensitive rats with similar to hypertension via impaired natriuresis also. How- 45 results. The initial interpretation of these experi- ever, ETB receptor knockout mice are salt-sensitive, ments was that the transplanted abnormality had to and this abnormality does not reverse with ETA be one affecting renal sodium handling. However, blockers,54 suggesting an antinatriuretic mechanism when Dahl SS and Dahl SR rats are subject to in their hypertension. marked increase in salt intake over 96 hours, their Despite the findings above, there is definite evi- body weight, plasma volume and cardiac output do dence that the pressor action of excess ETA vaso- not differ significantly, despite a substantial increase constriction plays a role in several models of experi- of BP in Dahl SS only.46 These findings may be inter- mental hypertension in the rat (see below). The preted as an abnormal adaptation of total peripheral predominant vasoconstrictor role of endothelin in resistance to volume expansion in salt-sensitive rats, these experiments, as opposed to a possible pre- as if their kidneys produced a vasoconstrictor humo- dominant vasodilator-natriuretic role in normal ani- ral factor. These transplantation experiments and mals, may be accounted for by overexpression of haemodynamic measurements suggest that the vascular endothelin in certain forms of experimental search for the major causes determining salt-sensi- hypertension.55,56 In normal human volunteers, tivity of BP should be focused on renal factors. Can- most,57,58 but not all59 investigators have docu- didate mechanisms or substances should be those mented that there is tonic ETA-vasoconstriction, capable of producing renal or systemic vasoconstric- while all agree on the existence of ETB-vasodilation tion, subtle changes in natriuresis with shift in the in the control of regional blood flow.57,58 Analogous pressure-natriuresis curve, or a combination of both. observations have been made for the control of total peripheral resistance, which is diminished by ETA 60–62 Endothelins and BP regulation and increased by ETB blockers. The relative quantitative contribution of these two components The three 21-aminoacid endothelin peptides are to the regulation of normal BP has not been estab- potent vasoconstrictors. Although their name lished. In essential hypertension, vasoconstriction derives from their original discovery in the endo- of regional (forearm) vascular beds by ETA seems to 58,59 thelium, they are now known to be ubiquitous. be enhanced, and dilation by ETB abolished or Endothelial endothelin, or ET-1 acts on two recep- reversed.59 These changes may be due to increased 63,64 tors, ETA and ETB, located in adjacent endothelium ET-1 in resistance vessels of hypertension, the and vascular smooth muscle. The vasoconstrictor, synthesis of which is shifted from the endothelium proliferative and hypertrophic actions of ET-1 on to the vascular smooth muscle cells of the tunica vascular smooth muscle are exerted via both these media.64 47 receptors. In contrast, endothelial ETB stimulation Bosentan (a combined ETA/ETB receptor blocker) acts as an opposing vasodilator, by enhancing the exerted significant antihypertensive action in a large synthesis of relaxing compounds, such as nitric double-blind clinical trial,65 indicating that the oxide and prostaglandins.48 In the kidney, locally pressor component of the endothelin system is ton- generated endothelin produces ETA-mediated vaso- ically present in human hypertension. However, the constriction and antinatriuresis49 but also very magnitude of BP reduction by bosentan in this trial important ETB-mediated diuresis and natriuresis via was barely greater than that observed in normotens- inhibition of vasopressin-induced synthesis of cyc- ive humans. This could be interpreted as a lack of lic AMP,50 and of the activity of tubular Na+/K+ specific, exaggerated contribution of endothelin to ATPase.51 the vasoconstriction of essential hypertension. Endothelins were discovered and described pre- Alternatively, a larger antihypertensive effect in a dominantly as pressor agents. Whether this is true subset of patients may have been masked or diluted in the normal animal has recently become more con- by recruitment of an unselected hypertensive popu- troversial. For example, mice heterozygous for the lation, since an endothelin-dependent subset of knockout of the ET-1 gene were unexpectedly found essential hypertensive patients has not yet been to be hypertensive, not hypotensive.52 This paradox well characterised. was initially interpreted as an endothelin-inde- In summary, due to the multiplicity of their pendent, sympathetic-driven phenomenon, due to actions, the endothelins are obvious candidate sub- concomitant ventilatory abnormalities in these ani- stances for a causative or intermediate role in salt- mals. However, mice engineered for diminished sensitive hypertension. This could be related to expression of the ETB receptor gene (which have excessive ETA vasoconstriction or antinatriuresis, to neither the ventilatory abnormalities of the ET-1 defective ETB vasodilation or natriuresis, or to a knockout, nor the colonic abnormalities of the ETB combination of any of these four mechanisms. receptor knockout) also develop hypertension.53 Therefore, the seemingly counterintuitive view that Endothelin and experimental salt- in normal animals the endothelin system may be sensitive hypertension predominantly vasodilator has been recently Early research on the participation of endothelin in strengthened. There are no direct studies to estab- models of experimental hypertension in the rat

Journal of Human Hypertension Endothelin and salt-sensitive hypertension F Elijovich and CL Laffer 462 made it clear that overexpression of endothelial that plasma endothelin tends to be higher in hyper- endothelin and participation of this peptide in tensive patients with certain clinical and biochemi- development and maintenance of hypertension and cal characteristics, including African-American vascular remodelling were much more important in race,85 low PRA,86 obesity,87 and insulin resist- DOC-salt rats than in SHR.55,66,67 These observations ance.88 Because these characteristics are also clinical were extended by other laboratories and it can now and biochemical correlates of the salt-sensitive be concluded that endothelin plays more of a vaso- hypertension phenotype, it is not surprising that constrictor and hypertrophy-promoting role in salt- plasma endothelin was found to be elevated in dependent (eg, DOC-salt, aldosterone-induced, Dahl hypertensive patients classified as salt-sensitive SS, insulin-resistant, and 1K-1C Goldblatt with a chronic dietary protocol.89 Consistent with models),55,66,68–70 than in salt-independent or renin- this, we found that racial differences in endothelin dependent forms of experimental hypertension (eg, levels are not present when equal proportions of SHR, 2K-1C Goldblatt, and L-NAME-induced salt-sensitive and salt-resistant African American hypertension).55,56,67,70 The only exception seems to and Caucasian individuals are included in the be the hypertension produced by exogenous admin- study.84 istration of angiotensin II, in which endothelin par- Participation of endothelin in salt-sensitive essen- ticipates in BP maintenance and vascular hypertro- tial hypertension could be mediated by a deficit in phy.71 Continuous stimulation of endothelin its natriuretic actions, or by an exaggerated contri- production by exogenous angiotensin infusion72 bution to the control of vasoconstriction, parti- may outweigh the complex interactions between cularly during changes in salt balance. We will intermittent activation of the endogenous renin- review the evidence for these mechanisms in human angiotensin and endothelin systems, and thus salt-sensitive hypertension. explain this exception. Urine endothelin, salt-balance and salt- Endothelin and hypertension in humans sensitive hypertension in humans With the advent of sensitive and specific radio- Filtered circulating endothelin is degraded by neu- immunoassays, increased plasma levels of immuno- tral endopeptidase.90 Only negligible amounts of reactive endothelin were reported in several forms radiolabelled endothelin are recovered in the urine of secondary hypertension in humans, including after injection in the systemic circulation.91 There- cyclosporine-induced hypertension in transplant fore, urine endothelin predominantly reflects pro- patients,73 erythropoietin-induced hypertension in duction of local, renal endothelin. Fractional dialysis patients74 and sleep apnea-induced hyper- excretion of endothelin is about 1.0 in normal tension.75 In pre-eclampsia, concomitant increase of human beings,77,92 indicating that renal production plasma endothelin and endothelial elastase supports of this peptide is of a magnitude similar to that of participation of endothelial dysfunction in the its filtered load. hypertension.76 Patients with chronic renal failure Several lines of evidence support a major role for have increased local renal production of endothelin, tubular endothelin in the regulation of salt excretion which correlates negatively with glomerular fil- in normal and hypertensive subjects. Hence, the tration rate.77 In addition, their elevated endothelin excretory circadian rhythms for endothelin and plasma levels78 exert an unequivocal pressor action, sodium exhibit a parallel pattern,93 there are posi- as demonstrated by the antihypertensive effect of an tive correlations between natriuresis and urine 93,94 ETA/ETB receptor blocker in patients with renal dis- endothelin, between fractional excretion of ease.54 Analogously, in patients with endothelin- sodium and that of endothelin,92 and between secreting malignant haemoangioendotheliomas, changes in natriuresis and urine endothelin pro- removal of the tumour led to resolution of the hyper- duced by a salt load.95 Interrelationships between tension.79 the urine excretion of sodium, the natriuretic pep- Reports on plasma endothelin in essential hyper- tide adrenomedullin, and endothelin,95 and tension have been controversial, with increased increased urine excretion of endothelin after a salt levels published by some,80,81 but not all investi- load92–95 also support participation of this peptide gators.82,83 It has now become evident that two in regulation of salt balance, although the latter reasons account for this apparent controversy. First, observation has not been confirmed by all labora- plasma endothelin tends to be elevated when hyper- tories.96,97 tension is severe, with concomitant end organ dam- The basal urine excretion rate of endothelin is age. We have confirmed, in 47 essential hyperten- diminished in hypertension, more so in salt-sensi- sive patients, that strong correlations between mean tive patients, and there is a negative correlation ambulatory BPs and plasma endothelin were due to between urine endothelin and mean arterial press- a few individuals with very high BPs and endothelin ure (MAP) in normotensive and hypertensive sub- levels, while the relationship between these para- jects.96 These observations are consistent with a role meters was much more scattered in the mid-range for endothelin in salt-sensitivity of BP via dimin- for their respective values.84 The second reason is ished natriuresis. An analogous mechanism has

Journal of Human Hypertension Endothelin and salt-sensitive hypertension F Elijovich and CL Laffer 463 been proposed in SHR, since this strain exhibits (correlations between plasma endothelin and urine impairment of renomedullary synthesis of endo- microalbumin in salt-sensitive patients only97). thelin.98,99 Inability to confirm decreased absolute In contrast to these observations, the group of Bra- levels of endothelin excretion in hypertensive gulat et al did not detect differences in plasma endo- patients by other authors93,97 may reflect differences thelin between salt-sensitive and salt-resistant in endothelin plasma levels and renal function of hypertensive patients, after 1 week of low (50 mmol) the studied subjects. When these parameters are and 1 week of high (250 mmol) salt intake.102 It is taken into consideration, it is unequivocal that frac- noteworthy that this observation was made in tional excretion of endothelin is reduced almost patients who had unequivocal endothelial dysfunc- three-fold in essential hypertensive patients com- tion, as assessed by forearm blood flow responses to pared to normotensive controls.77 acetylcholine and urine excretion of nitrates. Con- Finally, in response to a salt load, hypertensive sistent with this observation, we did not detect dif- patients with delayed natriuresis have less of an ferences in plasma endothelin of salt-sensitive and increase in urine endothelin compared to those with salt-resistant patients when the measurements were the characteristic exaggerated natriuresis of hyper- carried out before initiation of an acute protocol of tension.93 Therefore, reduced basal urine endothelin salt loading and unloading, that is, while they still and blunted renal endothelin responsiveness to a were on their habitual salt intake.84 salt load most likely participate in the pathogenesis The differential effects of salt loading and salt of human salt-sensitive hypertension. deprivation on plasma endothelin of salt-sensitive and salt-resistant patients also vary, probably Plasma endothelin, salt-balance and salt- depending on the patient population and the experimental protocol employed. Thus, Ferri et al89 sensitive hypertension in humans reported increases in plasma endothelin from a low In contrast to the relatively well characterised prov- (20 mmol) to a high salt diet (220 mmol) given for enance of urine endothelin, the origin of circulating, 14 days. Although the increase was observed in both plasma endothelin is less well known. Conceivably, salt-sensitive and salt-resistant patients, the magni- a large part of it is of endothelial origin, reflecting tude was greater in the former (46%) than in the lat- very minimal spillover from the predominant ablu- ter (22%). In contrast, the patients of Bragulat et al102 minal (ie, toward the smooth muscle layer) secretion exhibited a slight (not significant) decrease of of endothelin by endothelial cells.100 Release of plasma endothelin when switching from one week organ-specific endothelin to the circulation, such as of 50 mmol NaCl intake to another week on that from the neurohypophysis101 is another poss- 250 mmol. There was no difference in the magni- ible source. Studies on the regulation of plasma lev- tude of this decrease between salt-sensitive (7%) els of endothelin of hypertensive subjects by salt and salt-resistant patients (9%). From these two balance have provided diverse results. When studies, it may be speculated that a sustained high patients are not characterised beyond a diagnosis of sodium intake may be required to produce and essential hypertension, sodium loading or detect impairment of endothelial function specific unloading does not seem to have a significant effect to salt-sensitive patients. on plasma levels of the peptide,92,95,96,102 an obser- We studied changes of plasma endothelin in vation that we have confirmed in our patients.84 response to acute manipulation of salt balance with Classifying patients into salt-sensitive and salt- an in-patient protocol. Salt-loading was achieved in resistant, Ferri et al reported 42–44% higher levels 24 h with a 160 mmol diet plus 300 mmol given as of plasma endothelin in the former group, 2 weeks intravenous saline over 4 h. Salt depletion was pro- after a 120 mmol NaCl diet.89,97 Although statisti- duced in the ensuing 24 h with a 10 mmol NaCl diet cally significant, these changes occurred near the and three 40 mg doses of furosemide.84 Essential lower limit of detection of the peptide, with values hypertensive patients were classified as salt-sensi- ranging between 0.65 and 1.30 pg/ml. Also, the tive, indeterminate or salt-resistant according to population of these studies was highly selected their BP response to salt-deprivation. There was no (with exclusion of females, severe hypertension, difference in plasma endothelin levels among the obesity, smoking, diabetes, and atherosclerosis), and three groups before the protocol, ie, during their experienced large attrition due to problems com- habitual salt intake, and there were no significant plying with the diets. It is therefore unclear whether changes in levels of the peptide in any of the three these findings can be extrapolated to all salt-sensi- groups due to salt-loading. In contrast, during acute tive hypertensive humans. Nonetheless, the authors salt-deprivation, salt-sensitive patients exhibited an speculated that increased plasma endothelin may be 18% increase in plasma endothelin while salt-resist- a marker of more severe endothelial dysfunction in ant subjects had a 10% decrease. Patients classified salt-sensitive patients than in their salt-resistant as indeterminate exhibited a minor change (+4%) counterparts (correlations between endothelin, von that was not different from zero (Figure 2). As a Willebrand factor and endothelial selectin89) and consequence of these opposite changes in endo- also that it may be involved in the causation of thelin in response to salt-deprivation, salt-sensitive endothelial damage, particularly in the kidney patients had significantly higher plasma levels in

Journal of Human Hypertension Endothelin and salt-sensitive hypertension F Elijovich and CL Laffer 464 Conclusion We have reviewed the possible role for urinary (ie, renal natriuretic) and circulating (most likely endothelial) endothelin in the pathogenesis of salt- sensitivity of BP. Figure 3 depicts a conceptual framework that unifies the work of the multiple laboratories discussed in the review. It is suggested that adeficit in the synthesis of renal tubular endothelin, analogous to that of the SHR strain, may be an initi- ating factor for salt-sensitivity of BP, via impaired natriuresis. Once salt-sensitive hypertension is Figure 2 Adapted from Elijovich et al (Circulation 2001; 103: established, exposure to chronic elevated salt intake 263–268), with permission of the publisher, Lippincott Williams may contribute to exaggerated hypertension, shear and Wilkins. Plasma endothelin (ET-1) levels in the salt replete stress, endothelial damage and further release of state (left panel), their changes during salt deprivation (middle panel) and levels after salt depletion (right panel) in essential endothelin, with spillover to the plasma and hypertension. Patients were classified as salt-resistant (SR), inde- increased circulating levels. In a seemingly para- terminate (IN) or salt-sensitive (SS) with an acute, in-patient pro- doxical way, episodic salt deprivation (be it due to tocol of salt-loading and unloading. Results of ANOVAs are indi- the circadian rhythm of salt intake, salt-restricted = cated in each panel. * different from the other two groups by diet or use of diuretics) also has the potential for contrasting of means. The changes of endothelin induced by salt deprivation were of opposite direction in SS vs SR patients, increasing circulating endothelin, particularly in resulting in higher endothelin levels in salt-depleted salt-sensi- low-renin essential hypertensives, via stimulation tive patients, compared with the other groups. by the sympathetic nervous system. At both extremes of salt balance, increased plasma (and perhaps tissue) endothelin will therefore con- the salt-depleted state, compared to the other two tribute to the vasoconstriction of salt-sensitive groups. We showed that these different responses hypertension and to its end-organ damage. Pro- were not due to classification of patients into groups gression of microalbuminuria and decreased glom- with arbitrary cut-offs, because the relationship erular filtration rate may switch the patient from this between salt-sensitivity of BP and change in plasma pattern peculiar to salt-sensitive hypertension to the endothelin due to salt deprivation was continuous pattern observed in chronic renal insufficiency, ie, throughout the patient population. Before initiation decreased clearance of circulating endothelin with of the protocol, there was a negative relationship further increase in plasma levels, and increased between plasma endothelin and PRA in all patients, as reported by others.94 During salt-deprivation, the changes in endothelin correlated with the changes in catecholamines. In view of this combination of findings, we speculated that in the presence of a blunted renin-angiotensin system (a characteristic of salt-sensitive hypertension), salt-deprivation leads to compensatory stimulation of other pressors, including catecholamine-driven endothelin secretion.103 In brief, baseline differences in plasma endothelin between salt-sensitive and salt-resistant patients are equivocal or at best, minor. However, salt-sensitive patients have exaggerated or abnormal increases in plasma circulating levels of the peptide in two situ- ations, chronic sustained severe salt-loading, and rapid salt deprivation. Although the mechanism of these responses seems to be different (salt-induced Figure 3 Depiction of a hypothetical sequence of events by which or pressure-induced endothelial damage for chronic endothelin (ET) participates in salt-sensitivity (SS) of hyperten- dietary loading vs compensation of a blunted renin- sion. The primary event is a decrease in the production of reno- angiotensin system during acute salt depletion), it is medullary endothelin that leads to diminished natriuresis and salt-sensitive hypertension. At both extremes of salt balance, clear that these observations may have therapeutic there are mechanisms that will lead to increase in plasma endo- implications because in both cases, increased thelin levels (see text). Increased circulating endothelin partici- plasma endothelin levels predict BP dependency on pates in further blood pressure (BP) elevation (vasoconstriction) the peptide, thus a therapeutic response to blockers and cardiovascular (CV) and renal damage (growth promoting effects and remodelling). Renal dysfunction closes the circle by of its receptors. Confirmation of this prediction increasing renal production and decreasing clearance of plasma = = awaits further trials with the forthcoming endo- endothelin. ETB B-type endothelin receptor, RAS renin-angio- thelin blockers in development by industry. tensin system, GFR = glomerular filtration rate.

Journal of Human Hypertension Endothelin and salt-sensitive hypertension F Elijovich and CL Laffer 465 renal production of vasoconstrictor vascular or mes- 14 Campese VM et al. Abnormal relationship between angial endothelin. This would close a vicious circle sodium intake and sympathetic nervous system of endothelin-induced end-organ damage in salt- activity in salt-sensitive patients with essential hyper- sensitive hypertension. tension. Kidney Int 1982; 21: 371–378. Proving whether decreased renal synthesis of 15 Gill JR Jr, Grossman E, Goldstein DS. High urinary dopa and low urinary dopamine-to-dopa ratio in salt- endothelin is a primary event in salt-sensitive sensitive hypertension. Hypertension 1991; 18: 614– hypertension will have to wait for experiments of 621. targeted, kidney-specific disruption of the 16 Raji A et al. Insulin resistance in hypertensives: effect expression of the ET-1, ETA or ETB genes. From the of salt sensitivity, renin status and sodium intake. J clinical standpoint, however, more important than Hypertens 2001; 19:99–105. establishing whether endothelin has a primary or 17 Musiari L et al. Early abnormalities in left ventricular secondary role, is to test the conceptual scheme diastolic function of sodium-sensitive hypertensive above in terms of its therapeutic implications. The patients. J Hum Hypertens 1999; 13: 711–716. hypothesis predicts that endothelin-receptor block- 18 Morimoto A et al. Sodium sensitivity and cardiovascular events in patients with essential hypertension. ers (perhaps more so the ETA specific) will have an enhanced therapeutic action on BP and more Lancet 1997; 350: 1734–1737. 19 Bihorac A et al. Association between salt sensitivity importantly on end-organ damage, in salt-sensitive and target organ damage in essential hypertension. Am hypertension, a contention that can easily be proven J Hypertens 2000; 13: 864–872. in direct fashion by appropriate clinical trials with 20 Campese VM. Salt sensitivity in hypertension: renal these agents. and cardiovascular implications. Hypertension 1994; 23: 531–550. References 21 Bigazzi R et al. Microalbuminuria in salt-sensitive patients: a marker for renal and cardiovascular risk 1 Weinberger MH, Fineberg NS. Sodium and volume factors. Hypertension 1994; 23: 195–199. sensitivity of blood pressure: age and pressure change 22 Weinberger MH et al. Salt sensitivity, pulse pressure, over time. Hypertension 1991; 18:67–71. and death in normal and hypertensive humans. Hyper- 2 Weinberger MH, Stegner JE, Fineberg NS. A compari- tension 2001; 37: 429–432. son of two tests for the assessment of blood pressure 23 Luft FC. Salt and hypertension at the close of the mil- responses to sodium. Am J Hypertens 1993; 6: 179– lennium. Wiener Klin Woch 1998; 110: 459–466. 184. 24 Bayorh MA et al. Possible mechanisms of salt-induced 3 Grim CE et al. An approach to the evaluation of genetic hypertension in Dahl salt-sensitive rats. Physiol Behav influences on factors that regulate arterial blood press- 1998; 65: 563–568. ure in man. Hypertension 1980; 2: I34–I42. 25 Matsuoka H et al. Asymmetrical dimethylarginine, an 4 Weinberger MH et al.Definitions and characteristics of endogenous nitric oxide synthase inhibitor, in experi- sodium sensitivity and blood pressure resistance. mental hypertension. Hypertension 1997; 29: 242–247. Hypertension 1986; 8: II127–II134. 26 Schwartzman ML et al. Cytochrome P450 4A 5 Kawasaki T et al. The effect of high-sodium and low- expression and arachidonic acid omega-hydroxylation sodium intakes on blood pressure and other related in the kidney of the spontaneously hypertensive rat. variables in human subjects with idiopathic hyperten- Nephron 1996; 73: 652–663. sion. Am J Med 1978; 64: 193–198. 27 Stec DE et al. Renal cytochrome P4504A activity and 6 Sullivan JM. Salt sensitivity: definition, conception, salt sensitivity in spontaneously hypertensive rats. methodology, and long-term issues. Hypertension 1991; 17: I61–I68. Hypertension 1996; 27: 1329–1336. 7 Falkner B, Kushner H. Interaction of sodium sensi- 28 Sechi LA. Mechanisms of insulin resistance in rat tivity and stress in young adults. Hypertension 1991; models of hypertension and their relationships with 17: I162–I165. salt sensitivity. J Hypertens 1999; 17: 1229–1237. 8 Rodriguez BL et al. Rise of blood pressure with age: 29 Manunta P, Barlassina C, Bianchi G. Adducin in essen- new evidence of population differences. Hypertension tial hypertension. FEBS Lett 1998; 430:41–44. 1994; 24: 779–785. 30 Ferrari P, Lovati E, Frey FJ. The role of the 11 beta- 9 Williams GH et al. Parallel adrenal and renal abnor- hydroxysteroid dehydrogenase type 2 in human malities in young patients with essential hypertension. hypertension. J Hypertens 2000; 18: 241–248. Am J Med 1982; 72: 907–914. 31 Schorr U et al. G-protein beta3 subunit 825T allele and 10 Ferri C et al. Urinary kallikrein and salt sensitivity in response to dietary salt in normotensive men. J Hyper- essential hypertensive males. Kidney Int 1994; 46: tens 2000; 18: 855–859. 780–788. 32 Wang DH, Li J, Qiu J. Salt-sensitive hypertension 11 Bellini C et al. The influence of salt sensitivity on the induced by sensory denervation: introduction of a new blood pressure response to exogenous kallikrein in model. Hypertension 1998; 32: 649–653. essential hypertensive patients. Nephron 1993; 65: 33 Tan DY, Meng S, Manning RD Jr. Role of neuronal 28–35. nitric oxide synthase in Dahl salt-sensitive hyperten- 12 Bakris G et al. Role of vasopressin in essential hyper- sion. Hypertension 1999; 33: 456–461. tension: racial differences. J Hypertens 1997; 15: 34 Rudd MA et al. Salt-induced hypertension in Dahl 545–550. salt-resistant and salt-sensitive rats with NOS II inhi- 13 Kojima S et al. A gender difference in the association bition. Am J Physiol 1999; 277: H732–H739. between salt sensitivity and family history of hyper- 35 Barba G et al. Effects of sodium intake on the pressor tension. Am J Hypertens 1992; 5:1–7. and renal responses to nitric oxide synthesis inhi-

Journal of Human Hypertension Endothelin and salt-sensitive hypertension F Elijovich and CL Laffer 466 bition in normotensive individuals with different progress in the clinic. Trends Pharmacol Sci 1998; 19: sodium sensitivity. J Hypertens 2000; 18: 615–621. 5–8. 36 Stec DE, Mattson DL, Roman RJ. Inhibition of renal 55 Lariviere R, Thibault G, Schiffrin EL. Increased endo- outer medullary 20-HETE production produces hyper- thelin-1 content in blood vessels of deoxycorticos- tension in Lewis rats. Hypertension 1997; 29: 315–319. terone acetate-salt hypertensive but not in spon- 37 John SW et al. Genetic decreases in atrial natriuretic taneously hypertensive rats. Hypertension 1993; 21: peptide and salt-sensitive hypertension. Science 1995; 294–300. 267: 679–681. 56 Sventek P et al. Vascular and cardiac overexpression 38 Roman RJ, Alonso-Galicia M, Wilson TW. Renal P450 of endothelin-1 gene in one-kidney, one clip Goldblatt metabolites of arachidonic acid and the development hypertensive rats but only in the late phase of two- of hypertension in Dahl salt-sensitive rats. Am J Hy- kidney one clip Goldblatt hypertension. J Hypertens pertens 1997; 10: 63S–67S. 1996; 14:57–64. 39 Azam M et al. Genetic mapping of soluble guanylyl 57 Verhaar MC et al. Endothelin-A receptor antagonist- cyclase genes: implications for linkage to blood press- mediated vasodilatation is attenuated by inhibition of ure in the Dahl rat. Hypertension 1998; 32: 149–154. nitric oxide synthesis and by endothelin-B receptor 40 Ferri C et al. Abnormal atrial natriuretic peptide and blockade. Circulation 1998; 97: 752–756. renal responses to saline infusion in nonmodulating 58 Taddei S et al. Vasoconstriction to endogenous endo- essential hypertensive patients. Circulation 1994; 90: thelin-1 is increased in the peripheral circulation of 2859–2869. patients with essential hypertension. Circulation 1999; 41 Gomez-Fernandez P et al. Hormonal proﬁle and par- 100: 1680–1683. ticipation of nitric oxide in salt-sensitive and salt- 59 Cardillo C et al. Role of endothelin in the increased resistant essential arterial hypertension. Nefrologia vascular tone of patients with essential hypertension. 2000; 20: 415–423. Hypertension 1999; 33: 753–758. 42 Kotanko P, Hoglinger O, Skrabal F. Beta 2-adrenocep- 60 Haynes WG et al. Systemic endothelin receptor block- tor density in ﬁbroblast culture correlates with human ade decreases peripheral vascular resistance and blood NaCl sensitivity. Am J Physiol 1992; 263: C623–C627. pressure in humans. Circulation 1996; 93: 1860–1870.

43 Tobian L et al. A comparison of the antihypertensive 61 Spratt JC et al. Systemic ETA receptor antagonism with potency of kidneys from one strain of rats susceptible BQ-123 blocks ET-1 induced forearm vasoconstriction to salt hypertension and kidneys from another strain and decreases peripheral vascular resistance in heal- resistant to it. J Clin Invest 1966; 45: 1080. thy men. Br J Pharmacol 2001; 134: 648–654. 44 Dahl LK, Heine M, Thompson K. Genetic inﬂuence of 62 Strachan FE et al. Systemic blockade of the endo- the kidneys on blood pressure: evidence from chronic thelin-B receptor increases peripheral vascular resist- renal homografts in rats with opposite predispositions ance in healthy men. Hypertension 1999; 33: 581–585. to hypertension. Circ Res 1974; 40:94–101. 63 Schiffrin EL et al. Enhanced expression of endothelin- 45 Bianchi G et al. Blood pressure changes produced by 1 gene in resistance arteries in severe human essential kidney cross-transplantation between spontaneously hypertension. J Hypertens 1997; 15:57–63. hypertensive rats and normotensive rats. Clin Sci 64 Rossi GP et al. Endothelin-1 and its mRNA in the wall Molec Med 1974; 47: 435–448. layers of human arteries ex-vivo. Circulation 1999; 99: 46 Greene AS et al. Role of blood volume expansion in 1147–1155. Dahl rat model of hypertension. Am J Physiol 1990; 65 Krum H et al. The effect of an endothelin-receptor 258: H508–H514. antagonist, bosentan, on blood pressure in patients 47 de Nucci G et al. Pressor effects of circulating endo- with essential hypertension. Bosentan Hypertension thelin are limited by its removal in the pulmonary Investigators. N Engl J Med 1998; 338: 784–790. circulation and by the release of prostacyclin and 66 Li JS, Lariviere R, Schiffrin EL. Effect of a nonselective endothelium-derived relaxing factor. Proc Natl Acad endothelin antagonist on vascular remodeling in Sci USA 1988; 85: 9797–9800. deoxycorticosterone acetate-salt hypertensive rats. 48 Schiffrin EL. Endothelin: potential role in hyperten- Hypertension 1994; 24: 183–188. sion and vascular hypertrophy. Hypertension 1995; 25: 67 Li JS, Schiffrin EL. Effect of chronic treatment of adult 1135–1143. spontaneously hypertensive rats with an endothelin 49 Wilkins FC et al. Systemic hemodynamics and renal receptor antagonist. Hypertension 1995; 25: 495–500.

function during long-term pathophysiological 68 Park JB, Schiffrin EL. ETA receptor antagonist prevents increases in circulating endothelin. Am J Physiol 1995; blood pressure elevation and vascular remodeling in 268: R375–381. aldosterone-infused rats. Hypertension 2001; 37: 50 Tomita K, Nonoguchi H, Marumo F. Effects of endo- 1444–1449. thelin on peptide-dependent cyclic adenosine mono- 69 d’Uscio LV et al. Structure and function of small phosphate accumulation along the nephron segments arteries in salt-induced hypertension. Hypertension of the rat. J Clin Invest 1990; 85: 2014–2018. 1997; 30: 905–911. 51 Zeidel ML et al. Endothelin, a peptide inhibitor of 70 Sventek P, Turgeon A, Schiffrin EL. Vascular endo- Na(+)-K(+)-ATPase in intact renal tubular epithelial thelin-1 gene expression and effect on blood pressure

cells. Am J Physiol 1989; 257: C1101–C1107. of chronic ETA endothelin receptor antagonism after 52 Kurihara Y et al. Elevated blood pressure and cranio- nitric oxide synthase inhibition with L-NAME in nor- facial abnormalities in mice deﬁcient in endothelin-1. mal rats. Circulation 1997; 95: 240–244. Nature 1994; 368: 703–710. 71 Moreau P et al. Angiotensin II increases tissue endo- 53 Ohuchi T et al. Elevation of blood pressure by genetic thelin and induces vascular hypertrophy: reversal by and pharmacological disruption of the ETB receptor in ET(A)-receptor antagonist. Circulation 1997; 96: mice. Am J Physiol 1999; 276: R1071–1077. 1593–1597. 54 Webb DJ et al. Endothelin: new discoveries and rapid 72 Ito H et al. Endothelin-1 is an autocrine/paracrine

Journal of Human Hypertension Endothelin and salt-sensitive hypertension F Elijovich and CL Laffer 467 factor in the mechanism of angiotensin II-induced patients with essential and renovascular hyperten- hypertrophy in cultured rat cardiomyocytes. J Clin sion. J Hum Hypertens 1997; 11: 795–800. Invest 1993; 92: 398–403. 89 Ferri C et al. Clustering of endothelial markers of vas- 73 Stockenhuber F et al. Plasma levels of endothelin in cular damage in human salt-sensitive hypertension: chronic renal failure and after renal transplantation: inﬂuence of dietary sodium load and depletion. impact on hypertension and cyclosporin A-associated Hypertension 1998; 32: 862–868. nephrotoxicity. Clin Sci 1992; 82: 255–258. 90 Abassi ZA et al. Role of neutral endopeptidase in the 74 Carlini R, Obialo CI, Rothstein M. Intravenous erythro- metabolism of endothelin. Hypertension 1992; 20: poietin (rHuEPO) administration increases plasma 89–95. endothelin and blood pressure in hemodialysis 91 Benigni A et al. Increased renal endothelin pro- patients. Am J Hypertens 1993; 6: 103–107. duction in rats with renal mass reduction. Am J Phy- 75 Phillips BG et al. Effects of obstructive sleep apnea on siol 1991; 260: F331–339. endothelin-1 and blood pressure. J Hypertens 1999; 17: 92 Saito I et al. Urinary endothelin and sodium 61–66. excretion in essential hypertension. Nephron 1993; 76 Greer IA et al. Endothelin, elastase, and endothelial 65: 152–153. dysfunction in pre-eclampsia. Lancet 1991; 337: 558. 93 Hwang YS et al. Circadian rhythm of urinary endo- 77 Zoccali C et al. Urinary and plasma endothelin 1 in thelin-1 excretion in mild hypertensive patients. Am essential hypertension and in hypertension secondary J Hypertens 1998; 11: 1344–1351. to renoparenchymal disease. Nephrol Dial Transplant 94 Malatino LS et al. Renal endothelin-1 is linked to 1995; 10: 1320–1323. changes in urinary salt and volume in essential 78 Takahashi K, Totsune K, Mouri T. Endothelin in hypertension. J Nephrol 2000; 13: 178–184. chronic renal failure. Nephron 1994; 66: 373–379. 95 Cuzzola F et al. Urinary adrenomedullin is related to 79 Yokokawa K et al. Hypertension associated with endo- ET-1 and salt intake in patients with mild essential hypertension. Am J Hypertens 2001; 14: 224–230. thelin-secreting malignant hemangioendothelioma. 96 Hoffman A et al. Urinary excretion rate of endothelin- Ann Int Med 1991; 114: 213–215. 1 in patients with essential hypertension and salt sen- 80 Saito Y et al. Increased plasma endothelin level in sitivity. Kidney Int 1994; 45: 556–560. patients with essential hypertension. N Engl J Med 97 Ferri C et al. Elevated plasma and urinary endothelin- 1990; 322: 205. 1 levels in human salt-sensitive hypertension. Clin 81 Kohno M et al. Plasma immunoreactive endothelin in Sci 1997; 93:35–41. essential hypertension. Am J Med 1990; 88: 614–618. 98 Hughes AK, Cline RC, Kohan DE. Alterations in renal 82 Schiffrin EL, Thibault G. Plasma endothelin in human endothelin-1 production in the spontaneously hyper- essential hypertension. Am J Hypertens 1991; 4: 303– tensive rat. Hypertension 1992; 20: 666–673. 308. 99 Kitamura K et al. Immunoreactive endothelin in rat 83 Shichiri M et al. Plasma endothelin levels in hyperten- kidney inner medulla: marked decrease in spon- sion and chronic renal failure. Hypertension 1990; 15: taneously hypertensive rats. Biochem Biophys Res 493–496. Comm 1989; 162:38–44. 84 Elijovich F et al. Regulation of plasma endothelin by 100 Wagner OF et al. Polar secretion of endothelin-1 by salt in salt-sensitive hypertension. Circulation 2001; cultured endothelial cells. J Biol Chem 1992; 267: 103: 263–268. 16066–16068. 85 Ergul S et al. Racial differences in plasma endothelin- 101 Kaufmann H, Oribe E, Oliver JA. Plasma endothelin 1 concentrations in individuals with essential hyper- during upright tilt: relevance for orthostatic hypoten- tension. Hypertension 1996; 28: 652–655. sion? Lancet 1991; 338: 1542–1545. 86 Letizia C et al. High plasma endothelin-1 levels in 102 Bragulat E et al. Endothelial dysfunction in salt- hypertensive patients with low-renin essential hyper- sensitive essential hypertension. Hypertension 2001; tension. J Hum Hypertens 1997; 11: 447–451. 37: 444–448. 87 Parrinello G et al. Central obesity and hypertension: 103 Kaddoura S et al. Endothelin-1 is involved in norepi- the role of plasma endothelin. Am J Hypertens 1996; nephrine-induced ventricular hypertrophy in vivo. 9: 1186–1191. Acute effects of bosentan, an orally active, mixed 88 Zaporowska-Stachowiak I, Gluszek J, Chodera A. Com- endothelin ETA and ETB receptor antagonist. Circu- parison of the serum insulin and endothelin level in lation 1996; 93: 2068–2079.

Journal of Human Hypertension